Abstract
Introduction: CNS relapses with DLBCL tend to be uncommon. The risk increases with certain DLBCL types and certain risk features. There is no consensus regarding the optimal approach to CNS prophylaxis for high-risk DLBCL patients. While some favor prophylaxis with high-dose systemic therapy or intrathecal chemotherapy (IT), some experts advocate combining both approaches. In the absence of randomized trials, the role of CNS chemotherapy prophylaxis remains controversial in DLBCL with contradictory comparative studies. The purpose of this meta-analysis is to evaluate the impact of CNS prophylaxis approaches on the clinical outcomes of high-risk DLBCL.
Methods: A review of the medical literature was conducted using online databases. Inclusion criteria consisted of DLBCL diagnosis, English language, IT or high-dose systemic prophylactic chemotherapy, and comparative studies reporting CNS recurrence rates. A frequentists and Bayesian network meta-analyses were conducted using the netmeta package and random-effects model.
Results: Twenty-one comparative studies with a total of 11,507 patients were included and analyzed. The relative risk (RR) of CNS recurrence was not statistically different between IT prophylaxis, high-dose chemotherapy, and no prophylaxis. However, the combination of IT prophylaxis and high-dose chemotherapy was found to be significantly associated with a reduced RR of CNS recurrence when compared to IT prophylaxis (RR=0.28, 95%CI 0.13-0.58), high-dose chemotherapy (RR=0.34, 95%CI 0.14-0.80), and no prophylaxis (RR=0.41, 95%CI 0.19-0.87).
Conclusions: This network meta-analysis is the first to compare the different CNS prophylactic approaches. It indicates that IT prophylaxis and high-dose chemotherapy each were not better than no prophylaxis. However, the combination of IT prophylaxis with high-dose chemotherapy was significantly superior to each approach alone as well as no prophylaxis. In the absence of randomized clinical trials, this network meta-analysis represents the most compelling data supporting the use of CNS prophylaxis in patients with high-risk DLBCL.
No relevant conflicts of interest to declare.
