|
|
| |||||||||||||||||||||||||||||||||
Background Children and young adults with early-stage non-Hodgkin's lymphoma have an excellent prognosis, but treatment is prolonged and is associated with many side effects. We performed two studies to determine whether therapy could be simplified.
Methods Between 1983 and 1991, we conducted two consecutive trials in children and young adults (age, <21 years) with early-stage non-Hodgkin's lymphoma. In the first trial, patients were treated for 9 weeks with induction chemotherapy consisting of vincristine, doxorubicin, cyclophosphamide, and prednisone, followed by 24 weeks of continuation chemotherapy with mercaptopurine and methotrexate. Half the patients were randomly assigned to receive involved-field irradiation. In the second trial, after the 9 weeks of induction chemotherapy, the patients were randomly assigned to receive 24 weeks of continuation chemotherapy or no further therapy.
Results A total of 340 patients were enrolled in the two trials, 12 of whom did not have complete remissions. One hundred thirteen patients received nine weeks of chemotherapy without radiotherapy, 131 received eight months of chemotherapy without radiotherapy, and 67 received eight months of chemotherapy with radiotherapy. At five years, the projected rates of continuous complete remission were 89, 86, and 88 percent for the three groups, respectively. At five years, event-free survival among the patients with early-stage lymphoblastic lymphoma was inferior to that among the patients with other subtypes of lymphoma (63 percent vs. 88 percent, P<0.001). Continuation therapy was effective only in patients with lymphoblastic lymphoma.
Conclusions A nine-week chemotherapy regimen without irradiation of the primary sites of involvement is adequate therapy for most children and young adults with early-stage, nonlymphoblastic non-Hodgkin's lymphoma.
Because of a desire to spare young patients with early-stage non-Hodgkin's lymphoma from the adverse acute and late effects of therapy, we conducted two consecutive studies designed to maintain the high cure rate while reducing the intensity and duration of therapy. The first trial demonstrated that an eight-month chemotherapy regimen without irradiation of the primary sites of involvement was as effective as chemotherapy plus radiotherapy.6 The second trial was designed to determine whether six months of continuation treatment with two drugs is a necessary component of the regimen. We report the results of this study, as well as the updated results of the first trial.
Methods
Patients
Children and young adults were enrolled in the first trial between April 1983 and May 1987 and were enrolled in the second trial between June 1987 and November 1991 at all institutions that are members of the Pediatric Oncology Group (see the Appendix). Previously untreated children and adults younger than 21 years of age who had biopsy-confirmed non-Hodgkin's lymphoma of any histologic type were eligible for enrollment.
Studies and procedures performed to determine the stage of disease included a history taking and physical examination, blood counts, bone marrow aspiration, lumbar puncture for examination of a cytocentrifuged specimen of cerebrospinal fluid, chest radiography, and radionuclide bone scanning. Additional studies, including computed tomography or magnetic resonance imaging, were performed in some patients, particularly those with primary tumors in the head and neck region or abdomen. The clinical stage was determined according to a staging system described previously.8 Patients meeting the criteria for early-stage disease were eligible for the trials. Specifically, eligible patients were those who had stage I disease, with lymphoma confined to a single nodal or extranodal area (whether resected or not), or stage II disease, with a single tumor and regional lymph-node involvement, two or more sites of disease confined to one side of the diaphragm, or lymphoma of the gastrointestinal tract that had undergone grossly complete (>90 percent) resection. Excluded were patients with more advanced disease (stage III or IV), including those with mediastinal or massive intraabdominal primary tumors, lymphoma on both sides of the diaphragm, or involvement of the bone marrow or central nervous system. The histologic findings were reviewed centrally, and the diagnostic specimens were classified according to the working formulation for clinical use.9
Randomization and Treatment
The design of the two trials is shown in Figure 1. In the first trial,6 patients were randomly assigned to receive either eight months of chemotherapy with radiotherapy or eight months of chemotherapy alone. Radiotherapy, which was administered during induction chemotherapy, consisted of irradiation of the involved field with a total dose of 27 Gy in 18 150-cGy fractions during a period of three and a half weeks. In the second trial, all patients received identical regimens of induction and consolidation chemotherapy for nine weeks without irradiation. Remission was assessed after the nine-week period by clinical examination and repeated staging studies. Patients with complete remissions at this time were then randomly assigned to receive an additional 24 weeks of chemotherapy with mercaptopurine and methotrexate, as administered in the first trial, or no further therapy.
|
|
We had demonstrated in the first trial that irradiation of primary sites can be safely omitted from the treatment of children with early-stage non-Hodgkin's lymphoma.6 The second trial was designed to investigate the contribution of 24 weeks of continuation therapy as used in the earlier trial. In addition, we planned a comparison of the efficacy of eight months of chemotherapy combined with radiotherapy (the standard treatment in the first trial) with nine weeks of chemotherapy alone (the least intensive of the treatments) to assess any loss of overall efficacy resulting from two successive reductions in treatment. This analysis was intended to protect against the possibility that two small decrements in efficacy, neither one significant alone, would result in a significant overall decrement in efficacy when combined.
Since the disease is rare, we elected to include patients from the first trial in the analysis of the second trial and used a study design in which two patients were randomly assigned to nine weeks of chemotherapy without radiotherapy for each patient assigned to eight months of chemotherapy without radiotherapy. Because the study question is negative (Is a less intensive therapy as effective?), we considered a one-sided P value of 0.10 or less in favor of eight months of chemotherapy as evidence of the efficacy of continuation therapy. We planned for a power of 90 percent to detect this difference. Allowing for the data already accrued, assuming a 95 percent rate of continuous complete remission at two years for the patients receiving continuation therapy and an 85 percent rate for those not receiving continuation therapy, and assuming an enrollment of 36 patients per year, we determined that an additional 183 patients would be required, with a final analysis planned two years after the completion of enrollment. The two-to-one randomization scheme increased the power obtained with the usual one-to-one randomization and also increased the power of the secondary comparison (eight months of chemotherapy with radiotherapy vs. nine weeks of chemotherapy without radiotherapy).
The analysis is based on follow-up data as of April 1996. All eligible patients were evaluated according to the treatment assigned at randomization. The dependent variable for efficacy was complete continuous remission — the time from a complete remission to the last contact, a relapse at any site, the development of a second cancer, or death, whichever came first. In the analysis of the outcome according to the histologic subtype of lymphoma, the dependent variable was event-free survival — the time from the initiation of therapy to treatment failure for any reason or the last contact, whichever came first. All comparisons were made with the use of the log-rank test,10 and life tables for continuous complete remission, event-free survival, and overall survival were constructed according to the method of Kaplan and Meier,11 with standard errors calculated as described by Peto et al.12 All comparisons of efficacy were conducted as one-sided tests. Comparisons of outcome according to histologic type were omnibus (all-sided) tests. One-sided confidence intervals for the outcome in the patients treated with nine weeks of chemotherapy without radiotherapy (the least intensive of the treatments) were also determined.
Results
Patients
Between April 1983 and May 1987, 150 patients were enrolled in the first trial, and between June 1987 and November 1991, 205 patients were enrolled in the second trial. Fifteen patients were found to be ineligible on review and were excluded from the analysis: in seven, the diagnosis of non-Hodgkin's lymphoma could not be confirmed, and eight did not have early-stage disease. The 340 eligible patients in the two trials included 243 males (71 percent) and 97 females (29 percent), with a median age of 10 years (range, 16 months to 20 years; mean, 10 years). One hundred thirty-eight patients had stage I non-Hodgkin's lymphoma, and 202 had stage II disease. The distribution of cases according to the primary site of tumor and histologic subtype is shown in Table 2.
|
A total of 311 patients were randomly assigned to treatment in the two studies (Figure 1). In the second trial, 113 patients were randomly assigned to receive nine weeks of chemotherapy alone, and 69 were assigned to receive eight months of chemotherapy. (The imbalance in the treatment assignments reflects the two-to-one randomization scheme.) In the analysis of the outcome, the 62 patients from the first trial who were assigned to treatment with eight months of chemotherapy without radiotherapy were included with the 69 patients assigned to this treatment in the second trial, so that a total of 131 patients were assigned to this treatment. Finally, 67 patients from the first trial were randomly assigned to receive eight months of chemotherapy and radiotherapy, and this treatment served as the standard against which the other treatments were evaluated.
Outcome
The outcome for all patients is shown in Figure 1, Figure 2A, Figure 2B, and Table 3. A total of 54 patients had adverse events. Twelve did not have complete remissions, 38 had recurrent lymphoma after a complete remission, 1 died of sepsis during a complete remission, and 3 had secondary leukemia (acute nonlymphocytic leukemia in 1 patient, biphenotypic leukemia in 1, and secondary T-cell acute lymphoblastic leukemia in 1 patient with small-noncleaved-cell lymphoma). Of the 38 patients who had relapses, 10 had recurrences at the primary site, 7 in the bone marrow, 5 in the central nervous system, 3 in the testis, and 13 in various nodal and extranodal sites.
|
|
Considering that all 142 eligible patients enrolled in the first trial had complete remissions, we were concerned that induction therapy was unsuccessful in 12 of 198 patients in the second trial. An analysis of the characteristics of the patients enrolled in the two trials failed to account for this difference, and it remains unexplained. To assess the likelihood that our conclusions would have been altered had we included in the analysis the 12 patients who did not have complete remissions, we analyzed the data with the 12 patients assigned to treatment in the two-to-one ratio prescribed by our design. Only if all 12 of these patients had been assigned to nine weeks of chemotherapy without continuation therapy (a circumstance with a probability of less than 1 percent) would the P value have been less than 0.10.
When all 340 eligible patients are considered together, the projected event-free survival at five years was 85±2 percent, and only 4 patients (of 211 at risk) had adverse events more than five years after the diagnosis (Figure 3A). Salvage therapy was apparently successful for more than half the patients who did not have complete remissions or had recurrent disease. The projected overall survival nine years after the diagnosis of non-Hodgkin's lymphoma was 92±3 percent (Figure 3B).
|
Side Effects
Fifteen of 131 patients (11 percent) had severe or life-threatening granulocytopenia during continuation therapy, and 2 had severe central nervous system toxic effects (hemiparesis in 1 patient and encephalopathy in the other). Two patients reported severe nausea and vomiting during continuation therapy; severe anemia, stomatitis, ileus, and marked elevations in serum aminotransferase concentrations were rare. Weekly blood counts were required in the patients receiving continuation therapy, and those with lymphomas originating in the head and neck region received four doses of intrathecal chemotherapy, in addition to the daily oral mercaptopurine and weekly methotrexate. The patients who received nine weeks of chemotherapy alone were spared these side effects and the additional diagnostic and therapeutic interventions associated with continuation chemotherapy.
Discussion
The successful treatment of non-Hodgkin's lymphoma in children and young adults has evolved empirically since the early 1970s, when most patients with the disorder died. Although the use of intensive multimodal regimens proved to be successful in improving the prognosis for such patients,1,2,4,5,13,14,15,16,17 it became evident that patients with an early stage of disease had a more favorable prognosis than those with more advanced disease and might be curable with a reduction in treatment.3,6,7 We performed studies to identify the components of therapy that are necessary for a cure in patients with early-stage non-Hodgkin's lymphoma, allowing us to eliminate the components that provide no benefit and only contribute to toxic effects. We have been cautious in choosing the modifications to be tested and conservative in designing our studies in an effort not to compromise the efficacy of therapy.
The results of our latest study extend the observation made by us and other investigators3,6,7,18 that patients with early-stage non-Hodgkin's lymphoma can be treated successfully with a short course of chemotherapy of lessened intensity without irradiation of the primary sites of involvement and without continuation therapy. This treatment strategy is likely to reduce the acute side effects and financial and psychological costs of therapy, as well as the risk of late side effects, although further follow-up is necessary to confirm that the excellent outcome will be sustained and that the incidence of late side effects will be reduced. It might be argued that the elimination of 24 weeks of therapy with mercaptopurine and methotrexate, as well as four doses of intrathecal chemotherapy, represents a very small gain, because these drugs are associated with few late effects. However, the continued use of therapy that cannot be demonstrated to confer a benefit is not justifiable.
Not all patients benefited equally from our approach. The event-free survival of the patients with early-stage lymphoblastic lymphoma — a histologic subtype that accounts for approximately 15 percent of early-stage non-Hodgkin's lymphomas in children and young adults — was inferior to that of the patients with other subtypes of lymphoma. Moreover, the patients with early-stage lymphoblastic lymphoma — unlike those with other subtypes — appeared to benefit from continuation therapy. Although salvage therapy for the patients with early-stage lymphoblastic lymphoma who have relapses appears to be excellent (at least in the short term), further follow-up is necessary to determine whether they are cured. Other investigators have also recognized that patients with early-stage lymphoblastic lymphoma have a less favorable prognosis than those with other histologic subtypes and have treated such patients with more intensive therapies designed for advanced-stage disease.7,17,19,20 However, the overall outcome does not appear to be much improved with this approach.
Supported in part by grants (CA-33603, CA-29139, CA-31566, and CA-30969) from the National Cancer Institute.
We are indebted to the physicians and nurses at the participating institutions who cared for the patients and to the clinical research associates who assisted with the submission and analysis of the data.
Source Information
From the Departments of Pediatrics and Radiation Oncology, Stanford University School of Medicine and the Lucile Salter Packard Children's Hospital, Stanford, Calif. (M.P.L., S.S.D.); the Department of Statistics, University of Florida, and the Pediatric Oncology Group Statistical Office, Gainesville, Fla. (J.J.S.); the Department of Pathology, Saint Jude Children's Research Hospital and the University of Tennessee, Memphis (C.W.B.); and the Department of Pediatrics, Northwestern University School of Medicine and Children's Memorial Hospital, Chicago (S.B.M.).
Address reprint requests to Dr. Link at the Department of Pediatrics, Stanford University School of Medicine, 300 Pasteur Dr., Stanford, CA 94305-5208.
References
The following members of the Pediatric Oncology Group participated in the study: Alberta Children's Hospital, Calgary, Alta., Canada; All Children's Hospital, St. Petersburg, Fla.; Baylor College of Medicine, Houston; Boston Floating Hospital, Boston; Bowman Gray School of Medicine, Winston-Salem, N.C.; Carolinas Medical Center, Charlotte, N.C.; Children's Hospital of San Diego, San Diego, Calif.; Children's Hospital of Michigan, Detroit; Children's Hospital of the Greenville Hospital System, Greenville, S.C.; Children's Memorial Hospital, Chicago; Children's Hospital of New Orleans, New Orleans; Cook–Fort Worth Children's Medical Center, Fort Worth, Tex.; Cross Cancer Institute, Edmonton, Alta., Canada; Dana–Farber Cancer Institute, Boston; Duke University Medical Center, Durham, N.C.; East Carolina University School of Medicine, Greenville, N.C.; Emory University School of Medicine, Atlanta; Fairfax Hospital, Fairfax, Va.; Hackensack Medical Center, Hackensack, N.J.; Johns Hopkins Hospital, Baltimore; Kaiser Permanente at San Diego, San Diego, Calif.; M.D. Anderson Cancer Center, Orlando, Fla.; Maine Children's Hospital, Portland; Massachusetts General Hospital, Boston; McGill University, Montreal; Medical College of Virginia, Richmond; Medical University of South Carolina, Charleston; Midwest Children's Cancer Center, Milwaukee; Mount Sinai Medical School, New York; Naval Medical Center, Portsmouth, Va.; Nemours Children's Clinic, Jacksonville, Fla.; Presbyterian Hospital, Charlotte, N.C.; Rhode Island Hospital, Providence; Roswell Park Memorial Institute, Buffalo, N.Y.; Sacred Heart Hospital, Pensacola, Fla.; Scott and White Memorial Hospital, Temple, Tex.; St. Christopher's Hospital for Children, Philadelphia; St. Francis Regional Medical Center, Wichita, Kans.; St. Jude Children's Research Hospital, Memphis, Tenn.; Stanford University School of Medicine, Stanford, Calif.; State University of New York at Syracuse, Syracuse; Swiss Pediatric Oncology Group, Bern, Geneva, and Lausanne, Switzerland; University of Alabama, Birmingham; University of Arkansas, Little Rock; University of California at Davis, Sacramento; University of California at San Diego, San Diego; University of Florida, Gainesville; University of Kansas Medical Center, Kansas City; University of Miami School of Medicine, Miami; University of Mississippi Medical Center, Jackson; University of Missouri Health Sciences Center, Columbia; University of New Mexico School of Medicine, Albuquerque; University of Oklahoma Health Sciences Center, Oklahoma City; University of Puerto Rico, San Juan; University of Rochester Medical Center, Rochester, N.Y.; University of South Alabama, Mobile; University of Texas Southwestern Medical School, Dallas; University of Texas at Galveston, Galveston; University of Vermont College of Medicine, Burlington; Walter Reed Army Medical Center, Washington, D.C.; Warren Clinics, Tulsa, Okla.; Washington University Medical Center, St. Louis; West Virginia University at Charleston, Charleston; West Virginia University at Morgantown, Morgantown; Wichita Community Clinical Oncology Program, Wichita, Kans.; and Yale University School of Medicine, New Haven, Conn.
| |||||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | 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. |
Previous
