- Top of page
- PATIENTS AND METHODS
We have examined the toxicity and overall outcome of the Medical Research Council UKALL R1 protocol for 256 patients with relapsed childhood acute lymphoblastic leukaemia (ALL). Second remission was achieved in over 95% of patients. Two patients died during induction and seven patients died of resistant disease. The overall actuarial event-free survival (EFS) at 5 years for all patients experiencing a first relapse was 46% (95% CI 40–52). Duration of first remission, site of relapse, age at diagnosis and sex emerged as factors of prognostic significance. Five-year EFS was only 7% for children relapsing in the bone marrow within 2 years of diagnosis, but was 77% for those relapsing without bone marrow involvement > 2.5 years from diagnosis. All analyses in this report are by treatment received. For those receiving chemotherapy alone, the 5-year EFS was 48%; for autologous bone marrow transplantation (BMT), the 5-year EFS was 47%; for unrelated donor BMT, it was 52%; and for related donor BMT, the 5-year EFS was 45%. The groups, however, were not comparable with respect to risk factor profile, and therefore direct comparison of EFS is misleading. Adjustment for time to transplant and prognostic factors was used to reduce the effects of biases between treatment groups, but did not suggest benefit for any particular treatment. There was failure of our planned randomization scheme in this trial with only 9% of those eligible being randomized, which highlights the difficulties in running randomized trials especially in patients who have relapsed from a previous trial. The optimal treatment for relapsed ALL therefore remains uncertain. Alternative approaches are clearly needed for those with early bone marrow relapse if outcome is to improve.
- Top of page
- PATIENTS AND METHODS
Relapse remains one of the major obstacles to cure in childhood ALL and, despite improvements in first-line therapy, continues to occur in 20–30% of patients. What constitutes the best treatment for these patients has been the subject of much debate over recent years, and a consensus has not, as yet, been reached. A recent review ( Chessells, 1998) addresses many of the issues involved.
There is, however, agreement in certain areas. Namely, that overall 5-year EFS in second remission is poor, with figures of < 15% for those relapsing early (on treatment or within 6 months of stopping treatment) in the bone marrow, and generally < 50% for those with later (up to 36 months from diagnosis) bone marrow relapse or isolated extramedullary relapse. Very late (more than 36 months from diagnosis) isolated extramedullary relapse, however, has a 5-year EFS rate of > 50% ( Behrendt et al, 1990 ; Henze et al, 1991 ; Miniero et al, 1995 ; Schroeder et al, 1995 ; Billett et al, 1997 ; Gaynon et al, 1998 ; Wheeler et al, 1998 ).
The prognostic factors that influence the length of second remission have been consistently found in many studies and include duration of first remission, site of relapse, age and immunophenotype. In most studies, white cell count, sex and intensity of primary treatment do not appear to have any prognostic significance with respect to duration of second remission ( Behrendt et al, 1990 ; Henze et al, 1991 ; Buhrer et al, 1993 ; Miniero et al, 1995 ; Schroeder et al, 1995 ; Billett et al, 1997 ; Gaynon et al, 1998 ; Wheeler et al, 1998 ). Combined bone marrow relapses appear to do better than isolated bone marrow relapse in some studies ( Buhrer et al, 1993 ; Gaynon et al, 1998 ), but not in others ( Schroeder et al, 1995 ). More recently, the presence and level of minimal residual disease, measured by molecular or immunological methods, during the course of treatment has been shown to be of prognostic significance, although the majority of these studies have been performed in patients with newly diagnosed acute leukaemia ( Cave et al, 1998 ; Ciudad et al, 1998 ). In addition, the rate of disease reduction during induction remission therapy is thought to provide additional prognostic information ( Lilleyman et al, 1997 ).
Chemotherapy alone after bone marrow relapse can be curative in some cases, with EFS rates at 5 years of 47% in this study and rates of 20–65% reported elsewhere. For patients with early bone marrow relapse, however, results with chemotherapy alone are poor, with 5-year EFS rates of ≈ 20% in this study and generally < 20% in others ( Behrendt et al, 1990 ; Henze et al, 1991 ; Sadowitz et al, 1993 ; Miniero et al, 1995 ; Uderzo et al, 1995a ; Rivera et al, 1996 ; Billett et al, 1997 ). For isolated extramedullary relapse, chemotherapy alone can offer a chance of cure, with a 5-year EFS rate of 60% in this study and > 40% in most other studies ( Uderzo et al, 1990 ; Winick et al, 1993 ).
Bone marrow transplantation has been accepted as a form of treatment for relapsed ALL for many years, although its exact role remains controversial. Matched sibling bone marrow transplantation in second remission has been reported to give an EFS of 20–64% at 2 years. Factors affecting outcome after allogeneic transplant include site of relapse, duration of first remission, disease status at transplantation and presence of acute or chronic GVHD. Relapse still occurs after sibling allogeneic BMT in ≈ 20% of cases, and there is a transplant-related mortality of ≈ 20% ( Brochstein et al, 1987 ; Butturini et al, 1987 ; Sanders et al, 1987 ; Dopfer et al, 1991 ; Barrett et al, 1994 ; Weisdorf et al, 1994 ; Pinkel, 1995; Uderzo et al, 1995a , b; Gordonet al, 1997 ; Bordigoni et al, 1998 ).
For isolated extramedullary relapses, chemotherapy alone may produce a cure, and the role of BMT is even more controversial. In general, late isolated CNS relapse can be cured with chemotherapy alone, whereas more intensive treatment with BMT may be of benefit for early relapse with results for allogeneic and autologous grafts appearing to be similar ( Winick et al, 1993 ; Borgmann et al, 1995a ; Messina et al, 1996 , 1998). Isolated late testicular relapses are generally associated with a good prognosis with chemotherapy and local radiotherapy, although early relapses, especially those occurring on treatment, do poorly ( Uderzo et al, 1990 ; Schroeder et al, 1995 ).
The question as to what is the best treatment for relapsed ALL thus remains uncertain. The MRC UKALL R1 tried to address the role of autologous transplantation, but failure of randomization between continuation chemotherapy and autologous BMT has made this impossible. The results did, however, confirm that duration of first remission and site of relapse were significant prognostic factors (P < 0.001 and P < 0.001 respectively), as shown in previous studies. Age at diagnosis and immunophenotype were also of some prognostic significance, although T-cell disease was not of statistical significance in this study after allowance for duration of first remission and site of relapse. Interestingly, this study revealed a worse prognosis for male patients (P = 0.02).
The induction regimen used in this study of an unselected group of 256 children with first relapse of ALL has given an excellent complete remission rate of over 95%, better than that in many other reports. The toxicity of remission induction therapy was mild, with only two induction deaths. The overall EFS in this study for all patients experiencing a first relapse of ALL was 46% at 5 years. This is a good salvage rate, similar to most reports and better than many.
Despite the difficulties in this trial caused by failed randomization and differing risk profiles in the treatment groups, there was no clear difference in outcome for patients receiving different treatment after relapse. For those receiving chemotherapy alone, the 5-year EFS rate was 48%, for autologous BMT it was 47%, for unrelated donor BMT it was 52% and for related donor BMT the 5-year EFS rate was 45%. These results, however, do not allow for the biases in selection of treatment. The accepted method of reliably assessing the relative effect of two treatments is by randomized trial with analysis by intention to treat. Evaluation of transplantation based on treatment received introduces selection and time to treatment biases that invalidate direct analysis. Where treatments are not randomized, there is likely to be a difference in the proportion of high-risk patients in different treatment groups. As patients who receive transplants must remain in remission long enough to receive them, those with a short remission duration are less likely to be transplanted in complete remission than those with long remissions, thus early relapses are excluded from transplant cohorts but are included in the chemotherapy cohort. Adjustment for these known factors will reduce, but not eliminate, the effects of systematic differences between the groups, and so any comparisons should be treated with caution. Adjustments for time to transplant and known prognostic factors did not suggest benefit for any one treatment, but reduced the EFS rate for unrelated donor BMT in particular.
The type of event causing treatment failure varied between the treatment groups, with relapse being the cause in ≈ 50% of those receiving chemotherapy or autologous BMT. There were fewer relapse deaths in the allogeneic transplant group, although more deaths occurred during second remission. The results for related and unrelated BMT were similar with respect to transplant-related mortality (11% and 17% respectively) and EFS.
The lack of randomization in this study was not entirely surprising. This has been seen in many other attempted randomized studies in which lack of confidence in a treatment modality leads to failure of randomization, especially where side-effects are markedly different in the two arms. The difficulties of persuading patients and parents to accept a randomization when running a relapse protocol are even greater, when a previous trial treatment has failed. Of importance is the fact that > 50% of cases of failed randomization were due to physician preference/clinical decision. Although the trial design was initially accepted by the clinicians involved, with time they became more certain about which treatments were best for individual patients and they became less enthusiastic about autologous transplantation in particular. Large proportions of cases were not randomized because of the availability of an unrelated donor for bone marrow transplantation. Unrelated donor BMT was not included in the original trial protocol. However, as the trial progressed, evidence of the efficacy and probable superior results of unrelated donor BMT compared with chemotherapy or autologous BMT became apparent. It is likely that this is the explanation for many of the cases of failed randomization. It is important to note, however, that there were no randomized trials of unrelated donor BMT compared with chemotherapy or autologous BMT on which to base these decisions, but rather attempted comparisons suggesting that sibling donor BMT was superior to autologous BMT and evidence that results with unrelated donor BMT had improved and were similar to those with sibling donor BMT. Unrelated donor BMT has subsequently been included as a randomization in the succeeding relapse trial (UKALL R2) for those patients without a sibling donor, and it is hoped that acceptance of randomization will be superior in this trial. It is nevertheless disappointing as these decisions were not based on reliable information and the role of autologous BMT in relapsed ALL will not be answered without a truly prospective randomized trial. The same may also be said of the role of BMT in late relapse.
In conclusion, this study has revealed a high second CR rate of 95% in patients with bone marrow relapse, and an overall 5-year EFS rate of 46% in an unselected group of 256 children with a first relapse of ALL. There is currently no difference in the outcome between patients receiving chemotherapy or transplantation, although the groups are not directly comparable because of significant differences in risk profile. Because of the difficulties in this trial, it is important to note that we could not establish from these results that there was any difference between treatment groups in EFS. The lack of randomization highlights the difficulty of running randomized trials, especially in patients who have relapsed. This report is of outcome by treatment received. A further report will provide unbiased evidence on the value of related donor transplant by comparing groups according to donor availability. Patients who relapsed more than 6 months off treatment had a 5-year EFS rate of > 60%, whereas those who relapsed on treatment or within 6 months of stopping treatment had a 5-year EFS rate of 30%. Site of relapse was another prognostic feature, with bone marrow relapse faring worse than extramedullary relapse. Alternative approaches to therapy are clearly needed for those high-risk patients with early relapse at any site, and especially for those with bone marrow involvement, as the present approach with chemotherapy or conventional conditioning before BMT is salvaging only very few. Allogeneic bone marrow transplantation offers the strongest antileukaemic effect and alternative conditioning regimens may be considered. The improved EFS with allogeneic BMT, however, is often offset by the high transplant-related mortality (TRM). High TRM may be reduced by the use of less intensive non-myeloablative conditioning regimens, while still utilizing the graft versus leukaemia effect. A further option includes the use of immunotherapy, such as donor lymphocyte infusion, after BMT. These studies need, if possible, to be randomized, and, given the numbers involved, need to be performed in an international setting.