Strong correlation between imaging response and survival among patients with primary central nervous system lymphoma: A secondary analysis of RTOG studies 83-15 and 88-06

Purpose: Primary central nervous system lymphoma (PCNSL) is the brain tumor with the most rapidly increasing incidence, yet little is known about its radiographic response to cranial irradiation. If traditionally used doses o f radiotherapy (-60 Gy) are associated with low rates of complete response and poor survival, then an argument can be made to consider dose escalation of radiotherapy. Alternatively, if poor survival rates are associated with high rates of complete response, there would be no reason to subject patie nts to higher radiation doses with increased risks of treatment-related mor bidity. The purpose of this analysis is to provide a detailed review of res ponse following cranial irradiation. Based on these findings, recommendatio ns are offered for future protocol design. Methods and Materials: Patients were treated on either RTOG 83-15 (whole br ain irradiation to 40 Gy followed by a 20 Gy boost to the tumor plus 2-cm m argin) or RTOG 88-06 (induction cyclophosphamide, doxorubicin, vincristine, dexamethasone [CHOD] followed by whole brain irradiation to 41.4 Gy and an 18 Gy tumor boost). Imaging surveillance (CT, MR) was required following s urgery, prior to the initiation of RT and following completion of RT. Compl ete response referred to the absence of enhancement on follow-up scans in c omparison to the pretherapy study. A tumor size reduction of at least 50% i n the product of the largest cross-sectional diameter and its largest perpe ndicular diameter was scored as a partial response. Results: Seventy-nine patients had scans available for central review. Comp lete response was achieved in 83% and 85% of patients treated on RTOG 83-15 and 88-06, respectively. The rates of partial response (14%, 11%) and radi ographic progression (3%, 4%) also were comparable between the studies. For survival analyses, data were aggregated from the two studies. The 4-year s urvival rates were 24% for complete responders versus 11% for other patient s (p = 0.0006). In multivariate analysis, only complete radiographic respon se (p < 0.0007), and high Karnofsky performance status (KPS greater than or equal to 70) (p < 0.005) were independently associated with increased rate s of 4-year survival. Conclusion: A high rate of complete radiographic response was observed foll owing moderate doses of cranial irradiation (alone or in combination with C HOD chemotherapy). Although complete responders had a statistically signifi cant survival advantage at 4 years when compared with partial responders an d nonresponders, the majority of patients who achieved complete response we re dead of disease by 4 years following treatment. Based on this analysis o f the RTOG database, there is no rationale for radiation dose escalation as a therapeutic strategy to combat PCNSL. Consequently, the radiotherapy com ponent of the current RTOG Phase II trial (RTOG 93-10) now includes relativ ely low total doses of hyperfractionated irradiation for patients without r esidual disease (36 Gy/1.2 Gy, twice a day) as well as a more aggressive ch emotherapy regimen. (C) 2000 Elsevier Science Inc.