5/17/2013

1

Myelodysplastic Syndrome: a pathologist’s perspective

Kajal Sitwala, MD, PhD

MYELODYSPLASTIC SYNDROMES: Part 1 – Example MDS cases coming into our practice over the last several weeks Part 2 – Review of MDS definition, features, morphology, and biology

Case 1: -64 y.o. man with longstanding low back pain (unrelated) -MRI showed abnormal bone marrow signal referred to Hematology -Had normocytic anemia dating back 15 years -Bone marrow biopsy showed fairly normal overall cellularity, but relative expansion in erythropoiesis. Iron stain showed many ring sideroblasts -Diagnosis of RARS (low-grade MDS) -Refractory Anemia with Ring Sideroblasts -Interval treatment supportive (erythropoietin to boost RBC production) -Follow-up bone marrow

Dyserythropoiesis in aspirate smear

Ring sideroblasts (special stain of aspirate smear) Hypercellularity; odd megakaryocytes, but not typical for MDS

5/17/2013

2

Diagnosis:

Persistent RARS -No increase in blasts -Frank dysplasia still limited to the erythroid lineage -Cytogenetic analysis showed same abnormality as before (+8) -Trisomy 8 can be seen in myeloid disorders including MDS and AML -For RARS, it’s worse prognosis to see it (versus normal karyotype) -Flow cytometry: main contribution is confirming no increase in blasts -Mild aberrancy of CD56 co-expression on maturing granulocytes and monocytes -(supportive finding, not enough to definitively diagnose MDS) -Recently, MDS-RARS shown to have strong association with gene mutation: -Haploinsufficiency of SF3B1

Case 2: -62 y.o. man with anemia and thrombocytopenia -Bone marrow showed RAEB-1 (high-grade MDS) -Refractory Anemia with Excess Blasts -Treatment/management course complicated by cold agglutinin disease and transfusion refractoriness -Recently, transfusion requirements became too severe to manage supportively, hospitalized for aggressive chemotherapy -Bone marrow performed to assess response

Peripheral blood: Granulocyte with mature, clumped chromatin but lack of granulation or nuclear lobation (Pseudo-Pelger-Huet)

Increased blasts in aspirate smear

Hypercellular but loose marrow (fibrosis), Dysplastic megakaryocytes

Diagnosis:

Persistent MDS, now best classified as RAEB-2 -In subsequent bone marrows, I only reclassify MDS if worse -improvements are described, but disease isn’t “downgraded” as a new MDS subtype -14% blasts in differential count of aspirate smear -Reticulin stain confirmed the presence of fibrosis -Historically normal karyotype in leukemia cells; not repeated with this specimen -Flow cytometry with 15% blasts, similar phenotype as prior cells -Patient given even more aggressive treatment (AML induction), however, disease is persisting

5/17/2013

3

Case 3: -83 y.o. woman with history of MDS dating back to 2008 -When seen here in 2010, classified as RCMD -Refractory Cytopenia with Multilineage Dysplasia -2010 bone marrow with similar cytogenetic abnormalities, still low-grade (no increase in blasts -Since then, CBC counts have held pretty steady with lenolidamide treatment -uniquely efficacious in cases involving chromosomal deletions on 5q -Some concern with low hemoglobin values, so bone marrow reassessed

Basophilic stippling in peripheral blood

Dyserythropoiesis in aspirate smear Small and hypolobated megakaryocytes in aspirate smear

Ring sideroblasts in aspirate smear

Diagnosis:

Persistent RCMD -Cytogenetics unchanged over past 5 years -46,XX,del(5)(q13q33),add(11)(q23)[19]/46,XX[1] -doesn’t fit for isolated 5q deletion (‘’5q-minus syndrome’’) but some features overlap -often, dramatic worsening (e.g. increase in blasts) corresponds with clonal evolution -Flow cytometry: increased basophils, light scatter changes (but no aberrant phenotype) -No increase in blasts -Later: discuss role of lenolidamide in 5q deleted cases

5/17/2013

4

A group of clonal hematopoietic stem cell

diseases characterized by cytopenias,

dysplasia, ineffective hematopoiesis, and

increased risk of developing acute

myeloid leukemia.

-Principally a disease of older adults -Disorder of HSCs and their microenvironment

MDS 1953 – Knudson hypothesis: cancer results from accumulated DNA mutations 1990’s – 2-hit model of leukemia

Adapted from WHO 2008

Class I mutations Class II mutations

FLT3-ITD FLT3-TKD

JAK2-V617F KIT mutations RAS mutations

PML-RARA AML1-ETO

CBF -MYH11 CEBPA mutations NPM1 mutations?

Proliferation advantage Survival advantage

Differentiation arrest clonogenicity

AML

General (simplified) model of myeloid malignancies

Proliferation advantage Survival advantage

Differentiation arrest clonogenicity

AML

Myeloproliferative disorders Myelodysplastic syndromes

CLASS I CLASS II

So MDS is a stem cell neoplasm: but what does “immortalization” look like? Clonogenic, but not rapid, proliferation of stem cells (There is also increased apoptosis) Inability to complete differentiation and leave marrow Hallmark of MDS – peripheral blood cytopenias PLUS bone marrow hypercellularity In contrast to MPDs… Rapid proliferation but complete differentiation – cells accumulate in both places (clinical presentation from tumor burden rather than loss of function) Or acute leukemia… Proliferation and lack of differentiation: blasts in bone marrow and blood

Programs of self-renewal and differentiation use some of the same molecules (transcription factors)

-Clonogenicity: (Can’t see whether a cell will keep dividing or not)

-Hypercellularity

-Maturation arrest: (Marrow is full of precursors anyhow)

-Skew toward immaturity As Hematopathologists, what we SEE is morphologic abnormalities that result From these molecular processes

Normal Erythroid Maturation (ASH image Bank)

5/17/2013

5

Normal Megakaryocyte making platelets (ASH image Bank)

Pictures of normal hematopoietic elements

A blood smear and a marrow, or just a single marrow picture

Normal Leukocytes in peripheral blood (ASH image Bank)

Predictions about dysplasia, based on concept of “maturation arrest” – 1. immaturity of any kind a. cells that are more like a precursor form b. failure to develop characteristics of final state 2. dyssynchrony between different elements/split personality as far as further cell division 3. just plain weird

1a. Immaturity

Excess blasts Erythroid immaturity (increase in early forms) Small megakaryocytes that have not divided the nucleus yet

Patient with RCMD (including 5q minus but with additional abnormalities) Patient with RCMD (including 5q minus but with additional abnormalities)

5/17/2013

6

Patient with RCMD (including 5q minus but with additional abnormalities) Patient with RCMD (including 5q minus but with additional abnormalities)

Patient with RAEB-1)

1b. Failure to finalize -lack of granulation in circulating neutrophils -lack of nuclear lobation in circulating neutrophils *mimic of harmless genetic state (Pelger-Huet anomaly)

Pseudo-Pelger Huet neutrophils with hypogranulation (ASH image Bank)

2. Dyssynchrony, or confusion about further division -Hemoglobinized cytoplasm with still immature nucleus -Nuclear budding in erythroid precursors -Megakaryocytes with completely separated nuclear lobes

5/17/2013

7

Patient with RAEB-1 Dysplastic erythroid precursor with hemoglobinization and large nucleus (ASH image Bank)

Dysplastic megakaryocytes with separation of nuclear lobes (ASH image Bank) Nuclear budding in erythroid precursor (ASH image Bank)

Patient with RCMD (core biopsy touch imprints)

3. Just plain weird -Ring sideroblasts -Megaloblastoid chromatin in erythroid precursors -Basophilic stippling in red blood cells -Dimorphic circulating red cell population -Vacuoles (especially erythroid precursors) *beware of MDS mimic – copper deficiency (sometimes caused by zinc toxicity)

5/17/2013

8

Patient with RCMD (including 5q minus but with additional abnormalities) Megaloblastoid chromatin (ASH image Bank)

Ring sideroblasts (ASH image Bank)

That “classification” was purely speculative, a tool to mentally account for dysplasia Are there any cases where we have pinpointed the connection between genes and dysplasia?

5q minus syndrome

Anemia – can be quite severe Normal to elevated platelet count; characteristic megakaryocytes Hypercellular marrow, variable erythroid dysplasia Female predominance – middle to older age

5q minus syndrome

1974 – Nature – clinical syndrome with chr 5 long arm deletion reported 2001 – WHO classification recognizes 5q- as distinct subtype of MDS 2002 – Blood – commonly deleted region narrowed to 40 genes by FISH/Southern 2007 – ASH plenary abstract – identification of candidate gene no biallelic deletions or point mutations in 40 genes – likely haploinsufficiency therefore reduction in gene expression could be exploited: RNA interference findings published: Ebert et al., Nature, January 2008 (Golub laboratory – Harvard)

5/17/2013

9

Ebert et al., Nature, January 2008

Megakaryocyte:erythroid ratio using markers CD41, glycophorin A

RPS14 participates in ribosomal synthesis RPS19 mutations found in 25% Diamond-Blackfan Anemia cases

Haploinsufficiency of ribosomal proteins: Phenotype of macrocytic anemia shows p53 dependence -Cell cycle arrest, failure to complete erythroid differentiation -Interestingly, 5q minus syndrome differs from other MDS by exhibiting erythroid hypoplasia Studies in mice prove the genotype:phenotype correlation for anemia Further research showed increased RPS14 levels after lenolidamide therapy But, doesn’t answer the whole question…megs are normal in those mice Also, DBA patients (RPS19) don’t have platelet or megakaryocyte problems

From Identification of miR-145 and miR-146a as mediators of the 5q– syndrome phenotype; Daniel T Starczynowski, et al Nature Medicine 16, 49–58 (2010) (obtained with permission)

microRNA: from non-coding parts of genome -regulate mRNA (coding parts of genome) However, another gene on 5q-, in fact the genes for two microRNAs, have

recently been shown to recapitulate the phenotype of small megs….

From Identification of miR-145 and miR-146a as mediators of the 5q– syndrome phenotype; Daniel T Starczynowski, et al Nature Medicine 16, 49–58 (2010) (obtained with permission)

Targets of these microRNAs are transcription factors (TiRAP, TRAF6), and these in turn Regulate level of interleukin-6 Increased IL-6 can recapitulate megakaryocyte phenotype Lenolidamide* reduces IL-6 levels But wait…didn’t they think lenolidamide worked by increasing RPS14 expression? But wait…there’s also the SPARC gene on 5q-, necessary for mouse hematopoiesis, lenolidamide increases its expression as well! *Thalidomide analog with many anti-cancer mechanisms in vitro and in vivo Genotype:phenotype correlations are real, but messy and redundant Hard to untangle, because of complexity of hematologic pathway, evolved in parallel to general embryogenic cascade

The genes used to regulate hematopoiesis are the same set of genes that determine body patterning during embryogenesis!

Fro

m W

ikip

edia

5/17/2013

10

The genes used to regulate hematopoiesis are the same set of genes that determine body patterning during embryogenesis! Human genes: -Used once to make our body (in utero), then over and over for the rest of our lives to make blood cells -HOXA9 is most upregulated gene in AML -Potent mediator of leukemia

cluster chromosome genes

HOXA chromosome 7 HOXA1, HOXA2, HOXA3, HOXA4, HOXA5, HOXA6, HOXA7, HOXA9, HOXA10, HOXA11, HOXA13

HOXB chromosome 17 HOXB1, HOXB2, HOXB3, HOXB4, HOXB5, HOXB6, HOXB7, HOXB8, HOXB9, HOXB13

HOXC chromosome 12 HOXC4, HOXC5, HOXC6, HOXC8, HOXC9, HOXC10, HOXC11, HOXC12, HOXC13

HOXD chromosome 2 HOXD1, HOXD3, HOXD4, HOXD8, HOXD9, HOXD10, HOXD11, HOXD12, HOXD13

Control

Hoxa9

Meis1

0

2000

4000

6000

8000

10000

12000

0 2 4 6 8day

ce

lls x

10

00

0

- 4-OHT

+ 4-OHT+ -

>-8-fold no change >8-fold

Slc18a1 Sox4

Auts2 Flt3

Dnajc10 A930001M12

Cd34 Pctk2 Amot Tgm3 Cradd

Rora Foxp1

B4galt4 Man1a

Lmo2 5830405N20

Pcnx Pdcd4 Pde7a

Fut8 Tox

Tcf4 Fndc3b

4931406I20 Map4k5

Bmp2k 9330182L06

Aff3 Camk2d

Il10rb 1110028C15

Pdk1 Usp12

Csf2ra Lhfpl2

Myo1e Ier3

4631426J05 Per2

Vcl Ctsc

Cpne2 Ifngr1 Dach1

Dfna5h Bpil2

Klf5 Crem

2900024C23 A130090K04

Hist1h1c Niban

Ebi2 Plxnd1 P2ry1 Trps1 Bcar3 Msr1

Frmd4b Ccl4

Osbpl3 Mrvi1 Gpr84 Ch25h

Rgs1 Gca

Mgll Zfp36l1

Hgf Id2

Thbs1

Cd34 Flt3

Gpr56

Csf2ra B2Galt6

Tgfbr3 Tlr4

72 96 120 h 72 96 120 h

Me

an+/

- 2

SD C

on

serv

atio

n S

core

Peak-widths from center of peak 0 1 2 3 4 5 6

Inpp5a gene exon

Meis1

Hoxa9

Vertebrate conservation

Hox Loci

MLL Polycomb

TALE factors

Common developmental genes (direct targets)

Body patterning/ Hematopoiesis/ Embryonic development Leukemogenesis

Context-specific downstream targets

WHO 2008 Myelodysplastic Syndromes:

Refractory cytopenia with unilineage dysplasia: Refractory anemia Refractory neutropenia Refractory thrombocytopenia Refractory anemia with ring sideroblasts

Refractory anemia with multilineage dysplasia

Refractory anemia with excess blasts

Myelodysplastic syndrome associated with isolated del(5q)

Myelodysplastic syndrome, unclassifiable

Childhood myelodysplastic syndrome provisional category: Refractory cytopenia of childhood

In the separate category of AML and Related Precursor Neoplasms:

Therapy-related myeloid neoplasms (includes cases meeting morphologic MDS criteria)

Transformation from MDS: AML with myelodysplasia-related changes

Also a separate category: Myelodysplastic/Myeloproliferative Neoplasms

Summary – A pathologist’s perspective on MDS -WHO classification is our guide to approaching the diagnosis -Threshold is key -CBC information is the predictor -This is how patients come to attention of hematology, then to us -Morphology and cytogenetics make up our toolset -Flow cytometry can sometimes contribute -FISH plays a more minor role than conventional karyotype -New studies support possible role for point mutations -The biology guides our understanding, and contextualizes morphology -And genetics holds (some of) the answers