|
|||||
|
|
||||||
© 1999 American Society for Clinical Oncology Low-Grade Stage III-IV Follicular Lymphoma: Multivariate Analysis of Prognostic Factors in 484 PatientsA Study of the Groupe d'Etude des Lymphomes de l'AdulteFrom the Groupe d'Etude des Lymphomes de l'Adulte, Créteil, Paris, Nantes, Rouen, Caen, Lyon, and Le Mans, France, and Yvoir, Belgium. Address reprint requests to Didier Decaudin, MD, PhD, Department of Hematology, Institut Curie, 26 rue d'Ulm, 75.248 Paris cedex 05, France; email didier.decaudin{at}curie.fr
PURPOSE: To identify the prognostic factors that influence overall survival (OS) in patients with stage III-IV follicular lymphomas and evaluate the clinical usefulness and the prognostic value of the International Prognostic Index (IPI). PATIENTS AND METHODS: Four hundred eighty-four patients with Ann Arbor stage III-IV follicular lymphomas treated in two phase III trials from 1986 to 1995 were screened for this study. All histologic slides were reviewed by two hematopathologists. The influence of the initial parameters on survival was defined by univariate (log-rank test) and multivariate (Cox model) analyses.
RESULTS: The poor prognostic factors for OS (age > 60 years, "B" symptom(s), CONCLUSION: Three prognostic factors for poor OS were identified: B symptoms, age greater than 60 years, and at least three nodal sites greater than 3 cm. The IPI was prognostic for OS, but in this population, a very low number of patients belonged to the high-risk groups.
WITH A MEDIAN survival of 10 years,1,2 follicular lymphomas (FLs) are widely considered as an indolent subtype of non-Hodgkin's lymphomas (NHL). However, some patients with FL have poor progression-free survival (PFS) and overall survival (OS),3,4 suggesting that initial characteristics that are significant predictors of poor PFS or OS may exist. Such parameters have been identified in patients with aggressive NHL and have been used to define the International Prognostic Index (IPI), now widely accepted.5 New therapeutic approaches, such as intensified chemotherapy with autologous stem-cell transplantation and immunotherapy using monoclonal antibodies, could be tested for initial treatment of FL with adverse prognostic factors. The autologous stem-cell transplantation therapy seems efficient but possesses significant toxicity,6 and the impact on survival of monoclonal antibodies has yet to be elucidated.7 Therefore, such treatments cannot be offered to all patients with FL, and prognostic factors must be determined. The aims of this study were to identify the prognostic factors that influence survival in a large group of patients with FL and evaluate the clinical usefulness and prognostic value of the IPI in these patients. To select patients for whom innovative therapies could be warranted and to homogenize the inclusion criteria for future trials, we studied prognostic factors in 484 patients with newly diagnosed Ann Arbor stage III-IV FL treated in two phase III trials from 1986 to 1995, which have been reported previously.8,9
Eligibility Criteria All patients who were included in two open-label, phase III trials (Groupe d'Etude des Lymphomes Folliculaires [GELF] 86) between October 1986 and May 1995 in 40 centers were selected for the present study. Patients were eligible for these trials if they had small-cell (< 5% large noncleaved cells) or mixed-cell (5% to 50% large noncleaved cells) follicular NHL confirmed by nodal biopsy, were younger than 70 years of age, had measurable disease, and presented with Ann Arbor stage II, III, or IV disease. The patients were treated according to the tumor burden status. A large tumor burden was determined by at least one of the following parameters: any nodal or extranodal tumor mass with a diameter of more than 7 cm (nodal sites were defined according to the Ann Arbor system); involvement of at least three nodal sites (superficial or not), each of which had a diameter of more than 3 cm; "B" symptom(s); splenic enlargement with inferior margin below the umbilicus line; compression syndrome defined by ureteral, orbital, or gastrointestinal compression, by pleural or peritoneal serous effusion (irrespective of cell content), or by epidural involvement; and leukemia. Liver involvement was defined by visualization of tumor mass(es) by computed tomography of the abdomen and/or by a liver biopsy. The two trials were approved by an ethics committee, and all patients enrolled gave informed consent. All slides were re-examined by two hematopathologists to confirm the diagnosis of FL and to ensure a uniform system of cytologic subclassification. Because only few patients in these trials had stage II disease and were not representative of localized follicular NHL, only stage III-IV patients with a confirmed diagnosis of FL were included in the study.
Staging and Initial Treatment Patients with a low tumor burden (group 1) were randomly assigned to one of three treatment arms: no treatment until progression (arm 1), prednimustine (200 mg/m2 orally for 5 consecutive days per month for 18 months; arm 2), or interferon alfa-2b (Intron A; Schering-Plough, Kenilworth, NJ; 5 million units M/d subcutaneously for 3 months, then 5 MU three times per week for 15 months; arm 3). Patients with a high tumor burden (group 2) received the CHVP regimen, which consisted of cyclophosphamide 600 mg/m2, doxorubicin 25 mg/m2, and teniposide 60 mg/m2, all given intravenously on day 1, and prednisone 40 mg/m2, given orally on days 1 to 5. Treatment consisted of an induction phase of six courses administered monthly, followed by a maintenance phase for responders and patients with stable disease that consisted of one cycle every 2 months for 1 year. Before chemotherapy began, the patients were randomly assigned to receive either chemotherapy alone or chemotherapy plus concomitant subcutaneous interferon alfa-2b 5 MU three times weekly for 18 months.
Analysis of Prognostic Factors Survival was estimated using the Kaplan-Meier method. The influence of the initial parameters on PFS and OS was defined by univariate analysis (log-rank test). The most significant covariates (P < .1) were introduced in a stepwise Cox model. Results are presented in terms of the odds ratio for covariates effect, the 95% confidence limits for the odds ratio, and associated P value.
Patient Characteristics From October 1986 to May 1995, 556 patients with newly diagnosed FL were entered onto the GELF 86 trials. Review of the pathologic results did not confirm the diagnosis of low-grade follicular NHL in 25 patients. Mantle-cell lymphoma and follicular large-cell lymphoma were excluded from the analysis. Eight patients were not included in the analysis because of missing data, and 39 stage II cases were not considered for the study. Therefore, the analysis included 484 stage III-IV patients, 178 (37%) of whom fulfilled the criteria of low tumor burden. Initial characteristics of the patients are listed in Table 1.
The IPI was applied in 457 patients for whom lactate dehydrogenase level data were available. In this study, the score was quoted between 1 to 5 because of the stage III-IV patients' selection. Two hundred twenty-two patients (49%) belonged to the low-risk group, 177 (39%) to the lowintermediate-risk group, 49 (11%) to the intermediatehigh-risk group, and nine (2%) to the high-risk group.
Patient Outcome With a median follow-up duration of 55 months (range, 5 to 126 months), PFS and OS rates were 32% and 69% at 5 years, respectively (Fig 1). The 5-year OS rate of the patients in groups 1 and 2 was 81% and 63%, respectively (P = .0003). In group 1, the OS rate was 84%, 72%, and 86% for arms 1, 2, and 3, respectively (difference not significant). In group 2, the OS rate was 55% and 69% in arms 1 and 2, respectively (P = .003).
Analysis of Prognostic Factors Multivariate analysis was conducted on two thirds of patients for whom all factors significant for poor OS in univariate analysis were known (299 cases). Three factors remained significant for poor OS: presence of B symptoms (risk ratio = 1.8; Fig 2), age greater than 60 years (risk ratio = 1.6; Fig 3), and at least three nodal sites greater than 3 cm (risk ratio = 1.71; Fig 4 and Table 2).
To evaluate the impact of treatment, a multivariate analysis was performed on patients with a high tumor burden who received the CHVP regimen plus interferon alfa-2b (arm B, group 2) using the previously included characteristics. The three preceding factors remained significant for poor OS (data not shown). To identify different prognostic groups of patients with FL, each of the three factors significant for poor OS in the multivariate analysis was quoted 1 point when present and 0 otherwise. A score between 0 and 3 was thus defined for each patient. This model was validated on the target population, which represented the remaining third of the patients entered onto the study (158 cases). Seventy-three patients (46%) had a score of 0, 64 (41%) had a score of 1, and 21 (13%) had a score of 2 to 3. The OS at 5 years was 74% (95% confidence interval, 61% to 87%), 66% (95% confidence interval, 53% to 79%), and 45% (95% confidence interval, 20% to 70%) for patients with a score of 0, 1, and 2 to 3, respectively (P = .008).
Analysis of the Prognostic Value of the IPI
The multivariate analysis conducted on the stage III-IV patients of the GELF trials between 1986 and 1995 (after histologic review and confirmation of the initial diagnosis of FL) identified three prognostic factors that influence survival: presence of B symptoms, age greater than 60 years, and at least three nodal sites greater than 3 cm. A large number of published studies have reported several prognostic factors for poor survival in patients with low-grade lymphoma. The majority of these series is heterogenous in terms of histologic characteristics, and few of the publications cover a large cohort of patients. Therefore, the power of those prognostic analyses is limited necessarily. In our series, the large number of patients with only FL validates the accuracy of the three prognostic factors identified by multivariate analysis. Several studies already have shown those factors to be significant with respect to poor OS (Table 3).3,4,10-20 Particularly, the three largest series showed the negative prognostic value of old age and the presence of B symptoms on OS.18-20 The presence of at least three nodal sites greater than 3 cm, which has never been reported as prognostic factor for low OS, was found to adversely influence OS in patients with a high tumor burden. Despite the fact that many variables related to stage IV disease, including liver involvement, multiple extranodal sites, and a compression syndrome, were highly significant for poor OS in the univariate analysis, we observed that none of these criteria remained significant in the multivariate analysis. Moreover, in contrast with several publications,10-13,19,20 it is interesting to note that no cytopathologic feature was prognostic for short OS, but follicular large-cell lymphomas were not included in this trial.
Patients with FL for whom curative radiotherapy could be proposed, ie, localized NHL, were excluded from the analysis. Moreover, our choice to select stage III-IV cases was based on three reasons: (1) the fact that stage I patients were not included in the GELF trials; (2) the low proportion of stage II patients (7%); and (3) the fact that neither intensified chemotherapy nor immunotherapy could at present be proposed to patients with newly diagnosed Ann Arbor stage I-II FL. Undoubtedly, our selection, and the fact that patients older than 70 years were not included in the trials, slightly attenuates the significance of the analysis. Moreover, in our series we were unable to evaluate the prognostic value of elevated serum beta-2 microglobulin level. Considering the fact that patients included in the two phase III trials have been treated according to five different strategies, it was important to evaluate the prognostic impact of assigned treatment. The outcomes from the two randomized trials were reported recently.8,9 We performed a multivariate analysis on patients with a high tumor burden status who received the CHVP regimen plus interferon alfa-2b and observed that the three prognostic factors previously identified remained significant for poor OS. This second multivariate analysis confirmed the results of the first multivariate analysis realized on the overall study population and showed that the three prognostic factors did not disappear with the addition of interferon alfa-2b. The IPI was applied to our patient population and shown to be significantly prognostic for PFS (P = .002) and OS (P = .0001). With the exception of the analysis by Avilès et al,21 all retrospective studies have shown that the IPI, which initially was defined for intermediate- and high-grade NHL, can be used to separate low-grade FL patients into subgroups with different OS rates (Table 4).21-26 However, in this population of FLs, the use of the IPI is limited by the very low number of FL patients with a poor performance status at diagnosis; the variability of initial staging, eg, liver or Waldeyer's ring biopsies, gastrointestinal tract endoscopy, and, thus, the number of extranodal sites detected; and, in almost all series, the limited number of patients in the intermediatehigh- and high-risk groups. Indeed, only 11% and 2% of our patients had scores of 3 and 4 to 5, respectively. The very low number of patients with a poor performance status could be explained by the exclusion of patients older than 70 years and the fact that this analysis was performed on patients included in two prospective randomized trials, unlike most other studies, which are retrospective. In our series, the clinical utility of the IPI is decreased by the fact that some factors used in the model were not significant for poor OS in univariate analysis (Eastern Cooperative Oncology Group performance status > 1 and elevated serum lactate dehydrogenase level), that only one criteria of the IPI (age > 60 years) was prognostic for poor OS in multivariate analysis, and that certain parameters were stronger for adverse prognosis than criteria included in the model (presence of B symptoms and at least three nodal sites > 3 cm).
Using the three prognostic factors reported here, we defined a score for each patient between 0 and 3. This score was highly prognostic for OS (P = .008); however, the use of this score is limited by the fact that a low proportion of patients (13%) presented with two or three prognostic factors at diagnosis. This result, based on a large cohort of patients with FL, shows that it remains difficult to discriminate patients with an aggressive FL at diagnosis for which new therapeutic approaches are warranted. In conclusion, in a large group of patients with FL, three prognostic factors for poor OS were identified: B symptoms, age greater than 60 years, and at least three nodal sites greater than 3 cm. The IPI was prognostic for OS, but, in our study population, a very low number of patients belonged to the intermediatehigh- and high-risk groups. Large-scale multicenter trials are necessary to define a new predictive model to discriminate subpopulations of FL patients and, in particular, those for whom intensive or costly therapies are warranted.
Participants in the Groupe d'Etude des Lymphomes de l'Adults Study The following clinicians and pathologists actively participated in this study: C. Allard, B. Audhuy, J.C. Barats, Y. Bastion, E. Baumelou, J. Benevent, F. Berger, A.M. Bertrand, M. Blanc, D. Bordessoule, A. Bosly, R. Bouabdallah, J. Brière, J.C. Brouet, S. Castaigne, T. Caulet, B. Christian, J.P. Clauvel, B. Coiffier, F. Cosnard, M.F. d'Agay, J. d'Anjou, L. Degos, A. Delannoy, T. De Revel, M. Diviné, C. Doyen, P. Dubigeon, G. Dupont, E. Dupuy, C. Duval, B. Duvert, D. Eychenne, J.P. Fermand, A. Ferrant, M. French, E. Frilay, J. Gabarre, F. Gaillard, F. Galateau, G. Garnier, P. Gaulard, C. Gisselbrecht, J. Hamels, J.L. Harousseau, M. Hayat, A. Herrera, Y. Kerneis, G. Laurent, C. Lavignac, F. Lejeune, G. Lepeu, M. Leporrier, G. Marit, J.P. Marolleau, C. Martin, C. Marty-Double, G. Merignargues, J.L. Michaux, J.M. Miclea, N. Milpied, P. Mineur, P. Moreau, M. Moriceau, F. Morvan, H. Noël, E. Oksenhendler, P. Oriol, P.Y. Péaud, M. Peuchmaur, B. Pignon, G. Pinon-Netter, J.F. Ramée, M. Raphaël, M. Raymond-Gelle, O. Reman, M. Renoux, F. Reyes, R. Riou, J.F. Rossi, F. Rumilly, B. Salles, G. Salles, J.F. Schwed, C. Sebban, J. Simony-Lafontaine, N. Straetmans, I. Tabah, B. Taine, G. Tertian, A. Thyss, H. Tilly, A.M. Touchais, P. Travade, and J. Troncy.
We thank Dr Catherine Sebban and Saadia Houga for technical assistance and Catherine Balmale for statistical analysis.
1. Bastion Y, Berger F, Bryon PA, et al: Follicular lymphomas: Assessment of prognostic factors in 127 patients followed for 10 years. Ann Oncol2:123-134, 1991 (suppl 2)
2.
Rosenberg SA: The low-grade non-Hodgkin's lymphomas: Challenges and opportunities. J Clin Oncol3:299-310, 1985
3.
Leonard RCF, Hayward RL, Prescott RJ, et al: The identification of discrete prognostic groups in low-grade non-Hodgkin's lymphoma: The Scotland and Newcastle Lymphoma Group Therapy Working Party. Ann Oncol2:655-662, 1991 4. Romaguera JE, Mc Laughlin P, North L, et al: Multivariate analysis of prognostic factors in stage IV follicular low-grade lymphoma: A risk model. J Clin Oncol9:762-769, 1991[Abstract]
5.
The International Non-Hodgkin's Lymphoma Prognostic Factors Project: A predictive model for aggressive non-Hodgkin's lymphoma: The International NHL Prognostic Factors Project. N Engl J Med329:987-994, 1993
6.
Freedman AS, Takvorian T, Neuberg D, et al: Autologous bone marrow transplantation in poor-prognosis intermediate-grade and high-grade B-cell non-Hodgkin's lymphoma in first remission. J Clin Oncol11:931-936, 1993 7. Maloney DG, Grillo-Lopez AJ, Bodkin DJ, et al: IDEC-C2B8: Results of a phase I multiple-dose trial in patients with relapsed non-Hodgkin's lymphoma. J Clin Oncol15:3266-3274, 1997[Abstract]
8.
Brice P, Bastion Y, Lepage E, et al: Comparison in low-tumor-burden follicular lymphomas between an initial no-treatment policy, prednimustine, or interferon alfa: A randomized study from the Groupe d'Etude des Lymphomes Folliculaires. J Clin Oncol15:1110-1117, 1997 9. Solal-Céligny P, Lepage E, Brousse N, et al: Doxorubicin-containing regimen with or without interferon alfa-2b for advanced follicular lymphomas: Final analysis of survival and toxicity in the Groupe d'Etude des Lymphomes Folliculaires 86 trial. J Clin Oncol16:2332-2338, 1998[Abstract] 10. Rudders RA, Kadin M, De Lellis RA, et al: Nodular non-Hodgkin's lymphoma: Factors influencing prognosis and indications for aggressive treatment. Cancer43:1643-1651, 1979[Medline] 11. Anderson T, De Vita VT, Simon R, et al: Malignant lymphoma II: Prognostic factors and response to treatment of 473 patients at the National Cancer Institute. Cancer50:2708-2721, 1982[Medline]
12.
Gallacher CJ, Gregory WM, Jones AE, et al: Follicular lymphoma: Prognostic factors for response and survival. J Clin Oncol4:1470-1480, 1986 13. Ezdinli EZ, Costello WG, Kucuk O, et al: Effect of the degree of nodularity on the survival of patients with nodular lymphomas. J Clin Oncol5:413-418, 1987[Abstract] 14. Steward WF, Crowther D, Mc William LJ, et al: Maintenance chlorambucil after CVP in the management of advenced stage, low-grade histologic type non-Hodgkin's lymphoma. Cancer61:441-447, 1988[Medline] 15. Lepage E, Sebban C, Gisselbrecht C, et al: Treatment of low-grade non-Hodgkin's lymphoma: Assessment of doxorubicin in a controlled trial. Hematol Oncol8:31-39, 1990[Medline] 16. Litam P, Swan F, Cabanillas F, et al: Prognostic value of serum beta-2 microglobulin in low-grade lymphoma. Am J Med114:855-860, 1991 17. Soubeyran P, Eghbali H, Bonichon F, et al: Low-grade follicular lymphomas: Analysis of prognosis in a series of 281 patients. Eur J Cancer27:1606-1613, 1991 18. Cameron DA, Leonard RCF, Mao J-H, et al: Identification of prognostic groups in follicular lymphoma. Leuk Lymphoma10:89-99, 1993[Medline] 19. Gospodarowicz MK, Bush RS, Brown TC, et al: Prognostic factors in nodular lymphomas: A multi-variate analysis based on the Princess Margaret Hospital experience. Int J Radiat Oncol Biol Phys10:489-492, 1994 20. Denham JW, Denham E, Dear KB, et al: The follicular non-Hodgkin's lymphomas II: Prognostic factorsWhat do they mean? Eur J Cancer 32A:480-490, 1996
21.
Avilès A: The International Index is not useful in the classification of low-grade lymphoma. J Clin Oncol12:2766-2770, 1994 22. Bastion Y, Coiffier B: Is the International Prognostic Index for aggressive lymphoma patients useful for follicular lymphoma patients? J Clin Oncol12:1340-1342, 1994[Medline]
23.
Martin AR, Weisenburger DD, Chan WC, et al: Prognostic value of cellular proliferation and histologic grade in follicular lymphoma. Blood85:3671-3678, 1995
24.
Hermans J, Krol ADG, van Groningen K, et al: International Prognostic Index for aggressive non-Hodgkin's lymphoma is valid for all malignancy grades. Blood86:1460-1463, 1995 25. Lopez-Guillermo A, Montserrat E, Bosch F, et al: Applicability of the International Prognostic Index for aggressive lymphomas to patients with low-grade lymphoma. J Clin Oncol12:1343-1348, 1994[Abstract]
26.
The non-Hodgkin's Lymphoma Classification Project: A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. Blood89:3909-3918, 1997 Submitted September 14, 1998; accepted March 31, 1999.
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||
|
Copyright © 1999 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|