- Top of page
Between 2000 and 2006, 85 adult BCR-ABL negative acute lymphoblastic leukaemia (ALL) patients between 18 and 60 years of age were treated using a modified paediatric regimen, which included high doses of asparaginase delivered weekly for 30 weeks during intensification. The complete response rate with induction therapy was 89%, and decreased with increasing age, mainly due to higher induction mortality. All post-induction treatments were delivered on an outpatient basis. The most common complications during intensification were infections (47%), osteonecrosis (32%), venous thromboembolism (23%) and neuropathy (22%). At a median follow-up of 4 years, the 5-year overall survival (OS) and relapse-free survival (RFS) were 63% and 71%, respectively. Significant adverse predictors for OS were age >35 years, high white blood cell count, MLL rearrangement, allogeneic stem cell transplantation in first complete remission and <80% of the planned asparaginase dose delivered during intensification. Patients aged ≤35 years had a 3 year OS of 83%, as compared to 52% for patients aged >35 years. We conclude that the administration of this paediatric regimen is feasible and has considerable activity in adult ALL, particularly in younger patients. Effective delivery of asparaginase dosing appears to be important in achieving an optimal antileukaemic effect.
Acute lymphoblastic leukaemia (ALL) accounts for approximately 20% of adult and nearly 80% of childhood acute leukaemias. There has been considerable improvement in results with childhood ALL in recent years, with CR rates of 95% and 5-year disease-free survivals of 80% with modern protocols (Schrappe et al, 2000; Silverman et al, 2001; Pui & Evans, 2006). However, results in adults with ALL are far inferior, with survival rates of 35–45% in patients aged 18–60 years (Annino et al, 2002; Thomas et al, 2004; Goldstone et al, 2008). This is partly related to a higher incidence of BCR-ABL (Philadelphia chromosome) positive disease in adults. However, even in BCR-ABL negative patients, results are inferior. Adult ALL is also characterized by a higher frequency of other poor risk features, such as high white blood cell count (WBC) presentation, other karyotypic abnormalities and a lower frequency of favourable risk features, such as hyperdiploidy and the ETV6-RUNX1 (TEL-AML1) fusion gene (Mancini et al, 2005; Rowe et al, 2005; Moorman et al, 2007).
Treatment regimens used by paediatric and adult oncologists also vary considerably. Paediatric regimens use more intensive doses of certain drugs, such as asparaginase, as compared to adult regimens. Recently, retrospective data have been published demonstrating that those adolescents and young adults with ALL who have been treated with a paediatric ALL regimen have superior event-free and overall survival as compared to adult-based therapy (Barry et al, 2007; Ribera et al, 2008). In light of encouraging results reported regarding the Dana Farber Cancer Institute paediatric protocol DFCI 91-01 in adolescent ALL (Silverman et al, 2001; Barry et al, 2007), we have been treating all newly diagnosed ALL patients aged 18–60 years with a modified version of this protocol, and now report our results.
- Top of page
We have demonstrated that treatment with this modified paediatric regimen is feasible in ALL patients between the ages of 18 and 60, with acceptable overall toxicity. As this was an unselected series of consecutive patients treated at a single institution, which is the regional centre for the treatment of acute leukaemia, selection bias would not be a factor. Our patient population is typical when compared to adult patients studied using other regimens; therefore, differences in outcome can be inferred to arise from our treatment strategy and not from differences in tumour biology. The OS and RFS obtained in this series are superior to results reported in Philadelphia chromosome negative patients using adult ALL regimens, including Leucémies Aiguës Lymphoblastiques de l’Adulte (LALA)-94 (41% 3-year DFS) (Thomas et al, 2004), Hyper-CVAD (45% 5-year OS) (Kantarjian et al, 2000) and Medical Research Council (MRC) UKALL XII/Eastern Cooperative Oncology Group (ECOG) E2993 (43% 5-year OS) (Goldstone et al, 2008). The results also compare favourably with other recent reports using paediatric regimens in young adults (Ribera et al, 2008; Huguet et al, 2009), and further support the concept that such paediatric regimens have superior activity as compared to adult protocols.
The most impressive results were seen in the young adult population up to 35 years of age; the 83% 3-year OS is comparable to results reported in the adolescent population using this regimen (Silverman et al, 2001; Barry et al, 2007). As determined in other reports for adult ALL, older patients had a significantly inferior survival. This was partly related to a higher induction mortality. The Group for Research on Adult Acute Lymphoblastic Leukaemia (GRAALL)-2003 study (Huguet et al, 2009), using another paediatric regimen, also reported a higher induction death rate in older patients (13% for patients aged 46–60 years). The 20% induction mortality in patients aged ≥50 years in our study can potentially be addressed by improved infection prophylaxis and surveillance, or by modifying the induction itself – one possibility would be to move the high-dose methotrexate to the post-remission phase in older patients, in order to reduce mucosal toxicity during induction.
Patients aged >35 years achieving CR also trended toward a lower RFS. This might be partly related to differences in disease biology in older ALL patients. However, older patients were also less likely to achieve 80% of the targeted asparaginase dosing during intensification, suggesting that an inability to deliver adequate dose intensity in older patients, due to decreased tolerance, may be an important factor in their inferior outcome. Nevertheless, the results in the over-35 age group also compare favourably with previously reported series (Annino et al, 2002; Thomas et al, 2004; Goldstone et al, 2008).
Certain other high-risk groups still appear to have inferior outcomes with this protocol, including patients presenting with high WBC and those with MLL rearrangements, particularly the t(4;11) karyotype. The latter was strongly associated with high presenting WBC. These high-risk patients may benefit from alloBMT in CR1 (Annino et al, 2002; Vey et al, 2006), or from modifications to the post-remission regimen, such as the incorporation of high-dose cytarabine or cyclophosphamide. However, this needs to be further addressed in prospective studies.
Acute lymphoblastic leukaemia occurring in patients treated for a prior malignancy, although previously reported (Pagano et al, 2000), is rare and has not been a well-recognized entity. Our analysis suggests that this does occur and may represent a poor prognosis subgroup (3/5 relapses). However, this needs to be fully addressed with a larger series, as our numbers are small and it is unclear whether this is a reflection of poor risk karyotype or an independent prognostic feature.
Among the distinguishing features of this protocol is the heavy reliance on high doses of asparaginase during the intensification phase of treatment. This agent forms the backbone of the intensification phase; in total, the protocol delivers at least four times more asparaginase than most adult protocols. Our results suggest that, as reported in the paediatric population (Silverman et al, 2001), treatment with at least 80% of the targeted dose for asparaginase remains critical for optimal outcomes. The 3-year RFS of 88% in patients receiving this target dosing is comparable to outcomes in the adolescent population. Our experience is that this target can be achieved in approximately 80% of patients up to 60 years of age. It is not possible to ascertain definitively whether the inferior outcomes with reduced asparaginase dosing were specifically related to this drug, or were a reflection of overall reduced dosing intensity, because asparaginase intolerant patients tended to receive less of other drugs as well. Furthermore, as only 12 patients did not receive 80% of the targeted dose, confirming the importance of asparaginase dosing as an independent prognostic factor would require a larger series of patients.
The presence of CNS leukaemia did not predict for an adverse outcome. Most patients experienced a rapid clearance of CSF blasts with intrathecal chemotherapy plus high-dose intravenous methotrexate during induction, and CNS recurrence was infrequent. These findings are in keeping with those reported in the LALA-94 trial (Thomas et al, 2004). In contrast, Lazarus et al (2006) reported that patients with CNS leukaemia had a somewhat inferior OS (29% vs. 38% for those without CNS disease). However, even in the latter study, CNS involvement was far less predictive of survival than other factors, such as age and WBC, by multivariate analysis.
Vincristine-induced neuropathy is a common and sometimes debilitating occurrence in these patients. There have been concerns that substituting vinblastine for vincristine could potentially adversely impact outcomes, possibly by increased myelosuppression necessitating reductions in doses of other agents, such as 6-MP. Our results demonstrated that the substitution of vinblastine does not adversely impact survival; there was actually a trend toward a superior RFS and OS in the vinblastine-treated group.
Osteonecrosis (ON) was diagnosed in nearly one-third of patients who received intensification. This rate is higher than reported in retrospective paediatric ALL series (Vaidya et al, 1998; Patel et al, 2008), and is probably related to the high doses of corticosteroids used in this protocol, possibly exacerbated by the high doses of asparaginase. However, the cumulative incidence may actually be higher, as magnetic resonance imaging scans were mainly performed in symptomatic patients and not routinely; ON is frequently asymptomatic in the early stages (Ojala et al, 1997). ON can progress to joint destruction, eventually requiring hip replacement surgery. Postulated mechanisms of ON include increased intra-osseous pressure causing vascular occlusion, impaired fibrinolysis (French et al, 2008) and cytokine effects (Assouline-Dayan et al, 2002). Future strategies need to focus on elucidating the pathogenesis of this complication; this may lead to useful pharmacological preventative strategies. Reducing corticosteroid doses, or using a different ‘paediatric’ regimen that contains lower doses, may decrease the frequency of ON and may therefore be more suitable to the adult population. However, it is unclear what impact such modifications would have on relapse rates. ON can have a substantial impact on patients’ quality of life following the completion of therapy. We have not evaluated quality of life issues in our patients, but such an evaluation is currently being planned.
Venous thromboembolism events were also common, primarily during intensification. This is probably related to the high doses of asparaginase, which is associated with an increased risk of VTE secondary to decreased levels of antithrombin III and other anticoagulants (Beinart & Damon, 2004). Although none of these events were fatal, they were potentially life threatening and a cause of significant morbidity. Prophylactic low-dose anticoagulation with low-molecular weight heparin may potentially reduce the incidence, but has not been studied prospectively in this setting. As the asparaginase could be successfully restarted in most cases with continued anticoagulation, VTE did not have a major adverse impact on the ability to administer optimal asparaginase dosing.
A recent MRC/ECOG study found that standard risk ALL patients had an overall survival benefit of approximately 10% with alloBMT in CR1 as compared to a no donor group (Goldstone et al, 2008). However, there was no benefit seen in high-risk patients. Standard risk patients included those under age 35 years; the OS in the non-transplanted group of younger standard risk patients in that report (52% 5-year OS) was lower than that obtained in our series (83% 3-year OS which was maintained at 5 years, as shown in Fig 1). This suggests that the net benefit of alloBMT versus chemotherapy may be critically dependent on the chemotherapy regimen used. Our data do not definitely address this question, as the number of transplanted patients in this series was small and the group was selected. However, it would be difficult to improve on our survival results in patients aged <35 years with alloBMT, given the inherent treatment-related risks associated with the latter. Therefore, we do not favour transplanting standard risk young patients in CR1 treated with this protocol. However, this question would need to be evaluated in a proper comparative trial. It is possible that other approaches to evaluate patients at higher risk of relapse, such as minimal residual disease (MRD) evaluation (Bassan et al, 2009), could identify a subpopulation of patients that may benefit from BMT, but this requires further evaluation. MRD monitoring was not performed in our patients.
In summary, our results demonstrate the feasibility of using this modified paediatric regimen to treat adults with ALL up to the age of 60 years. Patients up to the age of 35 years appear to have particularly favourable results; outcomes are superior to previously reported series in adult ALL. Future strategies should focus on reducing some of the complications associated with the intensification portion of the regimen, particularly infections and osteonecrosis, as well as improving results in high-risk patients. Nevertheless, the overall results indicate that this is a highly effective antileukaemic regimen for adult ALL.