The article by Drs. Stein and Tallman is an excellent summary indicating that several different approaches may lead to the cure of acute promyelocytic leukemia (APL). The goal then becomes to use the treatment least likely to be associated with short-term or long-term toxicity. The principal short-term toxicity of standard therapy (all-trans retinoic acid [ATRA] + anthracyclines, hereafter referred to as “AIDA”) is treatment-related mortality (TRM) occurring during post-remission therapy. Such TRM has been reported to occur in 19% of patients older than 60 years.[1] It is in these patients that arsenic trioxide(Drug information on arsenic trioxide) (ATO) + ATRA is potentially most useful. Indeed, the ongoing European trial (F. Lo Coco, chair) randomizing newly diagnosed patients between AIDA and ATO+ATRA is limited to patients younger than 60 years of age, with older patients receiving ATO+ATRA. A long-term toxicity observed in about 1% of patients followed for a median of 51 months is development of myelodysplastic syndrome (MDS).[2] While it is plausible that MDS is part of the natural history of cured APL, MDS in this setting is probably more attributable to anthracyclines, which are well known to be associated with secondary MDS and acute myelocytic leukemia (AML). In principle, then, use of ATO+ATRA might avert secondary MDS. However, it is noteworthy that patients treated with this combination have been followed for considerably less time than patients treated with AIDA, and hence it is possible that ATO+ATRA may have longer-term toxicity than we anticipate today. Another very active drug in APL is gemtuzumab ozogamicin (GO, [Mylotarg]), which has been combined with ATO+ATRA in patients with white blood cell (WBC) counts greater than 10,000/μL (“high-risk”)[3] and used alone as therapy for molecular relapse.[4] Although the future of GO is in doubt (given its withdrawal from the commercial market in June 2010), other anti-CD33 antibody-toxin conjugates are being developed. Presumably the effectiveness of GO in APL reflects the CD33-rich nature of APL.
While it is possible that the increased death rate noted by Stein and Tallman in patients in whom initiation of ATRA is delayed is a result of worse disease rather than delayed treatment, a randomized trial to test this possibility is unethical. Hence we wholeheartedly concur with their proposal that ATRA be routinely available for immediate treatment of suspected APL and that its administration be based on clinical suspicion and/or review of a peripheral smear. We would point out, however, that ATO is a more effective APL drug than ATRA.[5] Thus, while oral ATRA used alone only very rarely results in cure of APL, ATO regularly does so.[6] Hence it might be reasonable to also have a supply of ATO on hand to either use alone or with ATRA in cases of suspected APL, given that ATO (+/− ATRA) is hardly more toxic than ATRA alone.[3] While the logistical obstacles to doing this are perhaps prohibitive today, this may not be true in the future, given the likely availability of oral ATO, as mentioned by Stein and Tallman. We certainly agree that in patients who present with circulating promyelocytes there is no need for an initial bone marrow test, or perhaps a marrow test at any time. The latter point may be more controversial, given data suggesting that the ability to detect molecular relapse, based on reappearance of the diagnostic PML-RAR rearrangement, may be reduced with use of blood rather than marrow.[7] On the other hand, if the rate of hematologic relapse is < 5%, as reported by investigators from MD Anderson who studied ATO+ATRA in low-risk patients (WBC < 10,000/μL),[3] there may be little need for molecular monitoring.
The authors point out the disparate results of Italian and Japanese studies regarding the value of maintenance therapy in APL. Such discrepancies seem to be a general theme regarding maintenance therapy in AML. One possible explanation is that the need for maintenance may depend on the “intensity” of consolidation therapy. Beyond this, however, is the need to recognize that there is almost certainly no single correct answer to the maintenance question. Rather, as suggested by Stein and Tallman, some patients may need maintenance and others may not, based on results of the molecular monitoring studies noted above. Indeed, although large randomized trials in APL are indispensable, a weakness of these trials is their focus on a single answer. It seems likely that as more is learned about the heterogeneity of AML, there will be greater emphasis on trials limited to a particular class of patient. Given that the number of such patients may be small, we may be forced to raise the conventional but arbitrary acceptable levels for false-positive (5%) and false-negative (20%) results.
A very important point previously made by Dr. Tallman is the much greater success seen when standard treatment for APL is employed in “academic centers” rather than in general practice, as represented by data from the Surveillance, Epidemiology and End Results (SEER) program and the New York State Cancer Registry.[8] This issue has profound implications for the relevance of academic practices to community practices, by which the majority of APL patients are likely treated. Perhaps the discrepancy will be resolved with more emphasis on earlier use of ATRA. However it remains possible that the types of patients seen in academic vs community practices are fundamentally different.
We note that when ATO+ATRA without chemotherapy was first used in newly diagnosed APL, it was known that the cure rate was 90% with AIDA in patients with WBC counts less than 10,000/μL. Nonetheless it was believed that unless new approaches were tried, the cure rate might never rise above 90%. Although use of ATO+ATRA without chemotherapy seemed to be based on solid clinical data, and although stopping rules were set so as to accept a high probability of a false-negative result,[9] the thought process motivating the ATO+ATRA study in patients for whom existing therapy was typically curative is perhaps worthy of discussion, particularly as cancer therapy becomes increasingly successful.
Financial Disclosure: The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
A hematologically stable CML patient developed a solitary, exquisitely painful ulceration, 7 cm × 5 cm, located on the mid-medial foreleg. There was never any evidence of venous insufficiency. The patient denied the possibility of a spider bite. What is the likely cause of this lesion?
