Involved Field Radiation Therapy Following Autologous Stem Cell Rescue for Hodgkin and Non-Hodgkin Lymphoma - Is It Efficacious?.

Introduction: Patients with recurrent or refractory Hodgkin (HL) and non-Hodgkin lymphoma (NHL) treated with high dose chemotherapy and autologous stem-cell rescue (ASCR) commonly relapse (post-ASCR) (70–90%) in sites of previous disease. Although adjuvant involved field radiation therapy (IFRT) to sites of previous recurrence might be expected to enhance local control, translation of this into improved disease-free (DFS) or overall survival (OS) is controversial. We sought to evaluate IFRT following ASCR in terms of patterns of recurrence, OS and DFS.

Methods and Materials: All 281 patients with recurrent or refractory HL and NHL who underwent ASCR between 5/92 and 7/03 were analyzed. Disease stratification at the time of transplant included HL, 24%, aggressive NHL (ANHL) 62%, and indolent NHL (INHL) 14%. Most, 46% underwent ASCR after 1^st relapse, 18% after 2^nd relapse, 4% after 3^rd relapse and 26% had refractory disease at ASCR. IFRT was administered to 129 patients (46%). Physician and patient choice determined which patients received IFRT. Dose ranged from 20–36 Gy depending on response to salvage therapy before ASCR and the presence of visible imaging abnormalities following ASCR. For end point analysis 39 patients (14%) including 11 HL, 23 ANHL, and 5 INHL had insufficient data and were excluded.

Results: Mean follow-up was 3 years (.3 – 12). The median age at ASCR was 45 years (8 – 73). Male to female ratio was 1.3:1. Thirty five percent of patients had prior RT. On univariate analysis, OS and DFS following IFRT for ANHL was superior (or approached this statistically) at 5 years. For HL, improved OS approached significance, whereas DFS did not despite an apparent benefit by disease-free percentages. IFRT appears to be disadvantageous in INHL, but patient numbers were very small. OS and DFS at 5 years are in the table.

On multivariate analysis, for ANHL, IFRT was protective (p = .002, Hazard Ratio = 0.39) and bulky disease was adverse (p = .04, HR = 2.04). For HL, an advantage of IFRT did not reach statistical significance (p = .63, HR = .8). For INHL, IFRT was associated with an inferior outcome (p = .23, HR = 4.9).

Conclusion: Recognizing that bias exists in patient selection for IFRT post-ASCR (in both directions), a survival benefit appears to exist for patients with ANHL, and potentially for those with HL. The absence of a difference in DFS may relate to relapses in other sites or competing events with censored data. Longer follow-up or larger patient numbers are necessary to confirm a long- term improvement. INHL did not benefit from IFRT as might be expected since this group is more likely to have occult disseminated, chemotherapy insensitive disease. Determination of specific patterns of disease recurrence is in progress.