A Novel Prognostic Index for Ocular Adnexal Lymphoma

Background: Ocular adnexal lymphoma (OAL) is a rare malignancy with a heterogeneous patient population. We sought to identify prognostic factors for OAL and to incorporate them into a novel prognostic index by using a large national database.

Methods: We identified patients diagnosed with histologically confirmed OAL in 1973-2013 using all registries of the Surveillance, Epidemiology, and End Results database. Patients with regional or distant disease at diagnosis were excluded. Indolent histologies were defined as follicular, marginal zone, small lymphocytic not otherwise specified, and lymphoplasmacytic. We determined the association between OAL histology, orbital site, and overall survival (OS) using adjusted proportional hazards analysis and Kaplan-Meier estimates. Relative survival (RS) was calculated relative to a matched cohort of patients categorized by race, gender, and age using the SEER*Stat Program.

Results: Among the 3,070 patients included, 941 (30.6%) had conjunctival tumors, 418 (13.7%) had lacrimal gland tumors, and 1711 (55.7%) had orbital-not otherwise specified tumors. Indolent histology was present in 2,073 (67.5%) patients. The 10-year OS and RS were 61% and 86%, respectively. In univariable and multivariable models, conjunctival site location (P<0.001) and indolent histology (P<0.001) were both associated with superior OS. Three primary OAL prognostic groups were identified: Group I (indolent histology, conjunctival location), Group II (indolent histology, non-conjunctival location or aggressive histology, conjunctival location), and Group III (aggressive histology, non-conjunctival location). The adjusted 10-year hazard ratios for OS were 1.4 (95% CI, 1.1-1.7, P<0.001) and 3.1 (95% CI, 2.5-4.0, P<0.001) for groups II and III, respectively, compared to group I. Groups I, II, and III had 10-year OS of 75%, 61%, and 37%, respectively, and 10-year RS of 96%, 89%, and 55%, respectively.

Conclusions: Both conjunctival location and aggressive histology were independent predictors of survival. Our proposed prognostic index could have implications on OAL staging and management, though it must first be externally validated.