A JH CME Information: Acute myeloid leukemia: 2014 update on risk-stratification and management Author: Elihu Estey M.D. CME Editor: Ayalew Tefferi M.D.

If you wish to receive credit for this activity, please refer to the website: www.wileyhealthlearning.com

䊏 Accreditation and Designation Statement: Blackwell Futura Media Services is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. Blackwell Futura Media Services designates this journal-based CME for a maximum of 1 AMA PRA Category 1 CreditTM. Physicians should only claim credit commensurate with the extent of their participation in the activity.

䊏 Educational Objectives Upon completion of this educational activity, participants will be better able to: 1. recognize the need to integrate post- and pre-treatment prognostic information 2. identify how such risk stratification informs treatment decisions in AML, particularly the choice between standard and investigational therapy and the option of allogeneic hematopoietic cell transplantation

䊏 Activity Disclosures

No commercial support has been accepted related to the development or publication of this activity. Author: Elihu Estey, M.D. has no relevant financial relationships to disclose. CME Editor: Ayalew Tefferi, M.D. has no relevant financial relationships to disclose. This activity underwent peer review in line with the standards of editorial integrity and publication ethics maintained by American Journal of Hematology. The peer reviewers have no conflicts of interest to disclose. The peer review process for American Journal of Hematology is single blinded. As such, the identities of the reviewers are not disclosed in line with the standard accepted practices of medical journal peer review. Conflicts of interest have been identified and resolved in accordance with Blackwell Futura Media Services’s Policy on Activity Disclosure and Conflict of Interest. The primary resolution method used was peer review and review by a non-conflicted expert.

䊏 Instructions on Receiving Credit This activity is intended for physicians. For information on applicability and acceptance of continuing medical education credit for this activity, please consult your professional licensing board. This activity is designed to be completed within one hour; physicians should claim only those credits that reflect the time actually spent in the activity. To successfully earn credit, participants must complete the activity during the valid credit period, which is up to two years from initial publication. Additionally, up to 3 attempts and a score of 70% or better is needed to pass the post test. Follow these steps to earn credit:       

Log on to www.wileyhealthlearning.com Read the target audience, educational objectives, and activity disclosures. Read the activity contents in print or online format. Reflect on the activity contents. Access the CME Exam, and choose the best answer to each question. Complete the required evaluation component of the activity. Claim your Certificate.

This activity will be available for CME credit for twelve months following its launch date. At that time, it will be reviewed and potentially updated and extended for an additional twelve months.

C 2014 Wiley Periodicals, Inc. V

doi:10.1002/ajh.00040

American Journal of Hematology, Vol. 89, No. 11, November 2014

1063

ANNUAL CLINICAL UPDATES IN HEMATOLOGICAL MALIGNANCIES

AJH Educational Material

A JH

Acute myeloid leukemia: 2014 Update on risk-stratification and management Elihu H. Estey* Overview: Evidence suggest that even patients aged 70 or above benefit from specific AML therapy. The fundamental decision in AML then becomes whether to recommend standard or investigational treatment. This decision must rest on the likely outcome of standard treatment. Hence we review factors that predict treatment related mortality and resistance to therapy, the latter the principal cause of failure even in patients aged 70 or above. We emphasize the limitations of prediction of resistance based only on pretreatment factors and stress the need to incorporate post-treatment factors, for example indicators of minimal residual disease. We review various newer therapeutic options and considerations that underlie the decision to recommend allogeneic hematopoietic cell transplant. C 2014 Wiley Periodicals, Inc. Am. J. Hematol. 89:1064–1081, 2014. V

䊏 Diagnosis Acute myeloid leukemia (AML) results from accumulation of abnormal blasts in the marrow. These cells interfere with normal hematopoiesis, thus contributing to the bone marrow failure that is the most common underlying cause of death. AML blasts can escape into the peripheral blood, and infiltrate organs, most ominously CNS and lung. Diagnosis rests on demonstration that the marrow or blood has >20% blasts of myeloid lineage. Blast lineage is assessed by multiparameter flow cytometry, with CD33 and CD13 being surface markers typically expressed by myeloid blasts.

䊏 Disease Overview It is likely that many different mutations, epigenetic aberrations, or downstream abnormalities can produce the same clinical picture. However these differences are responsible for the very variable response to therapy, which is AMLs principal feature. The fundamental therapeutic decision in AML is between standard or investigational therapy, the latter preferably in the context of a clinical trial. Standard therapy has generally been considered as “3 1 7,” most often 3 days of idarubicin (12 mg/m2 daily days 1–3) or daunorubicin (60 mg/m2 daily days 1–3) and 7 days of cytarabine (100–200 mg/m2 daily days 1–7) followed by several similar courses should CR be achieved. Since the results of a trial are, by definition, unknown, the decision between standard and the investigational approaches rests on the likely success rate with the former; the lower the chance of success the greater the impetus for investigational treatment. Hence before updating newer therapies, this article will update attempts to better refine prognoses following standard therapy, noting that often these attempts also result in identification of new therapeutic targets.

䊏 Risk Stratification Although a decision to recommend specific anti-AML treatment is usually axiomatic in academic centers, only 1/3 of American patients aged 65 receive such therapy, with the median age of AML patients begin 65–70 [1]. A trial randomizing older patients between purely palliative therapy and specific AML therapy is unlikely to be conducted. However, data in 864 patients aged 70–79 from the very complete Swedish Acute Leukemia Registry indicates that patients had lower early death rates (by day 30) and lived longer beyond 30 days if they resided in regions of the country where a higher proportion of patients received 3 1 7 rather than strictly palliative approaches [2]. There were no regional differences in performance status and 3 1 7 was superior regardless of performance status. Nonetheless, the effects of treatment (3 1 7 vs. palliative) were potentially confounded with the effects of other unknown/unreported covariates. Even in regions where 75% of patients received 3 1 7 only approximately 20% of patients were alive at 2 years. No account was made of “quality of life.” Still, the results suggest that a predisposition to reject specific AML treatment in 70- to 79-year olds is not appropriate. More recently, Wetzler et al. noted the efficacy of 3 1 7 in some patients aged 80–89 (median 82) [3]. Although obviously a minority, many of the patients who will plausibly benefit can be identified by readily available molecular testing.

Treatment-related morality (TRM) versus resistance to treatment as causes of failure Therapeutic failure in newly diagnosed AML results either from treatment-related mortality (TRM) or “resistance”. TRM refers to death of the patient before a remission can be achieved or to death in remission. Resistance denotes relapse from remission or failure to achieve remission despite surviving the time required to do so. Because this time may vary according to the biologic characteristics of the patient’s AML, there will Division of Hematology, University of Washington and Member, Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington

Conflict of interest: Nothing to report. *Correspondence to: Elihu H. Estey, Division of Hematology, University of Washington and Member, Clinical Research Division, Fred Hutchinson Cancer Research Center, Seattle, Washington. E-mail: [email protected] Received for publication: 14 August 2014 Am. J. Hematol. 89:1064–1081, 2014. Published online: in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/ajh.23834 C 2014 Wiley Periodicals, Inc. V

1064

American Journal of Hematology, Vol. 89, No. 11, November 2014

doi:10.1002/ajh.23834

ANNUAL CLINICAL UPDATES IN HEMATOLOGICAL MALIGNANCIES

AML: 2014 Update on Risk-Stratification and Management

TABLE I. Author’s Modification of ELN Prognostic System [9] Prognostic group

Subsets

CR rate; relapse w/3 1 7 w/o HCT

Comments

“Favorable”

Inv (16) or t(16;16);t(8;21) NK and NPM11/FLT3 ITD-;NK and CEBPA1/1

>80–90%; 35–40%

Intermediate-1

NK and NPM-/FLT3 ITD- or CEBPA1/-

50–80%; 50–60%

Intermediate-2

Cytogenetic abnormalities not in “favorable” or adverse groups; FLT3 ITD1b 25,-7,5q-,abn 3q,17p,11q (other than 9;11), t(6;9), complex; Insufficient metaphases for analysis

40–60%; 70–80%; 50–80%; 70–80%

Outcomes worse in older patients, with secondary AML, CKIT mutationsa, and poor response to initial therapy (e.g. MRD) Outcomes worse in older patients, with secondary AML, and poor initial response (e.g. MRD) Outcomes worse in older patients, with secondary AML, and poor initial response (e.g. MRD) Outcomes worse in older patients, with secondary AML, poor initial response (e.g. MRD)

“Adverse”

90%

a

With “allelic ratio” >25% (Ref. 12). Relapse rate previously thought worse with higher allelic ratio; now this is in doubt (Refs. 12 and 13). NK, normal karyotype; 1, abnormality present; 2, abnormality absent (“wild-type”); 1/1, double mutation; allelic ratio, proportion of abnormal alleles.

b

necessarily be overlap between TRM and resistance. Nonetheless 28 to 30 days is a frequently-used criterion for TRM in patients given 3 1 7 and is supported by the observation that the weekly death rate falls sharply once 4 weeks have elapsed from start of such therapy suggesting that patients who die in this time frame are qualitatively different from those who do not [4]. Similarly suggestive are data that different factors are associated with TRM and resistance [4]. Distinguishing TRM from resistance seems useful because different strategies might be employed in patients at high risk of TRM vs. those at high risk of resistance. Experience suggests that patients, perhaps because of the connotations associated with “chemotherapy”, are often more concerned with the possibility of TRM than of resistance. However even in patients in their 70s and 80s resistance is often the greater problem. In patients aged 75 the rates of death within the first 30 days of therapy are very similar to those of not entering remission despite surviving these 30 days [5], while, even in patients aged 70 with performance status 2 to 4 at time of CR, the risk of relapse is threefold that of death in CR [6], the distinction between TRM and resistance perhaps being more objective in patients in remission than in patients undergoing induction. Furthermore, it is widely accepted that the major advance in management of AML in the past 30 years has been the advent of better supportive care, particularly more effective anti-bacterial and anti-fungal agents. This can be quantified by decreases in TRM after induction therapy with 3 1 7 or anthracycline 1 higher doses of cytarabine from 18% in SWOG and 16% at MD Anderson (MDA) in 1991 to 1995 to 3% at SWOG and 4% at MDA in 2006 to 2009 [7]. This decrease was independent of the effect of known covariates such as performance status, albumin, creatinine, de novo vs. secondary AML. It is probably also unrelated to selection bias such that older patients who would have received anthracyline 1 cytarabine in the earlier period often received less “intense” therapies such as azacitidine or decitabine in the more recent period and hence would not have been included in the analysis. While it is very plausible that TRM would assume more significance if a greater proportion of patients received intense therapies, the data noted above suggest that newer therapies must be more effective not just less toxic.

Factors associated with TRM Although there may be a tendency to view TRM primarily in terms of age, performance status is more important than age in predicting TRM [4]. Other factors noted in the preceding paragraph contribute as well leading to the development of multivariate models

doi:10.1002/ajh.23834

predicting TRM [4]. Removing age from these models has only minimal effects on their predictive ability [4], suggesting age is a surrogate for other covariates. Thus, the influence of various co-morbidities associated with aging (such as those comprising the hematopoietic cell transplantation co-morbidity index as described below) on TRM is under investigation. A model predicting TRM is available online [8], although the predictions likely are overestimates given the recent falls in TRM noted above [7].

Factors associated with resistance: The European Leukemia Net (ELN) system and the prognostic limitations of its principal components Because it is more complex than TRM, the benefit of simultaneously examining several variables is even more important in assessing the risk of resistance following standard therapy. For example, the ELN system combines information about cytogenetics and the mutational status of the NPM1, FLT3, and CEBPA genes to define four groups with very different prognoses [9]. The “favorable” group includes patients with either inv 16, t (8; 21), mutant NPM1 without FLT3 internal tandem duplications (ITD) (NPM1/FLT3ITD-), or mutant CEBPA. An “adverse” group consists of patients with either 25, 27, 5q-, abnormalities of 3q, 17p, 11q (except t9;11), t(6;9), or 3 cytogenetic abnormalities not including translocations (complex karyotype). An intermediate 1 (int-1) group comprises patients with a normal karyotype (NK) and with the other genotypic combinations of NPM1 and FLT3 ITD (1/1, 2/2, 2/1) and an intermediate 2 (int-2) group consists of patients with t(9;11) and cytogenetic abnormalities not noted above. Likelihood of resistance, particularly as manifested by relapse, increases from best to int-1 to int-2 to worst. Recent data suggest that the approximately 10% of patients in whom cytogenetic analysis is unsuccessful belong in the adverse group [10] and only patients with double CEBPA mutations in the “favorable best group,” with single CEBPA mutations considered as int-1 [11]; some would place patients who are FLT3 ITD1 particularly without an NPM1 mutation in the int-2 group, as shown in an author’s modification of the ELN (Table I). The negative effect of FLT3 ITD is usually mediated through an increase relapse rate rather than a decrease in CR rate. In general the risk of relapse has been considered sufficiently high to justify allogeneic hematopoietic cell transplant (HCT) in the adverse and int-2 groups (relapse rates within 1–2 years without HCT >90% and 70– 80% respectively with FLT3 considered as int-2) and probably in the int-1 group (relapse rate without HCT 50–60%) but not in the favorable group (relapse rate without HCT 30–40%) [15].

American Journal of Hematology, Vol. 89, No. 11, November 2014

1065

ANNUAL CLINICAL UPDATES IN HEMATOLOGICAL MALIGNANCIES

Estey TABLE II. Prognostic Index for Survival Based on 2,483 Patients Aged  60 Treated in MRC Trials as Reported by Wheatley et al. in Ref. 19; see Fig. 1 for Relation with Survival Parameter Cytogenetic group WBC count Performance status Age AML type Total score

Score 15 not adverse, 5 5 adverse, 25 unknown 1  10, 2 5 10–49.9, 3 5 50–99.9, 4  100 0, 1, 2, 3, 4 1 5 60–64, 2 5 65–69, 35 70–74, 4 75 1 5 de novo, 2 5 secondary 4–6 5 “good”, 7–8 5 standard, 91 5 poor

Most of the data concerning the effect of NPM and FLT mutations has been derived from younger patients with NK de novo AML. Older patients are more likely to exhibit resistance, even after accounting for the association of older age with worst prognosis cytogenetics and/or an antecedent—hematologic disorder or therapy-related AML (“secondary AML”) and have a narrower range of outcomes; for example only about 20–25% of patients aged >65 with inv(16) or t(8;21), generally the most favorable cytogenetic findings, will be alive 2 to 3 years after diagnosis, largely due to relapse [16]. The more circumscribed outcomes in older patients motivated the MRC to examine the prognostic relevance of NPM1 and FLT3 ITD in 820 older patients (median age 68) treated intensively and 492 (median age 75) treated non-intensively [17]. NPM1, but not FLT3 ITD, was less common with increasing age. Thirteen percent of patients had the “favorable” NPM1/FLT3 ITD- genotype. In the intensively treated group, NPM1/FLT3—patients relapsed more slowly than patients with the other three genotypes leading to better survival, which was statistically significant after adjusting for age, secondary disease, cytogenetics, WBC, and performance status and similar to the trend seen in non-intensively treated patients. However 3-year survival rates ranged from only 15 to 26% in the former to 0 to 13% in the latter group. Information about NPM1 and FLT3 ITD resulted in a change in prognostic group (“favorable,” intermediate, unfavorable) in 11% of patients from status based only on the covariates noted immediately above in this paragraph [17]. SWOG found no effect of NPM1/ FLT32 status on CR, relapse or survival rates in the 25% of 58 patients aged >65 with this genotype, largely because their outcomes were poor enough to be indistinguishable from those in similarly aged patients with the other three genotypes [18]. Under these circumstances and given the somewhat limited amount of predictive ability provided by NPM and FLT3 (see below) the author also recommends the classification system developed by Wheatley et al. based on 2,483 patients aged 60 with newly-diagnosed AML given standard treatment for decisions regarding standard vs. investigational therapy [19] (Table II, Fig. 1). The ability to predict outcome, in particular resistance, is fundamental to decision making in AML. A recent analysis of 4,601 patients treated on SWOG, MDA, HOVON, or MRC protocols examined the ability to predict resistance using multivariate models including (a) cytogenetics, (b) de novo vs. secondary AML (c) age and (d) FLT3 ITD and NPM1 status, with an NPM1 mutation occurring in up to 50% of patients aged 50% of alleles affected) may not affect relapse rate, as previously thought [13], it does affect survival and relapse-free survival [14].

Beyond mutations In the 2013 update we noted that inconsistencies noted in the effects of some mutations, for example IDH, might relate to the presence/absence of other mutations, as noted above, or to inter-patient differences in expression of the mutations (and their ultimate translation into protein) due to epigenetic and/or micro RNA (miRNA) features [23]. Additionally un-mutated genes may be overexpressed. An example is SPARC [28]. CALGB found that expression of this gene was “up-regulated” in NK patients whose gene expression signatures

doi:10.1002/ajh.23834

ANNUAL CLINICAL UPDATES IN HEMATOLOGICAL MALIGNANCIES

AML: 2014 Update on Risk-Stratification and Management

Figure 1. Survival according to Wheatley score in patients aged 60 with newly diagnosed AML. The prognostic index derived from the MRCs AML11 trial was tested in patients in the AML 14 trial, as shown here.

were associated with unfavorable outcome and down-regulated in association with the favorable NPM1 mutation. This suggested a fundamental role for SPARC, whose overexpression was subsequently found associated with poorer outcome in NK AML, promotion of AML growth in murine models, and susceptibility to pharmacologic inhibition [28], as is miR-155 whose overexpression conveys a poor prognosis in NK AML [29]. Of particular interest Marcucci et al. recently described the integration of genetic and epigenetic data [30]. doi:10.1002/ajh.23834

Specifically they identified differentially methylated regions (DMRs) uniquely associated with prognostic mutations (e.g. FLT3, NPM) in NK AML. Those DMRs found to be most prognostic were associated with seven genes in which lower expression was associated with better prognosis. The lower a weighted summary score of the seven gene expression levels the better was CR rate, survival, and disease-free survival as verified in several independent data sets and independent of other prognostic factors [30] (Fig. 5). American Journal of Hematology, Vol. 89, No. 11, November 2014

1067

ANNUAL CLINICAL UPDATES IN HEMATOLOGICAL MALIGNANCIES

Estey

Figure 2. Top, survival in patients with complex karyotype (CK) according to presence of monosomal karyotype (MK); bottom, survival according to CK, MK, and TP53 status (see Ref. 21). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Incorporating post-treatment information/minimal residual disease (MRD) Given the currently limited predictive ability using only pretreatment covariates, as quantified by AUC values [20], it seems reasonable to examine the influence of post-treatment variables. Several can envisioned: a. Response to induction therapy: While it is well-known that patients who fail initial induction therapy have on average much shorter survival than patients who obtain a remission, the term “remission” has been expanded in recent years to include not only CR ( 1,000, platelet count > 100,000) but also CRp (as with CR but with platelet count < 100,000) and CRi (marrow with 200 marrow cells are enumerated [31]). Using SWOG, MDA, and ECOG data, Walter et al. suggested that after accounting for covariates such as age and cytogenetics and for time to achieve response, CRps are less likely to be maintained than are CRs and are likely associated with shorter survival [32]. Recent Fred Hutchinson Cancer Research Center data suggest the same is true when CRi is combined with CRp (Fig. 6) [33]. It remains to be seen if the benefit of CR also applies to patients receiving less intense therapies. b. PCR: Because relapse occurs in the great majority of patients who achieve CR, it has long been presumed that such patients have MRD undetectable by morphology. PCR is probably the most sensitive means to detect such MRD, having a sensitivity of 1 in 10(5) cells [34]. It can be used to detect (a) fusion transcripts (such as those associated with t[15;17], inv (16) or t(8;21) [PML-RARa, CBFB-MYH11, and RUNX1/RUNXT respectively], (b) mutated genes such as NPM1, DNMT3A, CEBPA, and IDH1/2, and (c) overexpressed genes such as WT1. The sensitivity and specificity for predicting relapse of PCR monitoring of MRD either when CR is achieved or during follow-up is

1068

American Journal of Hematology, Vol. 89, No. 11, November 2014

Figure 3. Influence of mutations on prognosis in patients age 3 log reduction in PCR transcripts as the cutoff best discriminating patients who will and will not relapse (Fig. 7) [37]. c. Multiparameter flow cytometry (MFC): MFC relies on the expression of aberrant surface antigens on AML blasts [34]. Most commonly doi:10.1002/ajh.23834

ANNUAL CLINICAL UPDATES IN HEMATOLOGICAL MALIGNANCIES

Figure 4. Effect of CD25 expression (positive 5 worse prognosis) on the “integrated classification scheme of Fig. 3; patients were age

Acute myeloid leukemia: 2014 update on risk-stratification and management.

Evidence suggests that even patients aged 70 or above benefit from specific AML therapy. The fundamental decision in AML then becomes whether to recom...
2MB Sizes 2 Downloads 8 Views