current patient management of chronic myeloid leukemia in latin america : a study by the latin...

Current Patient Management of ChronicMyeloid Leukemia in Latin AmericaA Study by the Latin American Leukemia Net (LALNET)

Jorge Cortes, MD1; Carmino De Souza, MD2; Manuel Ayala-Sanchez, MD3; Israel Bendit, MD4; Carlos Best-Aguilera, MD5;

Alicia Enrico, MD6; Nelson Hamerschlak, MD7; Katia Pagnano, MD8; Ricardo Pasquini, MD9; and Luis Meillon, MD10

BACKGROUND: Treatment recommendations have been developed for management of patients with chronic myeloid

leukemia (CML). METHODS: A 30-item multiple-choice questionnaire was administered to 435 hematologists and

oncohematologists in 16 Latin American countries. Physicians self-reported their diagnostic, therapeutic, and disease

management strategies. RESULTS: Imatinib is available as initial therapy to 92% of physicians, and 42% of physicians

have access to both second-generation tyrosine kinase inhibitors. Standard-dose imatinib is the preferred initial ther-

apy for most patients, but 20% would manage a young patient initially with an allogeneic stem cell transplant from a

sibling donor, and 10% would only offer hydroxyurea to an elderly patient. Seventy-two percent of responders per-

form routine cytogenetic analysis for monitoring patients on therapy, and 59% routinely use quantitative polymerase

chain reaction. For patients who fail imatinib therapy, 61% would increase the dose of imatinib before considering

change to a second-generation tyrosine kinase inhibitor, except for patients aged 60 years, for whom a switch to a

second-generation tyrosine kinase inhibitor was the preferred choice. CONCLUSIONS: The answers to this survey

provide insight into the management of patients with CML in Latin America. Some deviations from current recom-

mendations were identified. Understanding the treatment patterns of patients with CML in broad population studies

is important to identify needs and improve patient care. Cancer 2010;116:4991–5000. VC 2010 American Cancer

Society.

KEYWORDS: chronic myeloid leukemia, imatinib, dasatinib, nilotinib, tyrosine kinase inhibitors, guidelines, Latin

America, survey.

Imatinib is the established standard of care for initial treatment of chronic phase chronic myeloid leukemia (CML).1Although most patients have a favorable outcome, some patients are initially refractory or develop acquired resistance.2

Second-generation tyrosine kinase inhibitors provide alternative therapeutic options for CML patients who fail imatinib.With the availability of more effective therapeutic options, adequate use of therapies and proper monitoring have becomeincreasingly important to optimize patient outcomes. The European Leukemia Net has summarized the current under-standing of the management of CML and provided recommendations to guide physicians on how best to treat and moni-tor their patients to help accomplish the best possible outcome for patients with CML.1 These recommendations considerthe optimal management of patients under ideal circumstances that may include wide availability of all therapeutic andmonitoring tools. Recent studies have analyzed how patients in Europe and the United States are treated.3 However, otherareas of the world may have different approaches affected by regional factors such as availability of drugs and tests, culturalelements, and financial considerations. The goal of this study was to ascertain how physicians throughout Latin Americaperceive and use the available CML therapies, and to identify strategies used to diagnose and manage the disease.

DOI: 10.1002/cncr.25273, Received: October 14, 2009; Revised: December 1, 2009; Accepted: January 12, 2010, Published online July 21, 2010 in Wiley Online

Library (wileyonlinelibrary.com)

Corresponding author: Jorge Cortes, MD, Department of Leukemia, The University of Texas M. D. Anderson Cancer Center, 1500 Holcombe Boulevard, Unit 428,

Houston, TX 77030; Fax: (713) 794-4297; [email protected]

1Department of Leukemia, The University of Texas M. D. Anderson Cancer Center, Houston, Texas; 2Hematology and Hemotherapy Center, University of Campi-

nas, Campinas, Sao Paulo, Brazil; 3Department of Hematology, Hospital of Specialties National Medical Center ‘‘La Raza’’, Mexico City, Mexico; 4Department of He-

matology and Hemotherapy, Medical School at University of Sao Paulo, Sao Paulo, Brazil; 5Department of Hematology, Western General Hospital, Office of Health

of Jalisco, Guadalajara, Jalisco, Mexico; 6Division of Bone Marrow Transplantation, Italian Hospital La Plata, La Plata, Argentina; 7Department of Hematology and

Bone Marrow Transplantation, Albert Einstein Hospital, Sao Paulo, Sao Paulo, Brazil; 8Hematology and Hemotherapy Center, University of Campinas, Campinas,

Sao Paulo, Brazil; 9Department of Internal Medicine, Federal University of Parana, Curitiba, Parana, Brazil; 10Department of Hematology, Hospital of Specialties,

National Medical Center ‘‘Siglo XXI’’, Mexico City, Mexico

Cancer November 1, 2010 4991

Original Article

MATERIALS AND METHODSA questionnaire was developed to assess Latin Americanphysicians’ self-reported CML diagnostic, treatment, andmonitoring strategies. Hematologists and oncohematolo-gists were recruited from a large database and invited, byeither phone or e-mail, to participate in the study. Theanonymous and confidential questionnaire was created bythe Latin American Leukemia Net, and applied by anindependent agency through Internet or telephone inter-views. Before beginning, participants were informed thatthe questionnaire would take 20 to 25 minutes tocomplete. The 30-item multiple-choice questionnaire isshown in Figure 1. Before beginning the questionnaire,physicians were surveyed to determine their study eligibil-ity. Participants were considered eligible if they had pre-scribed imatinib for any condition, had treated �5 CMLpatients outside the context of a clinical trial in the past 2years, and were treating at least 2 CML patients at thetime of the survey.

Six-hundred fifty physicians from 16 Latin Ameri-can countries were contacted between May and August2008, and 435 (67%) completed the survey. Respondingphysicians were distributed among countries as follows:Brazil (n¼ 100), Mexico (n¼ 75), Argentina (n¼ 50),Colombia (n¼ 50), Venezuela (n¼ 40), Peru (n¼ 30),Chile (n¼ 20), Panama (n¼ 9), Nicaragua (n¼ 9), ElSalvador (n¼ 8), Costa Rica (n¼ 8), Guatemala (n¼ 8),Honduras (n¼ 8), Bolivia (n¼ 7), Ecuador (n¼ 7), andUruguay (n¼ 6).

RESULTS

Physician and Practice Characteristics

The first section of the questionnaire was aimed at defin-ing the characteristics of the participating physicians andtheir practices. Most respondents characterized their med-ical practice as nonacademic (85%), including combinedmedical settings (35%), state hospitals (25%), private hos-pitals (23%), or other settings (2%). Ninety-three percentof the survey participants described their specialty as he-matology, with the reminder classified as onco-hematol-ogy. This correlates with 82% of respondents describingtheir clinical practice as primarily composed of hemato-logic patients (including patients with both benign andmalignant disease), whereas 10% of the respondents’ clin-ical practice was comprised of patients with solid tumorsand hematologic malignancies, 7% mentioned only treat-ing patients with hematologic malignancies, and theremaining 1% of respondents described their practice as

primarily comprised of patients with solid tumors. At thetime of the survey, most physicians were directly involvedin the management of 2 to 5 CML patients (31%) or 6 to10 CML patients (30%), whereas another 18% of partici-pants managed >20 CML patients. Forty-two percent ofparticipants had >20 years of clinical experience sincecompleting their specialty training, and 56% had �10years.

Availability of Tyrosine Kinase Inhibitors

Among all responders, 85% reported that imatinib wasavailable to them only as front-line therapy for CML intheir specific country, whereas 8% reported it wasapproved only in the setting of interferon (IFN)-a failure,and 7% reported it available both as front-line therapyand after IFN-a failure. At the time of the survey, 79%and 45% of physicians reported approval of the second-generation tyrosine kinase inhibitors dasatinib and niloti-nib in their country, respectively, with 42% reportingapproval of both agents. At least 1 of these second-genera-tion tyrosine kinase inhibitors had received approval in allof the countries represented in this analysis. Overall, 66%of the participants answered that the majority of theirCML patients received state coverage for imatinib ther-apy. Private insurance, organizations or charity, and self-pay accounted for the remaining 18%, 10%, and 5% ofimatinib coverage, respectively.

General Disease Management

Participants were then asked several questions regardingtheir current management of CML. For initial workup ofa patient with suspected CML, physicians reported mostcommonly performing cytogenetic analysis (78%), bonemarrow aspiration (68%), and bone marrow biopsy(53%) as initial diagnostic tests. Quantitative polymerasechain reaction (QPCR) (41%), BCR-ABL mutation anal-ysis (39%), and fluorescent in situ hybridization (FISH)(26%) were performed less frequently (Fig. 2). In mostinstances, conventional cytogenetic or FISH analyses wereperformed at a nearby institution or hospital within theircountry (40%), whereas commercial laboratories and in-house hospital laboratories were used in 26% and 25% ofcases, respectively. A smaller proportion (8%) sent cytoge-netic or FISH samples to a hospital or institution inanother country. Similar results were obtained when therespondents were queried on where they obtained aQPCR or BCR-ABLmutation test.

Physicians were also asked what features they con-sidered representative of advanced phase CML. The

Original Article

4992 Cancer November 1, 2010

Figure 1. Questions and answers in the Internet and phone interviews are shown. CML indicates chronic myeloid leukemia; CP,chronic phase; FISH, fluorescence in situ hybridization; PCR, polymerase chain reaction; qRT-PCR, quantitative reverse transcrip-tase PCR; CyR, cytogenetic response; SCT, stem cell transplantation; WBC, white blood cell count; Ph, Philadelphia chromosome;CHR, complete hematologic response; CCyR, complete CyR; MCyR, major CyR; MMR, major molecular response; QPCR, quantita-tive PCR; IFN, interferon; TKI, tyrosine kinase inhibitor.

criteria most frequently used to define accelerated phasewas ‘‘blasts �10% in peripheral blood and/or bone mar-row,’’ identified by 61% of respondents. Other com-monly identified criteria for accelerated phase were ‘‘blasts�15% in peripheral blood and/or bone marrow’’ (49%)and ‘‘clonal evolution during the course of therapy’’(48%). Blasts �20% in peripheral blood and/or bonemarrow were considered to represent blast phase by 66%of responders, whereas 44% considered �30% blasts todefine this stage. Only 33% identified extramedullary dis-ease as a representative feature of blast phase CML.

Initial Therapeutic Strategy

The next sets of questions in the survey were aimed atgathering information on the preferred diagnostic andtreatment options. For this purpose, several hypotheticalclinical scenarios were presented to the participants.Regardless of the patient’s age or donor status, imatinib400 mg/d was found to be the preferred front-line therapyamong Latin American physicians for newly diagnosed,chronic phase CML patients (Fig. 3). For example, 67%and 75% of respondents scored imatinib (400 mg/d)as their primary choice of therapy for a newly diagnosed

20-year-old patient with either a related or a matchedunrelated donor (MUD) available, respectively. Alloge-neic stem cell transplantation was the second most favoredoption (20%) for the patient with a sibling donor, whereashigh-dose imatinib was the second choice in the patientwith an available MUD. Although 80% of physicianslisted imatinib 400 mg/d as their first choice of therapyfor an 80-year-old patient with newly diagnosed CML,10%would only use hydroxyurea.

Respondents were also presented with a case of apatient with chronic phase CML with a white blood cell(WBC) count of 225� 109/L. In this case, 63% ofrespondents would use hydroxyurea until the WBC countdecreased significantly, and then start imatinib. Only17% would initiate therapy with imatinib, and 3% woulduse imatinib and hydroxyurea simultaneously.

Monitoring During Treatment

To monitor patients with CML treated with imatinib,besides complete blood counts, 72% reported routinelyusing cytogenetic analysis, 59% QPCR, 30% mutationanalysis, and 19% FISH (Fig. 4). Cytogenetic analyses arerepeated every 6 months by 54% of participants, whereas

Figure 2. Initial diagnostic evaluation for patients with a clinical picture compatible with chronic myeloid leukemia (CML) isshown (N¼435). Question 3: When evaluating a patient with a white cell count (eg, 65,000) and a morphologic picture compati-ble with CML, which of the following initial diagnostic tests would you typically perform? PCR indicates polymerase chain reac-tion; FISH, fluorescent in situ hybridization.

Original Article


31% repeat it every 3 months and 9% only annually.Reverse transcriptase QPCR was reported to be per-formed every 6 months by 41% and every 3 months by31%. Thirteen percent of participants reported never

using QPCR. Mutation analysis was used by 33% ofphysicians when a patient lost or failed to achieve a hema-tological response, whereas 26% identified loss or failureto achieve a complete cytogenetic response as the reason

Figure 4. Tests routinely performed to monitor response to imatinib therapy are shown (N¼435). Question 4: In addition toperipheral blood counts, which of the following do you routinely use in monitoring response to imatinib therapy? PCR indicatespolymerase chain reaction; FISH, fluorescent in situ hybridization.

Figure 3. Preferred front-line treatment in different hypothetical case scenarios (N¼435). SCT, stem cell transplantation; CML,chronic myeloid leukemia.

CML Survey in Latin America/Cortes et al


for performing this test. A�2-fold rise in BCR-ABL tran-script levels was identified as a reason to performmutationanalysis by 19% of participants. Study participants werealso given 2 specific case scenarios being managed withimatinib to investigate any differences in monitoringapproach: a 50-year-old patient with a sibling donor and a70-year-old patient. No differences in monitoring strat-egies were identified, with cytogenetic analysis being usedroutinely by 64% and 65%, respectively, and QPCR by55% and 47%.

Management of Suboptimal Response

Physicians were then presented with different clinical sce-narios to determine whether they take into considerationa difference between failure to respond to therapy andsuboptimal response, and the criteria used to classify thesestates. Two thirds (67%) of participants considered thedistinction between failure and suboptimal response to beclinically relevant. Among the physicians who didacknowledge a difference, 37% identified no major cyto-genetic response at 6 months as a criterion to definesuboptimal response. Other criteria frequently chosen todescribe a suboptimal response include no cytogeneticresponse (ie, 100% Philadelphia chromosome positive[Phþ]) at 6 months, no complete cytogenetic response at6 months, and no complete hematologic response (CHR)at 3 months. When presented with a patient having a sub-

optimal response, 55% of participants would assess for aBCR-ABLmutation.

We then assessed the timing that physicians considerrelevant for treatment decisions. Failure to achieve a CHRat 3 months was considered a criterion to change therapyby 45% of respondents, whereas 29% would consider thisonly after 6 months (Fig. 5). Similarly, for patients still100% Phþ at different times, 42% and 24% stated theywould consider changing treatment at 3 and 6 months,respectively. However, for a patient achieving only a par-tial cytogenetic response, 34% would consider changing ifthis was the best response at 6 months, compared with22% if this was at 12 months, and 20% at 3 months.

When asked about the preferred course of action formanaging a 50-year-old patient with a matched siblingdonor and a suboptimal response to imatinib, 72% of par-ticipants would increase the imatinib dose (46% to 600mg/d and 26% to 800 mg/d), whereas 16% would changetherapy to dasatinib and 3% to nilotinib. Among thoseopting for a dose increase, 54% would try this strategy fora minimum of 3 months, 38% for 6 months, and 4% for12 months.

Management of Imatinib Failure

When asked to identify clinical scenarios that definedimatinib treatment failure, loss of complete cytogeneticresponse (39%), no CHR at 3 months (39%), loss ofCHR (39%), no cytogenetic response at 6 months (38%),

Figure 5. Timing for treatment change is shown (N¼435). CML indicates chronic myeloid leukemia.

Original Article


loss of a major molecular response (36%), and no majorcytogenetic response at 6 months (34%) were most fre-quently reported (Table 1). Presented with hypotheticalcases of imatinib failure, age of the patients influenced thechoice of therapy. When evaluating the options for a 50-year-old and a 35-year-old patient, increasing the dose ofimatinib was preferred by 61% (37% to 600 mg/d, 24%to 800 mg/d) for the older patient and 48% for theyounger patient (30% to 600 mg/d and 18% to 800 mg/d). In both cases, second-generation tyrosine kinase inhi-bitors would be the preferred option for only 24% and22%, respectively, whereas autologous stem cell trans-plantation would be offered more frequently to theyounger (12%) than the older (7%) patient. However, inthe case of a patient of more advanced age (60 years old),the preferred course of action was reported to be a switchof the treatment to a second-generation tyrosine kinaseinhibitor, regardless of donor status (82% in an olderpatient with a matched sibling donor, 85% in an olderpatient with anMUD).

Participants were also presented with a hypothetical35-year-old patient with a matched sibling who lostmajor molecular response but remained in completecytogenetic response while on imatinib 400 mg/d. In thissetting, 54% of physicians would perform a mutationanalysis. Of these, 40% would select a new tyrosine ki-nase inhibitor based on mutations detected, and 19%

would increase the dose of imatinib if no mutation wasidentified. Interestingly, in the case of a 50-year-oldpatient failing imatinib treatment, only 27% of respond-ents who would preferably switch treatment to a second-generation tyrosine kinase inhibitor would performBCR-ABL mutations.

Imatinib Toxicity

Imatinib was perceived to be overall well tolerated. Overhalf (58%) of the study participants reported that only1% to 2% of their patients were intolerant to imatiniband would need to change therapy because of intolerance.Conversely, 4% of physicians felt that >10% of theirpatients would be intolerant to imatinib. Nonetheless,imatinib is occasionally associated with significant toxic-ity. The most frequently observed imatinib toxicitiesencountered in the participants’ patients included neutro-penia (75% of responders), thrombocytopenia (74%),fluid retention (67%), nausea/vomiting (66%), anemia(64%), periorbital edema (61%), muscle cramps (57%),fatigue (57%), and weight gain (56%) (Table 2). Themost common cause of treatment interruptions were neu-tropenia, thrombocytopenia, and anemia (in 33%, 30%,and 11%, respectively). Physicians reported permanenttreatment discontinuation for toxicity in 5% to 12% ofparticipants. Physicians ranked the 3 most common toxic-ities that have caused them to permanently discontinue

Table 1. Considering Suboptimal Response or Failure to Imatinib Treatment (N ¼ 435)

Criteria SuboptimalResponse toTreatment

Failure toTreatment

At 6 months, no major cytogenetic response (ie, Phþ >35%) 37% 34%

At 6 months, no cytogenetic response (ie, Ph 100%) 29% 38%

At 6 months, no complete cytogenetic response (ie, Phþ >0%) 26% 26%

At 3 months, no CHR 26% 39%


Loss of CHR 17% 39%

Loss of CCyR 17% 39%


At 12 months, no major cytogenetic response (ie, Phþ >35%) 16% 29%

Loss of MCyR 16% 36%

At 18 months, no major molecular response (ie, <3 log reduction

in BCR-ABL transcript levels)

15% 26%


At 12 months, no major molecular response (ie, <3 log reduction

in BCR-ABL transcript levels)

12% 23%


�1 log rise in QPCR while in CCyR 10% 17%

2-fold rise in QPCR while in CCyR 9% 21%

Ph indicates Philadelphia chromosome; CHR, complete hematologic response; CCyR, complete cytogenetic response;

MCyR, major cytogenetic response; QPCR, quantitative polymerase chain reaction.

Question 17: In a patient with CML receiving therapy with imatinib, which of the following would cause you to decide that

the patient had experienced failure to therapy and should receive an alternative treatment?



imatinib therapy in their patients (Table 2). Neutropeniaand thrombocytopenia were most commonly ranked asthe first and second most frequent toxicities. Other toxic-ities that physicians identified as leading to treatment dis-continuation included fluid retention (67%), nausea andvomiting (66%), and anemia (64%).

Patient case studies were then presented to assessphysicians’ approach to imatinib-associated hematologi-cal toxicity. For a 45-year-old patient who developed aneutrophil count of 1.2� 109/L while receiving imatinib(400 mg/d), participants were roughly equally dividedamong those who would stop therapy until neutrophilcount recovered to �1.5� 109/L and then would restartat a reduced dose (32%), those who would discontinuetherapy until neutrophil count recovered to �1.5 � 109/L and then would restart imatinib at the same dose(21%), and those who would decrease the imatinib dosageto 300 mg/d without treatment interruptions (20%). Fur-thermore, most participants reported considering a treat-ment interruption in patients when the neutrophil countreached either <0.5� 109/L (33%) or <1.0� 109/L(43%). The therapeutic management of imatinib-relatedthrombocytopenia was also evaluated, and most physi-cians reported discontinuing imatinib therapy whenplatelet count reached either <20� 109/L (26%) or<50� 109/L (49%); 22% of them would discontinuetreatment with platelet counts <75� 109/L (8%) and<100� 109/L (14%).

DISCUSSIONImatinib and other second-generation tyrosine kinaseinhibitors for the treatment of CML have greatlyimproved patient outcomes, and have served as importantexamples of the clinical benefit of targeted therapies. Rec-ommendations have been developed for the optimal man-agement of patients with CML. The extent to which theserecommendations are followed in practice is not known.Elements such as economic limitations, educational dif-ferences, and availability of drugs and laboratory tests mayaffect the extent to which these recommendations are fol-lowed. Recently, Kantarjian and colleagues3 surveyed 956eligible physicians from the United States and Europe toassess patterns of practice. In that study, the investigatorsconcluded that although the practice patterns amongthose physicians in general were aligned with currentguidelines and recommendations, important differencesdo exist. Because there are considerable differences inhealthcare access and practice between developing coun-tries and the United States and European countries, weinvestigated the practice patterns of physicians through-out 16 Latin American countries.

Availability of diagnostic and monitoring tools var-ied depending on the type of institution. For example,nonacademic physicians were more likely to send FISHand cytogenetic samples to another institution or hospitalwithin their country, whereas academic physicians weremore likely to use an in-house hospital or institution

Table 2. Most Frequently Observed Imatinib Toxicities (N ¼ 435)

Toxicity No. ofRespondents

ToxicitiesEncountered

CausedInterruptionof Imatinib

CausedDiscontinuationof Imatinib

Neutropenia 328 75% 33% 12%

Thrombocytopenia 320 74% 30% 11%

Fluid retention 293 67% 5% 1%

Nausea and vomiting 288 66% 8% 6%

Anemia 280 64% 11% 7%

Periorbital edema 266 61% 7% 2%

Muscle cramps 247 57% 7% 3%

Fatigue 246 57% 6% 1%

Weight gain 245 56% 4% —

Diarrhea 205 47% 6% 4%

Bone aches 199 46% 5% 3%

Rash 183 42% 12% 9%

Liver dysfunction 140 32% 21% 14%

Pleural effusion 47 11% 15% 13%

Pericardial effusion

or pericarditis

36 8% 17% 6%

Congestive heart failure 33 8% 27% 6%

Question 30: Which, of the following imatinib toxicities have you encountered in your patients with chronic myeloid leuke-

mia? Which of these toxicities that you have encountered have caused you to interrupt, dose reduce, or discontinue ima-

tinib therapy?

Original Article


laboratory. These practices are very similar with thoseobserved in the US/Europe study.3 However, an impor-tant difference revealed in the current study is that nearly10% of Latin American physicians are required to rely oninstitutions based in other countries to perform thesebasic disease management tests.

Some monitoring practices were notable. Amongthem, 39% of responders indicated they would use muta-tional analysis at the time of diagnosis, and 55% wouldtest for BCR-ABL kinase domain mutations when manag-ing a patient with a suboptimal response to imatinib treat-ment. These rates may reflect the intent more than actualpractice, because most physicians indicated that they donot have direct access to mutational analysis at their insti-tution. Interestingly, the US/European study by Kantar-jian and colleagues reported that US respondents ingeneral were not familiar with BCR-ABL mutation tests.3

That survey was done >3 years ago, when the availabilityand understanding of the clinical significance of such testswere at an early stage. With the broader use of second-generation tyrosine kinase inhibitor this has clearlychanged.

Regarding the management of patients, the surveyreflects the change in practice that resulted from the intro-duction of tyrosine kinase inhibitors, with imatinib widelyfavored as initial therapy. Interestingly, 20% of respond-ers would select a stem cell transplant to manage a 20-year-old patient with a matched related sibling. The use oftransplant has decreased significantly in recent years, andnearly all patients throughout the world are offered imati-nib as initial therapy. However, investigators in Mexicohave published on the favorable results with their stem celltransplant approach and have emphasized the potentialcost advantages of a transplant over the long-term use ofimatinib.4 The cost of a standard dose of imatinib in LatinAmerica is similar to that in the United States, althoughthere is great variability based on the variability in accessprograms available in different countries (eg, state cover-age, the Gleevec International Patient Assistance Program[GIPAP]). The study by Ruiz-Arguelles et al reported thatthe median cost of a nonmyeloablative transplant (first100 days) in Mexico was US$18,000, and US$30,000 fora conventional allograft. Subsequent costs are highly vari-able, depending on complications. In contrast, the me-dian cost of standard-dose imatinib in that country wasreported as US$100.4 Thus, the cost of the first 100 daysof transplant would cover 180 days of imatinib. Long-term comparisons of the costs would depend on the com-plications associated with transplant, but it was suggested

that a successful transplant with no or minimal long-termcomplications could have an economical edge. Despitethis potential advantage, there are several reasons whytransplant may not have been considered as initial therapyin more patients. These include the availability of donorsas well as the local availability of transplantation or the ex-perience with this procedure, which may not be as favor-able as those reported in other places. Wide availabilityand coverage of imatinib for all patients in need such asoccurs for most patients in Brazil, Costa Rica, Panama,Uruguay, and Venezuela according to the respondersfrom these countries would likely influence the selectionof this therapy, particularly if it is perceived to be effectiveand nontoxic. Interestingly, 10% of participants wouldtreat an 80-year-old patient only with hydroxyurea.

The approach at changing therapy for patientsreceiving imatinib suggested some impatience in waitingfor an adequate response. Forty-two percent of respondersstated they would consider a change of therapy if there wasno cytogenetic response after 3 months of therapy orno complete cytogenetic response at 6 months. Theseapproaches would be more aggressive than what is recom-mended by the European Leukemia Net.1 Regardingpatients meeting definitions for failure, 48% and 61% ofresponders indicated that their preferred course of actionwould be imatinib dose escalation for patients aged 35years and 50 years, respectively, with <25% deciding toswitch to a second-generation tyrosine kinase inhibitor.The recommendations from the European Leukemia Netpublished in 2006 listed this approach as 1 of the possiblealternatives to consider. However, soon after that theresults with dasatinib and nilotinib after imatinib failurehave established these agents as the treatment of choice forsuch patients, with evidence from a randomized trial sug-gesting that the probabilities of a favorable outcome afterchange to second-generation tyrosine kinase inhibitorwould be superior to that with dose escalation of imati-nib.5 However, second-generation tyrosine kinase inhibi-tors were not yet widely available throughout LatinAmerica at the time of this survey. Among responders,14% reported not having dasatinib, and 44% did not havenilotinib available for their patients. In this case, dose esca-lation is an adequate option, and recent reports suggestthat with this approach complete cytogenetic response canbe achieved in approximately 40% of patients, particularlythose who lost a cytogenetic response to imatinib.6,7 De-spite initial concerns about the durability of response,some of the responses are indeed durable, with event-freesurvival at 2 years of 85%.7 Interestingly, change to a



second-generation tyrosine kinase inhibitor was greatlyfavored (80% of responders) for an older patient (aged 60years) over dose escalation. One possible explanation forthis difference could be a concern about tolerability ofhigher-dose imatinib among older patients. Of note, stemcell transplant was selected as second-line therapy by veryfew physicians, even for the younger patients.

In this study, a high rate of Latin American physi-cians (93%) reported that they preferred to conduct fre-quent visits with their patients to monitor for imatinib-associated toxicities. This rate was similar to the 90% ofUS physicians and 97% of European physicians whoreported a similar practice.3 In this study, neutropeniaand thrombocytopenia were some of the most commontoxicities triggering treatment discontinuation for LatinAmerican physicians. Comparatively, a smaller propor-tion of US/European physicians rated neutropenia (6%and 11% of US and European respondents) or thrombo-cytopenia (8% and 10% of US and European respond-ents) as leading to treatment discontinuation. In contrast,congestive heart failure, pericardial effusion or pericardi-tis, pleural effusions, and liver dysfunction were the mostfrequently cited toxicities leading to treatment discontinu-ation for US/European physicians.

We conclude that the management of patients withCML frequently deviates from the recommendationspublished in the literature. The causes of these deviationsare variable and should be investigated. Regional econom-ical, cultural, and other factors should be considered andintegrated into guidelines that may be applicable todifferent areas of the world, with the aim of improving theoutcome of all patients with CML.

CONFLICT OF INTEREST DISCLOSURESThis work was supported in part by an unrestricted grant fromNovartis Pharmaceuticals. J.C. received research support fromNovartis, BMS, and Wyeth; M.A. received research supportfrom BMS; C.B. received research support from Novartis; K.P.received research support from Novartis; R.P. received honorariafrom Novartis, Bristol Myers, and Schering.

REFERENCES1. Baccarani M, Saglio G, Goldman J, et al. Evolving concepts

in the management of chronic myeloid leukemia: recommen-dations from an expert panel on behalf of the European Leu-kemiaNet. Blood. 2006;108:1809-1820.

2. Deininger M, Buchdunger E, Druker BJ.The development ofimatinib as a therapeutic agent for chronic myeloid leukemia.Blood. 2005;105:2640-2653.

3. Kantarjian HM, Cortes J, Guilhot F, Hochhaus A, BaccaraniM, Lokey L.Diagnosis and management of chronic myeloidleukemia: a survey of American and European practice pat-terns. Cancer. 2007;109:1365-1375.

4. Ruiz-Arguelles GJ, Tarin-Arzaga LC, Gonzalez-Carrillo ML,et al. Therapeutic choices in patients with Ph1 (þ) chronicmyelogenous leukemia living in Mexico in the tyrosine kinaseinhibitors (TKI) era: stem cell transplantation or TKI’s? BoneMarrow Transplant. 2008;42:23-28.

5. Kantarjian H, Pasquini R, Levy V, et al. Dasatinib or high-dose imatinib for chronic-phase chronic myeloid leukemia re-sistant to imatinib at a dose of 400 to 600 milligrams daily:two-year follow-up of a randomized phase 2 study (START-R). Cancer. 2009;115:4136-4147.

6. Kantarjian HM, O’Brien S, Cortes JE, et al. Complete cyto-genetic and molecular responses to interferon-alpha-basedtherapy for chronic myelogenous leukemia are associated withexcellent long-term prognosis. Cancer. 2003;97:1033-1041.

7. Jabbour E, Kantarjian HM, Jones D, et al. Imatinib mesylatedose escalation is associated with durable responses in patientswith chronic myeloid leukemia after cytogenetic failureon standard-dose imatinib therapy. Blood. 2009;113:2154-2160.

Original Article


current patient management of chronic myeloid leukemia in latin america : a study by the latin...

Documents