AUTHORS
Matthew Gromowsky, MD, Elizabeth Lyden, MS2, Yun Kyoung Tiger, MD, PhD3, Michele D. Stanchina, DO,MS4, Juan Pablo Alderuccio, MD5, Eiman Elmileik, MD6, Thomas Ollila, MD7, Shristi Upadhyay Banskota, MD8, Smriti I Sharma, MD9, James Olen Armitage, MD10, Matthew A. Lunning, DO, FACP11, Julie Vose, MD12, Christopher D'Angelo, MD, MS13
AFFILIATIONS
1 University of Nebraska Medical Center, Omaha, Nebraska, United States
2 University of Nebraska Medical Center, Omaha, Nebraska, United States
3 Mayo Clinic, Rochester, Minnesota, United States
4 University of Miami, Miami, Florida, United States
5 University of Miami, Miami, Florida, United States
6 Brown University, Providence, Rhode Island, United States
7 Brown University, Providence, Rhode Island, United States
8 Avera, Sioux Falls, South Dakota, United States
9 Nebraska Cancer Specialists, Omaha, Nebraska, United States
10 University of Nebraska Medical Center, Omaha, Nebraska, United States
11 University of Nebraska Medical Center, Omaha, Nebraska, United States
12 University of Nebraska Medical Center, Omaha, Nebraska, United States
13 University of Nebraska Medical Center, Omaha, Nebraska, United States
ABSTRACT
Background: Despite effective frontline and consolidation therapy for primary central nervous system lymphoma (PCNSL), up to 50% of patients suffer disease relapse. Surveillance imaging is often recommended following completion of primary therapy, but there is limited data to guide interval and duration of surveillance. It remains unclear whether patients whose relapse is determined through routine surveillance prior to symptom onset experience improved treatment outcomes or a quality-of-life benefit such as improved long-term neurologic outcomes. Similar studies have investigated the role of surveillance imaging in the past, but results including survival or quality of life benefit from asymptomatic detection have been variable with limited investigation (Langner-Lemercier 2016, Mylam 2017, Puhakka 2024). We hypothesized that asymptomatic relapses detected by surveillance imaging would have similar outcomes to symptomatic relapses in this period including progression-free survival (PFS) and overall survival (OS), and that relapses would be uncommon greater than 2 years from frontline therapy.
Methods: We performed a multi-center retrospective analysis of patients with newly diagnosed PCNSL from 2004 to 2022. Demographics, disease characteristics, treatment, and surveillance imaging data were collected. Survival analysis was performed only among patients with CNS relapse following frontline therapy. Date of relapse was defined by imaging.
Results: We analyzed 148 patients from 4 academic centers with a median follow-up of 59.5 months from diagnosis among living patients. After excluding patients with refractory disease, 45 patients were found to have relapsed CNS disease after an initial response to frontline therapy and were included in the analysis. 25 (56%) patients had documented complete response, 17 (38%) had partial response, and 1 (2%) had stable disease. 35 patients were symptomatic at relapse, and 10 were asymptomatic detected by surveillance imaging.
Median follow-up in the cohort post-relapse was 42 months. There were no significant differences in characteristics between patient groups including response to therapy, age, sex, or performance status. Frontline treatment was methotrexate-based combination therapy in 42/45 patients. 26 patients received consolidation therapy, with 11 receiving thiotepa-autologous stem cell transplantation and 6 receiving cytarabine-based therapy. Patients with symptomatic relapse included cognition (49%), motor (26%), sensory (9%), speech (3%), vision (13%), and gait (17%) deficits. 15 patients relapsed within 1 year (10 symptomatic), 14 patients relapsed between 1-2 years (12 symptomatic), 12 relapsed 2-5 years (9 symptomatic), and 4 relapsed >5 years (4 symptomatic) after the initiation of frontline therapy. At the time of relapse, at least 34 patients (76%), were actively undergoing surveillance imaging with brain MRI, and 5 patients (11%) had unclear ongoing surveillance; 6 (13%) were not receiving active surveillance imaging.
Following relapse, the median PFS and OS for patients asymptomatic compared to symptomatic at the time of relapse was 8.5 vs 7.8 months (PFS p=0.62) and 19.8 vs 27.3 months (OS p=0.74), respectively. Among symptomatic patients, symptoms resolved completely in 29% (10/35) of symptomatic patients, partially in 17% (6/35), and had unclear resolution in 20% (7/35) of subjects. 34% (12/35) of patients experienced persistent symptoms with no resolution. Treatment for relapsed disease included chemotherapy (including retreatment with methotrexate) (29/45, 64%), ibrutinib (5/45, 11%), lenalidomide (1/45, 2%), radiation alone (1/45, 2%), and not reported (6/45, 13%). 3 subjects did not pursue treatment.
Conclusion: Relapses largely occurred within 2 years (29/45) of initiation of frontline therapy and 22% (10/45) of relapses were asymptomatic, including 25% (3/12) of relapses occurring in years 2-5 and no asymptomatic relapses detected after year 5. Up to 34% (12/35) of subjects experienced persistent symptoms at relapse despite therapy, although there were no significant differences in PFS or OS between symptomatic and asymptomatic presentations and the small sample size currently limits interpretability of these findings. Data collection is ongoing with additional sites participating in an effort to expand the sample size and help determine the utility of surveillance imaging, especially in later years post frontline therapy.
DISCLOSURES
Matthew Gromowsky, MD: No relevant conflicts of interest to declare.