Safety of Upfront Locoregional Therapy in Conjunction with Atezolizumab and Bevacizumab for Unresectable Hepatocellular Carcinoma

Purpose/Objective(s)

The use of atezolizumab and bevacizumab for advanced hepatocellular carcinoma (HCC) is a recently established standard of care. Addition of locoregional therapies (LRT) such as external beam radiotherapy (RT), radioembolization (Y90), transarterial chemoembolization (TACE), and/or radiofrequency ablation (RFA) may hold promise for further improving outcomes, but the safety of these approaches with atezolizumab/bevacizumab has not been established. This study assessed the safety of upfront LRT in conjunction with this immunotherapy regimen in HCC.

Materials/Methods

We reviewed all patients initiating atezolizumab with or without bevacizumab (IO) for HCC from July 2020 to January 2022 at a single institution. Patients receiving LRT (RT, Y90, TACE, and/or RFA) within 30 days of IO initiation (LRT+IO) were identified alongside a comparison arm of all patients initiating IO alone. Bevacizumab was typically held until after LRT. Treatment-related toxicities were recorded according to CTCAE definitions from the date of treatment initiation until discontinuation of IO or last follow up. Baseline, 1 month, and 3-month Child-Pugh (CP) score, ALBI score, transaminases, and total bilirubin were additionally recorded and analyzed using a repeated-measure mixed effects model.

Results

50 HCC patients initiating atezolizumab were identified, 49 (98%) of these in combination with bevacizumab. Median follow up was 6.5 months (range 1.2-20), median age was 64 (34-79). 37 (74%) patients had BCLC stage C disease and 12 (34%) BCLC stage B. The majority of patients (86%) had CP class A liver function. Baseline characteristics were balanced between arms on univariate analysis (p>0.05). Of 27 (54%) patients in the LRT+IO arm, 9 received RT alone, 4 received RT combined with Y90 or TACE, 10 received Y90 alone, 3 received TACE alone, and 1 received RFA. The majority of the LRT+IO arm (63%) received LRT after the first infusion of atezolizumab. Grade 3 toxicities were observed in 2 (7.4%) patients in the LRT+IO arm, compared to 2 (8.6%) patients in the IO-only arm (p=0.87); grade 2 toxicity incidence was 63% in the LRT+IO arm versus 43% in the IO-only arm (p=0.17). The receipt of LRT+IO was associated with transient increase in CP and ALBI scores at 1 month relative to IO alone (median score increase 1 and 0.34, respectively, p<0.01); this association did not persist at 3 months. There was no treatment-related liver failure or grade 4/5 toxicity reported in either arm. LRT+IO was not associated with increased ALT, AST, or total bilirubin at 1 or 3 months.

Conclusion

In this single institution retrospective cohort study, upfront locoregional therapy with atezolizumab and bevacizumab for HCC was not associated with a significantly greater incidence of grade 2/3 toxicity or clinically significant liver dysfunction relative to the initiation of atezolizumab and bevacizumab alone. Further studies are needed to verify the safety of combination therapy and to assess clinical benefit.