Evaluating the Impact of Combination of Nivolumab and Ipilimumab to Liver/Lung SBRT on Local Control for Colorectal and Pancreatic Cancer: A Pooled Analysis of Two Prospective Trials

Abstract

Immunotherapy (IO) has emerged as an effective treatment option for many metastatic cancers, however, its role in pancreatic or colorectal cancer is limited to those with microsatellite instability. It has been suggested that radiation (RT) and combination IO may work synergistically to increase the immune system’s exposure to tumor antigens, potentiating an abscopal effect. However, the impact of combination IO to stereotactic body radiation therapy (SBRT) on local control of irradiated lesions has not been elucidated. In this study, we combined two prospective phase II single arm studies to compare the impact of SBRT alone (NCT01239381) or with PD-L1/CTLA4 blockade (NCT03104439) on the growth of colorectal and pancreatic liver and lung metastases. Eligible patients had metastatic colorectal or pancreatic adenocarcinoma. Those treated on NCT01239381 received liver SBRT alone (30-50 Gy in 5 fractions), while those on NCT03104439 received combination IO 6 weeks prior to and concurrently with 24 Gy in 3 fractions to liver or lung metastases. Tumor genotyping was performed to assess for KRAS and TP53 mutations. Target lesions were measured on CT scans prior to and following RT. Peak response was measured via modified RECIST criteria as the largest tumor reduction at any time. Wilcoxon Rank Sum tests were performed to evaluate % tumor reduction with regards to combination IO, primary tumor site, metastasis site, and KRAS and TP53 mutation status. There were 50 eligible patients from NCT01239381 and 41 from NCT03104439, of whom 42 and 39 were evaluable, respectively. Of these 81 patients, 52 and 29 had colorectal and pancreatic cancer, respectively. Of those with molecular analysis, 50% (33/66) had a KRAS mutation and 55% (34/61) had a TP53 mutation. Median follow up was 78 days in the IO arm (driven by distant progression) and 160 days in the non-IO arm (P<0.0001). Mean BED in the IO arm was 43.2 Gy vs. 79.6 Gy in the SBRT alone arm (P<0.0001). Mean target size was 3.8 cm in the IO arm, and 5.0 cm in the SBRT alone arm (P=0.12). There was no difference with respect to overall response rate (38.5% vs. 52.4%, P=0.27) and disease control rate (92.3% vs. 92.9%, P>0.99). There was no significant difference in the % tumor reduction between the IO and non-IO arms (-21% vs. -41%, P=0.08). Among the entire cohort, there was no significant difference in response of irradiated lesion with respect to primary tumor type (P= 0.54), KRAS status (P=.67), TP53 status (P=0.10), or metastatic site (P=0.58). Despite receiving approximately half the BED dose compared to liver SBRT alone, patients treated with IO and RT had similar response rates, suggesting a potential synergistic effect in the irradiated lesions. Future studies should look to validate these findings and explore optimal RT dose/fractionation schemas in combination with IO. These results support ongoing randomized studies of IO with RT in locally advanced rectal (NRG GI002) and pancreatic (NCT03563248) cancers.