The limits of limited stage lymphoma.

In the wildly popular novel by Dan Brown, Angels and Demons, Robert Langdon, an academician from Harvard University, is inadvertently swept up in a plot to use antimatter to destroy civilization. Langdon sleuths unpublished manuscripts in the Vatican archives for clues to find the culprit and interrupt the sinister plot. In this issue of the Journal of Clinical Oncology, Horning et al dust off data from a 1984 Eastern Cooperative Oncology Group (ECOG) study in an attempt to find a lead to prevent death from lymphoma. The plan almost works. ECOG 1484 attempted to determine whether low-dose consolidative radiotherapy (RT) adds efficacy to systemic chemotherapy (cyclophosphamide, doxorubicin, vincristine, prednisone [CHOP]) in patients with limited-stage, aggressive-histology, non-Hodgkin’s lymphoma (NHL). Of 399 patients registered, 352 (88%) were eligible and assessable and were randomly assigned to receive low-dose RT or observation alone after eight cycles of CHOP chemotherapy (CHOP[8]). However, only 282 patients (71% of initial registrations) were registered for the consolidation step, of which 243 (61%) were subsequently eligible and assessable. Among these 243 patients were 172 complete responders (or 43% of the initial patients registered) who then actually received RT (93 patients) or observation (79 patients). An intent-to-treat analysis was also performed to partially mitigate any potential bias. Efficacy was measured using a variety of end points, including disease-free survival; time to lymphoma progression, which censored the data for cause of death; failure-free survival, which included any cause of death; and overall survival. In this study, low-dose consolidative RT did not affect survival, whereas there was some marginal improvement in disease-free survival, time to lymphoma progression, and failure-free survival. From a patient’s perspective, efficacy is measured by relief of symptoms or prolongation of survival. In ECOG study 1484, RT did neither, as all patients randomized had already achieved a complete remission (CR) (with any symptoms from lymphoma presumably relieved by the chemotherapy), and RT had no effect on survival. Importantly, the authors do not advocate combined treatment in patients who achieve CR with chemotherapy alone. This study is old—a fact that the authors note. But all published studies of limited-stage, aggressive histology NHL are relatively old. Having reported ECOG 1484, however, the authors raise an important and timely issue as investigators plan new studies of limited NHL. “Limited disease” must be more precisely defined. Limited disease and its pseudonyms, “early-stage disease,” “low-stage disease,” and “localized disease,” have been used as definitions in every published study to date. It is, therefore, no wonder that Horning et al remark that, “Differences in outcomes of various studies are always challenging to explain.” There is extreme heterogeneity within the group of patients described as having “limited-stage” lymphoma, and not only are comparisons across this heterogeneous population difficult, but treatments designed for one subgroup may not be appropriate for another subgroup. Major studies of limited disease, published or otherwise presented, have selected patients from the relatively broad stage I and stage II categories based on the Ann Arbor staging system. That staging system, designed primarily for use in Hodgkin’s lymphoma, has prognostic value but lacks precision when applied to the spectrum of diseases comprising non-Hodgkin’s lymphomas. Most of the differences in comparing the results among studies can be attributed to the imprecise nature and erratic application of this staging system. Stage I-II aggressive histology of NHL is too broad a category; thus, outcome varies considerably. However, both the current study and past reports of stage I-II disease are remarkably consistent when adjusted for stage. Consequently, consistent application of staging criteria in the future design of studies intended to define therapy for limited disease would likely improve comparability and allow cross-comparisons. Stage I-II aggressive histology of NHL can be easily divided into three subgroups using different average outcome, based on survival. Importantly, each of these three subgroups likely requires different therapeutic strategies to JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 22 NUMBER 15 AUGUST 1 2004