Update Analysis of Central Nervous System Multiple Myeloma Prognosis and Survival: A Real-World Multi-Institutional Study of the Greek Myeloma Study Group

Central Nervous System Multiple Myeloma (CNS-MM) is a rare but aggressive complication of MM, and its treatment remains an unmet clinical need. Herein we present a real-world national analysis of incidence, characteristics, prognostic factors and outcome of a large cohort of patients with CNS-MM.

We evaluated 53 patients with CNS-MM (M/F: 23/30; median age: 59, range: 20-96 years; IgG: 27, IgA: 10, light-chain: 11, non-secretory: 4, IgD: 1; ISS1: 22, ISS2: 13, ISS3: 18; RISS1: 15, RISS2: 22, RISS3: 16) out of 4352 MM patients (1.2%), who were treated within the Greek Myeloma Study Group centers from 2000 to 2022 (23% of patients were diagnosed after 2016). A control group of MM patients matched for age, gender, and stage (ISS/RISS) who were treated during the same period was used for comparisons. Presence of soft tissue plasmacytomas (STP) and/or plasma cell leukemia (PCL; circulating plasma cells, cPCs: ≥5%) was determined and analyzed as extramedullary involvement (EMI).

Median time to CNS-MM from MM diagnosis was 28 months (range 0-149 months); 6/53 (11.3%) patients experienced CNS-MM at initial MM diagnosis. Clinical manifestations of CNS-MM included visual disturbances/diplopia: 32%, paresis/paraplegia: 22%, lethargy/confusion: 37%, headache: 22%, cranial nerve palsy: 13% and dysarthria: 11%. PCL was present in 16/53 (30%) CNS-MM patients, while 29/53 (54.7%) CNS-MM patients had bone and/or STP vs. 13/53 matched controls (p=0.001); EMI was found in 24/53 (45%) patients with CNS-MM vs. 4/53 (7%) in controls (p<0.001). 35/53 (66%) CNS-MM patients had leptomeningeal infiltration (LMI) and 34% had intraparenchymal lesions. Median number of cells in the cerebrospinal fluid (CSF) was 60/μL (range 10-2550) and almost all of them (80%, range: 20-100%) were clonal plasma cells. Out of 53 CNS-MM patients, 14 (26.4%) had only CNS involvement without marrow infiltration by PCs; 51% had ≥1 high-risk molecular features (i.e., del17p, t4:14, t14:16, 1q+), and this was marginally higher vs. controls (36%; p=0.15). CD56 negative marrow PCs did not differ between groups (33% vs. 37%). A binary logistic regression analysis demonstrated that EMI was the strongest prognostic factor for the development of CNS-MM (p=0.01; HzR: 5.7).

Median number of lines of treatment before the development of CNS-MM was 2 (range 0-9). First line anti-myeloma treatment in patients who developed CNS-MM included conventional chemotherapy (CT) or thalidomide-based regimens in 19 patients, bortezomib-based triplets in 27 and daratumumab-based regimens (DBR) in 7; this did not differ from controls. Overall response rate (ORR) in first-line therapy was similar between CNS-MM and controls (85% for both groups; CR 41% vs. 36%, respectively). After a median follow-up of 150 months post-MM diagnosis (95% CI: 107-192), 46 CNS-MM patients (87%) deceased (MM progression: 34, infection/sepsis: 10, cerebral hemorrhage: 2). Median overall survival (OS) of CNS-MM patients vs. matched controls from initial MM diagnosis was 43 months (95% CI: 31-54) vs 60 months (95% CI: 34-86) (p=0.002; Figure).

Regarding CNS-MM therapy, 17 patients received IMiD-based regimens and 26 received PI-based regimens, 6 received CT and 4 received DBR. Patients with CNS-MM and marrow infiltration by PCs displayed an ORR of 31% only; CNS symptoms improved in 74% of all patients. Median OS of CNS-MM patients from the time of CNS involvement was 4 months (95% CI: 2.6-5.3), and it was longer in 12 patients treated after 2016 [18 months (95% CI: 7-29) vs. 2 months (95% CI: 1-3)]. Cerebrospinal fluid (CSF) infusions were performed in 28 patients (53%); 9 patients (19%) underwent radiotherapy. A multivariate analysis showed that EMI, either at diagnosis or during MM course, was the strongest negative predictor for post CNS-MM OS (p=0.008; HR: 3.4 95% CI: 1.4-8.0). Moreover, LMI was a negative predictor (p=0.02; HR: 2.7, 95% CI: 1.2-6.3) for OS, whereas treatment after 2016 had a positive impact (p=0.02; HR: 0.28, 95% CI: 0.12-0.8); CSF infusions did not affect OS.

In conclusion, our analysis representing one of the largest real-world series of CNS MM patients after 2016, showed that, OS of CNS-MM remains poor however, there is a positive impact of treatment after 2016, reflecting the efficacy of modern anti-myeloma therapy; EMI strongly predicts for CNS-MM and induces a 3-fold increase in the probability of post CNS-MM death indicating a potential need for CNS prophylaxis in these patients.