Stage III follicular lymphoma: Long-term follow-up and patterns of failure

Purpose: To analyze the long-term outcomes and pattern of failures for Stage III follicular lymphomas. Methods and Materials: A retrospective review of all patients with Stage III follicular lymphoma presented to our institution between 1978 and 1993 was performed. One hundred ten patients were eligible and form the basis of this analysis. Fifty-seven patients were male. The median age was 57 years (range: 21-82 years). The treatments were as follows: chemotherapy alone (CTX), 39 patients; combined modality with chemotherapy and radiation therapy (CMT), 69; radiation therapy alone, 2. Radiation therapy fields were as follows: regional, 13 patients; extended, 6; subtotal, 44; and central lymphatic, 8. The median number of chemotherapy cycles was 8, and 98 patients received doxorubicin-containing regimens. Enough information was available to calculate the International Prognostic Index for 107 patients. The following prognostic factors were examined for predictive value in overall survival (OS) and freedom from progression (FFP) by univariate and multivariate analyses: International Prognostic Index, gender, lactate dehydrogenase (LDH) (normal vs. elevated), B symptoms, performance status, β-2 microglobulin, presence or absence of a bulky disease, age (≤ vs. >60 years), number of sites of involvement, treatment (CTX vs. CMT), and pathology. To minimize patient selection biases given the nature of the retrospective analysis, the patterns of relapse were analyzed only for the patients who achieved a complete response. Results: The median follow-up for the alive patients was 9.5 years (range: 1.1-19.5 years). Complete response was achieved in 80 patients: 24 of 39 patients in the CTX group (62%), 54 of 69 patients in the CMT group (78%), and 2 of 2 patients in the radiation therapy alone group. The actuarial 5- and 10-year OS rates were 65% and 42%, respectively. The 5- and 10-year FFP was 42% and 26%, respectively. Significant prognostic factors by multivariate analyses were age and LDH for OS and LDH for FFP. For complete responders, 5-year freedom from recurrence in the original sites of involvement (with or without recurrence in the new sites) was 43% for CTX and 60% for CMT patients (p = 0.03, Wilcoxon). Five-year freedom from isolated recurrence in the original sites of involvement was 60% for CTX and 69% for CMT patients (p = 0.17). The 5-year overall FFP was 43% for CTX patients and 56% for CMT patients (p = 0.06). Conclusion: Combined modality treatment seems to give an advantage in terms of disease control in the primary sites compared to chemotherapy alone, though the advantages in OS and FFP were not statistically significant in our patient population. By multivariate analyses, LDH was a significant prognostic indicator for OS and FFP, whereas age was for OS.