|
|
| |||||||||||||||||||||||||||||||
Background Therapy for childhood lymphoblastic leukemia has evolved during the past three decades, but key questions about what are the least toxic, most effective forms of treatment remain unanswered because of the lack of comprehensive follow-up information.
Methods To assess long-term outcome in the series of clinical trials conducted at St. Jude Hospital, we compared the results of treatment typical of four eras: exploratory combination chemotherapy (era 1, 1962 to 1966; 91 patients), regimens for the control of meningeal leukemia (era 2, 1967 to 1979; 825 patients), limited intensification of therapy (era 3, 1979 to 1983; 428 patients), and extended intensification of therapy (era 4, 1984 to 1988; 358 patients). ("Intensification" refers to strategies of systemic chemotherapy that are more aggressive than conventional ones.) The major end points were survival and event-free survival; we also calculated the relative risk of treatment failure and the rate of relapse or death after treatment ended (post-treatment failure rate).
Results The probability of event-free survival improved significantly in each successive era (P<0.001 by the log-rank test), reaching 71 percent in era 4. There was a decrease of approximately 50 percent in the risk of treatment failure from one era to the next in each subgroup of patients defined according to different combinations of the leukocyte count, race, age, and sex. Leukemia appeared to be eradicated in patients who remained in complete remission for three years or more after treatment in era 4. The incidence of death due to nonleukemic causes remained 4 to 6 percent despite the trend toward more intensive treatment. An estimated 765 patients (45 percent) are long-term survivors; most of them (80 percent) have no health problems related to leukemia or its treatment.
Conclusions The development and successful application of preventive therapy for meningeal leukemia, followed by the intensification of systemic chemotherapy, has progressively improved the rate of cure of childhood lymphoblastic leukemia, with relatively few adverse sequelae.
To meet the need for more comprehensive information about the least toxic, most effective forms of therapy for children with acute lymphoblastic leukemia, we reviewed data on more than 1700 patients enrolled in 11 consecutive clinical studies conducted from 1962 through 1988 at St. Jude Children's Research Hospital. Our study addressed questions that are important for planning future treatment and that are best answered with follow-up information collected over many years. For example, have rates of cure improved progressively since the development of effective antileukemic therapy in the late 1960s, or has a plateau been reached? Are traditional definitions of cure still relevant given recent advances in treatment? Has remission-retrieval therapy contributed to long-term survival? Finally, are second cancers more common with contemporary treatment?
Methods
From 1962 through 1988, 1702 consecutive patients in all risk categories who were less than 18 years of age and who had lymphoblastic or undifferentiated leukemia were enrolled in 11 treatment studies at St. Jude Children's Research Hospital. The diagnosis was based on morphologic evaluation of Wright's-stained smears of bone marrow and negative staining for myeloperoxidase (<3 percent positive blasts). Since 1968, each protocol has been approved by an institutional review board and by the National Cancer Institute. Written informed consent was obtained for all patients.
Definitions of remission, failure of induction, relapse, meningeal leukemia, and clinical risk groups, as well as guidelines for ending therapy, have been reported elsewhere4,8,9,10,11,12,13,14. We did not consider the immunologic and cytogenetic features of acute lymphoblastic leukemia in this analysis, since they were not routinely determined before 1980. All patients identified since 1980 as having surface-immunoglobulin-positive mature B-cell leukemia were excluded from the study cohort.
Long-Term Follow-up
For the first 10 years after diagnosis, all patients were followed by physicians at St. Jude Children's Research Hospital; thereafter, the patients' own physicians forwarded pertinent medical records to us. Only 3 of the 1702 patients were lost to follow-up. Since 1984, patients who survived for two or more years after the completion of treatment have been seen in the After Completion of Therapy Clinic at this center15. The purpose of these visits is to monitor the patients' health and to provide medical care for late adverse effects, such as retarded growth, thyroid dysfunction, learning disability, infertility, second cancers, and neurologic and psychosocial problems.
Statistical Analysis
Survival, event-free survival, and estimates of cumulative risk were computed by the method of Kaplan and Meier. The associated standard errors were calculated by the method of Peto et al.,16 and the 95 percent confidence intervals for cumulative risk estimates were calculated after logarithmic transformation17. The log-rank test was used to compare survival curves. We defined event-free survival as complete remission in a surviving patient without relapse at any site and without the development of a life-threatening second cancer (such as a brain tumor). Patients who survived without leukemia for at least three years after the cessation of therapy were classified as long-term survivors. Relative risks of treatment failure were calculated on the basis of estimates generated by proportional-hazards analysis that included the initial leukocyte count, race, age, sex, and treatment as prognostic variables18. Rates of relapse, development of a second cancer, or death after treatment ceased (post-treatment failure rates), with 95 percent confidence intervals, were estimated by standard life-table methods. Only patients who were in continuous complete remission when treatment ceased were included in these estimates. All analyses were based on records that were up to date through July 14, 1992.
Results
Treatment Eras
The St. Jude therapy program for childhood acute lymphoblastic leukemia spans four eras (Table 1). The first, represented by studies 1 through 4 (1962 to 1966),8,9 was characterized by efforts to prolong the duration of hematologic remission with the use of combination chemotherapy. Effective treatment for subclinical central nervous system leukemia had not yet been developed. The second era, comprising studies 5 through 9 (1967 to 1979),10,11,12,13,14 was characterized by the administration of 2400 cGy of irradiation to the central nervous system and the use of intrathecal chemotherapy to prevent overt meningeal leukemia. Study 10 in era 3 (1979 to 1983)19,20 explored the relation between systemic exposure to methotrexate and treatment outcome and introduced a new class of drugs, the epipodophyllotoxins, to eliminate leukemia resistant to standard agents. The fourth era (study 11, 1984 to 1988)4 was marked by the intensification of early therapy for all patients, a wider selection of cytoreductive agents, and the alternating use of non-cross-resistant pairs of drugs during the post-remission period. "Intensification" refers to strategies of systemic chemotherapy that are more aggressive than conventional ones (Table 1).
|
Characteristics of the Patients
The presenting features of the patients enrolled in these treatment studies were similar in all four eras (Table 2). Two exceptions were the higher proportions of children with leukocyte counts above 100,000 per cubic millimeter in eras 1 and 4 and the higher proportion of black children in era 1. Altogether, 75 percent of the patients were older than 1 year of age but younger than 10, and 78 percent had leukocyte counts below 50,000 per cubic millimeter. There was a higher proportion of boys, which is consistent with findings in other large cohorts of children with acute lymphoblastic leukemia25.
|
Event-free survival improved progressively from era 1 through era 4 (Figure 1) (P<0.001). The vast majority of patients treated in era 1 either relapsed or had other adverse events within two years after diagnosis. The five long-term survivors, now 23 to 26 years old, had favorable presenting features: age of 2 to 11 years and leukocyte counts below 10,000 per cubic millimeter. With the introduction of effective therapy for subclinical meningeal leukemia in 1967, the estimated five-year event-free survival rate increased from 9 percent to 36 percent. However, randomized comparative trials during this era resulted in the exclusion of some patients from preventive meningeal therapy or combination chemotherapy. The intensification of methotrexate treatment, the introduction of the epipodophyllotoxins, and the use of effective prophylaxis against Pneumocystis carinii pneumonia during era 3 were associated with an additional gain of 17 percent, and in era 4 the reinforcement of early treatment, followed by the rotational use of non-cross-resistant pairs of drugs, boosted the estimated rate of event-free survival to 71 percent. These improvements extended to virtually all subgroups of patients (Table 3).
|
|
|
Changes in treatment within the therapy program influenced the pattern as well as the frequency of treatment failure (Table 5). Early advances in prophylactic therapy for meningeal leukemia led to a decrease in the rate of meningeal relapse (from 44 percent to 13 percent), but the frequency of hematologic relapse remained high (34 percent). With limited intensification of therapy in era 3, the proportion of patients who had hematologic relapses declined to 23 percent. This improvement continued in era 4, when only 12 percent of the patients had a hematologic relapse as their first adverse event -- a proportion not appreciably different from the 5 percent rate of isolated meningeal relapse in that cohort. The incidence of testicular relapse, once substantial, decreased to less than 1 percent with better control of systemic leukemia.
|
Post-treatment Failure and Long-Term Survival
The effectiveness of leukemia therapy is ultimately gauged by the proportion of patients who remain well after the cessation of treatment. Since 1962, 919 of 1702 patients (54 percent) have had their therapy stopped during continuous complete remission, usually after 2.5 years. Among these 919 patients, 718 (78 percent) are still alive without a relapse, a second cancer, or another adverse event. Analysis of the rates of post-treatment failure according to era (excluding era 1, during which only 12 patients in continuous complete remission had their treatment discontinued) revealed a relatively high probability of relapse or death in the first year after the cessation of therapy (8 to 13 percent for all eras), which declined to 1 percent or less by the fifth year of era 2, the fourth year of era 3, and the third year of era 4. Thus, the likelihood of having a late adverse event decreased with each succeeding era.
An additional 104 patients ceased therapy while in complete but not continuous remission after successful retrieval therapy for an initial hematologic, meningeal, or other relapse; of these, 47 remain free of leukemia. The best example of the contribution of patients in second or subsequent complete remission to the group of long-term survivors is in era 3. Although the estimated 10-year event-free survival for this subgroup was only 49 percent (Figure 1), the observed 10-year survival was 67 percent (curve not shown), reflecting the ability to treat successfully a substantial proportion of patients whose initial treatment would not be considered intensive by current standards. Altogether, an estimated 765 patients (45 percent) have survived for at least three years after the completion of treatment and are probably cured.
Delayed Effects of Therapy
Delayed treatment effects were investigated in patients who were followed for a median of 18 years after diagnosis (range, 2 to 28). The majority of these children (80 percent) are leading normal lives without obvious health problems.
Malignant solid tumors were diagnosed in 29 patients (18 long-term survivors) treated in era 2 or 3 and were the first adverse events in 20 patients. Twelve of the second cancers were brain tumors, all of which occurred in patients who had undergone cranial irradiation as a major component of central nervous system prophylaxis. After five years of follow-up, the Kaplan-Meier estimates of the risk for the development of a malignant solid tumor were 0.7 percent (95 percent confidence interval, 0.2 to 2.3) among patients treated in era 2 and 0.3 percent (95 percent confidence interval, 0.04 to 2.2) among those treated in era 3. At 10 years, the respective risk estimates were 1.8 percent and 3.0 percent. Overall, the risk estimates for these two eras were not significantly different by log-rank analysis.
Among the 27 patients in whom secondary acute myeloid leukemia developed, this disease occurred as the first adverse event in 21 patients who were in continuous complete remission. The five-year Kaplan-Meier estimates of the cumulative risk of this complication were 0.5 percent for era 2, 2.3 percent for era 3, and 3.6 percent for era 4. The 10-year estimates were 0.8 percent for era 2 and 3.2 percent for era 3. Log-rank analysis revealed significantly higher risks of acute myeloid leukemia for patients treated in eras 3 and 4 than for those treated in era 2 (P = 0.03 and P = 0.01, respectively). In contrast to the high survival rate among patients treated for solid tumors that developed after the treatment of acute lymphoblastic leukemia (18 of 29), only 2 of the 21 patients with secondary acute myeloid leukemia survive.
Specific abnormalities of growth and neuropsychological function have been described elsewhere27,28.
Discussion
This study demonstrates continued improvement in treatment outcomes during four consecutive eras of clinical trials of treatment for childhood acute lymphoblastic leukemia. We attribute these gains to a series of modifications of treatment, beginning with the introduction of effective therapy to prevent overt meningeal leukemia10 and culminating in the intensification of early treatment for all patients, regardless of their risk status4. Additional factors, such as improvement in antimicrobial therapy and advances in intensive care, probably contributed to more favorable outcomes by decreasing the number of deaths due to infectious and toxic causes and by shortening interruptions of chemotherapy29.
Cure of leukemia should mean permanent recovery from the disease; however, the inability to distinguish between a true disease-free state and one in which residual leukemic cells remain has made it difficult to apply this definition with confidence. The results of our study indicate that a patient's risk of treatment failure becomes negligible (less than 1 percent) after three or four years of event-free survival after the cessation of therapy, or perhaps after two years if the favorable trend seen in era 4 continues. The occasional relapses that occur more than three years after treatment suggest that cure may never be a certainty. For the great majority of patients, however, the disease appears to be eradicated if complete remission lasts as long as 5.5 years after diagnosis.
With the publication of the results of study 5,10 it appeared that intensified therapy, as perceived in the early 1970s, would cure approximately half of patients with acute lymphoblastic leukemia. The survival rates were no better in the four subsequent trials conducted during era 2, despite satisfactory control of meningeal leukemia. The lack of improvement in outcome may be explained in part by the inferior results produced by some regimens used in randomized trials during this period. Although the results of studies conducted in era 3 indicated that intensification of chemotherapy retarded the emergence of drug resistance, hematologic relapse continued to be the main cause of death. Thus, beginning in 1984, we treated all patients, regardless of their risk status, with intensified therapy. Since the failure to control residual leukemia during clinical remission favors the development of drug-resistant blasts,30,31 we administered ostensibly non-cross-resistant pairs of drugs on a rotating schedule to 88 percent of the patients. The superior outcome (Figure 1), together with our ability to administer more than 90 percent of the therapy in an outpatient setting,4 demonstrates the feasibility and efficacy of this approach. Other investigators, using different agents, dosages, and schedules of administration, have reported similar gains with the intensification of treatment,1,2,3 although some have reported little or no benefit32.
In era 3, 37 percent of the patients who had relapses became long-term survivors after successful remission-retrieval therapy, a result that boosted the overall estimated survival at 10 years to 67 percent. This outcome raises a perplexing question: Should intensive, multidrug therapy be administered to all patients at diagnosis or be reserved for those who do not respond to treatment or have a very high risk of early relapse? Preliminary results from the Pediatric Oncology Group indicate that patients with hyperdiploid B-cell-progenitor acute lymphoblastic leukemia, characterized by more than 50 chromosomes per leukemic cell, have a better prognosis and will probably respond well to antimetabolite-based therapy, with its lower risk of genotoxic effects33. We have reported similar findings with different forms of reduced therapy34. Thus, high rates of cure may be attainable in carefully selected subgroups of patients without the use of alkylating agents, topoisomerase II inhibitors, or irradiation.
As treatment plans have evolved toward more intensive systemic chemotherapy tailored to the patient's risk group, preventive meningeal irradiation has been either omitted or reduced in dosage. Indeed, reports from the Pediatric Oncology Group35 and the Children's Cancer Group36 indicate that most children with B-cell-progenitor acute lymphoblastic leukemia can safely receive intrathecal chemotherapy as a primary treatment for subclinical meningeal leukemia. We contend, however, that with today's therapy, selected patients will still benefit from central nervous system irradiation and intrathecal chemotherapy.
Serious late effects could prove to be limiting factors in future applications of intensified therapy for acute lymphoblastic leukemia. The Dana-Farber study under protocol 81-01 (from 1981 through 1985) resulted in excellent control of leukemia (7-year event-free survival, 72 percent),37,38 but the large cumulative dosages of doxorubicin given to high-risk patients were associated with cardiomyopathy in more than half the patients who survived for 5 to 15 years6.
Approximately 80 percent of our 765 long-term survivors are free of obvious health problems related to leukemia or its treatment. The most disturbing delayed complication has been the development of second cancers in 56 patients. The cumulative incidence of solid tumors has not, however, increased with the use of more intensive chemotherapy -- a trend that requires longer follow-up for confirmation. The incidence of secondary acute myeloid leukemia, less than 3 percent overall, increased during eras 3 and 4 at least in part because of the use of epipodophyllotoxins7. We have demonstrated that the risk of epipodophyllotoxin-related acute myeloid leukemia depends largely on the schedule of drug administration7. Hence, for patients at higher risk, we currently use a treatment schedule that has not been associated with an increased risk of secondary leukemia (estimated risk, 1.6 percent).
The development of effective therapy for children with acute lymphoblastic leukemia is one of the undisputed successes of modern clinical hematology. Once a universally fatal disease, acute lymphoblastic leukemia is highly curable today with the use of multidrug regimens. However, certain subgroups of patients, including infants no more than one year old and patients with leukocyte counts above 100,000 per cubic millimeter, continue to fare poorly39. Promising approaches that may further improve the outcome of therapy include strategies that specifically target recurring genetic lesions,40 earlier therapeutic intervention when minimal residual disease is detected in patients with apparently complete responses,41 and modification of treatment on the basis of a better understanding of variables that affect cytotoxicity42.
Supported in part by grants (CA 20180 and CA 21765) from the National Cancer Institute and by the American Lebanese Syrian Associated Charities.
We are indebted to the many physicians and nurses who participated in this research program, especially Drs. Rhomes J. Aur, H. Omar Hustu, Luis Borella (deceased), and Ching-Hon Pui, and Ms. Adynel Wood, P.N.P., for their contributions; and to Mr. John Gilbert, senior editor, for his critical review of the manuscript and for medical editing.
Source Information
From the Departments of Hematology-Oncology (G.K.R., W.M.C.) and Biostatistics (M.L.H.), St. Jude Children's Research Hospital, and the Department of Pediatrics, University of Tennessee, Memphis, College of Medicine (G.K.R., W.M.C.), both in Memphis; the M.D. Anderson Cancer Center, Houston (D.P.); and the Memorial Sloan-Kettering Cancer Center, New York (J.V.S.).
Address reprint requests to Dr. Rivera at 332 N. Lauderdale, P.O. Box 318, Memphis, TN 38101-0318.
References
100 x 109/L. J Clin Oncol 1987;5:1015-1021.
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
