1050 - Splenic Metabolic Activity as Biomarker in Cervical Cancer

Background

Inflammation has a significant impact on cervical cancer (CC) development and therapy response. We sought to determine if splenic metabolic activity reflects host immune status and could improve prognostic and predictive categorization of CC.

Methods

Ninety-two consecutive patients with FIGO stage IB1 to IVB CC who received neo-adjuvant (chemo)radiation (NA-CRT) with curative intent were included. PET scans were performed at diagnosis and after completion of the NA-CRT. PET images were retrospectively assessed for pretreatment spleen-to-liver SUV ratio (preSLR), posttreatment spleen-to-liver SUV ratio (postSLR) and ΔSLR (post-pre) on Oasis Nuclear Medicine Workstations. (Δ)SLR was calculated as both (Δ)SLR_maxand (Δ)SLR_mean. The prognostic (DFS) and predictive (pCR) abilities of these variables together with established predictors were calculated by C-index and ROC analysis, respectively. The optimal long-rank statistic and Youden index determined cutoff values. Multivariate Cox proportional hazard regression models were ranked based on their Akaike information criterion (AIC). Clinicopathological differences between patients with low or high SLR were performed by chi-square and Mann-Whitney U tests.

Results

For preSLR_maxand preSLR_mean, association with DFS was found for preSLR_max>0.92 with HR = 2.25 (95% CI (1.08-4.66); p = 0.026) and for preSLR_mean>0.94 with HR = 2.79 (95% CI (1.33-5.86); p = 0.005), respectively. The selected multivariate model consisted of three factors: preSLR_max, ΔSLR_maxand parametrial invasion (dichotomized; HR = 6.40 95% CI (2.70-15.20); p < 0.001). The model’s prognostic ability was quite favorable compared to FIGO staging (C-index 0.69 vs. 0.64). Further, uni- and multivariate analyses suggest that both low preSLR_maxand low preSLR_mean influence pCR. Patients (n = 31) with high preSLR_maxhad a higher density of CD3⁺,CD4⁺, CD8⁺, CD20⁺(77.8% vs. 36.4%; p = 0.036), CD68⁺ (88.9% vs. 40.9%; p = 0.015), CD163⁺, FoxP3⁺and PD-L1⁺immune cells, as well as PD-L1⁺tumor cells (85.7% vs. 55.6%; p = 0.019) in the primary tumor using IHC; the same trends were observed for preSLR_mean

Conclusions

SLR is a promising prognostic and predictive biomarker in CC and is associated with the tumor immune infiltrate.

Clinical trial identification

Editorial acknowledgement

Legal entity responsible for the study

Ghent University Hospital.

Funding

Research Foundation-Flanders (FWO).

Disclosure

E.A. De Jaeghere: Travel / Accommodation / Expenses: PharmaMar. F. Laloo: Honoraria (institution): Bayer. K. De Man: Travel / Accommodation / Expenses: Bayer. H. Denys: Honoraria (institution), Travel / Accommodation / Expenses: Pfizer; Honoraria (institution), Research grant / Funding (institution), Travel / Accommodation / Expenses: Roche; Honoraria (institution), Travel / Accommodation / Expenses: PharmaMar; Travel / Accommodation / Expenses: Teva; Honoraria (institution), Travel / Accommodation / Expenses: AstraZeneca; Honoraria (institution): Eli Lilly and Company; Honoraria (institution): Novartis; Honoraria (institution): Amgen; Honoraria (institution): Tesaro. K. Vandecasteele: Travel / Accommodation / Expenses: PharmaMar. All other authors have declared no conflicts of interest.