Abstract
Introduction BsAbs have emerged as a promising therapeutic option for patients with LBCL, but their outcomes may be impacted by a wide range of clinical and biological characteristics. The aim of this study was to analyze the role of extranodal involvement, in terms of number and specific sites, on the results of BsAbs in this patient population.
Methods A retrospective multicentric analysis was conducted on a cohort of patients with LBCL treated with CD20-CD3 BsAbs, either as monotherapy or in combination. Clinical and biological variables were collected, including number of extranodal sites, specific organ involvement, and treatment characteristics. Progression-free survival (PFS), and overall survival (OS) were evaluated by Cox regression. A p-value <0.05 was considered statistically significant.
Results Our cohort included a total of 86 patients (61% male) with a median follow-up of 13.3 months [IQR 6-37]. The median number of prior treatments was 2 [IQR 1-2]. BsAbs treatment categories included: BsAbs as monotherapy in 46 (53%) patients, BsAbs combined with chemotherapy in 27 (32%), and BsAbs combined with co-stimulators in 13 (15%) patients.
Extranodal involvement was present in 69 (80%) of patients (45% ³2 extranodal sites). Extranodal involvement included: muscular (n=21, 24%), bone (n=17, 20%), serosal (n=16, 19%), skin (n=13, 15%), lung (n=12, 14.0%), liver (n=11, 13%), gastrointestinal (n=10, 12%), renal (n=8, 9%), adrenal (n=5, 6%), gynecological (n=5, 6%), ENT (n=5, 6%), bone marrow (n=3, 4%) and other (n=4, 4%).
In the full cohort, the presence of extranodal involvement (n=69) was associated with a shorter PFS (median of 5.6 vs. 72.7 months (HR=3.5 [IQR 1.5-8.1], p=0.004)) and a shorter OS (median of 14 months vs. non-reached (NR) (HR 3 [IQR 1.16-7.50], p=0.02)). Furthermore, the involvement of ³2 extranodal sites showed an even shorter mPFS: 3.8 months vs 11.9 months (HR=1.7, p=0.04). As the number of extranodal sites increased, there was a higher risk of progression and/or death. Considering the number of involved extranodal sites as a quantitative discrete variable, it was also associated with shorter PFS (HR=1.4, p<0.001) and OS (HR=1.5, p=0.002).
Specific sites correlated with worse outcomes: muscle involvement (n=21) was associated with a shorter mPFS (3.2 vs. 11.9 months (HR=2.3, p=0.003)) and mOS (10.9 months vs. NR (HR=2.6, p=0.002)); extramedullary bone involvement (n=17) was associated with a shorter mPFS of 2.9 vs. 10 months (HR=2, p=0.016) and mOS of 8.1 vs. 41.9 months (HR=2.4, p=0.006); serosa involvement (n=16) was associated with a shorter mPFS of 2.2 vs. 8.5 months (HR=2, p=0.02), but not with OS (HR=1.4, p=0.3); and gastrointestinal involvement (n=10) was associated with a shorter mPFS of 1.3 vs. 8.5 months (HR=2.8, p=0.005), but not with OS (HR=1.9, p=0.14). If we analyzed muscle and skin involvement like a single group (musculoskeletal) we achieve even more robust outcome prediction with a mPFS of 4.5 months (HR 2.5, p < 0.001) and a mOS of 10 months (HR 3.3, p < 0.001).
Additionally, the number of prior treatment lines, analyzed as a discrete variable, was also associated with worse survival outcomes. In the univariable analysis, it significantly impacted both PFS (HR: 1.41; p=0.001) and OS (HR: 1.47; p=0.001).
A multivariable Cox analysis for PFS and OS was performed including relevant clinical variables such as age, IPI, presence of extranodal involvement, number of extranodal sites, number of treatment lines received, muscle, bone, serosa and gastrointestinal involvement. Musculoskeletal (HR=1.9, p=0.03) and gastrointestinal involvement (HR=3.6, p=0.002) maintained a shorter PFS whereas musculoskeletal (HR=3.3, p=0.002) and gastrointestinal (HR=2.9, p=0.034) were associated with a worse OS.
Conclusions Our data suggest that extranodal involvement may influence the response to BsAbs in LBCL, with reduced PFS or OS depending on the site involved, especially when looking at musculoskeletal and gastrointestinal involvement. These findings support the importance of considering specific extranodal sites when assessing the patient's eligibility for BsAb therapy. Future studies will need to explore if these differences are related to the total tumor volume or specific genetic alterations that condition efficacy at these sites.
Conflicts of interest: None declared.
