Study Population
A total of 94 patients with advanced HCC who received rego-sintilimab or regorafenib as second-line treatment were assessed for eligibility during the study period. Eleven patients were excluded because they met the exclusion criteria (Fig. 1). Finally, 83 patients were included in this study (rego-sintilimab group, n = 48; regorafenib group, n = 35). The baseline characteristics between the two groups were not significantly different (Table 1).
In detail, 37.5% (18/48) of patients in rego-sintilimab group and 34.3% (12/35) in regorafenib group had Child-Pugh B liver function (P = .763). 41.7% (20/48) of patients in rego-sintilimab group and 37.1% (13/35) in regorafenib group received lenvatinib in first-line treatment (P = .678). Median time on first-line treatment was 8.0 months (IQR, 4.6–12.3) in rego-sintilimab group and 7.9 months (IQR, 5.2–13.7) in regorafenib group (P = .941). Mean daily dose of regorafenib was 108.3 mg ± 23.3 in rego-sintilimab group and 114.3 mg ± 24.0 in regorafenib group (P = .260). Median duration of second-line treatment with rego-sintilimab or regorafenib was 5.8 months (IQR, 3.5–9.4) in rego-sintilimab group and 3.9 months (IQR, 2.4–6.8) in regorafenib group.
Safety
Treatment-related AEs are shown in Table 2. No treatment-related mortality occurred. The overall incidence of AEs was similar between rego-sintilimab group and regorafenib group (any grade: 93.8% vs 89.6%, P = .661; grade 3/4: 39.6% vs 31.4%, P = .445). There was a higher incidence of any grade rash in rego-sintilimab group than in regorafenib group (22.9% vs 5.7%, P = .033). Incidences of other AEs were not significantly different between the two groups. AEs of any grade with an incidence over 20% in rego-sintilimab group included fatigue, hand–foot skin reaction, rash, hypertension, hypothyroidism, hyperbilirubinemia, diarrhea, and anorexia. Grade 3/4 AEs with an incidence over 5% in rego-sintilimab group included hand–foot skin reaction, hypertension, hyperbilirubinemia.
Treatment was discontinued because of AE in 12.5% (6/48) of patients in rego-sintilimab group and in 8.6% (3/35) in regorafenib group (P = .833). Grade 4 skin toxicity was observed in 1 patient after 2 cycles of rego-sintilimab and grade 4 hyperbilirubinemia in 1 patient with immune-related hepatitis confirmed by liver biopsy after 6 cycles of rego-sintilimab, and both recovered from discontinuation of sintilimab and treatment with glucocorticoid. Other 4 cases of treatment discontinuation in rego-sintilimab group were caused by grade 3 AEs, i.e. rash in 1 patient, pneumonitis in 1 patient, and gastrointestinal hemorrhage in 2 patients.
Treatment Response
Tumor responses of the two groups were shown in Table 3. Two patients (5%) in rego-sintilimab group versus no patients in regorafenib group had a complete response. The DCR for rego-sintilimab group was significantly higher than that observed in regorafenib group (72.9% vs 48.6% by RECIST 1.1, P = .024; 72.9% vs 51.4% by mRECIST, P = .044). The ORR for rego-sintilimab group was significantly higher than that observed in regorafenib group when assessed by mRECIST (33.3% vs 14.3%, P = .049), but there was only a nonsignificant better ORR trend in rego-sintilimab group compared with regorafenib group when assessed by RECIST 1.1 (22.9% vs 8.6%, P = .085).
ROC curve analyses
ROC curve analyses for predicting 1-year survival revealed that the optimal cut-off values for NLR, PLR, and neutrophils were 3.56 (area under the curve [AUC], .698; 95% CI, .579–.818; P = .004), 119 (AUC, .656; 95% CI, .530–.782; P = .022) and 4.53*109/L (AUC .699; 95% CI, .575–.822; P = .003), respectively. Thus, these variables were clarified into NLR ≤ 3.5/> 3.5, PLR ≤ 120/> 120, and neutrophils ≤ 4.5*109/L/> 4.5*109/L. The AUC for lymphocyte and platelet were not statistically significant (P = 0.121 and 0.284, respectively), and thus their cut-off values were not determined by the ROC curve.
Progression-free survival
The median follow-up duration was 12.8 months (IQR, 7.0–16.5) for the rego-sintilimab group and 8.7 months (IQR, 5.3–12.4) for the regorafenib group. During follow-up, 41 of 48 (85.4%) patients in the rego-sintilimab group and 34 of 35 (97.1%) patients in the regorafenib group experienced tumor progression. The median PFS in rego-sintilimab group vs in regorafenib group were 5.1 months (95% CI, 3.4–6.8) vs 3.0 months (95% CI, 2.3–3.7), respectively (P = .001) (Fig. 2).
According to the uni- and multi-variate analyses, regorafenib alone (HR, 2.186; 95% CI, 1.353–3.532; P = .001), α-fetoprotein (HR per 104 ng/mL, 1.044; 95% CI, 1.019–1.069; P = .001), and NLR > 3.5 (HR, 1.769; 95% CI, 1.093–2.863; P = .020) were independent prognostic factors for PFS (Table 4).
Overall survival
During the follow-up period, 32 of 48 (66.7%) patients in rego-sintilimab group and 27 of 35 (77.1%) patients in regorafenib group died. The 6-, 12-, and 24-month OS rates were 83.3%, 54.1%, and 36.3% (median OS, 13.3 months [95% CI, 9.5–17.1]), respectively, in rego-sintilimab group and 76.2%, 31.4%, and 0% (median OS, 9.1 months [95% CI, 5.6–12.5]), respectively, in regorafenib group (P = .001) (Fig. 3). According to the uni- and multi-variate analyses, regorafenib alone (HR, 2.141; 95% CI, 1.178–3.890; P = .012), Child-Pugh B liver function (HR, 2.225; 95% CI, 1.301–3.804; P = .003), and NLR > 3.5 (HR, 1.897; 95% CI, 1.075–3.348; P = .027) were independent prognostic factors for OS (Table 5).
Subgroup analyses showed that, in patients with Child-Pugh A (16.4 vs 11.5 months; P = .005) or B (8.8 vs 6.4 months; P = .032) liver function, the median OS was significantly improved in rego-sintilimab group compared with regorafenib group (Fig. 4a and 4b). In patients with an NLR ≤ 3.5, the median OS was significantly improved in rego-sintilimab group compared with regorafenib group (16.3 vs 11.5 months; P = .012), whereas it was not significantly different between the two groups in patients with an NLR > 3.5 (8.4 vs 7.0 months; P = .288) (Fig. 4c and 4d).