1
REVIEW OF LITERATURE
I. Diffuse Large B-cell Lymphoma
Classification
Diffuse large B-cell lymphoma (DLBCL) includes a variety of
different intermediate-to high-grade lymphomas derived from mature B
cells. All DLBCL cases share a diffuse growth pattern and predominant
large tumor cell size, defined as being at least twice as large as a normal
lymphocyte or with a nuclear size equal to or greater than that of a normal
macrophage. Large cell lymphomas of T cell, natural killer cell, or unclear
lineage are classified separately. Other terms used for DLBCL include
centroblastic or immunoblastic lymphoma, lymphosarcoma, and
reticulosarcoma. In the recent WHO classification 2008 of hematopoietic
malignancies, specific subtypes of DLBCL are defined based on the cell of
origin, presumed pathogenesis (e.g., viral association), or their predominant
site of involvement (Table 1), whereas those that do not fit any of the
individual entities are considered DLBCL, not otherwise specified (NOS).
The recognized subtypes of DLBCL in the current WHO classification
represent a fraction of the heterogeneity observed in this common tumor, so
further changes in subclassification will likely occur in the coming years
(Swerdlow, 2008).
2
Incidence, Demographics, and Clinical Features
DLBCL is the most common subtype of adult lymphoma
worldwide, comprising 30–40% of cases in Western countries and
approximately 50% of lymphomas in China and many parts of Asia.
DLBCL arises in all age groups, but occurs most commonly in the elderly,
with approximately 25,000 new cases per year in the United States
(Rueffer, 2001). DLBCL represents about 54.67% of NHL cases in
national cancer institute (NCI), cairo (Mokhtar et al, 2007). Along with the
Burkitt lymphoma (BL), DLBCL comprises the majority of all childhood
B-cell lymphomas and is also the most common type of lymphoma
associated with genetic and acquired immunodeficiency states. DLBCL can
also arise rarely as a therapy-associated malignancy, most commonly
following breast cancer or Hodgkin’s lymphoma (Rueffer, 2001).
Table (1). Subclassification of diffuse large B-cell lymphoma.
Categories in the 2008 WHO classification: Alternative classification:
1)DLBCL, NOS
2)DLBCL, specific histogenic variants
Primary mediastinal DLBCL
T-cell/histiocyte-rich DLBCL
ALK+ DLBCL
1)Morphologic variants
Centroblastic
Immunoblastic
Anaplastic
Plasmablastic
3
Continued table (1)
3)DLBCL, distinctive extranodal variants
Primary DLBCL of the CNS
Primary cutaneous DLBCL, leg type
Intravascular DLBCL
4)DLBCL, primarily associated with viral
infection
EBV-associated DLBCL of the elderly
Lymphomatoid granulomatosis
DLBCL associated with chronic
inflammation
Plasmablastic lymphoma
Primary effusion lymphoma
Large B-cell lymphoma arising from
HHV8+ multicentric Castleman disease
5)DLBCL, unclassifiable
intermediate between DLBCL and Burkitt
lymphoma
intermediate between DLBCL and
classical Hodgkin lymphoma
2)Immunophenotypic variants
GCB-like
Non-GCB
3)Molecular subgroups
Lymphoma: GCB-like vs.
ABC-like
Stroma: Immune response
vs.angiogenic types
4
DLBCL can present with localized or generalized lymphadenopathy
and is also the most common lymphoma type in nearly every extranodal
site. Symptoms are highly dependent on the site(s) of presentation. Routine
clinical evaluation requires anatomic staging, commonly using the Ann
Arbor system, which is combined with laboratory data to derive the
International Prognostic Index (IPI) (Table 2) (Armitage, 2005). DLBCL is
clinically aggressive, but potentially curable due to its high proliferation
rate, especially in those tumors presenting at a low stage with a low IPI
score. However, each DLBCL subtype has variable patterns of treatment
response, relapse, and progression. Multi-agent combination chemotherapy
with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP)
is most commonly used for DLBCL and shows long-term remission rates
in up to 40% of patients. The addition of the anti-CD20 antibody
(Rituximab) has improved overall survival by 10–15%. However, DLBCL
still accounts for nearly 10,000 cancer deaths per year in the United States
(Dong, 2010). Extensive molecular stratification and risk prediction
modeling of DLBCL have been actively investigated in recent years in an
attempt to predict therapy response and relapse as well as to better define
mechanisms of transformation. DLBCL may also arise as large cell
transformation of a low-grade B-cell malignancy, such as follicular
lymphoma (FL), marginal zone lymphoma (MZL), nodular lymphocyte
predominant Hodgkin lymphoma (NLP-HL), classical Hodgkin lymphoma
(CHL), or chronic lymphocytic leukemia/small lymphocytic lymphoma
(CLL/SLL). Such transformed DLBCL shares many of the same genetic
features as de novo DLBCL and is usually treated similarly. In some cases,
the underlying low-grade lymphoproliferative neoplasm may be
unrecognized at diagnosis, but can be suspected if cytogenetic studies
5
reveal the genetic characteristics of a particular low-grade B-cell
lymphoma (Swerdlow, 2008).
Table (2): Clinical parameters affecting prognosis in lymphoma.
International Prognostic Index (IPI)
Unfavorable prognostic factors
Age >60 yrs
Poor performance status (ECOG ≥2)
Extranodal involvement ≥2 sites
High Ann Arbor stage (III or IV)
High lactate dehydrogenase (LDH)
Risk group Score (5 factors)
All patients > 60yrs
Low
Low-intermediate
High-intermediate
High
0 or 1
2
3
4 or 5
0
1
2
3-4
Eastern Cooperative Oncology Group (ECOG) performance status grading
0 Fully active without restriction
1 Restricted in physically strenuous activity
2 Ambulatory and capable of all self-care but unable to carry out any work
activities.
6
3 Capable of only limited self-care, confined to bed or chair more than 50% of
waking hours.
Continued table (2)
4 Completely disabled. Cannot carry on any self-care. Totally confined to bed or
chair.
Ann Arbor staging system for lymphoma
I Involvement of a single lymph node region or lymphoid structure (eg, spleen,
thymus, Waldeyer’s ring)
II Involvement of two or more lymph node regions on the same side of the
diaphragm
III Involvement of lymph regions or structures on both sides of the diaphragm
IV Involvement of extranodal site(s) beyond that designated E
For all stages
A No symptoms
B Fever (>38oC), drenching sweats, weight loss (10% body weight over 6
months)
For stages I to III
E Involvement of a single, extranodal site contiguous or proximal to known nodal
site.
Morphological Evaluation of DLBCL
Nodal DLBCL usually partially or completely effaces the
architecture with frequent extra-capsular extension and grows as either
7
sheets or as individual cells dispersed in a dense background of non-
neoplastic/inflammatory cells. Various degrees of sclerosis (in slowly
growing tumors) and necrosis (in rapidly-growing tumors) are present.
Extra-nodal DLBCL diffusely invades and replaces normal structures. In
cytologic preparations, such as fine needle aspirate (FNA) smears and
touch imprints, DLBCL can be readily recognized because of its large cell
size, prominent nucleoli, and irregular nuclear contours with nose-like
protrusions or indentations (Dong, 2008). Morphology subtypes of DLBCL
include centroblastic, immunoblastic, plasmablastic and anaplastic.
Centroblastic DLBCL, the most common morphologic variant, resembles
GC centroblasts with their round nuclei, vesicular, fine chromatin, and 2–4
distinct membrane-bound nucleoli, and discernible amphophilic cytoplasm.
DLBCL with polylobate nuclei are often grouped with these cases are more
common at certain extranodal sites, such as bone. Immunoblastic DLBCL
are those where at least 90% of tumor cells resemble reactive
immunoblasts with their prominent single, central nucleolus, and moderate
amounts of amphophilic cytoplasm. Plasmablastic DLBCL has eccentric
nuclei with prominent nucleoli, and abundant, often basophilic cytoplasm.
Anaplastic DLBCL shows large polygonal, spindle- or bizarre-shaped
tumor cells and often angulated or multilobated nuclei. Reed–Sternberg
(RS)-like cells are common in this variant as well as atypical mitoses and
frequently admixed inflammatory cells raising the differential diagnosis of
CHL (Swerdlow, 2008).
Immunophenotyping
Pan-B cell markers used for diagnosis of DLBCL by flow cytometry
(FCM) include CD19, CD20, CD22, and CD79a, whereas CD20, CD79a,
8
and PAX5 are commonly used for immunohistochemistry (IHC) in fixed
sections. Clonality can be confirmed by demonstrating immunoglobulin
(Ig) light chain restriction, defined as Igκ + cells greater than 4X the
number of Igλ + cells, or Igλ + cells more than twice the number Igκ+
cells, using FCM, or by in situ hybridization (ISH) or IHC in fixed sections
in some cases. Absence or dim expression of one or more pan-B antigens is
a characteristic feature of some DLBCL subtypes at diagnosis or status-
post treatment, particularly CD20 loss after use of rituximab. For these
cases, other markers such as Ig heavy and/or light chains, CD22, CD79a,
PAX5, and CD138 may help. CD138 is particularly useful in detecting
plasmablastic DLBCL (which often lack CD20) and these cases often also
show cytoplasmic Ig light chain restriction detectable by IHC or ISH.
Given its crisp nuclear staining pattern, the B-cell marker PAX5 can
highlight tumor cases in cases with marked sclerosis or crush artifact, such
as primary mediastinal B-cell lymphoma (PMBL). However, in DLBCL
with extensive necrosis, CD20 immunostaining is usually retained, whereas
nuclear markers, including PAX5, BCL1/cyclin D1, BCL6, MUM1, and
Ki-67, will generally fail (Dunphy, 2010). Many other markers that are
routinely used in IHC are variably expressed in DLBCL. Strong CD30
expression is characteristic of DLBCL with a sinusoidal distribution, those
associated with EBV, and cases with anaplastic morphology. About 10% of
de novo DLBC are positive for CD5 and DLBCL may also aberrantly
express T/NK cell antigens, such as CD2, CD7, CD8, and CD56. DLBCL
associated with EBV may be identified by in situ hybridization for EBV-
encoded RNAs (EBER), or expression of EBV latent membrane protein-1
(LMP1) (Rosenwald, 2002).
9
Molecular Testing
Molecular genetic assays may be useful in demonstrating B-cell
clonality in difficult cases of DLBCL. However, although essentially all
cases of DLBCL have clonally rearranged immunoglobulin genes, IGH
polymerase chain reaction (PCR) assays can give a false-negative result in
up to 15–20% of cases due to somatic mutations in the IGH primer binding
sites which prevent amplification. The monoclonal nature of DLBCL can
also be established by detecting chromosomal translocations involving IGH
and/or BCL6 genes that occur at high frequency in DLBCL. Translocations
involving the BCL6 gene at chromosome 3q27 occur in about 30% of
DLBCL, but they are not subtype specific since they occur in 5–10% of
follicular lymphoma and rarely in T-cell lymphomas (Sahai et al, 2011).
About 30% of DLBCL have the t (14;18)(q32;q21) translocation producing
the IGH/BCL2 fusion, which may or may not be related to transformation
from an underlying follicular lymphoma. Finally, MYC translocations
involving the immunoglobulin loci, while present in all cases of BL, also
occur in 10–15% of DLBCLs or aggressive B-cell lymphomas with mixed
features. Use of "break-apart" FISH probes which detect IGH, MYC, or
BCL6 translocations irrespective of the partner genes increases the
detection frequency for these translocations and may be considered as an
alternative clonality assay, especially when the IGH PCR fails to detect
clonality (Dong, 2008).
10
Molecular and Immunophenotypic Subclassification
Gene expression profiles (GEP) established using large-scale gene
expression arrays have led to recognition of several DLBCL subtypes with
distinct clinical outcomes. The best studied are the germinal center B cell
(GCB)-like and the activated B cell (ABC)-like patterns (Rosenwald,
2002). The GCB and ABC signatures reflect fundamental differences in
expression of hundreds of genes that largely match GCB and post-GC
states of B-cell maturation. The GCB-like type highly expresses many
genes that are also expressed in non-neoplastic GCB, and usually shows
ongoing immunoglobulin gene somatic hypermutation (Lossos, 2000),
whereas the ABC-like type shows expression of growth factor and
signaling molecules seen in post-GC B cells where somatic hypermutation
has ceased. DLBCL with plasmacytoid differentiation and most EBV-
associated DLBCL have ABC-like features, whereas primary mediastinal
large B-cell lymphoma (PMBL) has a distinct GEP pattern. De novo CD5+
DLBCL may have an either GCB- or ABC-like GEP. As a group, the
GCB-like DLBCL has a significantly better clinical outcome than the
ABC-like type among patients treated with CHOP, with 5-year survival
rates of 60 and 35%, respectively (Rosenwald, 2002). Some recent studies
have shown that these outcome differences also apply to R-CHOP-treated
patients (Lenz, 2008). In addition, GCB- and ABC-like DLBCL correlate
with particular genomic changes reflective of distinct patterns of oncogenic
progression For example, a subset of GCB cases show IGH/BCL2
rearrangements or amplification of the REL locus at chromosome 2q15,
11
features not seen in ABC-like DLBCL. In contrast, ABC-like cases display
trisomy 3 and amplification of a small region on chromosome19,
corresponding to upregulation of the ABC-associated genes FOXP1 and
SPIB. These cases may also exhibit gains of chromosome 18q spanning the
BCL2 and MALT1 loci, as well as deletions of chromosome 9p involving
the CDKN2A (p16/INK4a) locus (Lenz, 2008). A number of studies have
reported IHC-based correlates of the GCB vs. ABC molecular schema for
DLBCL, most often using stains for CD10, BCL6, and MUM1 in the
commonly used Hans scoring system (Hans et al, 2004). GCB-like cases
are positive for CD10 (with or without BCL6) and negative for MUM1,
whereas ABC-like DLBCLs express MUM1 and lack CD10. Most studies
agree that MUM1 (which is expressed in approximately 35% of DLBCL) is
mutually exclusive with CD10 expression. BCL6 can be expressed in either
GCB or ABC cases as a result of an activating promoter mutation or a
chromosomal translocations and in those cases its expression will not
reflect cell of origin. Likely for these reasons and due to differences in the
quality and scoring criteria of the immunostains, IHC studies have not
clearly shown differences in outcome stratification of DLBCL. Recent
studies have added additional GCB-like markers such as HGAL (Lossos,
2003) and LMO2 (Natkunam et al. 2008) and other ABC-like markers
such as FOXP1(Banham, 2005) which significantly improved outcome
prediction in patients treated with either CHOP or R-CHOP. Nevertheless,
there is not yet a consensus on the routine use of immunophenotyping for
treatment stratification or on what markers would constitute the best
routine IHC panel to determine prognosis and guide therapy selection. GEP
studies of DLBCL also identified distinct molecular signatures related to
tumor microenvironment that predict survival independently (Sahai et al,
2011). Using cohorts of both CHOP and R-CHOP-treated patients, a
12
“stroma-1” signature, enriched for genes responsible for increased host
response, predicted a significantly better prognosis reminiscent of GCB-
like DLBCL. It was also associated with histologic evidence of increased
deposition of extracellular proteins and a prominent histiocytic infiltrate. A
“stroma-2” signature rich in genes highly expressed in endothelial cells and
in those encoding angiogenic factors and regulators predicted a poor
prognosis and correlated with increased angiogenesis seen in the biopsies
(Lenz, 2008).
Prognostic Factors in DLBCL
In addition to an ABC-like GEP and IPI clinical factors, high-risk
genetic changes in DLBCL include complex karyotype, MYC
translocation, and P53 loss (or p53 mutation). DLBCL with a MYC/IGH
translocation do worse than those carrying a MYC translocation with other
partner genes. Relapsed DLBCL frequently acquires additional genetic
abnormalities and unstable karyotypes. Relapsed DLBCL that has been
previously treated with rituximab may show a higher false-negative rate in
nuclear medicine scans (Shipp et al. 2002).
Differential Diagnosis
DLBCL with predominantly medium-sized cells, especially if
cytoplasmic vacuoles are present on smears should be distinguished from
Burkitt lymphoma (BL), and cases with an anaplastic morphology may
mimic Classic Hodgkin lymphoma (CHL) and/or anaplastic large cell
lymphoma (ALCL). None of the commonly used pan-B markers are
13
specific for B-cell tumors. CD20 and CD79a can be expressed in some T-
cell lymphomas. Neuroendocrine carcinomas (Merkel cell carcinoma and
small cell carcinoma) and CD19+ acute myeloid leukemias express PAX5.
MUM1 is expressed in CHL and T-cell lymphoma, and CD138 is
expressed in a number of non-hematopoietic neoplasms (Dunphy, 2010).
Histogenetic Variants of DLBCL:
1) Primary Mediastinal Large B-cell Lymphoma (PMBL):
PMBL comprises 6–10% of DLBCL and it is twice as common in
woman (median age, 35years). PMBL is believed to arise from a
specialized population of thymic B cells and warrants separate recognition
because of its distinct clinical, morphologic, and genetic features.
Lymphomas presenting with concurrent involvement of distant lymph
nodes or bone marrow are excluded from this entity. It usually responds
well to initial treatment and has a relatively good prognosis. PMBL is
composed of intermediate to large cells with frequently irregular or
multilobate nuclear contours and moderate amounts of clear cytoplasm
embedded in fibrosis with a reticulated or alveolar pattern. Sclerosis may
be absent in rapidly growing tumors, whereas zonal necrosis is a common
finding. PMBL cells express CD20, PAX5, and BCL6, with MUM1
positivity in up to 75% supporting an origin from post-GC activated B cells
in most cases. More than 50% of cases show at least focal CD30 reactivity,
but the staining is usually less uniform compared to CHL. Furthermore, the
transcription factors BOB1 and OCT2 are positive, whereas they are
usually negative or dim in CHL. CD10 and CD15 are usually negative, and
EBER is always negative. FCM has only limited diagnostic value because
14
of the tissue sclerosis and the frequent absence of surface immunoglobulin
on the tumor cells. (Winter et al. 2006).
2) T-cell/Histiocyte-Rich Large B-cell Lymphoma (TCHR-LBCL):
TCHR-LBCL is a DLBCL variety characterized by numerous non-
neoplastic T cells and/or histiocytes. It comprises less than 10% of DLBCL
and is most common in middle-aged men (median age, 49 years). It often
involves deep lymph nodes in the retroperitoneum, as well as the liver,
spleen, and bone marrow. Patients usually have high clinical stage at
presentation, and tumors are largely refractory to R-CHOP chemotherapy.
In lymph node, TCHR-LBCL shows diffuse effacement of the architecture,
resulting in an abnormal appearance even on needle core biopsy. Whether
in lymph node or extranodal sites, the neoplastic cells typically comprise
less than 10% of the cellularity and are dispersed amid dense populations
of small T cell and/or sheets of large epithelioid histiocytes. Small B cells,
eosinophils, plasma cells, and neutrophils are usually completely absent.
Tumor cell morphology is variable, sometime resembling Reed–Sternberg
cells, with other cases having multilobate nuclei resembling the popcorn
cells of NLPHL. Given the T-cell predominance, tumor cells may be
missed by FCM so IHC is recommended in all cases (Boudova, 2003).
CD20 and BCL6 are uniformly positive, MUM1 is variable, and CD5,
CD10, CD30, and BCL2 are expressed in only a minority of cases. TCHR-
LBCL is likely a heterogeneous disorder, with some cases develop as a
transformation of preexisting NLPHL, and the two tumors can even coexist
in the same lymph node. Therefore, distinction of these two entities can be
difficult whenever NLPHL acquires a diffuse growth pattern and has few
associated small B cells. Spread of NLPHL outside of lymph nodes to liver
15
or bone marrow is a sign of transformation to TCHR-LBCL (Swerdlow,
2008).
3) DLBCL Expressing ALK:
This is an extremely rare lymphoma comprising much less than 1% of
DLBCL that has also been reported as ALK+ plasmablastic lymphoma. It
occurs in all age groups, but is most common in young to middle-aged
men. Despite having been reported at extranodal sites, it is generally an
aggressive nodal lymphoma and is insensitive to rituximab due to lack of
CD20 expression. ALK+ DLBCL displays a sinusoidal distribution or
forms large tumor nodules both grossly and microscopically mimicking
carcinoma and melanoma. The lymphoma displays monotonous
immunoblast-like cytology, but the cell size is much larger than regular
immunoblasts. Like other lymphomas with sinusoidal localization, ALK+
DLBCL may have prominent membrane villous projections. ALK+
DLBCL is negative for B-cell or T-cell markers, as well as CD30 (unlike
ALK + ALCL) and CD45, but expresses granular cytoplasmic ALK and
plasma cell antigens such as CD138, MUM1, and EMA (Linderoth et al,
2003).
4) Extranodal Variants of DLBCL:
a. Primary DLBCL of the Central Nervous System (CNS):
DLBCLs may present in the CNS from systemic spread (often from
PMBL), in immunosuppressed patients (often HIV+), or in
immunocompetent patients. The latter is known as primary DLBCL of the
CNS. Extracranial metastasis and bone marrow involvement are very rare.
16
Biopsy usually shows peri-vascular parenchymal infiltration by clusters of
tumor cells and subtle meningeal spread, but the limited sampling afforded
by brain biopsies will sometimes show only necrosis, foamy histiocytes,
and rare CD20+ cells. Tumor cells are positive for BCL2, MUM1 (~90%),
and BCL6 (60–100%) with mutated IGH consistent with a post-GC origin
(Montesinos-Rongen, 2009).
b. Primary Cutaneous DLBCL, Leg Type:
Most DLBCLs involving the skin are localized to the dermis (Stage
IE), have an indolent course and can be easily treated with excision and/or
radiotherapy. However, there is an aggressive variant of primary skin
DLBCL typically presenting as bluish nodules on the lower extremities (in
contrast to the trunk, head, and neck involvement seen more commonly in
indolent cases), which has been named “leg type”. Such cases comprise
about 20% of primary cutaneous B-cell lymphoma and commonly occur in
elderly women having a 5-year survival of only 50%. The presence of
multiple tumor nodules and rapid systemic spread are adverse risk factors.
Both immunoblastic and centroblastic morphology are common, and
epidermal or adnexal involvement is absent. Tumor cells display molecular
signatures consistent with ABC-like DLBCL (Hoefnagel, 2005) and
express BCL2, BCL6, MUM1, and FOXP1, but lack CD10 expression
commonly seen in most of the cutaneous DLBCL of GCB origin. Deletions
of chromosome 9p21 spanning the CDKN2A tumor suppressor are seen in
67% of cases. Translocations involving IGH, BCL6, and MYC may be
seen (Dunphy, 2010).
17
c. Intravascular DLBCL:
This is an exceedingly rare type of extranodal DLBCL with a
characteristic intravascular pattern of growth that can be diagnosed
incidentally during biopsy for other reasons. It often presents with
symptoms related to vasculitis and thrombosis in various organs such as
brain (stroke and infarct), lung (pneumonia and pulmonary embolism),
kidney (infarction or renal insufficiency), adrenal (endocrine disorders), or
skin. Although there is often preferential involvement of some organs, it is
a systemic disease with bone marrow involvement in nearly all cases that
can be demonstrated by immunostaining. Delays in diagnosis due to the
subtlety and focal nature of the intravascular infiltrates often lead to a
terminal disease with death before chemotherapy is initiated. Most cases
have an immunoblastic or anaplastic appearance and express BCL2 and
MUM1, consistent with a post-GC/ABC-like origin (Dong, 2008).
5) Variants of DLBCL Associated with Viral Infections:
a. The Role of Herpes viruses in DLBCL:
Viral-associated DLBCL are the most common malignancies in
patients with an underlying immunodeficiency, whether it is related to
primary immunodeficiencies, chronic iatrogenic immunosuppression,
human immunodeficiency virus (HIV) infection and other chronic
infections, transplantation, or age-related declines in immune function. The
two primary viral agents involved in lymphomagenesis are the gamma
18
herpesviruses, EBV and human herpesvirus 8 (HHV8). EBV is endemic
among all populations worldwide, with sero-positivity rates of over 90% in
adults in the United States. However, geographic differences in genetic
strains of the virus, age of primary infection, and patient socioeconomic
status lead to differences in the profile of EBV-associated malignancies in
different countries. HHV8 infects up to 5% of the population in the United
States, with higher percentages in Europe, and up to 40–70% sero-
positivity rates in some parts of Africa. HHV8-associated malignancies are
rare in Western countries, except in the context of severe
immunodeficiency (Dunphy, 2010). Nearly all EBV- and HHV8-
associated DLBCLs are MUM1+ and CD10-, consistent with a post-
GC/ABC-like origin. Plasma cell differentiation is also very common.
Most of these lymphomas are clinically aggressive with median survivals
of only months to 1–2 years. EBER (small nuclear RNAs associated with
the EBV) detected by ISH is the most useful marker to demonstrate EBV
infection in the tumor cells, and latency associated nuclear antigen
(LANA)-1 detected by IHC is the most useful marker of HHV8 infection.
Detection of LMP1 (EBV latent membrane protein 1) by IHC in EBV +
lymphomas is typically associated with reactivation of EBV replication
(Dong, 2008).
b. Acquired Immunodeficiency Syndrome (AIDS)-Associated
DLBCL:
Prior to the advent of highly active anti-retroviral therapy, there was
a 110- to 200-fold increase in the incidence of lymphoma in HIV-infected
patients, the majority of which were EBV-associated DLBCL and BL.
Currently, in the United States, there has been a marked reduction in the
incidence of AIDS-associated lymphomas, which are now limited to
19
patients who are refractory to, or noncompliant with, therapy. Involvement
of various extranodal sites is frequently seen, and tumor cell morphology
ranges from polymorphous B-cell infiltrates with oligoclonal IGH gene
rearrangements to mono-morphous high-grade immunoblastic DBLCL
similar to the range of features seen in post-transplant tumors (Dunphy,
2010).
c. Lymphomatoid Granulomatosis (LyG):
LyG is an extranodal EBV + B-cell lymphoproliferative disorder that
characteristically presents with multifocal angiocentric lesions, most
commonly involving lung (bilateral pulmonary nodules in middle and
lower lobes), skin (multiple nodules or ulcers), CNS, and less commonly
kidney and liver. Although most common in middle-aged men, it affects a
wide age range and is associated with an immuno-compromised state in
many patients. The EBV + large cells typically have a perivascular
distribution with vascular invasion, and are regularly positive for CD30 as
well as both EBER and LMP1. Prognosis of LyG depends on histological
grade (Table 3), with disease progression manifested by an increased
number of EBV + tumor cells and a decreased number of reactive CD4+ T
cells (Dong, 2008).
Table (3): Histological grading of lymphomatoid granulomatosis.
Features Grade 1 Grade 2 Grade 3
Polymorphic
background
Dominant Significant Focal
Necrosis Focal Frequent Extensive
Tumor cells Rare (need IHC) Occasional, may be Frequent, may be
20
< 5/hpf in Clusters 5–20/hpf,
may be up to 50/hpf
in aggregates
> 50/hpf
Outcome Wax & wane, may
Regress
May have durable
response to therapy
Same as EBV +
DLBCL
d. DLBCL Associated with Chronic Inflammation:
The outgrowth of an EBV + DLBCL in response to unremitting
chronic suppurative inflammation was first recognized in Japan in patients
who had a history of pyothorax. Such pyothorax-associated lymphoma
(PAL) typically occurred after 10–20 years of persistent inflammation.
Similar lymphomas have now been reported following chronic
osteomyelitis, chronic skin ulcers, and around protheses and metallic
implants. The tumor infiltrates dense fibrotic linings or capsule of cavities
and joint spaces and can show just focal nests of EBER + and LMP1+ large
B cells. Surgical debulking with complete tumor resection has been
reported to improve survival, but the overall prognosis is poor, possibly
due to delays in diagnosis (Dong, 2005).
e. Plasmablastic Lymphoma:
Plasmablastic lymphoma is typically an EBV + extranodal,
extramedullary lymphoma with immunoblastic or plasmablastic
morphology that is negative for pan-B-cell antigens. Some patients may
have a low-level serum para-protein (usually IgG kappa or lambda), but the
tumor cells only infrequently exhibit cytoplasmic immunoglobulin
expression by IHC. Nodal involvement is very rare and tends to be found in
elderly patients without an identifiable immunodeficiency state.
Plasmablastic lymphoma should be distinguished from plasmablastic
transformation of myeloma, as the latter usually has the characteristic triad
21
of lytic bone lesions, plasmacytosis, and prominent monoclonal
gammopathy, as well as expression of cytoplasmic immunoglobulin, and is
typically negative for EBV (Dong, 2005).
f. Primary Effusion Lymphoma (PEL):
PEL, also known as body-cavity lymphoma, is a highly aggressive
extranodal tumor invariably associated with HHV8 infection. A vast
majority of cases are also positive for EBV. PEL presents with massive
serous effusions in one or more of the large body cavities (i.e., pleura
space, pericardium, or peritoneum) without localized masses, adenopathy,
or organomegaly (Nador, 1996). PEL cells exhibit immunoblastic or
anaplastic morphology and are often extremely pleomorphic. They lack
most pan-B-cell markers, but express CD45, MUM1, and CD138. CD30,
EMA, and CD79a are variably expressed (Dong, 2008).
g. EBV + DLBCL of the Elderly:
There is an increasing recognition that age-related declines in T-cell
surveillance can lead to emergence of nodal and extranodal systemic EBV
+ DLBCL in elderly patients, and are termed senile EBV-associated
lymphoproliferative disorders. While this entity is defined in the 2008
WHO classification as a disease of patients > 65 year-old, it does
occasionally occur in younger patients. The boundaries of this EBV-
associated DLBCL are currently under investigation, but the diagnosis
should only be made when a tumor does not fit other entities. This
lymphoma should be suspected whenever a DLBCL has areas containing
Reed-Sternberg-like cells, polymorphic infiltrates with plasmacytoid
features, and zonal necrosis (Oyama, 2003).
22
h. Large B-cell Lymphoma Arising from HHV8+ Multicentric
Castleman Disease (MCD):
This is primarily a nodal IgM plasmablastic lymphoma arising from
MCD and is the only known lymphoma subtype exclusively associated
with HHV8 without EBV co-infection. In a minority of patients, it may
disseminate to the GI tract, liver, lungs, or evolve into a leukemic phase.
The early lesions manifest as monotypic IgM + Igl + plasmablastic
proliferations localized to the mantle zones of MCD follicles, which then
evolve into microscopic aggregates (micro-lymphoma) and eventually
frank lymphoma with confluent sheets of tumor cells. Unlike EBV +
extranodal plasmablastic lymphoma, HHV8+ plasmablastic lymphoma
arising from MCD exhibit intense cytoplasmic expression of IgM and
lambda light chain (rarely kappa). The tumor cells express CD20 (weak)
and MUM1, but lack PAX5 and CD138 (Du MQ, 2001).
6) B-cell Lymphoma, Unclassifiable, with Features Intermediate
Between DLBCL and Burkitt lymphoma (BL):
This category comprises aggressive B-cell lymphomas that are
difficult to classify. They display features closely resembling BL but vary
in their morphology, immunophenotype and genetics. For example, a BL-
like lymphoma may show more variable cytomorphology, such as
increased number of large cells and more irregular nuclear contours. Cases
with morphology typical of BL may have strong BCL2 expression and
harbor both t(14;18) and t(8;14), that may be either de novo or represent
23
transformation of follicular lymphoma. Indeed, when routine FISH analysis
is performed, up to 2.5% of DLBCL may have both t(8;14) and t(14;18),
and such "double-hit" cases have very poor outcome. Rare B-ALL/BL
hybrid cases may have MYC translocation and classical BL morphology,
but display a mixed phenotype (e.g. surface Ig+ and CD10+ but with TdT
expression). Cases with typical BL morphology and immunophenotype but
no identifiable MYC translocation may be diagnostically challenging,
though they might be best considered as BL, especially in young patients
who might be expected to benefit from the intensive chemotherapy given
for BL (Dong, 2008).
7) B-cell Lymphoma, Unclassifiable, with Features Intermediate
Between DLBCL and CHL:
This category covers so-called gray-zone lymphoma and large B-cell
lymphoma with Hodgkin-like features. It mostly reflects the diagnostic
overlaps between PMBL and CHL in young adults presenting with a
mediastinal mass though similar lymphomas have been reported at other
sites as well. The classification difficulties arise when there is a sheet-like
proliferation of large lymphocytes that show a CHL-like immunophenotype
(CD20−,CD15+), or conversely the histologic appearance of nodular
sclerosis CHL but with tumor cells positive for CD20, CD79a, and/or
CD45, and negative for CD15. In most cases, the tumor cells will be larger
and more pleomorphic than is typical for PMBL but lack a classical Reed-
Sternberg appearance, and may strongly express the B-cell transcription
factors PAX5, BOB1, and OCT2 (like PMBL) along with CD30. In some
cases that have separate areas resembling CHL and DLBCL, the tumor may
be better classified as composite lymphoma. These borderline cases
probably reflect a shared biology, since microarray studies have shown
24
overlap in the molecular signature between CHL and PMBL (Traverse-
Glehen, 2005).
Treatment of DLBCL:
Diffuse large B-cell lymphoma (DLBCL) is an aggressive NHL in
which survival without treatment is measured in months. For treatment
purposes, patients with DLBCL are generally classified as having either
limited stage disease or advanced stage disease based upon whether or not
the tumor can be contained within one irradiation field (Swerdlow et al,
2008).
Limited stage disease (usually Ann Arbor stage I or II)
Limited stage DLBCL can be contained within one irradiation field.
This population accounts for 30-40 % of patients with DLBCL. Limited
stage DLBCL is treated primarily with combined modality therapy
consisting of abbreviated systemic chemotherapy (three cycles), the
recombinant anti-CD20 antibody rituximab, and involved field radiation
therapy. Alternatively, full course (six to eight cycles) systemic
chemotherapy plus rituximab without radiation therapy may be used (Sehn
et al. 2005).
Advanced stage disease (usually Ann Arbor stage III or IV)
Advanced stage DLBCL cannot be contained within one irradiation
field. This population accounts for 60-70 % of patients with DLBCL.
Advanced stage DLBCL is treated primarily with systemic
chemotherapy plus the recombinant anti-CD20 antibody rituximab.
25
Patients with bulky (>10 cm) stage II disease and patients with stage
IIB disease have a less favorable prognosis than those with non-bulky stage
II disease without systemic B symptoms. Many clinicians treat such
patients in a similar fashion to those with advanced stage disease.
Rituximab containing regimens
With the advent of combination chemotherapy with
cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) or
CHOP-like regimens, disease-free survival rates of 35 to 45 % at four years
have been realized (Fisher et al. 1993).
Survival has been further improved with the addition of rituximab to
standard CHOP-based therapy (R-CHOP). The addition of rituximab to
CHOP-based therapy results in an approximately 10 to 15 % overall
increase in survival beginning at one year from initiation of therapy in
patients of all ages with almost no increase in toxicity (Sehn et al. 2005).
In addition to stage of disease, survival with CHOP-based therapy
varies with the presence or absence of certain clinical features. The
International Prognostic Index (IPI) provides information on patient
outcomes with CHOP-based therapy based upon the patient's age,
performance status, stage, number of extranodal sites, and LDH level. For
patients treated with R-CHOP, the original IPI classifies patients into four
risk categories with three-year overall survival rates of 91, 81, 65, and 59
% for patients with IPI scores of 0-1, 2, 3, or 4-5, respectively (Ziepert et
al. 2010).
26
II. DNA methyaltion and epigenetics in cancer
Classic genetics alone cannot explain the diversity of phenotypes
within a population. Monozygotic twins despite their identical DNA
sequences can have different phenotypes and different susceptibilities to a
disease. The concept of epigenetics offers a partial explanation of these
phenomena. It was first introduced by C.H. Waddington in 1939 to name
"the causal interactions between genes and their products, which bring the
phenotype into being". Epigenetics was later defined as heritable changes
in gene expression that are not due to any alteration in the DNA sequence
(Fraga et al, 2005).
The best-known epigenetic mechanism is DNA methylation. The
initial finding of global hypomethylation of DNA in human tumors was
soon followed by the identification of hypermethylated tumor-suppressor
genes, and then, more recently, the discovery of inactivation of micro RNA
(mi RNA) genes by DNA methylation. These and other demonstrations of
how epigenetic changes can modify gene expression have led to human
epigenome projects and epigenetic therapies. Moreover, DNA methylation
occurs in a complex chromatin network and is influenced by the
modifications in histone structure that are commonly disrupted in cancer
cells (Siato and Jones, 2006).
Epigenetic in normal cells:
27
DNA methylation has critical roles in the control of gene activity and
the architecture of the nucleus of the cell. In humans, DNA methylation
occurs in cytosines that precede guanines; these are called dinucleotide
CpGs. There are CpG-rich regions known as CpG islands, which span the
5′ end of the regulatory region of many genes. These islands are usually not
methylated in normal cells. The methylation of particular subgroups of
promoter CpG islands can, however, be detected in normal tissues
(Herman, 2005).
DNA methylation is one of the layers of control of certain tissue-
specific genes, such as MASPIN (a member of the serine protease inhibitor
family) and germ-line genes such as the MAGE genes (silent in almost all
tissues except malignant tumors). Genomic imprinting also requires DNA
hypermethylation at one of the two parental alleles of a gene to ensure
monoallelic expression, and a similar gene-dosage reduction is involved in
X-chromosome inactivation in females (Feinberg et al, 2002).
The hypermethylation of repetitive genomic sequences probably
prevents chromosomal instability, translocations, and gene disruption
caused by the reactivation of transposable DNA sequences. Cells that lack
the stabilizing effect of DNA methylation because they have spontaneous
defects in DNA methyltransferases (DNMTs) or experimentally disrupted
DNMTs have prominent nuclear abnormalities (Espada et al. 2007).
DNA methylation occurs in the context of chemical modifications of
histone proteins. Histones are not merely DNA-packaging proteins, but
molecular structures that participate in the regulation of gene expression
(Esteller, 2007). They store epigenetic information through such post-
translational modifications as lysine acetylation, arginine and lysine
28
methylation, and serine phosphorylation. These modifications affect gene
transcription and DNA repair. It has been proposed that distinct histone
modifications form a "Histone code". Acetylation of histone lysines, for
example, is generally associated with transcriptional activation (Jenuwein,
2006). The functional consequences of the methylation of histones depend
on the type of residue –lysine (K) or arginine- and the specific site that the
methylation modifies (e.g., K4, K9, or K20). Methylation of H3 at K4 is
closely linked to transcriptional activation, whereas methylation of H3 at
K9 or K27 and of H4 at K20 is associated with transcriptional repression.
What emerges from these findings is a flexible but precise pattern of DNA
methylation and histone modification that is essential for the physiologic
activities of cells and tissues (Karpf, 2006).
DNA Hypomethylation in Tumors
The low level of DNA methylation in tumors as compared with the
level of DNA methylation in their normal tissue counterparts was one of
the first epigenetic alterations to be found in human cancer. The loss of
methylation is mainly due to hypomethylation of repetitive DNA sequences
and demethylation of coding regions and introns regions of DNA that allow
alternative versions of the messenger RNA (mRNA) which is transcribed
from a gene. A recent large-scale study of DNA methylation with the use
of genomic microarrays has detected extensive hypomethylated genomic
regions in gene-poor areas. During the development of a neoplasm, the
degree of hypomethylation of genomic DNA increases as the lesion
progresses from a benign proliferation of cells to an invasive cancer (Fraga
et al, 2007).
29
Three mechanisms have been proposed to explain the contribution of
DNA hypomethylation to the development of a cancer cell: generation of
chromosomal instability, reactivation of transposable elements, and loss of
imprinting. Hypomethylation of the DNA can favor mitotic recombination,
leading to deletions and translocations, and it can also promote
chromosomal rearrangements. This mechanism was seen in experiments in
which the depletion of DNA methylation by the disruption of DNMTs
caused aneuploidy. Hypomethylation of DNA in malignant cells can
reactivate intragenomic endoparasitic DNA, such as L1 (long interspersed
nuclear elements), and Alu (recombinogenic sequence) repeats. These
undermethylated transposons can be transcribed or translocated to other
genomic regions, thereby further disrupting the genome (Eden, 2003).
The loss of methyl groups from DNA can also disrupt genomic
imprinting. In the hereditary Beckwith–Wiedemann syndrome (a syndrome
characterized by exomphalos, macroglossia, and gigantism), for example,
there is loss of imprinting of IGF2 (the insulin-like growth factor gene) and
an increased risk of cancer. Loss of imprinting of IGF2 is also a risk factor
for colorectal cancer, and disrupted genomic imprinting contributes to the
development of Wilms’ tumor. In animal models, mice with a loss of
imprinting of IGF2 or overall defects in imprinting have an increased risk
of cancer (Kaneda and Feinberg, 2005).
Normally, certain testis specific genes (genes that encode melanoma
antigens, or specific proliferation-linked genes) are silent in somatic cells
because promoter-regions are methylated. In some cancer cells, by
contrast, these promoter regions undergo demethylation, and the usually
repressed genes become expressed. Two notable examples of the
30
hypomethylation mechanism are the activation of PAX2 (a gene that
encodes a transcription factor involved in proliferation and other important
activities of cells) and the activation of the let-7a-3 miRNA gene, which
has been implicated in endometrial and colon cancer (Sakatani et al.
2001).
The hypomethylation of DNA can have unpredictable effects. The
progeny of a mouse deficient in DNA methylation and a Min mouse
(Mouse strain which has a genetic defect in the adenomatous polyposis coli
"APC" gene and is prone to colon adenoma) have fewer tumors than one
would expect; by contrast, another DNMT-defective mouse strain has an
increased risk of lymphoma (Gaudet, 2003). Moreover, hypomethylation
suppresses the later stages of intestinal tumorigenesis but promotes early
precancerous lesions in the colon and liver through genomic deletions
(Yamada, 2007).
Hypermethylation of Tumor-Suppressor Genes
Hypermethylation of the CpG islands in the promoter regions of
tumor-suppressor genes is a major event in the origin of many cancers. The
initial reports of hyper-methylation of the CpG islands in the promoter
region of the retinoblastoma tumor- suppressor gene (Rb) were followed
by the findings that hypermethylation of the CpG island was a mechanism
of inactivation of the tumor-suppressor genes VHL, p16INK4a
, hMLH1 (a
homologue of MutL E. coli), and BRCA1 (breast-cancer susceptibility
gene1) (Herman, 2005).
31
Hypermethylation of the CpG-island promoter can affect genes
involved in the cell cycle, DNA repair, the metabolism of carcinogens, cell-
to-cell interaction, apoptosis, and angiogenesis, all of which are involved in
the development of cancer. It can occur at different stages in the
development of cancer and in different cellular networks, and it interacts
with genetic lesions. Such interactions can be seen when hypermethylation
inactivates the CpG island of the promoter of the DNA-repair genes
hMLH1, BRCA1, MGMT (O6-methylguanine–DNA methyltransferase),
and the gene associated with Werner’s syndrome (WRN) (a very rare,
autosomal recessive disorder characterized by the appearance of premature
aging and telomere instability). In each case, silencing of the DNA-repair
gene blocks the repair of genetic mistakes, thereby opening the way to
neoplastic transformation of the cell (Agrelo et al. 2006).
The profiles of hypermethylation of the CpG islands in tumor-
suppressor genes are specific to the cancer type (Agrelo and Wutz, 2009).
Each tumor type can be assigned a specific, defining DNA
“hypermethylome.” Such patterns of epigenetic inactivation occur not only
in sporadic tumors but also in inherited cancer syndromes, in which
hypermethylation can be the second lesion in Knudson’s two-hit model of
how cancer develops. Recently devised epigenomic techniques have
revealed maps of hypermethylation of the CpG islands that suggest the
occurrence of 100 to 400 hypermethylated CpG islands in the promoter
regions of a given tumor (Esteller, 2007).
The mechanism through which CpG islands become
hypermethylated in some types of cancer but not in others is still unclear.
Inactivation of a particular gene by methylation could give certain tumor
32
types a growth advantage. CpG islands can have a location within a
particular nucleotide sequence that allows them to become
hypermethylated (Weber et al, 2005), or they can be located in a
chromosomal region that is subject to large-scale epigenetic dysregulation
(Esteller, 2007). In addition, there is a mechanism in which modifications
of histones mark a gene for hypermethylation. This marking occurs in the
binding site of the methyltransferase enhancer of zeste drosophila
homologue 2 (EZH2)( a component of the polycomb family of gene-
silencing proteins) (Hellebrekers et al, 2006), in histones of stem cells with
unmethylated gene promoters (Widschwendter et al, 2007) and in the
histone-associated silencing of p16INK4a
in colon-cancer cells ( (Bachman,
2003).
Histone Modifications in Cancer Cells
The most reliable method for detecting changes in histones is mass
spectrometry, which is highly specialized but time-consuming (Esteller,
2007). Moreover, histone modifications occur in different histone proteins,
histone variants (e.g., H3.3), and histone residues such as lysine, arginine,
and serine. These modifications also involve different chemical groups
(e.g., methyl, acetyl, and phosphate) and have different degrees of
methylation (e.g., monomethylation, dimethylation, and trimethylation).
Acetylation and methylation of histones have direct effects on a variety of
nuclear processes, including gene transcription, DNA repair, DNA
replication, and the organization of chromosomes. Generally, histone
acetylation is associated with transcriptional activation but the effect of
33
histone methylation depends on the type of amino acid and its position in
the histone tail (Bernstein, 2007). Hyper-methylation of the CpG islands
in the promoter regions of tumor-suppressor genes in cancer cells is
associated with a particular combination of histone markers: deacetylation
of histones H3 and H4, loss of H3K4 trimethylation, and gain of H3K9
methylation and H3K27 trimethylation (Jones, 2007).
Epigenetic Factors and miRNA
“miRNAs” are short, 22-nucleotide, non-coding RNAs that regulate
gene expression by sequence-specific base pairing in the 3′ untranslated
regions of the target mRNA. The result is mRNA degradation or inhibition
of translation. Patterns of miRNA expression are tightly regulated and play
important roles in cell proliferation, apoptosis, and differentiation (He and
Hannon, 2004). The number of human genes known to lose activity as a
result of the binding of a miRNA to the untranslated regions of the mRNA
is growing rapidly (Bueno et al, 2008).
The miRNA expression profiles differ between normal tissues and
tumor tissues and among tumor types (Lu J, 2005). Down-regulation of
subgroups of miRNAs, a common finding, implies a tumor-suppressor
function for miRNAs, as in the examples of downregulated let-7 and miR-
15/miR-16, which target the RAS and BCL2 oncogenes, respectively
(Johnson et al, 2008). DNA hypermethylation in the miRNA 5′ regulatory
region is a mechanism that can account for the down-regulation of miRNA
in tumors. In colon-cancer cells with disrupted DNMTs, hypermethylation
of the CpG island does not occur in miRNAs. The methylation silencing of
miRNA-124a also causes activation of the cyclin D–kinase 6 oncogene
34
(CDK6), and it is a common epigenetic lesion in tumors (Saito and Jones,
2006).
Epigenetics in Cancer Management
The DNA-methylation and histone-modification patterns associated
with the development and progression of cancer have potential clinical use.
DNA hyper-methylation markers are under study as complementary
diagnostic tools, prognostic factors, and predictors of responses to
treatment. For instance, the glutathione S-transferase gene (GSTP1) is
hypermethylated in 80 to 90% of patients with prostate cancer (Cairns and
Adams, 2004), but it is not hypermethylated in benign hyperplastic prostate
tissue. Thus, the detection of GSTP1 methylation could help to distinguish
between prostate cancer and a benign process (Costa et al, 2007).
Hypermethylation of CpG islands can be a marker of cancer cells in
all types of biologic fluids and biopsy specimens, making detection of
GSTP1 methylation in urine, a possible clinical application. Analysis of
hypermethylation of the CpG island has potential diagnostic applicability
for carriers of high-penetrance mutations in tumor-suppressor genes. For
example, identification of DNA hyper-methylation in a breast-biopsy
specimen from a carrier of a BRCA1 mutation could be useful when the
pathological diagnosis is uncertain, because hypermethylation of the CpG
island is an early event in the development of cancer (Esteller, 2007).
Analysis of several hypermethylated genes detects twice as many
tumor cells in breast ductal fluids as conventional cytologic analysis (
Mehrotra et al, 2006), and hypermethylated genes can be found in
35
exfoliated cells at different stages in the development of cervical cancer
(Feng et al. 2006). The application of DNA-hypermethylation markers as
tumor markers in routine clinical practice will require rapid, quantitative,
accurate, and cost-effective techniques and objective criteria for selection
of the genes that are applicable to different tumor types. Hypermethylation
of a tumor-suppressor gene and DNA hypermethylome profiles can be
indicators of the prognosis in patients with cancer. Hypermethylation of the
death-associated protein kinase (DAPK), p16INK4a, and epithelial
membrane protein 3 (EMP3) has been linked to poor outcomes in lung,
colorectal, and brain cancer, respectively (Esteller, 2007).
Prognostic dendrograms similar to those used in gene-expression
microarray analyses, with the use of a combination of hypermethylated
markers and CpG-island microarrays, have been developed. These
epigenomic profiles are complementary to profiles of gene-expression
patterns and can be developed with DNA extracted from archived material
(Laird et al. 2004). The hypermethylation of particular genes is potentially
a predictor of the response to treatment. The methylation-associated
silencing of the gene for the DNA-repair protein MGMT in gliomas is an
example (Esteller and Herman, 2002). MGMT reverses the addition of
alkyl groups by the alkylating agents to the guanine base of DNA. Two
studies have shown that the hypermethylation of MGMT is an independent
predictor of a favorable response of gliomas to carmustine (BCNU) or
temozolomide (Hegi et al, 2009).
The potential of the methylation status of MGMT and other DNA-
repair genes to predict the response to chemotherapy has also been seen
with cyclophosphamide (with the MGMT gene), cisplatin (with the
hMLH1 gene)(Strathdee, 2007), methotrexate (with the reduced folate
36
carrier [RFC] gene), and irinotecan (with the WRN gene) (Agrelo et al,
2006).
Epigenetic Therapy of cancer
Unlike mutations, DNA methylation and histone modifications are
reversible. Epigenetic alterations allow the cancer cell to adapt to changes
in its micro-environment, but dormant, hypermethylated tumor-suppressor
genes can be re-activated with drugs. It is possible to re-express DNA
methylated genes in cancer cell lines by using demethylating agents and to
rescue their functionality (Yoo and Jones, 2006). DNA demethylating
drugs in low doses have clinical activity against some tumors. Two such
agents, 5-azacytidine (Vidaza) and 5-aza-2′-deoxycytidine (decitabine),
have been approved as treatments for the myelodysplastic syndrome and
leukemia (Oki and Issa, 2007). However, these demethylating agents have
not yet been shown to have clinical activity against solid tumors.
Histone deacetylase (HDAC) inhibitors can induce differentiation,
cell-cycle arrest, and apoptosis in vitro, although it has not been possible to
pinpoint a specific mechanism that explains these effects (Ropero et al.
2004). The first drug of this type, suberoylanilide hydroxamic acid
(vorinostat), has been approved by the Food and Drug Administration
(FDA) for the treatment of cutaneous T-cell lymphoma (Marks et al,
2009). The efficacy of HDAC inhibitors in the treatment of other tumors is
limited. The nonspecific effects of DNA demethylating agents and HDAC
inhibitors could have unintended consequences with regard to gene
expression, and as a paradoxical result, they could have growth-promoting
effects on a tumor. However, there are prospects for directed epigenetic-
specific therapy with the use of transcription factors that target particular
37
gene promoters (Moore and Ullman, 2003). For instance, the engineered
zinc finger proteins target unique sequences in the MASPIN promoter;
these proteins not only reactivate the epigenetically silenced gene but also
inhibit tumor growth in vitro (Beltran et al, 2008).
DNA methylation in Non-Hodgkins lymphoma
Hematological neoplasms are known to have different
hypermethylation profiles than those of other solid tumors. The three major
forms of lymphoid/hematopoietic malignancies; non-Hodgkin’s
lymphomas, Hodgkin lymphoma & multiple myeloma show overlapping
but individual patterns of methylation (Takahashi1, 2004). To date, there
haven’t been extensive studies about aberrant promoter methylation of
tumor suppresser genes (TSGs) in NHL; only a limited numbers of TSGs
have been tested and their analysis has been restricted to certain types of
NHLs (Esteller, 2007). In a recent study, the prevalence of aberrant
promtor methylation was explored. The selected eight TSGs are known to
be involved in cell cycle regulation (p16, COX2), DNA repair (MGMT),
apoptosis (DAPK, RUNX3), angiogenesis inhibitor (THBS1), invasion and
metastasis (CDH1) and cell proliferation (MT1G). The methylation status
was examined in all the enrolled lymphoma cases and this was analyzed
specifically according to the cellular origins (B-cells or T/NK-cells) of the
NHLs. This study suggests that aberrant CpG island methylation is a
frequent event in NHLs, and diffuse large B-cell lymphomas show
overlapping but distinct methylation profiles (Kim1 et al, 2008).
38
Methods of detection of DNA methylation
DNA methylation can be detected by the following assays currently used in
scientific research:
Methylation-Specific PCR (MSP): MSP is based on a chemical
reaction of sodium bisulfite with DNA that converts unmethylated
cytosines of CpG dinucleotides to uracil or UpG, followed by
traditional PCR. However, methylated cytosines will not be converted
in this process, and primers are designed to overlap the CpG site of
interest, which allows one to determine methylation status as
methylated or unmethylated (Herman et al, 1996).
The MethyLight method: is based on MSP, but provides a
quantitative analysis using real-time PCR. Methylated-specific
primers are used, and a methylated-specific fluorescence reporter
probe is also used that anneals to the amplified region. Quantitation
is made in reference to a methylated reference DNA. A modification
to this protocol to increase the specificity of the PCR for successfully
bisulfite-converted DNA (ConLight-MSP) uses an additional probe
to bisulfite-unconverted DNA to quantify this non-specific
amplification (Rand et al. 2002).
Melting curve analysis (Mc-MSP): This method amplifies
bisulfite-converted DNA with both methylated-specific and
unmethylated-specific primers, and determines the quantitative ratio
of the two products by comparing the differential peaks generated in
a melting curve analysis. A high-resolution melting analysis method
that uses both real-time quantification and melting analysis has been
39
introduced, in particular, for sensitive detection of low-level
methylation (Kristensen et al. 2008).
Whole genome bisulfite sequencing: also known as BS-Seq, which
is a high-throughput genome-wide analysis of DNA methylation. It is
based on aforementioned sodium bisulfite conversion of genomic DNA,
which is then sequencing on a Next-generation sequencing platform.
The sequences obtained are then re-aligned to the reference genome to
determine methylation states of CpG dinucleotides based on
mismatches resulting from the conversion of unmethylated cytosines
into uracil (Yuanxin Xi and Wei Li, 2009).
The HELP assay: which is based on restriction enzymes' differential
ability to recognize and cleave methylated and unmethylated CpG DNA
sites (Batbayar et al, 2006).
ChIP-on-chip assays: which is based on the ability of commercially
prepared antibodies to bind to DNA methylation-associated proteins
like MeCP2 (Pillai and Chellappan, 2009).
Methylated DNA immunoprecipitation (MeDIP): analogous to
chromatin immunoprecipitation. Immunoprecipitation is used to isolate
methylated DNA fragments for input into DNA detection methods such
as DNA microarrays (MeDIP-chip) or DNA sequencing (MeDIP-seq)
(Mohn et al, 2009).
Pyrosequencing of bisulfite treated DNA: This is sequencing of an
amplicon made by a normal forward primer but a biatenylated reverse
primer to PCR the gene of choice. The Pyrosequencer then analyses the
sample by denaturing the DNA and adding one nucleotide at a time to
the mix according to a sequence given by the user. If there is a mis-
match, it is recorded and the percentage of DNA for which the mis-
40
match is present is noted. This gives the user a percentage methylation
per CpG island (Tost and Gut, 2007).
Molecular break light assay for DNA adenine methyltransferase
activity: an assay that relies on the specificity of the restriction enzyme
DpnI for fully methylated (adenine methylation) GATC sites in an
oligonucleotide labeled with a fluorophore and quencher. The adenine
methyltransferase methylates the oligonucleotide making it a substrate
for DpnI. Cutting of the oligonucleotide by DpnI gives rise to a
fluorescence increase (Yan et al, 2007).
Methyl Sensitive Southern Blotting: is similar to the HELP assay,
although uses Southern blotting techniques to probe gene-specific
differences in methylation using restriction digests. This technique is
used to evaluate local methylation near the binding site for the probe
(Moore, 2009).
41
III. DAPK and MT1G genes
DAPK (death-associated protein kinase 1)
Death-associated protein kinase (DAPK) is a tumor suppressor gene
(mediator of apoptotic systems). DAPK was discovered in the mid 1990s
during genetic screening in which an antisense library was used to identify
genes necessary for interferon (IFN)-induced death in HeLa cells (Deiss et
al, 1995). Subsequent sequence and activity analysis indicated that DAPK
encoded a Ca2+/calmodulin (CaM) regulated Ser/Thr kinase, with a
catalytic domain highly homologous to that of myosin light chain kinase
(MLCK). This 160 kDa protein bears an interesting multi-domain structure,
including ankyrin repeats and the death domain (Figure 1). DAPK is
necessity for cell death and not limited to IFN-signaling; numerous studies
have demonstrated that DAPK activity is required for the induction of cell
death by multiple death signals, including those generated by death
receptors, cytokines, matrix detachment, and hyperproliferation (Jang et al,
2002). Transcriptional silencing of death-associated protein kinase
(DAPK) occurs in many diverse types of human cancers through promoter
hypermethylation at prevalences ranging from 7% in liver tumors to 84%
42
in B-cell non-Hodgkin’s lymphoma (Lehmann et al, 2009). The ubiquitous
silencing of this gene implies a critical role for it in cancer development.
Loss of expression of DAPK confers a selective growth advantage for
cancer cells that may drive tumor aggressiveness and progression. This
gene has a CpG island extending 2500 bp from the translational start site;
however, studies characterizing its transcriptional regulation have not been
conducted. Two transcripts for DAPK were identified that code for a single
protein, while being regulated by two promoters. The previously identified
DAPK transcript designated as exon 1 transcript was expressed at levels 3-
fold greater than the alternate exon 1b transcript. Deletion constructs of
promoter 1 identified a 332 bp region containing a functional CP2-binding
site important for expression of the exon 1 transcript. While moderate
reporter activity was seen in promoter 2, the region comprising intron 1 and
containing a HNF3B- binding site sustained expression of the alternate
transcript (Pullin et al, 2009). Sequencing the DAPK CpG island in tumor
cell lines revealed dense, but heterogenous methylation of CpGs that
blocked access of the CP2 and HNF3B proteins that in turn, was associated
with loss of transcription that was restored by treatment with 5-aza-2-
deoxycytidine (Toyooka et al, 2003).
Figure (1): Schematic diagram of DAP-kinase protein structure. The 160
kDa actin microfilament-associated Ca2+/calmodulin (CaM)-regulated
Serine/Threonine kinase bears a multiple domain structure. The catalytic
43
and the calmodulin regulatory domains determine substrate specificity and
regulation of kinase catalytic activity, respectively. The non-catalytic
association domains, involved in subcellular localization or interactions
with other proteins, include the 8 ankyrin repeats, two nucleotide-binding
P-loops, a cytoskeleton-binding region, and a death domain.
Phosphorylation by RSK at Ser289 triggers a suppression of DAPK
proapoptotic function (Anjum et al., 2005). The autophosphorylation site
was mapped to Ser308 within the CaM-regulatory domain (Shohat et al.,
2002). ERK phosphorylates DAPK at Ser735, which stimulates DAPK-
mediated apoptosis (Chen et al., 2005).
Metallothionein 1G
Zinc is an essential trace element as a component of several
metalloenzymes involved in critical physiologic processes, including cell
growth and proliferation, osteogenesis, immunity, and antioxidant activity
(Platz and Helzlsouer, 2001). The bioavailability of zinc is controlled by
metallothioneins, a class of low molecular weight proteins with metal-
binding and antioxidant properties (Andrews, 2000). Human
metallothioneins are encoded by a family of genes located at chromosome
16q13 and some of the isoforms seem expressed in an organ-dependent
manner (Coyle et al. 2002). The MT-IF and MT-IG genes are
differentially regulated in two human hepatoma cell lines ( HepG2 and
Hep3B2) and a human lymphoblastoid cell line ( WI-L2 ) in response to
the heavy metals cadmium, zinc and copper, and glucocorticoids
(Gedamu et al, 1987). MT1G hypermethylation is more frequent in
prostate cancer that spread beyond the prostate capsule. MT1G promoter
hypermethylation was found in 29 of 121 prostate cancer, 5 of 39 HGPIN
(High grade PIN), 3 of 29 benign prostatic hyperplasia, and 0 of 13
44
normal prostate tissue samples (Henrique et al, 2005). Melatonin has
been shown to bind to the MT1G protein-coupled receptor (GPCR) in
MCF-7 breast cancer cells to modulate the estrogen response pathway
suppressing estrogen-induced estrogen receptor alpha (ER alpha)
transcriptional activity, blunting ER/DNA binding activity and
suppressing cell proliferation ( Kiefer et al, 2005). Loss of expression of
MT1G is accompanied by hypermethylation in the 5' regions of these
genes (Huang et al, 2003). Hypermethylation, but not LOH, is associated
with the low expression of MT1G and CRABP1 in papillary thyroid
carcinoma (Huang et al, 2003). VEGF stimulation also led to the
increased acetylation E2F1 as well as the histones in the hMT1G
promoter region (Joshi et al, 2005). Combined treatment with the DNA
methyltransferase inhibitor 5-aza-2'-deoxycytidine (5-Aza-dC) and the
histone deacetylase inhibitor trichostatin A (TSA) resulted in
demethylation and re-expression of the MT1G gene in the cell line K2
(Huang et al, 2003). Human metallothionein 1G (hMT1G) promoter is
upregulated by E2F1 upon VEGF stimulation of human aortic endothelial
cells (Joshi et al, 2005).The MTI-F and MT-IG gene promoters were also
functional in human chondrocytes (su et al, 1996).
45
MATERIAL AND METHODS
I) Study material:
This retrospective study included 70 cases of diffuse large B-cell
NHL lymph node biopsies, which were obtained from the Pathology
Department, NCI, Cairo during the period from 2003-2009. Paraffin blocks
of the studied cases were recruited and those with scanty or exhausted
material were excluded. All relevant clinico-pathologic data of the patients
were obtained from the pathology reports & clinical records including age,
sex, type of specimen, diagnosis, number of extra-nodal sites, PS, B-
symptoms, LDH level, stage, IPI, type of treatment, response and survival.
From each paraffin block, one H&E slide was prepared to
confirm the diagnosis and to assess the neoplastic to non-neoplastic cell
ratio. Only cases with more than 80% neoplastic cells in the examined
sections were included in the study.
46
II) Immunohistochemistry:
Three sections (5um thick each) were cut from the paraffin blocks
onto positive charged slides and stained for CD10 (Clone 56C6, ready to
use, Dako, Denmark), BCL6 (clone PG-B6p, dilution range 1:10-1:20,
Dako, Denmark) and MUM1 (clone MUM1p, dilution range 1:25-1:50,
Dako, Denmark) in order to sub-classify the cases into germinal center
(GCB) and non-germinal center B-cell (Non-GCB) prognostic types (Hans
et al, 2004). This was done according to the following algorithm (Figure 2).
A case of reactive lymphoid hyperplasia was used as a positive control.
Figure (2): The immunohistochemical algorithm for identification of
prognostically important subgroups of DLBCL proposed by Hans et al
2004.
Immunostaining Procedures
The Avidin-Biotin immunoperoxidase technique (ABC) was used.
The following steps were done (Hsu et al, 1981):
47
1-Deparaffinization and rehydration: paraffin embedded tissues were de-
paraffinized in xylene 3 times for 5-10 minutes each, and re-hydrated
through graded alcohol series (100%, 90 %, 70% ) for 5 minutes each,
then immersed in tap water for 5 minutes.
2-Antigen retrieval: Tissue sections were treated in a microwave oven for
5minutes at 700w, in antigen retrieval citrate buffer (PH 6), and then
sections were left to cool.
3- Endogenous peroxidase blocking: The slides were immersed in 3%
hydrogen peroxide solution for 10 minutes, and then washed in PBS (PH
7.2) for 2 x 5 minutes.
4-Primary antibody: Slides were then put in a wet chamber; two drops of
the primary antibodies were added on each slide to cover the entire tissue
section. They were incubated for one hour at room temperature and then
washed in PBS for 3 x 5 minutes.
5- Secondary antibody (Link): Tissues were covered with 2 drops of
UltraVision biotinylated goat anti-polyvalent secondary antibody,
incubated at room temperature for 10 minutes and then washed in PBS for
3x3 minutes.
6- Streptavidin/ Peroxidase (Label): Drops of streptavidin peroxidase
were added to cover the sections, incubated at room temperature for 10
minutes, and then washed in PBS for 2x5 minutes.
7-Substrate/ Chromogen: Drops of chromogen (DAB: 3′3-
diaminobenzidine HCl) were added to cover the entire sections, incubated
at room temperature for 10 minutes, and then washed in PBS for 5 minutes.
8-Counterstaining: The slides were placed in hematoxylin bath for 1
48
minute, and then washed in a water bath 5 times and tap water for 5
minutes.
9- Cover slipping: Sections were dehydrated through 70%, 90%,
100% alcohol and then put in xylene for 5 minutes. Few drops of
DPX (Di-N-Butyle Phthalate in Xylene) were added and slides were
covered by cover slips, and allowed to dry for few minutes.
Cases were considered positive for CD10, BCL6 & MUM1 if
30% or more of the tumor cells in the section were stained with the
antibody. The intensity of staining was not used to determine
positivity because the variability in tissue fixation and processing
appeared to affect the intensity of staining (Hans et al, 2004).
III) Molecular studies:
1) DNA extraction:
DNA extraction was done for each case according to Shi et al (2002)
as follows:
Three sections 10 microns thick were taken from each paraffin block
in an autoclaved plastic micro-tube (1.5ml).
One ml xylene was added to the micro-tube containing the tissue
sections for 30 min for two changes.
Ethanol (100% and 75%) was added for 30 min (two changes).
Washing with PBS for 15 min was done with two changes.
49
50 µl, Lysis buffer were added (Proteinase K 20 mg/ml, 1 M Tris-
HCl solution 10 µl, 0.5 M EDTA 2 µl, 10% SDS 100 µl), and
completed with distilled water to a final volume of 500ul in each
tube.
Overnight incubation was done at 52C until all tissue fragments
were dissolved completely.
Phenol: chloroform: isopropanol (25:24:1) was added to the de-
waxed tissue as 500 ul in each tube.
Mixing with vortex.
Centrifugation at 12,000 X g for 10 min.
The supernatant fluid was transferred to another autoclaved micro
tube using a 100-µl pipette.
An equal (One) volume of chloroform was added to the supernatant
& mixed by vortexing.
Centrifugation was done at 12,000 x g for 5 min.
The upper aqueous supernatant was carefully removed of to another
micro tube.
0.1 volume of 3 M sodium acetate was added to the tube followed by
(two volumes) 100% ice-cold ethanol to precipitate the DNA.
The tube was incubated at -20˚C overnight.
The DNA was precipitated by centrifugation at 12,000 x g at 4˚C.
50
The supernatant fluid was discarded and the DNA was washed once
with 75% ethanol, left to dry on a filter paper.
The final yield of DNA was dissolved in 50 µl of the storage buffer
(10mM Tris HCL + 1mM EDTA, PH 8) or distilled water after
complete drying in a hood.
The DNA concentration was calculated by Spectrophotometry
through the following equation:
(Unknown DNA) mg/ml = 50 mg/ml x Measured A260 x dilution
factor.
2) Bisulfite Modification:
The methylation status of a DNA sequence can best be determined
using sodium bisulfite. Incubation of the target DNA with sodium bisulfite
results in conversion of unmethylated cytosine residues into uracil, leaving
the methylated cytosines unchanged. Therefore, bisulfite treatment gives
rise to different DNA sequences for methylated and unmethylated DNA.
Bisulfite modification was done for each sample according to Herman et al,
1996:
1. DNA (1ug) in a volume of 50 ul was denatured by NaOH (final
concentration, 0.2 M) for 10 min at 37°C.
2. 30 ul of 10 mM hydroquinone (Sigma) and 520 ul of 3 M sodium
bisulfite (Sigma) at pH 5, both freshly prepared, were added and mixed,
and samples were incubated under mineral oil at 50°C for 16 hr.
51
3. Modified DNA was purified using the Wizard DNA purification resin
according to the manufacturer (Promega) and eluted into 50 ul of water.
4. Modification was completed by NaOH (final concentration, 0.3 M)
treatment for 5 min at room temperature, followed by ethanol
precipitation.
5. DNA was resuspended in water and used immediately or stored at -
20°C.
3) Methylation specific PCR (MSP):
The modified DNA was used for methylation specific PCR reaction
for detection of DAPK and MT1G promoter methylation using the
primer sequences, PCR conditions and the product size for both genes
as illustrated in the (Table 5). All the PCR amplifications were
performed in the PTC 200 thermal cycler (MJ research, Waltham, MA,
USA) using:
Bisulphite modified DNA (30-50 ng)
primers (10 p mol each)
dNTPs (1mM each)
10X standard PCR buffer (Qiagen)
0.5 U of Hot Start Taq plus DNA polymerase (Qiagen) in a volume of
20.
PCR cycles:
52
The reactions were hot-started at 95C for 5 min
Followed by 35 cycles of
94 C for 30 sec
60 C for 30sec
72 C for 30 sec
A final extension step 72 C for 10 min
The PCR products were electrophori zed in 2% ethidium bromide-
stained agarose gels and were then visualized in the photo-documentation
system. For each MSP reaction, both positive (a sample known to harbor
promoter methylation of the tested gene) and negative (distilled water
without template DNA) controls were used.
Table (4): Primer sequences and PCR conditions for methylation-
specific PCR analysis.
Primer
Name
Primer sequence (5′-3′) Produc
t
Size
(pb)
Annealing
Temp. (ºc) Forward Reverse
DAPK M GGATAGTCGGATCGAGTTAACGTC CCCTCCCAAACGCCGA 98 60
U GGAGGATAGTTGGATTGAGTTAATGTT CAAATCCCTCCCAAACACCAA 98 60
MT1G M TGCGAAAGGGGTCGTTTTGC GCGATCCCGACCTAAACTATACG 93 59
U GTGAGTTGGTGTGAAAGGGGTT CCACACCACCCACAATCCCA 113 59
. m, methylated sequence; u, unmethylated sequence.
53
IV) Statistical methods:
The data was coded and entered using the statistical package SPSS
version 15. The data was summarized using descriptive statistics:
mean, standard deviation, minimal and maximum values for
quantitative variables and number and percentage for qualitative
values. Statistical differences between groups were tested using Chi
square test for qualitative variables, independent sample t test for
quantitative normally distributed variables while Non-parametric
Mann Whitney test and kruskal-Wallis test were used for quantitative
variables which aren’t normally distributed. Correlations were done
to test for linear relations between variables. Kaplan-Meier Survival
Analysis was used. P-values less than or equal to 0.05 were
considered statistically significant.
54
RESULTS
linical features:C-1
Age and sex:
The age of patients showed a wide distribution range (between 20 and
85 years). The median age was 54 years with a standard deviation of 13.91.
Twenty two percent of the patients were above 60 years (Figure 3). Males
represented 58.6% of cases (41/70) and females represented 41.4% (29/70).
Male to female ratio was 1.4.
Figure (3): The age distribution in DLBCL patients.
B-symptoms:
B-symptoms were present in 38/53 (71.7%) of the cases and
absent in 15/53(28.3%). Seventeen cases had no data as regards B-
symptoms (Figure 4).
55
Figure (4): The B-symptoms occurrence in the DLBCL cases.
Performance status (PS):
As regards PS, 37/63 (58.7%) of the cases showed good PS (<2) and
26/63 (41.3%) showed poor PS (≥2) according to ECOG performance
status criteria (Figure 5), with missing data in 7 cases.
Figure (5): The performance status in the DLBCL cases.
LDH level:
Most of cases 38/55 (69.1%) showed high LDH serum level and 15/55
(30.9%) showed normal LDH serum level (Figure 6), with missing data in
15 cases.
56
Figure (6): The LDH level in the DLBCL cases.
Stage:
Forty nine out of the sixty five cases (74.2%) who have
presented in advanced stages (III and IV) in contrast to 16/65 (24.2%)
in early stages (I and II) (Figure 7). Five cases had missing data.
Figure (7): The stage distribution in the DLBCL cases.
57
Number of extra-nodal sites affection:
The number of extra-nodal sites (including soft tissue, hepatic,
mesenteric, bone marrow and others) affection was <2 in 60/66
(90.9%) and ≥ 2 in 6/66 (9.1%) of cases with missing data in 4 cases.
International prognostic index (IPI):
As regards the international prognostic index, Cases with low/low-
intermediate IPI (0-2) and high-intermediate/high IPI represented (3-5)
represented 32.1% and 67.9% of cases, respectively (Figure 8).
32.10%
67.90%
0.00%
10.00%
20.00%
30.00%
40.00%
50.00%
60.00%
70.00%
80.00%
Low IPI High IPI
Figure (8): The international prognostic index (IPI) in the cases.
:Histopathologic features-2
All cases were DLBCL, NOS category. The centroblastic, anaplastic,
immunoblastic and plasmablastic morphologic variants were 88.5%, 7.5%,
4% and 0%, respectively (Figure 9).
58
Figure(9): DLBCL cases: (A) a case of DLBCL with dominant
centroblastic morphologic variant (H&E, 40X), (B) a case of
DLBCL with dominant immunoblastic morphology (H&E, 40X), (C)
a case of DLBCL with high turn-over & tangible-body macrophages
forming starry-sky appearance (H&E, 40X) & (D) a case of DLBCL
with anaplastic morphology showing giant cells (H&E, 20X)
59
:Immunophenotypic classes andmmunophenotyping I -3
CD10
Positive membranous immunostaining for CD10 was reported in
20/65 (30.8%) of the cases whereas 45/65 (69.2%) were negative.
Bcl6
Positive nuclear stain for BCL6 was detected in 39/65 (60%) of
the cases and negative stain was reported in 26/65 (40%) of cases.
Mum1
Positive nuclear stain for MUM1 was detected in 41/65 (63%) of the
cases while 24/65 (37%) were negative (Figure 11). Five cases were
excluded from immunostaining due to inconclusive results.
The non-germinal center B-cell phenotypic class predominate in the
cases as 38/61 (62.3%) showed immunostaining results consistent with
non-GCB pattern according to Hans algorithm, whereas 23/61 (37.7%) of
the cases showed germinal center B-cell pattern (Figure 10). Nine cases
were unclassifiable.
62.3%
37.7%
Class
Non-GCB
GCB
Figure (10): The phenotypic classes; germinal-center B-cell (GCB) &
Non-Germinal center B-cell (Non-GCB) subtypes.
60
Figure (11): Immunostain cases (A & B) two cases of DLBCL shows
positive nuclear stain with MUM1 (40X). (C) a case of DLBCL
shows positive nuclear stain for BCL6 (20X).
61
:methylation patternpromoter hyperDNA -4
By methylation-specific PCR, 29/70 (41.4%) of the cases showed
DAPK PM. On the other hand, 33/70 (47.1%) showed evidence of MT1G
PM. No PM detected for DAPK and MT1G in the control cases (Figures 12-
15).
Figure (12): DAPK methylation pattern.
Figure (13): MT1G methylation pattern
62
Figure (14): A photomicrograph for an ethidium bromide-stained
4% agarose gel showing: A) DAPK PM: case 1 & case 2 show pure
methylated DAPK gene & case 3 shows mixed pattern (methylated
& unmethyated) DAPK gene. (B)MT1G PM: case 1 & case 3 show
mixed methylated & unmethylated MT1G gene & case 2 shows pure
methylated MT1G gene (Gel electrophoresis, photoducmentation)
Figure (15): Summary of the
methylation analysis of
DAPK (D) and MT1G (M) in
DLBCL cases. The filled
boxes indicate the presence
of methylation and the open
boxes indicate the abscence
of methylation.
63
Treatment and response: -5
CHOP was given to 61/64 (95.3%) and no treatment was given in
3/64 (4.7%). No data was available about the treatment in 6 cases.
Complete remission (CR) occurred in 28/45 (62.2%) of the cases and non-
complete remission (Non-CR) (including partial remission (21.7%),
sustained disease (13%) and progressive disease (2.1%)) occurred in 17/45
(37.8%) of the cases with missing data in 25 cases (Figure 16).
62.2%
37.80%
0%
20%
40%
60%
80%
CR Non-CR
Response
Response
Figure (16): The treatment response in the cases (CR: complete
remission & Non-CR including PR: partial remission, SD: sustained
disease & PD: progressive disease).
vival:up and sur-Follow -6
Relapse occurred in 27.5% of the cases (Figure 17). After two years
follow-up, 41/58 (70.7%) of cases were alive and 17/58 (29.3%) were
dead. The cause of death in all cases, except one, was disease related (16
disease related and 1 unrelated). Twelve cases had lost follow up. The
overall survival (OS) ranged from 1-80 months. The two years OS was
51.5% (Figure 18). The disease free survival (DFS) ranged from 2-75
months. The two years DFS was 60% (Figure 19).
64
Figure (17): The occurrence of relapse in the DLBCL cases.
Figure (18): overall survival (OS) among DLBCL cases.
65
Figure (19): Disease free survival (DFS) among DLBCL cases
with complete remission (CR).
7- Correlations and relationships:
a) Univariate analysis:
1) Correlation between studied DAPK and MT1G promotor
hypermethylation:
Both genes were methylated in 20% of cases and both were
unmethylated in 31.4% of cases (P=0.873) (Table 6).
66
Table (5): Correlation of MT1G and DAPK promotor
hyermethylation
DAPK Total P-value
U M
MT1G U 22(31.4%) 15(21.4%) 37 0.873
M 19(27.1%) 14(20%) 33
Total 41 29 70
(U: unmethylated- M: methylated).
2) Correlations DAPK and MT1G PM and phenotypic classes with the
clinical parameters:
As regards DAPK promoter hypermethylation, a statistically significant
correlation was detected with resistance to therapy, P-value was 0.005
(Table7).
As regards the MT1G promoter hypermethylation, statistically
significant correlation was detected with stage IV & occurrence of relapse
(P-value 0.016 and 0.049, respectively) (Table 8).
As regards phenotypic classification of DLBCL (GCB & Non-GCB);
the Non-GCB was significantly correlated with poor IPI and occurrence of
relapse (P-values 0.024 and 0.030, respectively) (Table 9).
3) Survival correlations:
The overall survival was significantly correlated with age groups, B-
symptoms, number of extra-nodal sites, PS and IPI. The disease free
survival is significantly correlated with age, B-sypmtoms and number of
extra-nodal sites (Table 10).
67
Table (6): Correlation of DAPK methylation pattern with other
clinical & pathological parameters.
P-value
DAPK
Methylated
No (%)
Unmethylated No (%)
0.327
22(84.6)
4 (15.4)
30 (73.2)
11 (26.8)
Age:
≤60
˃60
0.107
9(30)
20(70)
20(41.4)
21(58.6)
Sex:
Female
Male
0.432
6(30.8)
16(69.2)
11(33.3)
22(66.7)
LDH:
normal
high
0.500
9(36)
16(64)
6(21.4)
22(75.6)
B-symptoms: absent
present
0.083
23(92)
2(8)
37(90.2)
4(9.8)
Extra-nodal sites: ≤1
≥2
0.092
5(20)
20(80)
11(26.8)
30(73.2)
Stage:
Early(I,II)
Advanced(III,IV)
0.591
13(59)
9(41)
24(58.5)
17(41.5)
PS:
<2
≥2
0.553
7(33.3)
14(66.7)
10(45.5)
22(54.5)
IPI:
Low/Low intermediate
High/High intermediate
0.005
12(60)
8(40)
5(20)
20(80)
Response:
Non-CR
CR
0.597
12(63.2)
7(36.8)
13(61.9)
8(38.1)
Relapse:
No relapse
Relapsed
0.704
17(65.4)
9(34.6)
28(73.7)
10(26.3)
CD10:
negative
positive
0.822
11(42.3)
15(57.7)
14(36.8)
24(63.2)
BCL6:
negative
positive
0.338
12(46.2)
14(53.8)
11(28.9)
27(71.1)
MUM1: negative
positive
0.194
13(48.1)
14(51.9)
10(26.3)
28(73.7)
Class: GCB
Non-GCB
Significant level (P-value<0.05).
68
Table (7): Correlation of MT1G methylation pattern with other
clinical & pathological parameters.
P-value
MT1G
Methylated
No (%)
Unmethylated
No (%)
0.350
26(81.2)
6(18.8)
26(74.3)
9(25.7)
Age:
≤60
˃60
0.333
16(48.5)
17(51.5)
13(35.1)
24(64.9)
Sex:
Female
Male
0.607
8(30.8)
18(69.2)
9(31)
20(69)
LDH:
normal
high
0.500
9(34)
16(64)
6(21.4)
22(75.6)
B-symptoms: absent
present
0.277
26(86.7)
4(13.3)
34(94.4)
2(5.6)
Extra-nodal sites: <2
≥2
0.315
9(30)
21(70)
7(19.5)
29(80.5)
Stage:
Early(I,II)
Advanced(III,IV)
0.016
17(56.7)
10(28.6) Stage: IV
0.318
15(51.7)
14(48.3)
22(64.7)
12(35.3)
PS:
<2
≥2
0.311
10(59)
17(61)
7(27)
19(73)
IPI:
Low/Low intermediate
High/High intermediate
0.278
6(28.6)
15(71.4)
11(45.8)
13(54.2)
Response:
Non-CR
CR
0.049
10(50)
10(50)
15(75)
5(25)
Relapse:
No relapse
Relapsed
0.254
18(62)
11(38)
27(77.1)
8(22.9)
CD10:
negative
positive
0.567
12(41.4)
17(58.6)
13(37.1)
22(62.9)
BCL6:
negative
positive
0.456
9(31)
20(69)
14(40)
21(60)
MUM1: negative
positive
0.106
13(44.8)
16(55.2)
10(27.8)
26(72.2)
Class: GCB
Non-GCB
Significant level (P-value<0.05).
69
Table (8): Correlation of DLBCL phenotypes with other clinical &
pathological parameters.
P-value
Immunophenotypic class
Non-GCB
No (%)
GCB
No (%)
0.104
27(73)
10(27)
19(90.5)
2(9.5)
Age:
≤60
˃60
0.387
15(40.5)
22(59.5)
11(47.8)
12(52.2)
Sex:
Female
Male
0.130
7(24.1)
22(75.9)
8(44.4)
10(55.6)
LDH:
normal
high
0.588
10(31.2)
22(68.8)
5(27.8)
13(72.2)
B-symptoms: absent
present
0.395
34(94.4)
2(5.6)
21(100)
0 (0.0)
Extra-nodal sites: ≤1
≥2
0.077
6(17.1)
29(82.9)
8(38.1)
13(61.9)
Stage:
Early(I,II)
Advanced(III,IV)
0.353
19(57.6)
14(42.4)
14(66.7)
7(33.3)
PS:
<2
≥2
030.0
5(16)
26(84)
10(59)
7(41)
IPI:
Low/Low intermediate
High/High intermediate
0.293
11 (47.8)
12(52.2)
5(33.3)
10(66.7)
Response:
Non-CR
CR
0.030
9(50)
9(50)
13(86.7)
2(13.3)
Relapse:
No relapse
Relapsed
0.151
24(65)
13(35)
11(47.8)
12(52.2)
DAPK
Unmethylated
methylated
0.206
23(62.2)
14(37.8)
11(47.8)
12(52.2)
MT1G
Unmethylated
methylated
Significant level (P-value<0.05).
70
Table (9): Correlation of OS and DFS with other parameters.
Disease free survival (DFS) Overall survival (OS)
P-value Median 2years DFS P-value Median 2years OS
550.0
080.0
Age:
28.1±4.6 68.8% 20±3.725
57.7% ≤60
6±0.82 0% 5±1.309 16.7% ˃60
0.407
0.472 Sex:
28.1±8.3 75% 15±6.103 47% Female
21.7±7.275 44.4% 18±1.363 50% Male
0.643
0.764
LDH:
19.4±14.6 66.7% 32.8±4.372 66.7% Normal
27.2±5.7 66.7% 18±0.074 50.7% High
0.002
0.01
B-symptoms: 36±7.1 87.5% 34.7±11.010 63.6% absent
20±1.992 33.3% 18±3.566 39% present
0.001
0.046
Extra-nodal sites: 27.2±5.1 64.7% 20±4.6
57.1% ≤1
2±0.01 0% 15±10.954 0% ≥2
0.713
0.125 Stage:
22.4±9.3 66.7% 27±6.67
71.4% Early(I,II)
21.7±7.9 58.3% 16±2.191 41.7% Advanced(II,IV)
0.465
060.0
PS:
19.4±7.7 45.5% 25.6±5.85
71.4% <2
28.1±3.82 83.3% 12±7.255 31.3% ≥2
0.954
210.0
IPI:
30±2.286 100% 30.133±2.383 83.3% Low
22.4±6.27 60% 25.567±7.071 36% High
0.603
0.501 Class:
27.2±6.285 75% 17.9±12.8
55.6% GCB
28.1±6.267 70% 18±3.2 47.7% Non-GCB
0.195
0.989 DAPK:
27.2±7.2 46.2% 21±6.042 52.6% Unmethylated
21.7±3.404 100% 16±4.137 41.7% Methylated
0.247
0.283 MT1G:
28.1±4.954 71.4% 18±0.111 53.8% Unmethylated
22.4±9.107 50% 18±4.577 44.4% Methylated
71
b) Multivariate analysis:
1) Correlations DAPK and MT1G PM, phenotypic class and IPI with
response to treatment:
DAPK and MT1G PM, phenotypic class and IPI were together
significantly correlated with response to treatment. DAPK PM was
independently significantly correlated with the response.
Table (10): Correlation of DAPK, MT1G, phenotypic class and IPI
with response.
Response
P-value
DAPK
MT1G
Phenotypic class
IPI
0.007
0.872
0.294
0.203
Constant 0.004
2) Correlations DAPK and MT1G PM, phenotypic classes and IPI
with occurance of relapse:
DAPK and MT1G PM, phenotypic class and IPI were together
significantly correlated with relapse. Phenotypic class independently
showed border line significant correlation with relapse.
72
Table (11): Correlation of DAPK, MT1G, phenotypic class and IPI
with relapse.
Relapse
P-value
DAPK
MT1G
Phenotypic class
IPI
0.133
0.823
0.060
0.155
Constant 0.020
3) Correlations DAPK and MT1G PM, phenotypic class and IPI with
overall survival (OS):
DAPK and MT1G PM, phenotypic class and IPI were together non-
significantly correlated with OS. IPI independently showed border line
significant correlation with OS.
Table (12): Correlation of DAPK, MT1G, phenotypic class and IPI
with OS.
Overall survival (OS)
P-value
DAPK
MT1G
Phenotypic class
IPI
0.591
0.893
0.231
0.058
Constant 0.329
73
4) Correlations DAPK and MT1G PM, phenotypic class and IPI
with disease free survival (DFS):
DAPK and MT1G PM, phenotypic class and IPI were together non-
significantly correlated with DFS but non of the former independently
correlated with DFS.
Table (13): Correlation of DAPK, MT1G, phenotypic class and IPI
with DFS.
Disease free survival (DFS)
P-value
DAPK
MT1G
Phenotypic class
IPI
0.456
0.644
0.950
0.327
Constant 0.550
5) Correlations DAPK and MT1G PM and phenotypic class and
IPI with IPI:
DAPK and MT1G PM and phenotypic class were together
significantly correlated with IPI. Phenotypic class independently
showed significant correlation with IPI.
74
Table (14): Correlation of DAPK, MT1G and phenotypic class with
IPI.
International prognostic index (IPI)
P-value
DAPK
MT1G
Phenotypic class
0.499
0.613
0.002
Constant 0.025
75
DISCUSSION
Diffuse large B-cell lymphoma (DLBCL) comprises the largest
subtype of non-Hodgkin’s lymphomas (NHL), accounting for
approximately 32% of all lymphomas seen throughout the world (Xu et al,
2001). DLBCL represents about 54.67% of NHL cases in national cancer
institute (NCI), cairo (Mokhtar et al, 2007).
DLBCL is characterized by a significant spectrum of morphologic,
immunophenotypic and molecular genetic features. The treatment of
patients with DLBCL has been guided traditionally by clinical parameters
such as the Ann Arbor Staging and International Prognostic Index (IPI).
Although the IPI represents the most widely accepted prognostic model,
there is still a marked variability in patient's outcome within identical IPI
subgroups, reflecting the heterogeneity of this malignancy. Recent
application of DNA microarray, real-time reverse transcription polymerase
chain reaction, and tissue array immunohistochemistry makes the
development of new classifications possible based on molecular profiling.
The molecular classification of DLBCL may lead to grouping of specific
disease entities sharing similar biologic features, clinical behavior, and
outcome (Morgensztern and Lossos, 2005).
Beyond genetic alterations, identification of novel epigenetic
alterations, such as aberrant promotor methylation may yield better
76
diagnostic, prognostic and therapeutic information. (Jones and Baylin,
2007).
The profiles of promotor hypermethylation of the CpG islands in
tumor-suppressor genes are specific to the tumor type (Agrelo and Wutz,
2009). Thus, each tumor type can be assigned a specific, defining DNA
“hypermethylome.” Such patterns of epigenetic inactivation occur not only
in sporadic tumors but also in inherited cancer syndromes, in which
hypermethylation can be the second lesion in Knudson’s two-hit model of
how cancer develops. Recently devised epigenomic techniques have
revealed maps of hypermethylation of the CpG islands that suggest the
occurrence of 100 to 400 hypermethylated CpG islands in the promoter
regions of a given tumor (Esteller, 2007).
Hematological neoplasms are known to have different
hypermethylation profiles than those of other solid tumors ( Takahashi1,
2004). In a recent study, the prevalence of aberrant promtor methylation of
eight TSGs that are known to be involved in cell cycle regulation (p16,
COX2), DNA repair (MGMT), apoptosis (DAPK, RUNX3), inhibition of
angiogenesis (THBS1), invasion and metastasis (CDH1) and cell
proliferation (MT1G). This study demonstrated that aberrant CpG island
methylation is a frequent event in NHLs especially diffuse large B-cell
lymphomas with overlapping but distinct methylation profiles (Kim1 et al.
2008).
77
DAPK and MT1G genes were the most commonly affected gene by
promoter hypermethylation in DLBCL according to Kim et al 2008. Thus
we attempted in the current study to determine the prevalence of DAPK
and MT1G PM in a well characterized cohort of DLBCL from Egypt (70
cases) and to assess their prognostic and predictive value.
Methylation specific PCR was done using high molecular weight DNA
extracted from formalin fixed, paraffin embedded tissues. Inspite of the fact
that DNA extract from fresh tissue is better in quality and quantity, but the
absence of tissue banking system in NCI makes availability of fresh tissue
is difficult. Another disadvantage of using paraffin embedded tissue is that
purification of DNA extracted from paraffin embedded sections is limited
by the length of fixation, with longer fixation times leading to a steep
increase in cross-links between proteins and nucleic acid. The result of
extensive cross-linking is often evidenced by limitations of purification of
the final DNA product. However, using of paraffin embedded tissue is
widely used all over the word due to conveniently & availability and to
allow retrospective studies with full data and long follow-up.
We chose the methylation specific PCR method as it is the cheapest
and the most widely used method that fits our limited resources. Bisulfite
modification is a principle step in this technique. The latter requires low
PH (PH 5) to be accomplished which again affect the integrity of the DNA
increasing the possibility of false negative results. We tried to overcome
this problem by increasing the starting amount of DNA entering the
78
bisulfite modification reaction, avoiding prolonged time of DNA in low PH
and finally by re-introducing of the PCR products of negative cases to
another PCR reaction simulating semi-nested PCR in which some cases
which were false negative in the first PCR reaction turned into positive in
the second PCR.
The available clinical data were collected from the clinical records but
we faced a major problem due to the large fraction of missing data as
regards clinical parameters and survival rates. We tried to solve these
statistical problems by compiling possible parameters as early (I&II) and
advanced (III&IV) stages, low/low intermediate IPI and high/high
intermediate IPI and CR and non-CR. OS and DFS were divided into above
and below median to allow for multivariate analysis. Missing data were
excluded from calculation.
The median age of our studied cases was 54 year, 77.6% of the Cases
were ≤60 years and 22.4% of the cases were >60 years. The median age
reported in the current study is slightly higher than that reported in an early
study performed in NCI (Cairo) by Abdel Aty et al (2004) including 90
cases. In this study, the median age was 49 years, 90% of cases were ≤60
years and 10% of cases were >60 years. Similarly, the NCI pathology
registry showed comparable median age (47.2 years), Mohktar et al (2007).
This slight increase in the median age suggests a shift in the incidence of
DLBCL cases to an older age group. On the other hand, the median age of
DLBCL is much higher in western countries since the WHO (2008)
79
reported a median age in the 7th
decade which is in accordance with the
median age reported by Ott et al (2010) (69 years) in a large series of 949
cases.
Our study shows a slight male predilection with a male to female ratio
of 1.4:1. Similar finding was detected in the study of Abdel Aty et al
(2004) as well as in the NCI pathology registry (2007). However, some
studies showed a reversed ratio with slight female sex predilection (1:1.25),
Broyde et al (2009).
B-symptoms were reported in 28% of our studied cases. This is in
agreement with Amara et al (2008) who reported B-symptoms in 26% of
their studied cases. However, Abdel Aty et al (2004) and Mosaad et al
(2008) from the NCI, Cairo, reported higher percentages (54.4% and 55%,
respectively). However, lower percentage (20%) was reported in the WHO
(2008).
A high LDH serum level was reported in 69.1% of the cases in the
current study. The results of Abdel Aty et al (2004) are lower than our
study since they found high LDH serum level in 57.2% of cases. The
literature shows high LDH serum level in variable percentages including
higher percentage (84.8%) in the study of Amara et al (2008) and lower
percentage (38%) reported by Shu-Nan et al, (2009).
80
The majority of our cases (74.2%) had advanced stages (stages III &
IV). This finding is comparable to Liu et al (2008), who reported 77.9% of
advanced stages in DLBCL cases from China. Other studies performed on
Egyptian patients from the NCI, Cairo, Abdel Aty et al (2004) showed
higher percentage (95.6%) while Mosaad et al (2008) showed lower
percentage (57%) in their studied cases.
Out of the 70 studied cases, 41.3% had poor performance status
(PS≥2), which is comparable with other studies (Abdel Aty et al, 2004),
(Amara et al, 2008) and (Broyde et al, 2009).
The number of extra-nodal sites affected were 2 in 9.1% of the
studied cases, which is in agreement with Abdel Aty et al (2004) (10%).
However, higher frequencies were reported by Amara et al (2008) (15.5%)
and Broyde et al (2009) (30%).
As regards the international prognostic index, cases with low/low-
intermediate IPI and high-intermediate/high IPI represented 32.1% and
67.9%, respectively. Our data were comparable to those of Abdel Aty et al
(2004) in which low/low-intermediate IPI and high-intermediate/high IPI
represented 21.1% and 78.9%, respectively. Low/low-intermediate IPI and
high-intermediate/high IPI represented 71.7% and 28.3%, respectively,
according to Amara et al (2008), 46% and 54%, respectively, as reported
by Liu et al (2008) and 56.2% & 43,8%, respectively as shown by Broyde
81
et al (2009). Thus, our study and the study of Abdel Aty et al (2004)
showed the highest percentage of high-intermediate/high IPI (67.9%) and
this is related to high percentages of advanced stages and poor PS in
Egyptian patients.
Our data shows that CD10, BCL6 and MUM1 were positive in 30.8%,
60% and 63% of our cases, respectively. According to the
immunophenotyping results, we subclassified our studied cases into
germinal center B-phenotypic class (GCB) and non- germinal center B-
phenotypic class (Non-GCB). The GCB class represented 37.7% (23/61) of
the cases. Hans et al 2004 reported a comparable percentages 27.6%
(42/152), 55.9% (85/152) and 46.7% (71/152) as regarding CD10, BCL6 &
MUM1, respectively. Xu et al 2001 showed higher percentage 43.4%
(23/53) positive to CD10. Joanaa et al 2010 found positive CD10 in lower
percentage 13.6% (9/66) and comparable percentages 53% (35/66) and
54.5 (36/66) for BCL6 and MUM1, respectively.
The Percentage of GCB class shows variation in the literatures ranging
from 22% (Chen et al, 2010) up to 73% (Joanna et al, 2010). Some
studies showed racial difference in the phenotypic class distribution, in this
regard, Chen et al (2010) compared distribution of GCB DLBCL and Non-
GCB DLBCL in Chinese and Western cases. They found that Chinese
cases showed lower percentage of GCB class 22% (27/124) in contrast to
Western cases 52.6% (60/114).
82
The current study addresses the possible role of DAPK and MT1G
promotor hypermethylation (PM) in the development and progression of
DLBCL. DAPK PM was detected in 41.4% (29/70) of the studied cases but
not in the control cases.
Literature review shows variable incidence of DAPK PM ranging
from 22.2% up to 94.3%. In hematolymphoid malignancies, Gutierrez et al
(2003) reported lower percentage (30%) of DAPK PM in acute
lymphoblastic leukemia (ALL) cases while higher percentages of DAPK
PM were reported by Huang et al (2007) (55%) in gastric lymphoma,
Amara et al (2007) (74%) in Tunisian DLBCL cases, Kim et al (2008)
(76.1%) (35/46) in Korean DLBCL cases and much higher percentage
(94.3%) was reported by Choung et al (2012) in Korean ocular MALT
lymphoma. This variation in DPAK PM may be related to: 1) differences in
the detection methods, 2) differences in the nature of specimen (fresh or
paraffin embedded tissues), 3) differences in tumor type and subtypes as
well as 4) population difference that includes variable etiologic,
carcinogenic and genetic factors.
Other non-hematologic malignancies show variable percentages of
DAPK PM. So, Kato et al (2008) reported DAPK PM in 22.2% of gastric
carcinoma, Xian-Lan et al (2008) reported DAPK PM in 65.4% cervical
carcinoma cases and Hee Jung Park et al (2010) reported 28.1% of DAPK
PM in urothelial carcinoma cases.
83
The current study showed MT1G PM in 47.1% (33/70) of the studied
cases but not in the control cases. The available data in the literature show
variable incidences of MT1G PM ranging from 24% up to 76.1%. There
are only two published manuscripts reporting on MT1G PM in lymphoma;
Kim et al (2008) reported higher percent (76.1%) of MT1G PM in Korean
DLBCL cases while Choung et al (2012) reported comparable percentage
(48.6%) of MT1G PM in Korean ocular MALT lymphoma cases.
In other non-hematologic malignancies, Henrique et al (2005) reported
MT1G PM in 24% of prostatic carcinoma cases, Ferrario et al (2008)
reported MT1G PM in 60% of papillary carcinoma cases, Kanda et al
(2009) reported MT1G PM in 60.4% of hepatocellular carcinoma cases and
Luis et al (2010) reported MT1G PM in 55% of hepatoblastoma cases.
Correlation of the promoter hypermehtylation status with the standard
clinico-pathological prognostic factors of the patients was attempted when
there is available clinical data. DAPK PM showed significant correlation in
both univariate and multivariate analysis with resistance to therapy as non-
CR was reported in 60% of the cases with DAPK PM compared to 20% in
the unmethylated cases. On the other hand, CR reported in 40% of cases
with DAPK PM compared to 80% in the unmethylated cases. Amara et al
(2008) showed comparable results regarding response to therapy where
81% of DLBCL cases resistant to chemotherapy showed DAPK PM. The
study of Amara et al (2008) represents the only study which correlated
DAPK PM with patients response to treatment in lymphoma, however,
84
other studies on different solid tumors revealed comparable results
including the study of Kato et al (2008) who reported a response rate to
chemotherapy of 21.4% and 44.8% in gastric carcinoma with methylated
and unmethylated DAPK, respectively. All these results confirm that
DAPK PM seems to play a significant role in the acquisition of resistance
to chemotherapy. This could be explained by the fact that DAPK has a pro-
apoptotic function which is blocked by its PM which in turn increases the
resistance to chemotherapy. Therefore, DAPK PM could be considered a
possible predictive factor that can predict response to treatment in hemato-
lymphoid and other solid malignancies.
As regards the correlation between survival rates and DAPK PM, we
found that the median OS in the current study for methylated and
unmethylated DAPK cases was 16 months and 21 months, respectively,
while the median DFS was 21 months and 27 months, respectively.
Although in our study DAPK PM was correlated with shortened OS and
DFS, this correlation did not reach significant level. This might be related
to the large fraction of cases with lost follow up (20/70 cases) and the short
period of follow-up (2 years). In multivariate analysis by Amara et al
(2008) DAPK PM was found to be an independent prognostic factor in
predicting shortened OS (P=0.001) and DFS (P=0.024). Similarly,
Hoffmann, et al. (2009) reported that preoperative measurement of
methylated DAPK in peripheral blood DNA may contribute to better
estimate of postoperative survival chances of patients with esophageal
carcinoma, especially adenocarcinoma. Therefore, we recommend a larger
sample size and a longer follow-up period (5years) to get more accurate
results.
85
On the other hand, DAPK PM in our cases showed no significant
correlation with the other clinic-pathologic features of the patients
including age, sex, LDH level, B-symptoms, IPI, Ann Arbor Stage, number
of extra nodal sites affection, PS, relapse and phenotypic class. Our data in
this regard are comparable to Amara et al (2008).
As regards MT1G PM; there were significant correlations in
univariate and multivariate analysis with stage IV and the incidence of
relapse. Stage IV was reported in 56.7% of methylated MT1G and in
28.6% of unmethylated MT1G. Similarly, the incidence of Relapse was
50% in methylated MT1G and 25% in unmethylated MT1G. MT1G PM is
not widely studied in hematologic malignancies thus comparison of our
data with other studies was not valid, however all previous correlations
considering other solid malignancies confirm that MT1G PM increases
tumor aggressiveness. For instance in prostatic carcinoma, significant
correlation was detected in higher stages by Henrique et al (2005) as stages
T3 & T4 were 65.5% (19/29) in methylated MT1G & 42.4% (39/92) in
unmethylated MT1G (P=0.049). Similarly, Sakamoto et al (2010) reported
that MT1G PM showed a significant correlation with poor prognosis of
patients with hepatoblastoma. Kanda et al (2009) reported that MT1G PM
indicating a poorer prognosis in hepatocellular carcinoma than the negative
group, although the difference was not significant (p<0.0978).
On the other hand, MT1G PM showed no significant correlation with
age, sex, LDH level, B-symptoms, IPI, extra-nodal sites affection, PS,
response to therapy, phenotypic class or survival rates.
The DLBCL phenotypic classes in the current study showed significant
correlations with IPI and relapse rate (in both univariate and multivariate
86
analysis). A high/high intermediate IPI (3-5) was reported in 84% of non-
GCB class and 41% of GCB class, whereas, low/low intermediate IPI (0-2)
was reported in 16% of Non-GCB class and in 59% of GCB class. Relapse
rate in non-GCB and GCB classes was 50% and 13.3%, respectively. On
the other hand, phenotypic classes showed near significant correlation
(P=0.077) with stage since advanced stages (Stages III&IV) were reported
in 83% and 62% of Non-GCB and GCB, respectively, while early stages
(Stages I&II) were reported in 17% and 38% of Non-GCB and GCB,
respectively.
No significant correlations were detected with age, sex, LDH level, B-
symptoms, PS, response, DAPK, MT1G or survival rates. Comparing our
data regarding the phenotypic classes with those of Joanna et al (2010), the
later showed significant correlation of DLBCL phenotypic classes with
stage, IPI, response and overall survival as follows: Early stages (I&II) in
GCB and Non-GCB were 89% (16/18) and 42% (20/48), respectively,
while advanced stages (III&IV) in GCB and Non-GCB were 11% (2/18)
and 58% (28/48), respectively (P= 0.001). Similarly, a low IPI in GCB and
non-GCB were 72% (13/16) and 40.5% (17/42), respectively. High IPI was
detected in 17% (3/16) of the GCB class and in 59.5% (25/42) of the non-
GCB class (P=0.008). CR in GCB and non-GCB were 94% (17/18) and
54% (26/48), respectively. Non-CR in GCB and non-GCB were 6% (1/18)
and 46% (22/48), respectively (P=0.003). The 5-year OS for the GCB
group was 83% compared with only 30% for the non-GCB group. No
significant correlation detected with age groups, sex, B-symptoms, LDH
level and number of extra-nodal sites. Our study showed no correlation
with response to therapy and survival in relation to phenotypic class which
may be related to the large fraction of cases with lost follow up (20 out of
87
70 cases) and the short period of available follow up (2years) in contrast to
Joanna et al (2010) (5years survival).
In the current study, 62.2% of our cases showed complete remission
after treatment. Our result in this context is comparable to Xu et al (2001)
(66.7%), Abdel Aty et al (2008) (61.1%) and Amara et al (2008) (56%) but
slightly lower than Mosaad et al (2008) in which CR was reported in 78.6%
of cases.
Similarly, the high rate of relapse (27.5%) in our studied cases after CR
is higher than Xu et al (2001) who reported a much lower rate of relapse
(7.7%) but our finding is comparable to Abdel Aty et al (2008) (34.5%).
The two years overall survival (OS) reported in the current study
(51.5%) is lower than those reported by Abdel Aty et al (2004) 80%,
Amara et al (2008) 55% and Mosaad et al (2008) 63%.
Our two years DFS (60%) is comparable to those reported by Abdel
Aty et al (2004) 54.4%, Amara et al (2008) 46% and Mosaad et al (2008)
60%. The low OS in the current study could be attributed to the high
percent of high IPI cases and cases with non-GCB phenotypic class.
88
In our study, the OS was significantly correlated in the current study
with age groups, B-symptoms, number of extra-nodal sites, PS and IPI.
Disease free survival was significantly correlated with age, B-symptoms
and number of extra-nodal sites only.
89
CONCLUSIONS
Our work is a preliminary study which needs to be verified in a
larger study containing more cases with complete clinical data and longer
follow-up. However, according to this study we can conclude the
following:
1) Promotor methylation of DAPK and MT1G can be used as molecular
prognostic and predictive markers in specific cases of DLBCL. DAPK PM
can predict resistance to chemotherapy in DLBCL, whereas MT1G PM can
predict the incidence of relapse.
3) The Egyptian DLBCL shows high percentage of Non-GCB class
(62.3%) using the standard immunophenotyping panel (CD10, BCL6
and MUM1) denoting an aggressive behavior.
3) The IPI is still considered the strongest prognostic factor in DLBCL in
prediction of overall survival.
90
RECOMMENDATIONS
Studying of PM in a larger panel of genes using recent profiling
technique in order to identify more reliable predictive biomarkers.
Implementation of tissue banking system to provide fresh tissue for
molecular research.
Improvement of the hospital filling to provide complete set of
clinical data.
91
ANCESREFER
Abed Aty, K. Zekry, M. Abu Taleb et al (2004): "Evaluation of serum CD44 in NHL
patients and its relation to IPI and response to treatment." MD thesis, NCI, Cairo
university.
Agrelo, R., W. H. Cheng, et al. (2006): "Epigenetic inactivation of the premature aging Werner
syndrome gene in human cancer." Proc Natl Acad Sci U S A 103(23): 8822-7.
Agrelo, R. and A. Wutz (2009): "Cancer progenitors and epigenetic contexts: An existing
connection." Epigenetics 4(8): 21-31.
Al-Kuraya K, Narayanappa R and Siraj AK(2006): "High frequency and string prognostic
relevance of O6-methylguanine DNA methyltransferase silencing in diffuse large B-cell
lymphomas from the Middle East." Hum Pathol. Jun; 37(6):742-8.
Amara K, M. Trimeche1, S. Ziadi1, et al (2008):" Prognostic significance of aberrant promoter
hypermethylation of CpG islands in patients with diffuse large B-cell lymphomas",
Annals of Oncology 19: 1774–1786.
Andrews GK (2000): "Regulation of metallothionein gene expression by oxidative stress and
metal ions." Biochem Pharmacol 59:95–104.
Armitage (2005). "Staging non-Hodgkin lymphoma. ." CA Cancer J Clin 55: 368-76.
Bachman KE, P. B., Rhee I, et al. (2003): "Histone modifications and silencing prior to DNA
methylation of a tumor suppressor gene." Cancer Cell 3: 89-95.
Banham AH, C. J., Brown PJ, et al. (2005): "Expression of the FOXP1 transcription factor is
strongly associated with inferior survival in patients with diffuse large B-cell
lymphoma." Clin Cancer Res 11: 1065-72.
Batbayar Khulan,1 Reid F. Thompson, et al (2006): " Comparative isoschizomer profiling of
cytosine methylation: The HELP assay." Genome Res. August; 16(8): 1046–1055.
92
Beltran, A. S., X. Sun, et al. (2008): "Reprogramming epigenetic silencing: artificial
transcription factors synergize with chromatin remodeling drugs to reactivate the tumor
suppressor mammary serine protease inhibitor." Mol Cancer Ther 7(5): 1080-90.
Bernstein BE, M. A., Lander ES. (2007): "The mammalian epigenome." Cell 128: 669-681.
Boudova L, T. E., Delabie J, et al. (2003): "Nodular lymphocyte-predominant Hodgkin
lymphoma with nodules resembling T-cell/histiocyte-rich B-cell lymphoma: differential
diagnosis between nodular lymphocyte-predominant Hodgkin lymphoma and T-
cell/histiocyte-rich B-cell lymphoma." Blood 3753-8.
Broyde A, Boycov O, Strenov Y, et al (2009): " Role and prognostic significance of the Ki-67
index in non-Hodgkin's lymphoma" Am J Hematol .84(6):338-43.
Bueno, M. J., I. Perez de Castro, et al. (2008): "Genetic and epigenetic silencing of microRNA-
203 enhances ABL1 and BCR-ABL1 oncogene expression." Cancer Cell 13(6): 496-
506.
Cairns, P. and P. D. Adams (2004): "Clipped wings: epigenetic inactivation of APC." Cancer
Biol Ther 3(10): 965-6.
Christine P. Hans, Dennis D. Weisenburger, et al (2004): "Confirmation of the molecular
classification of diffuse large B-cell lymphoma by immunohistochemistry using a tissue
microarray" The American Society of Hematology, Blood, 103(1): 275-82.
Cloos PA, C. J., Agger K, et al. (2006): "The putative oncogene GASC1 demethylates tri- and
dimethylated lysine 9 on histone H3." Nature 442: 307-311.
Coyle P, Philcox JC, Carey LC, et al. (2002): "Metallothionein: the multipurpose protein". Cell
Mol Life Sci, 59:627–47.
Costa, V. L., R. Henrique, et al. (2007): "Epigenetic markers for molecular detection of prostate
cancer." Dis Markers 23(1-2): 31-41.
93
Deiss LP, Feinstein E, Berissi H, et al.(1995): " Identification of a novel serine/threonine kinase
and a novel 15-kd protein as po-tential mediators of the gamma interferon-induced cell
death". Genes Dev, 9: 15–30.
Delecluse HJ, A. I., Dallenbach F, et al. (1997): " Plasmablastic lymphomas of the oral cavity: a
new entity associated with the human immunodeficiency virus infection." Blood 89:
1413-20.
Dong HY, S. D., de Leval L, Tang Z, et al (2005): "Plasmablastic lymphoma in HIV-positive
patients: an aggressive Epstein–Barr virus-associated extramedullary plasmacytic
neoplasm." Am J Surg Pathol 29: 1633–41.
Dong HY, B. P., Liu Z, Gangi M. (2008): "PAX-5 is invariably expressed in B-cell lymphomas
without plasma cell differentiation." Histopathology 53: 278-87.
Dunphy, C. H. (2010): "Diffuse Large B-Cell Lymphomas". Molecular Pathology,10: 978-1007.
Du MQ, L. H., Diss TC, et al. ( 2001): "Kaposi sarcoma-associated herpesvirus infects
monotypic (IgM lambda) but polyclonal naive B cells in Castleman disease and
associated lymphoproliferativedisorders." Blood 97: 2130-6.
Eden A, G. F., Waghmare A, Jaenisch R. (2003): "Chromosomal instability and tumors
promoted by DNA hypomethylation." Science 300: :455-455.
Espada, J., E. Ballestar, et al. (2007): "Epigenetic disruption of ribosomal RNA genes and
nucleolar architecture in DNA methyltransferase 1 (Dnmt1) deficient cells." Nucleic
Acids Res 35(7): 2191-8.
Esteller, M. and J. G. Herman (2002): "Cancer as an epigenetic disease: DNA methylation and
chromatin alterations in human tumours." J Pathol 196(1): 1-7.
94
Esteller, M. (2007): "Epigenetic gene silencing in cancer: the DNA hypermethylome." Hum
Mol Genet 16 (1): 50-9.
Feinberg, A. P., H. Cui, et al. (2002): "DNA methylation and genomic imprinting: insights from
cancer into epigenetic mechanisms." Semin Cancer Biol 12(5): 389-98.
Feng, Y. Q., R. Desprat, et al. (2006): "DNA methylation supports intrinsic epigenetic memory
in mammalian cells." PLoS Genet 2(4): 65.
Ferrario C., Paola Lavagni, Manuela Gariboldi, et al. (2008):" Metallothionein 1G acts as an
oncosupressor in papillary thyroid carcinoma" Laboratory Investigation 88: 474–481.
Fisher RI, Gaynor ER, Dahlberg S, et al. (1993): "Comparison of a standard regimen (CHOP)
with three intensive chemotherapy regimens for advanced non-Hodgkin's lymphoma."
N Engl J Med, 328:1002.
Fraga, M. F., E. Ballestar, et al. (2005): "Epigenetic differences arise during the lifetime of
monozygotic twins." Proc Natl Acad Sci U S A, 102(30): 10604-9.
Fraga, M. F., R. Agrelo, et al. (2007): "Cross-talk between aging and cancer: the epigenetic
language." Ann N Y Acad Sci, 1100: 60-74.
Gaudet F, H. J., Eden A, et al. (2003): "Induction of tumors in mice by genomic
hypomethylation." Science, 300: 489-492.
Gedamu, L., Varshney, U., Jahroudi, N., et al. (1987): ''Structure and expression of the human
metallothionein genes" Experientia Suppl 52:361-72.
Gutierrez MI, Siraj AK, Khaled H, et al (2004): "CPG island methylation in Schistosoma and
non-Schistosoma associated bladder cancer", Pathol, 17(10):1268-74.
95
Hans, C. P., D. D. Weisenburger, et al. (2004): "Confirmation of the molecular classification of
diffuse large B-cell lymphoma by immunohistochemistry using a tissue microarray."
Blood 103(1): 275-82.
He L, Hannon GJ (2004): " MicroRNAs: small RNAs with a big role in gene regulation" Nat
Rev Genet 5(7):522-31.
Hegi, M. E., D. Sciuscio, et al. (2009): "Epigenetic deregulation of DNA repair and its potential
for therapy." Clin Cancer Res 15(16): 5026-31.
Hellebrekers, D. M., K. Castermans, et al. (2006): "Epigenetic regulation of tumor endothelial
cell anergy: silencing of intercellular adhesion molecule-1 by histone modifications."
Cancer Res 66(22): 10770-7.
Henrique, R., Jerónimo, C., Hoque, M.O., et al (2005): ''MT1G hypermethylation is associated
with higher tumor stage in prostate cancer.'' Cancer Epidemiol. Biomarkers Prev 14(5):
1274-8.
Herman, J. G. (2005): "Epigenetic changes in cancer and preneoplasia." Cold Spring Harb
Symp Quant Biol 70: 329-33.
Herman JG, G. J., Myöhänen S, Nelkin BD, et al (1996): "Methylation-specific PCR: a novel
PCR assay for methylation status of CpG islands." Proc. Natl. Acad. Sci. U.S.A. 93
(18): 9821–6.
Hiraga J, Kinoshita T, Ohno T(2006): "Promtor hypermethylation of the DNA-repair gene O6-
methylguanine-DNA methyltransferase and P53 mutation in diffuse large B-cell
lymphoma." Int J Hematol. 84(3):248-55.
Hoefnagel JJ, D. R., Basso K, et al. (2005): "Distinct types of primary cutaneous large B-cell
lymphoma identified by gene expression profiling." Blood 105: 3671-8.
96
Hoffmann, A.-C., Vallböhmer, D., Prenzel, et al (2009): "Methylated DAPK and APC promoter
DNA detection in peripheral blood is significantly associated with apparent residual
tumor and outcome." Journal of cancer research and clinical oncology, 135(9):, 1231-7.
Hsu SM, Raine L, and Fanger H (1981): "Use of avidin-biotin peroxidase complex (aBC) in
immunoperoxidase techniques: a comparison between aBC and unlabeled antibody
(PaP) procedures." J Histochem Cytochem, 29:577-80
Huang, Y., de la Chapelle, A., Pellegata, N.S. Int. J. (2003): ''Hypermethylation, but not LOH,
is associated with the low expression of MT1G and CRABP1 in papillary thyroid
carcinoma." Cancer 104(6): 735-44.
Jahroudi, N., Foster, R., Price-Haughey, J., et al (1990): '' Cell-type specific and differential
regulation of the human metallothionein genes. Correlation with DNA methylation and
chromatin structure." Biol. Chem 265(11): 6506-11.
Jang CW, Chen CH, Chen CC, et L (2002): "TGF-beta induces apoptosis through smad-
mediated expression of DAP-kinase." Nat Cell Biol 4:51–8.
Jenuwein, T. (2006): "The epigenetic magic of histone lysine methylation." FEBS J 273 (14):
3121-35.
Joanna J., Dorota J., Piotr P., et al (2010): "COMPARISON OF TWO DIFFERENT
IMMUNOHISTOCHEMICAL ALGORITHMS IDENTIFYING PROGNOSTIC
SUBGROUPS OF DLBCL", POL J PATHOL 3: 124–132.
Johnson, R. L., W. Huang, et al. (2008): "A quantitative high-throughput screen identifies
potential epigenetic modulators of gene expression." Anal Biochem 375(2): 237-48.
Jones PA and Baylin SB(2007): "The epigenomics of cancer", cell; 128(4) : 683-92.
97
Joshi, B., Ordonez-Ercan, D., Dasgupta, P., et al (2005): ''Induction of human metallothionein
1G promoter by VEGF and heavy metals: differential involvement of E2F and metal
transcription factors." S. Oncogene.
Kaneda, A. and A. P. Feinberg (2005): "Loss of imprinting of IGF2: a common epigenetic
modifier of intestinal tumor risk." Cancer Res 65(24): 11236-40.
Karpf, A. R. (2006): "A potential role for epigenetic modulatory drugs in the enhancement of
cancer/germ-line antigen vaccine efficacy." Epigenetics 1(3): 116-20.
Kato K., Satoru Iida, Hiroyuki Uetake, et al. (2008): " Methylated TMS1 and DAPK genes
predict prognosis and response to chemotherapy in gastric cancer", Int. J. Cancer: 122,
603–608
Kiefer, T.L., Lai, L., Yuan, L., et al (2005): "Differential regulation of estrogen receptor alpha,
glucocorticoid receptor and retinoic acid receptor alpha transcriptional activity by
melatonin is mediated via different G proteins." Pineal Res 38(4): 231-9.
Kim, Sun Och Yoon,∙Young A Kim, et al.(2008): "Diffuse Large B Cell Lymphoma Shows
Distinct Methylation Profiles of the Tumor Suppressor Genes among the Non-
Hodgkin’s Lymphomas." Lymphomas 42: 16-20.
Kristensen LS, M. T., Krypuy M, Dobrovic A (2008): "Sensitive Melting Analysis after Real
Time- Methylation Specific PCR (SMART-MSP): high-throughput and probe-free
quantitative DNA methylation detection." Nucleic Acids Res 36(7)(42).
Laird,C. D., N. D. Pleasant, et al. (2004): "Hairpin-bisulfite PCR: assessing epigenetic
methylation patterns on complementary strands of individual DNA molecules." Proc
Natl Acad Sci U S A 101(1): 204-9.
Lehmann,U. Berg-Ribbe I, Wingen LU et al. (2009): "Distinct methylation patterns of benign
and ma- lignant liver tumors revealed by quantitative methylation profiling." Clin.
Cancer Res., 11: 3654–3660.
98
Lenz G, W. G., Dave SS, et al. (2008): "Stromal gene signatures in large-B-cell lymphomas." N
Engl J Med Clin (Barc) 359: 2313–23.
Lenz G, W. G., Emre NC, et al. (2008): "Molecular subtypes of diffuse large B-cell lymphoma
arise by distinct genetic pathways." Proc Natl Acad Sci U S A 105: 13520–5.
Li J, Yan H, Wang K, et al (2007): "Hairpin fluorescence DNA probe for real-time monitoring
of DNA methylation". Anal. Chem. 79 (3): 1050–1056.
Linderoth J, Jerkeman M, Cavallin-Stahl E, et al (2003): "Immunohistochemical expression of
CD23 and CD40 may identify prognostically favorable subgroups of dif- fuse large B-
cell lymphoma: a Nordic Lymphoma Group Study." Clin Cancer Res.9:722–728
Lo, P. K., J. Mehrotra, et al. (2006): "Epigenetic suppression of secreted frizzled related protein
1 (SFRP1) expression in human breast cancer." Cancer Biol Ther 5(3): 281-6.
Lossos IS, A. A., Eisen MB, et al. (2000): " Ongoing immunoglobulin somatic mutation in
germinal center B cell-like but not in activated B cell-like diffuse large cell
lymphomas." Proc Natl Acad Sci U S A 97: 10209–13.
Lossos IS, A. A., Rajapaksa R, Tibshirani R, et al. (2003): "HGAL is a novel interleukin-4-
inducible gene that strongly predicts survival in diffuse large B-cell lymphoma." Blood
101: 433-40.
Lu J, G. G., Miska EA, et al. (2005): "MicroRNA expression profiles classify human cancers."
Nature 435: 834-838.
Mack, G. S. (2006) "Epigenetic cancer therapy makes headway." J Natl Cancer Inst 98(20):
1443-4.
Marks, H., J. C. Chow, et al. (2009): "High-resolution analysis of epigenetic changes associated
with X inactivation." Genome Res 19(8): 1361-73.
99
Mohn F, Weber M, Schübeler D, et al (2009): " Methylated DNA immunoprecipitation
(MeDIP)" Methods Mol Biol ; 507:55-64.
Mokhtar N, Gouda I and Adel I(2007): "Cancer pathology registry 2003-2004", Departement of
pathology, National Cancer Institute, cairo university, Egypt; 5:46-48.
Montesinos-Rongen M, S. R., Deckert M. (2009): " Primary lymphoma of the central nervous
system:just DLBCL or not?" Blood 113: 7-10.
Moore T. (2009): "Southern analysis using methyl-sensitive restriction enzymes ". Methods
Mol Biol. ;181:193-203.
Moore, M. and C. Ullman (2003): "Recent developments in the engineering of zinc finger
proteins." Brief Funct Genomic Proteomic 1: 342-355.
Mosaad, Khaled, Abdel Warith and N. Zekry (2008): "comparative study between classic
CHOP and CHOP+VP16 in treatment of diffuse large cell lymphoma", MD thesis, NCI,
Cairo university.
Nador RG, C. E., Chadburn A, et al. (1996): "Primary effusion lymphoma: a distinct
clinicopathologic entity associated with the Kaposi’s sarcoma-associated herpes virus."
Blood 88: 645-56.
Natkunam, Y., P. Farinha, et al. (2008): "LMO2 protein expression predicts survival in patients
with diffuse large B-cell lymphoma treated with anthracycline-based chemotherapy
with and without rituximab." J Clin Oncol 26(3): 447-54.
Oki, Y. and J. P. Issa (2007): "Treatment options in advanced myelodysplastic syndrome, with
emphasis on epigenetic therapy." Int J Hematol 86(4): 306-14.
Oyama T, I. K., Suzuki R, et al. (2003): "Senile EBV + B-cell lymphoproliferative disorders: a
clinicopathologic study of 22 patients." Am J Surg Pathol 27: 16-26.
Ozda H, T. A., Ahmed AA, et al. (2006): "Differential expression of selected histone modifier
genes in human solid cancers." BMC Genomics 7: 90-90.
100
Pillai S, Chellappan SP.(2009)." ChIP on chip assays: genome-wide analysis of transcription
factor binding and histone modifications"Methods Mol Biol. ;523:341-66.
Platz EA, Helzlsouer KJ. (2001): "Selenium, zinc, and prostate cancer." Epidemiol Rev, 23:93–
101.
Pogribny, I. P., V. P. Tryndyak, et al. (2007): "Epigenetic effects of the continuous exposure to
peroxisome proliferator WY-14,643 in mouse liver are dependent upon peroxisome
proliferator activated receptor alpha." Mutat Res 625(1-2): 62-71.
Pulling, L. C., Grimes, M. J., Damiani, et al (2009): "Dual promoter regulation of death-
associated protein kinase gene leads to differentially silenced transcripts by methylation
in cancer." Carcinogenesis, 30(12), 2023-30.
Rand K, Q. W., Ho T, Clark SJ, et al (2002): "Conversion-specific detection of DNA
methylation using real-time polymerase chain reaction (ConLight-MSP) to avoid false
positives." methods 27(2): 114-20.
Ropero, S., F. Setien, et al. (2004): "Epigenetic loss of the familial tumor-suppressor gene
exostosin-1 (EXT1) disrupts heparan sulfate synthesis in cancer cells." Hum Mol Genet
13(22): 2753-65.
Rosenwald A, W. G., Chan WC, et al. (2002): "The use of molecular profiling to predict
survival after chemotherapy for diffuse large-B-cell lymphoma." N Engl J Med 346:
1937–47.
Rueffer U, J. A., Franklin J, et al. (2001): "Non-Hodgkin’s lymphoma after primary Hodgkin’s
disease in the German Hodgkin’s Lymphoma Study Group: incidence, treatment, and
prognosis." J Clin Oncol Rep 19: 2026–32.
Siato, Y. and P. A. Jones (2006): "Epigenetic activation of tumor suppressor microRNAs in
human cancer cells." Cell Cycle 5(19): 2220-2.
Sakatani, T., M. Wei, et al. (2001): "Epigenetic heterogeneity at imprinted loci in normal
populations." Biochem Biophys Res Commun 283(5): 1124-30.
101
Sakamoto, L. H. T., DE Camargo, B., Cajaiba, M., et al (2010): "MT1G hypermethylation: a
potential prognostic marker for hepatoblastoma." Pediatric research, 67(4), 387-93.
Sahai, K., Gangadharan, V., Signh, H., et al (2011): "A study of the newer prognostic markers
in diffuse large B cell lymphomas." Medical Journal Armed Forces India, 67(1): 41-45.
Sambrook J, Fritsch EF, Maniatis T. (1989): "Molecular cloning. A laboratory manual." 2nd ed.
Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
Sehn LH, Donaldson J, Chhanabhai M, et al.(2005): "Introduction of combined CHOP plus
rituximab therapy dramatically improved outcome of diffuse large B-cell lymphoma in
British Columbia." J Clin Oncol , 23:5027-33
Shipp, M. a, Ross, K. N., Tamayo, et al (2002): "Diffuse large B-cell lymphoma outcome
prediction by gene-expression profiling and supervised machine learning." Nature
medicine, 8(1): 68-74.
Shiozawa E, Takimot M and Makino R (2006): "Hypermethylation of GPG island in P16 as a
prognostic factor for diffusw large B-cell lymphoma in a high risk group." Leuk Res.
30(7): 859-67.
Strathdee, G. (2007): "Epigenetic markers and response to chemotherapy in cancer." Dis
Markers 23(1-2): 43-9.
Su, S., Gedamu, L., Dehnade, F., et al (1996): 'Testing the eukaryotic promoters for efficient
expression of exogenous genes in chondrocytes and synoviocytes." FEBS Lett 395:
179-82
Swerdlow SH, C. E., Harris NL, et al. (2008): "WHO classification of tumors of hematopoietic
and lymphoid tissues." Lyon: IARC Press.
102
Takao Takahashi1, N. S., 2, Jyotsna Reddy1, Hisayuki Shigematsu1, et al (2004): "DNA
Methylation Profiles of Lymphoid and Hematopoietic Malignancies." clinical caNCER
RESEARCH 10: 2928-2935.
Tost J, Gut IG. (2007): "DNA methylation analysis by pyrosequencing" Nat Protoc. 2 (9):2265-
75.
Toyooka,S. et al. (2003): "Epigenetic down-regulation of death-association protein kinase in
lung cancers." Clin. Cancer Res., 9: 3034–3041.
Traverse-Glehen A, P. S., Gaulard P, et al. (2005): "Mediastinal gray zone lymphoma: the
missing link between classic Hodgkin’s lymphoma and mediastinal large B-cell
lymphoma." Am J Surg Pathobiology 29: 1411-21.
Weber, M., H. Hagege, et al. (2005): "Epigenetic regulation of mammalian imprinted genes:
from primary to functional imprints." Prog Mol Subcell Biol 38: 207-36.
Widschwendter, M., H. Fiegl, et al. (2007): "Epigenetic stem cell signature in cancer." Nat
Genet 39(2): 157-8.
Winter JN, Weller EA, Horning SJ, et al. (2006): "Prognostic significance of Bcl-6 protein
expression in DLBCL treated with CHOP or R-CHOP: a prospective correlative study".
Blood. 107: 4207–4213.
Xu, Y., McKenna, R. W., Molberg, K. H., et al (2001): "Clinicopathologic analysis of CD10+
and CD10- diffuse large B-cell lymphoma. Identification of a high-risk subset with
coexpression of CD10 and bcl-2." American journal of clinical pathology, 116(2): 183-
90.
Yamada, T. (2007): "Reliable safety assessment depends on species differences in epigenetic
mechanisms of gene regulation." Yakugaku Zasshi 127(3): 481-90.
103
Yoo, C. B. and P. A. Jones (2006): "Epigenetic therapy of cancer: past, present and future." Nat
Rev Drug Discov 5(1): 37-50.
Yuanxin Xi and Wei Li. (2009):" whole genome bisulfite sequence MAPping program" BMC
Bioinformatics , 10:232.
Ziepert M, Hasenclever D, Kuhnt E, et al. (2010): "Standard International prognostic index
remains a valid predictor of outcome for patients with aggressive CD20+ B-cell
lymphoma in the rituximab era." J Clin Oncol, 28:2373-80.