Outcomes of Patients (pts) with Primary CNS Lymphoma (PCNSL) Treated with the De Angelis-Lite Protocol (Modified DeAngelis Protocol with Omission of Consolidative Radiotherapy): The Roswell Park Comprehensive Cancer Center (RPCCC) Experience

Introduction: High dose methotrexate (HDMTX) based protocols are commonly used for the frontline treatment of PCNSL, however, outcomes remain suboptimal. The modified DeAngelis regimen was widely adopted based on an overall response rate (ORR) of 94%, median progression-free survival (PFS) of 24 months (mo), and overall survival (OS) of 36.9 mo (DeAngelis et al, JCO. 2002). Tolerability was an issue, especially in older adults, with 15% pts experiencing severe neurologic toxicity. Intending to decrease toxicities, we modified the protocol to omit RT while keeping the dose of methotrexate at 2.5 g/m2 [DeAngelis-Lite,(DLR)]. We report our outcomes with this regimen and describe factors associated with response and toxicities.
Methods: All immunocompetent PCNSL pts treated at RPCCC from 01/2002 to 12/2019 were identified retrospectively. Baseline clinical, pathological, and outcomes data was collected. Comparisons were made using the Mann-Whitney U or Fisher’s exact tests. OS and PFS were summarized using Kaplan-Meier methods and compared using the log-rank test. Hazard ratios (HR) were obtained from Cox regression models. All analyses were conducted in SAS at a significance level of 0.05.
Results: Of total 83 pts with PCNSL treated at RPCCC, 67 (81%) received the DLR, 13 (16%) received WBRT and 3 were treated on other protocols. Clinical characteristics of pts treated with DLR (n=67) include: median age at diagnosis 68yr (28-83), male (37, 55%), ECOG performance status (PS) 0-1 (30, 44%), 2-4 (37, 55%), cell of origin- GCB (7,10%), Non-GCB (45,67%), unclassified (14,20%) and miscellaneous (1,1%). 27 pts (40%) were 70-80 yrs and one pt was more than 80 yrs old. 15 (22%) pts had a creatinine clearance (CrCl) <60 ml/min, 32 pts (48%) had a high (3-5) IELSG score, 27 pts (40%) had a Charlson comorbidity index (CCI) ≥ 4. 34 pts (51%) had multiple lesions, 37 (55%) had >/=1 deep brain lesion.
The default starting dose of HDMTX was 2.5 g/m2. MTX was given at a lower dose (1.25-2) in 4 pts due to lower CrCl. The median time to clear the 1st dose of HDMTX was 46.6hrs (12-276). Only 42 (63%) pts were able to complete therapy, reasons for discontinuation were noninfectious adverse events (6,7%), infections (13,16%), progression of disease (6,7%), renal injury (3,4%), poor PS (5,6%) and death (3, 4%). Cytarabine consolidation was given to 38 pts (60%). None of the patients received consolidative WBRT or autologous stem cell transplant (ASCT).
After induction therapy, 42 pts (63%) had a complete response (CR), 11 pts (16%) had partial response (PR), 2 pts (3%) had stable disease & response could not be evaluated in 11 pts (16%). Median follow was of 49.5 (2-146) mo. The median PFS was 18.7 (9.9- 40.8) mo (figure 1) & the median OS was 24.2 months (12.1- 56.3 mo) (figure 2). 35 (52%) pts died, cause of death included disease progression (24%), treatment-related mortality (16%) & unknown (12%). 24 pts (29%) received second-line therapy of which 6 underwent ASCT in 2nd remission. While response rates did not differ by age, pts > 70 years had worse PFS (9.1 vs 24.8, p=0.019) & OS (NR vs 10.1 mo, p=0.001). A higher CCI was associated with decreased OS (NR for 0-1, 31.1 mo for 2-3, and 12.1 mo for 4+ score, p=0.022) (figure 3). IELSG score, number, and site of lesions and cell of origin did not impact outcomes.
64 pts (95%) developed grade 3/4 toxicities, majority of which were hematologic [anemia (58%), neutropenia (54%), thrombocytopenia (48%)]. Other grade 3/4 toxicities included neutropenic fever (16%), infection (52%), neurologic (27%), respiratory (24%) and renal (24%). The incidence of neurological toxicity was higher in pts with age>60, ECOG PS 2-4, and high IELSG score. Age >60 was also associated with an increased incidence of anemia and thrombocytopenia.
Conclusions:
PCNSL remains a challenging disease to treat, especially in older adults. DeAngelis-Lite, a 2.5 g/m2 HDMTX based induction regimen without consolidative RT or ASCT lead to a high ORR of 79%, however, PFS and OS were shorter than with previously reported HDMTX-based regimens highlighting the importance of consolidative strategies. Despite a lower dose of MTX, only 63% of pts were able to complete therapy suggesting that tolerability was still a major issue. The CCI emerged as a useful tool to predict OS in pts with PCNSL. Supplanting MTX with novel agents, especially in pts with high comorbidity indices could potentially improve outcomes and is currently under study in ongoing trials.