Abstract
Treatment of acute promyelocytic leukemia (APL) has evolved over recent years, resulting in a cure rate of 75-80%. However, the prognosis of older patients with APL remains poorer as compared to young adults because of substantial morbidity of either induction or consolidation therapy. We describe therapeutic results in a series of 34 consecutive APL patients aged over 60 years, with particular emphasis on those patients managed outside of clinical trials because of comorbidities at diagnosis. All patients were programmed to receive the GIMEMA AIDA protocol, based on all transretinoic acid as induction followed by chemotherapy as consolidation. The median age was 70 years. Twenty-three patients (68%) received the protocol, while 11 (32%) were given a personalized approach. The median age was 69 years for patients on protocol as opposed to 75 years for the remaining ones (p=0.02). Six patients (18%) died within two days of diagnosis; among these, only one was on the AIDA protocol. Overall, complete response (CR) was achieved in 68% of cases; the CR rate was 74% for patients on the protocol as opposed to 54 % for those not. The most frequent cause of death was cerebral hemorrhage. Patients accrued into the GIMEMA AIDA protocol achieved longer survival (median not reached vs. 10 months, p=0.03). In conclusion, our data demonstrate that at least 30% of older APL patients are not eligible to accrual in multicenter trials; furthermore, in this subset, the possibility of early death is substantial. However, when CR is achieved, a personalized consolidation approach can be adopted with the possibility of achieving long-term disease control.
Advanced age represents the major adverse prognostic factor in acute myeloid leukemia (AML). Poor therapeutic results in older patients with AML depend on either severe comorbidities, which result in inability to tolerate intensive chemotherapy, or adverse biological characteristics of leukemic cells (1-5). Acute promyelocytic leukemia (APL) has unique biological and molecular characteristics which have led to therapeutic approaches which differ significantly from those used in other subtypes of AML (6-9). Treatment strategies for APL have evolved over the last 15 years resulting in an increase in remission and survival rates for affected patients, with a cure rate of 75-80% (9). Of note, while hematological, cytogenetic and molecular features in older patients appear similar to those of young adults, the prognosis of APL in the elderly is poorer given that the combination of ATRA and anthracycline, which represents the current standard of induction therapy, can be contraindicated because of cardiomyopathy or other severe organ dysfunction (10-14). Furthermore, consolidation therapy is in turn associated with significant morbidity and mortality (10-14). Most information on the outcome of elderly patients with APL derive from data of multicenter studies in which selected patients fulfilling inclusion criteria were accrued (12-14). On the contrary, little is known with regard to frail patients who are not eligible for protocol enrollment (10). Finally, the clinical impact of age-related factors resulting in fulminant APL, defined as early death occurring as soon as at diagnosis or within a few days, remains unclear. In this study, we describe the clinical characteristics and therapeutic results from a series of 34 consecutive patients with APL aged over 60 years, with particular emphasis on the difference between patients accrued into clinical trials and those managed with a personalized approach because of comorbidities at diagnosis.
Patients and Methods
APL diagnosis. APL diagnosis was initially established according to the WHO criteria (15) and confirmed in all patients by molecular identification of the promyelocytic leukemia/retinoic acid receptor alpha (PML/RARalpha) hybrid gene by reverse transcriptase polymerase chain reaction (RT–PCR) and karyotypic detection of the t(15;17) translocation in leukemic cells. Immunophenotypic analysis was also concordant in all patients as previously described (16).
Treatment. All patients were programmed to receive, after informed consent, the GIMEMA 0493 protocol ‘AIDA’, started in 1993 and amended in 1997 for the elderly population in order to reduce the intensity of consolidation. Details of the therapeutic program have been described elsewhere (12); briefly, the induction phase consisted of oral all-trans retinoic acid (ATRA) (45 mg/m2/day) until hematological complete remission (CR) and intravenous idarubicin (IDA) (12 mg/m2/day) given on days 2, 4, 6 and 8. Consolidation consisted of three consolidation courses (first course: IDA 5 mg/m2+ Ara-C 1 g/m2 from days 1 to 4; second course: mitoxantrone 10 mg/m2 + etoposide 100 mg/m2 from days 1 to 5; third course: IDA 12 mg/m2 on day 1 + Ara-C 150 mg/m2 subcutaneously every 8 h from days 1 to 5 plus 6-thioguanine 70 mg/m2 every 8 h from days 1 to 5). Patients in molecular CR after three consolidation courses (as assessed by RT-PCR of PML/RARa, see below) were randomized into four arms including chemotherapy alone with 6-mercaptopurine and methotrexate (arm 1), ATRA alone (arm 2), alternating chemotherapy and ATRA (arm 3), or no further therapy (arm 4). Since March 1997, an age-adapted AIDA regimen with reduced consolidation was proposed to the GIMEMA centers for patients aged >60 years. This consisted, after an identical induction, of the single first consolidation course of the AIDA protocol. Patients in molecular CR after this cycle received 2 years of maintenance treatment with intermittent ATRA alone (45 mg/m2/day for 15 days every 3 months). Patients received low-dose steroid (6-methylprednisolone at 0.3-0.5 mg/kg/day) during induction therapy. Management of ATRA syndrome consisted of prompt discontinuation of ATRA and administration of dexamethasone (10 mg every 12 h for 4 days).
Statistical methods. Time to event analysis was performed using the Kaplan–Meier method (17), with differences compared by the log-rank test. Overall survival (OS) was calculated from the date of diagnosis to death from any cause or to the date of last follow-up. Disease free survival (DFS) was calculated from the achievement of CR to relapse and death in CR, or to the date of the last follow-up. Differences in the distribution of individual parameters among patient subsets were analyzed using the chi-square or Student's t-test. All statistical comparison used two-tailed p-values. The statistical analysis was carried out using the SPSS statistical program (SPSS software, version 13) .
Results
Patients' characteristics. Between January 2005 and December 2007, 34 consecutive patients with APL aged over 60 years were observed at 5 different hematologic institutions in the South of Italy. As summarized in Table I, the median age was 70 years (range 61-84 years); 23 patients were males and 11 females. According to WHO scale, 2 (6%) patients were classified as having performance status (PS) 0, 12 (35%) as having PS 1, 16 (47%) as PS 2, and 4 (12%) as PS 3. Twenty-seven patients (84%) were affected by a least one concomitant disease requiring specific treatment. Thirty-one patients (91%) had classical APL (M3), while in 3 (9%) a diagnosis of variant APL (M3v) was made. All patients were found to have t(15;17), in one of them a concomitant deletion of chromosome Y was detected. Molecular analysis demonstrated BCR1 hybrid gene in 22 patients (67%), BCR2 in one (3%) and BCR 3 in 11 patients (30%). According to Sanz et al.'s prognostic model (18), 10 patients (29%) were classified as being at low risk, 17 (50%) as intermediate and 7 (21%) as being at high risk. Twenty-three patients (68%) fulfilled the inclusion criteria of the GIMEMA AIDA protocol and were actually given the programmed treatment, while 11 (32%) received a personalized approach. Causes of exclusion by the protocols were: poor PS in 4 patients, severe cardiomiopathy in 7 patients and death occurring within 48 hours from diagnosis in 5. Five patients presented with two concomitant reasons for exclusion from aggressive treatment. In all the above patients, poor PS was judged as being unrelated to APL. In frail individuals, treatment depended on their physician's attitude; in particular, 10 patients received only ATRA, while one patient received ATRA + IDA (at reduced dose of 6 mg/m2) because of severe concomitant leukocytosis. Of note, the median age was 69 years for patients on protocol as opposed to 75 years for the remaining patients (p=0.02).
Treatment results. Table II summarizes relevant therapeutic results. Overall, 6 patients, accounting for 18% of the whole patient population, died within 2 days from diagnosis, all because of fatal cerebral hemorrhage, in spite of beginning ATRA as soon as on the day of admission. Among these, 4 out of 6 had been classified as being at high risk according to Sanz et al's criteria and only one had been enrolled into the GIMEMA AIDA protocol. As opposed to our experience in young adults, the probability of early death is significantly higher in the group of older patients [as mentioned, 6 out of 34 (18%) in comparison to 7 out of 151 in the age range 15-60 (5%), p=0.02].
Overall, CR was achieved in 23 out of 34 patients (68%). In particular, the CR rate was 74% for patients on protocol as opposed to 54% for the remaining 11 patients, p=0.43. Causes of death were cerebral hemorrhage in 10 patients, while 1 patient died because of uncontrolled ATRA syndrome. Among 6 patients achieving CR after therapy with ATRA only, 3 were judged as unfit for chemotherapy and were consolidated with gemtuzumab-ozogamycyn (GO) at 3 mg/m2 for a total of 4 doses, while the remaining 3 received cycle 1 of the AIDA protocol. All patients achieved molecular remission at the end of consolidation and all received maintenance therapy with ATRA (45 mg/m2 from day 1 to 15 every three months) for two years. On the contrary, all patients accrued into AIDA study were able to receive the programmed consolidation therapy. The median overall survival was 38 months, the median DFS was not reached, as shown in Figure 1. Patients accrued into the AIDA protocol achieved better results in terms of OS (median not reached vs. 10 months, p=0.03), as shown in Figure 2. Overall, 8 patients out of 23 (35%) relapsed; in 5 cases relapse was hematological, in 1 patient it was molecular. Six of them (75%) achieved a second CR with arsenic trioxide (ATO) (n=4) or ATO + ATRA (n=2). One patient experienced very late relapse at 71 months of the first CR achievement and is currently in second molecular remission.
Clinical and hematologic characteristics of the patients.
Discussion
APL represents a paradigm of successful targeted therapy in clinical hematology. The introduction in newly diagnosed patients of ATRA and, more recently, of ATO results in disease-free and overall survival of 70 and 80% at 5 years, respectively (19). In addition, the potential curability without any cytotoxic agent is currently under investigation, by therapeutic programs based on the combination of ATRA + ATO or single-agent ATO (20-24). Notwithstanding, some clinical challenges remain unresolved in APL; apart from relapse, the occurrence of death during induction therapy accounting for 5-7% of patients and the treatment of the disease in the elderly, namely in frail patients affected by significant comorbidities, still represent major obstacles to achieving a cure. In addition, early deaths occurring as soon as at diagnosis or within the first 24-48 hours from it should also be taken into account, given that most of these events are not considered in the results of multicenter trials. Different studies have focused on the treatment of older patients with APL; in particular, three larger series have been reported by the European APL Group, the GIMEMA group and the Spanish Group, respectively (12-14). In both the GIMEMA and European Group study, the frequency and severity of complications linked to intensive chemotherapy were highlighted, suggesting that in APL occurring in the elderly, less intensive induction and post-remission therapy allows significant reduction of severe treatment-related toxicity being equally effective (12). On the contrary in the Spanish study, excellent results, accompanied by a high degree of compliance and very good tolerance from administering to elderly patients the same protocol as that used for younger adults (ATRA with anthracycline monochemotherapy) were reported (14). Our analysis differs from the above multicenter studies in providing an unselected group of patients aged >60 years and does clearly demonstrate that, mainly in the age above 70 years, a relevant selection is made as to accrual into large multicenter studies. As matter of fact, in our series, 32% of patients were treated outside of a clinical protocol; worthy of note, the median age of these individuals was significantly higher (75 vs. 69 years, p=0.02), as compared to patients on protocol. Cardiomyopathy as well as poor PS were the main reasons accounting for exclusion.
Therapeutic results.
Overall, very early death due to fatal cerebral hemorrhage was the most frequent cause of failure to achieve CR and occurred more frequently in the older population as compared to young adult patients (18% vs. 5%, p: 0.02). In no patient was resistance to induction therapy observed, failure to achieve CR being only due to hemorrhagic events or uncontrolled ATRA syndrome; this suggests that, differently from other AML FAB subtypes, the poorer prognosis of APL at an advanced age is strictly dependent on the patient's characteristics rather than disease biology. Of note, a clear association between white blood cell count at diagnosis and occurrence of cerebral hemorrhage was found, suggesting that the percentage of high risk patients according to the score of Sanz et al. strongly influences early mortality rate in APL, namely in the older patient population (25). It is conceivable that age-related concomitant cerebral angiopathy can further account for the above fatal complication. Of note, in our series, 5 out of 6 cases of fulminant APL occurred in patients treated off-protocol. It is conceivable that a delay in APL diagnosis, probably due to a misinterpretation of pancytopenia as expression of myelodysplastic syndrome, may account for the high number of fatalities in this subset. On the contrary, in our series, there were no deaths following consolidation, which in other studies has been reported as main cause for poor prognosis in older APL patients (12, 13). The administration of a maximum of one cycle of chemotherapy or single-agent GO accounts for these favorable results. Worthy of note, after consolidation, all patients in our series achieved molecular remission, including three patients who received GO, confirming the potential utility of this agent as post-remission treatment of APL (26-29). In conclusion, the analysis of our data demonstrates that 30% of older APL patients are not eligible for multicenter protocols and receive a sub-optimal induction therapy. Furthermore, in this subset, the possibility of very early death occurring as soon as at diagnosis or within 24-48 hours is substantial. However, when CR is achieved, a personalized consolidation approach can be adopted, with substantial possibility of achieving long-term disease control, provided that molecular remission is obtained at the end of consolidation. The use of arsenic trioxide with or without ATRA, as recently suggested by European Leukemia Net recommendations (9), could represent a reasonable option for these patients.
Overall (A) and disease-free survival (B) of the whole patient population: median 38 months and not reached, respectively.
Overall survival according to inclusion in the GIMEMA AIDA protocol: median not reached vs. 10 months (p=0.03).
Footnotes
- Received September 23, 2009.
- Revision received February 9, 2010.
- Accepted February 16, 2010.
- Copyright© 2010 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
