In Patients Treated by Selective Internal Radiotherapy, Cellular In Vitro Immune Function Is Predictive of Survival

Abstract

In patients with liver malignancies, the cellular immune function was impaired in vitro after selective internal radiotherapy (SIRT). Because immunosuppression varied substantially, in the current study, we investigated in 25 SIRT patients followed up for ten years whether the lymphocyte function was correlated with survival. Peripheral blood mononuclear cells were stimulated with four microbial antigens (tuberculin, tetanus toxoid, Candida albicans and CMV) before therapy and at four time points thereafter, and lymphocyte proliferation was determined by H3-thymidine uptake. The median sum of the responses to these four antigens decreased from 39,464 counts per minute (CPM) increment (range 1080-204,512) before therapy to a minimum of 700 CPM increment on day 7 after therapy (0-93,187, p < 0.0001). At all five time points, the median survival in patients with weaker responses was 2- to 3.5-fold shorter (p < 0.05). On day 7, the median survival in patients with responses below and above the cutoff of a 2 CPM increment was 185 and 523 days, respectively (χ² = 9.4, p = 0.002). In conclusion, lymphocyte function could be a new predictor of treatment outcome after SIRT.

Keywords: ELISpot; immune checkpoint molecule; interferon-gamma; interleukin-2; lymphocyte proliferation; patient survival; selective internal radiotherapy.

Figure 1

Cellular response towards four microbial antigens in 25 patients treated by selective internal radiotherapy (SIRT). Panel (a) shows individual data prior to SIRT and one hour and on day 2, 7 and 28 thereafter, together with median and interquartile range. Red dots indicate values for patients with a short survival (<median of 369 days), and green dots those for patients with a long survival (≥median). Data were compared by 1-way ANOVA, and the comparison of data from all test points yielded a p value of <0.0001. p values in panel (a) indicate the significance of pairwise comparisons, corrected for multiple comparisons. Panel (b–f) indicate the results of a receiver operating characteristic (ROC) curve analysis of the cellular response towards four microbial antigens and of survival (<median vs. ≥median). Each of these panels provides information on area under the curve (AUC), cutoff, sensitivity, specificity and likelihood ratio. On day 7 (panel (e)), we considered two different cutoffs for the cumulative (cum.) antigen response, as we observed two maxima for the likelihood ratio. CPM incr., counts-per-minute increment, i.e., antigen-specific response minus the negative control.

Figure 2

Impact of the cumulative antigen response on the survival of 25 patients treated by selective internal radiotherapy (SIRT). Panel (a) shows the survival curves with respect to the cumulative (cum.) antigen response (towards four microbial antigens) prior to SIRT, and panel (b–f) show curves after SIRT. Please note that the analysis on day 7 after SIRT was performed with two different cutoff values, as explained in the legend to Figure 1e. Survival curves were compared by the Log Rank (Mantel–Cox) test. CPM increment, counts-per-minute increment, i.e., antigen-specific response minus the negative control.