Impact of age on the outcome of patients with chronic myeloid leukemia in late chronic phase: results of a phase II study of the GIMEMA CML Working Party

Abstract

To assess the effect of age on response and compliance to treatment in patients with chronic myeloid leukemia (CML) we performed a sub-analysis within a phase II trial of the GIMEMA CML Working Party (CML/002/STI571). Since the WHO cut-off age to define an older patient is 65 years, among the 284 patients considered, we identified 226 (80%) younger patients (below 65 years) and 58 (20%) older patients (above 65 years) before starting imatinib. Response rates (hematologic and cytogenetic) were lower in the older age group but the probabilities of progression-free survival and overall survival (median observation time 3 years) were the same. Moreover, among complete cytogenetic responders, no differences were found in the level of molecular response between the two age groups. As might be expected, older patients experienced more adverse events, both hematologic and non-hematologic: this worsened compliance did not, however, prevent a long-term outcome similar to that of younger patients.

Older age constitutes a poor prognostic variable in patients with chronic myeloid leukemia (CML): the negative effect of age on long-term survival has been consistently observed with most effective therapeutic modalities, both pharmacological (busulfan, hydroxyurea and interferon) and allogeneic transplantation.1–3 When recombinant interferon was the gold standard for CML treatment, poor compliance was clearly age-related.4–6 Older patients had significantly worse side effects from interferon, although they had rates of complete hematologic response, complete cytogenetic response and overall survival similar to those of younger patients. Their poor prognosis may have been due at least in part to poorer tolerability and inadequate treatment delivery. Older patients were reported to tolerate only lower doses of interferon and dose adjustments were required more frequently. Conversely, other groups7 reported no significant differences in interferon compliance between patients in different age groups, but the lower interferon dosage required in the elderly could have contributed to this finding.

The response to treatment and outcome of older patients with CML receiving effective treatment was not extensively investigated until the introduction of imatinib. Imatinib determines durable complete hematologic remissions in almost all Philadelphia chromosome-positive (Ph) patients with CML in early and late chronic phase.8–16 Moreover, a major cytogenetic response is currently achieved in more than 80% of patients with early chronic phase disease and in more than 50% of those with late chronic phase disease. 17 In their series, Cortes et al.18 showed that imatinib eliminated the negative effect of age on response and survival. Within the frame of a large phase II trial of the GIMEMA CML Working Party,19 which enrolled 284 late chronic phase patients treated with imatinib (400 mg daily) after interferon failure, we evaluated responses, progression-free and overall survival and compliance in patients younger and older than 65 years of age at enrollment.

Design and Methods

The general outline of the trial, (CML/002/STI571), inclusion criteria and response definitions have been published previously.19–21 Briefly, late chronic phase Ph CML patients were eligible to be enrolled in the trial if resistant or intolerant to interferon. Patients received 400 mg of imatinib once daily until disease progression or until treatment intolerance. A complete hematologic response to treatment was defined as normalization of peripheral blood counts (white cell count <10×10/L and platelet count <450×10/L), with a normal white blood cell differential (up to 5% bands or meta-myelocytes and occasional myelocytes). Cytogenetic studies were performed by standard banding techniques and at least 20 metaphases were analyzed. The cytogenetic response (CgR) was rated according to the proportion of Phmetaphases as complete (Ph 0), partial (Ph 1–35%), minor (Ph 36–65%), minimal (Ph 66–95) or none (Ph ≥96%). In patients who achieved a complete cytogenetic remission minimal residual disease was detected on peripheral blood samples by a standardized real time quantitative reverse transcriptase polymerase chain reaction method that was established in the framework of the UE concerted action and has been previously described.19–21

Statistic analysis

Means were compared with the t-test, and frequencies with the χ test or Fisher’s exact test, as appropriate. Overall survival and time to progression to accelerated or blastic phase were calculated by the product-limits method of Kaplan and Meier. The level of significance for all statistical tests was 0.05.

Discussion and Results

Hematologic, cytogenetic and molecular responses

Two hundred and eighty-four patients were treated with imatinib for chronic phase CML after treatment with interferon had failed. In accordance with the WHO, which defines patients ≥65 years as old, we stratified the whole series into two subgroups: 58 patients (20%) were 65 years of age or older and 226 (80%) were less than 65 years old at enrollment into the trial. The characteristics of the younger and older patients are compared in Table 1. The median age at enrollment was 74 (range 65–85) and 47 (range 17–63) years in the older and younger groups, respectively. The categories of enrollment were balanced between two age groups: a larger proportion of older patients than younger were enrolled because of intolerance to interferon treatment (52% versus 28%; p=ns). Table 2 presents the responses to treatment by age group: older patients had a significantly lower probability of complete hematologic response (CHR) and complete cytogenetic response (CCyR) compared to younger patients. In the older age group 31/58 patients (53%) obtained a CHR and 21/58 patients (36≥%) achieved a CCyR compared with 168/226 (74%) and 130/226 (58%) patients in the younger age group. Minimal residual disease was monitored by real time quantitative polymerase chain reaction analysis (PCR) in all patients who achieved a CCyR. Our purpose was to identify whether there was any difference in the amount of BCR-ABL transcript between patients in the two age groups, even though they were in cytogenetic remission. At baseline, the median BCR-ABL/ β2M% ratio was 0.1340 and 0.0892 in the older and younger patients, respectively. We observed no significant difference in the kinetics of BCR-ABL transcript level reduction between older and younger patients who achieved CCyR during imatinib treatment, as shown in Figure 1. In fact, the median levels of BCR-ABL/β2M % were not different at any of the check-points up to the last analysis performed after 4 years on imatinib, when both groups had a BCR-ABL/β2M% ratio of 0.0007 (median value).

Figure 1.Pattern of molecular response in peripheral blood samples by treatment time. The transcript level is expressed as the ratio of BCR-ABL to β2-microglobulin × 100 and the values are medians. There was no significant difference in the level of BCRABL transcript between the two age groups during imatinib treatment and after 48 months of treatment the BCR-ABL/β2M % ratio was 0.0007 in patients ≥65 years old and in patients <65 years old.

Table 1.Characteristics of younger and older chronic phase CML patients at enrollment.

Table 2.Hematologic and cytogenetic responses in younger and older chronic phase CML patients treated with imatinib after failure of interferon treatment.

Adverse effects

Older patients experienced more adverse events, both hematologic and non-hematologic, than did younger patients. In fact, older patients experienced more grade III and grade IV neutropenia (p=0.04) and thrombocytopenia (p=0.02) (Table 3). The incidence of grade III and IV extra hematologic adverse events was also significantly higher in older patients (12%) versus younger ones (6%) (p=0.001). Overall, 6% of older patients definitely abandoned imatinib due to adverse events as compared to 2% of younger ones. With a median follow-up of 36 months (range, 12–54 months), the rate of progression to accelerated and blastic crisis was 12% in older patients and 10% in younger ones. The progression-free survival and the overall survival were not different between the two age groups (Figure 2). Older age constitutes a poor prognostic factor for outcome in patients with Ph-positive CML.1–3 Significantly, the most widely employed staging systems for CML, the Sokal score and the Euro score22 include age within the variables that can negatively influence the survival of CML patients. Older age was associated with a higher incidence of poor performance status, hepatomegaly and anemia. Until recently, the response and survival of elderly CML patients managed with effective treatment modalities has not been widely explored. Our analysis is focused on investigating the influence of age on responses (hematologic, cytogenetic and – for the first time – molecular) in a subset of aging patients with CML in late chronic phase after unsuccessful interferon therapy. Older patients had a lower probability of CHR (53%) compared to younger ones (74%). Cortes et al.18 reported a higher probability of CHR (94%) in both groups of patients using the same criteria to define a CHR. The differences in CHR rates probably reflect a difference in the ability to assure high dose intensity between a single, very experienced center and a multi-institutional, national trial. The 53% CHR rate in older patients in our study, significantly lower than that reported by Cortes et al. (94%), might also be explained by the different threshold ages chosen for defining patiens as elderly (60 years by Cortes et al., 65 years in the present study). The probability of CCgR was lower in elderly patients (36%) than in younger ones (58%) (p=0.02). We investigated the kinetics and the level of molecular response in patients who obtained a CCgR and found no difference. Despite differences in the hematologic and cytogenetic response rates between the two groups, with a median follow-up of 36 monts, survival free from accelerated or blastic phase and overall survival of the two groups were the same. These results were obtained notwithstanding the higher incidence of adverse events (both hematologic and extrahematologic): it is, however, well known that most imatinib-related adverse events can be managed without jeopardizing treatment end-points significantly. In conclusion, this study demonstrates that the poor prognostic influence of older age in patients with CML in chronic phase appears to be minimized in those treated with imatinib. Our data confirm the results of Cortes et al. in patients treated in late CP after failure of interferon treatment: fewer responses but the same long-term outcome for older patients. Finally, in the era of targeted therapies in hematology and oncology, it would be reasonable to define old patients on the basis of partially or completely age-independent reproducible indicators of fragility rather than simply according to years of age.

Figure 2.Disease free-progression (A) and overall survival (B) rates by age groups in chronic phase CML patients treated with imatinib. The overall survival curves were superimposed (91%).

Table 3.Older patients experienced more hematologic and nonhematologic side effects than did younger patients.

Acknowledgments

the assistance of Katia Vecchi is gratefully acknowledged

Footnotes

• Funding: supported by COFIN 2003 (Molecular therapy of Ph-positive leukemias), by FIRB 2001, by the University of Bologna (grants 60%), by the Italian Association for Cancer Research (A.I.R.C.), by Fondazione del Monte di Bologna e Ravenna, by European LeukemiaNet founds, and by A.I.L. grants.
• Author Contributions GR: data analysis and interpretation, manuscript writing; II: data analysis and interpretation, manuscript writing; SB: data analysis and interpretation; FC: data analysis and interpretation; MA: data analysis and interpretation; DC: collection and assembly of data; AP: data analysis and interpretation; SS: data analysis and interpretation; FP: collection and assembly of data; GRC: collection and assembly of data; FI: collection and assembly of data; GA: collection and assembly of data; RL: collection and assembly of data; NT: data analysis and interpretation; FP: data analysis and interpretation; GS: data analysis and interpretation; MB: final approval of manuscript; GM: conception and design.
• Conflict of Interest The authors reported no potential conflicts of interest.

• Received May 12, 2006.
• Accepted August 25, 2006.