Patients
Between February 15, 2023 and June 28, 2023, we screened 38 patients and 25 of them were finally enrolled in this study and received at least one cycle of study treatment (ITT and safety set; Fig. 1). One patient refused to continue the treatment after the first cycle of treatment due to personal reasons. The remaining 24 patients had at least one post-treatment tumor assessment and were evaluable for treatment responses (Fig. 2).
Baseline characteristics of the enrolled patients are summarized in Table 1. The median age was 44 years (range, 24-60 years). All patients received at least one line of prior systemic therapy for advanced disease, with 3 (12%) received 2 lines and 6 (24%) received ≥ 3 lines. There were 5 (20%) patients previously received two or more different PD-1 inhibitors. Twelve patients (48%) had > 3000 copies/ml plasma EBV DNA level before treatment. The median interval of the last anti-PD-1 therapy to enrollment was 2.03 months (range, 0.7 – 16.03 months). Detail information of previous treatment regimens was shown in supplementary Table S1.
Efficacy
Of the 10 patients enrolled in the first stage, 7 patients obtained confirmed responses, which achieved the requestion of the first stage and the trial continued to full accrual. In the ITT set (n=25), 17 patients reached a confirmed objective response (ORR 68%, 95% CI, 48 - 88%), with 3 CR (12%), and 14 PR (56%) (Table 2, Fig. 2, Supplementary Fig. S1). Six patients maintained SD, one patient with PD and one patient was not evaluated. Interestingly, 20 patients (20/24, 83.33%) got EBV DNA level decrease ≥ 50% from baseline to the first post-treatment assessment, and 17 patients (17/24, 70.83%) decrease ≥ 90%. Patients with plasma EBV DNA level decrease ≥ 50% from baseline to the first post-treatment assessment had significantly higher ORR than those with plasma EBV DNA level decrease < 50% (17/20 vs. 4/4, p < 0.001). Furthermore, the EBV DNA level decrease, interval of the last administration of immunotherapy to the enrollment, and previous immunotherapy efficacy showed a numerically correlation with ordered response variables (PD, SD, PR to CR) decided by Cochran-Armitage trend test, although there was not a significantly difference (p > 0.05) (Supplementary Fig. S2).
To the date cutoff (March 25, 2024), the median follow-up was 10.2 months (range, 4.3 – 13.5 months). The mTTR was 1.7 months (range, 1.2 – 4.9 months). The median DoR was 9.1 months (95% CI, 3.8 – 14.5 months; Fig. 3A). The median PFS was 10.6 months (95% CI, 5.2 – 16.0 months; Fig. 3B). The median OS has not reached, and 12-month OS was 75.6% (Fig. 3C). Intriguingly, we observed that patients with EBV DNA level decrease ≥ 50% from baseline to the first post-treatment assessment had a significantly longer PFS (p=0.006) and OS (p=0.008, Supplementary Fig. S3) than those with plasma EBV DNA level decrease < 50%. However, no significant difference was observed between EBV DNA level at baseline and survival (PFS and OS). Up to the date cutoff, fourteen patients (14/25, 52%) discontinued the study treatment, with thirteen developed disease progression, and one declined further therapy after the first cycle of the study treatment (Fig. 1-2, Supplementary Table S2).
Safety
In the safety set, twenty-four patients (24/25, 96%) experienced at least one treatment-related AE (TRAE) of any grade. TRAEs with an incidence greater than 50% included hypoesthesia (n=19, 76%), anemia (n=18, 72%), decreased appetite (n=18, 72%), fatigue (n=18, 72%), leukopenia (n=17, 68%), nausea (n=16, 64%), increased thyroid-stimulating hormone (n=13, 52%, Table 3). Grade 3 or 4 TRAEs occurred in 12 patients (48%), the most common of which (the incidence of preferred term ≥ 2%) were anemia (n=6, 24%), neutropenia (n=6, 24%), thrombocytopenia (n=4, 16%), hypoesthesia (n=2, 8%), fatigue (n=2, 8%), leukopenia (n=2, 8%), rash (n=2, 8%), and febrile neutropenia (n=2, 8%). Fourteen patients (14/25, 56%) reported potentially immune-related AEs (irAEs). The most common irAEs (the incidence of preferred term ≥ 10%) include grade 1-2 TSH elevation (n=10, 40%), hypothyroidism (n=8, 32%), musculoskeletal (n=8, 32%), pruritus (n=7, 28%), mucositis (n=6, 24%), and rash (n=4, 16%). Only two patients experienced grade 3 irAEs, one with rash and another with lipase increased. Both of them were managed with supportive measurements and signs/symptoms subsided without sequelae.
During induction phase, twenty-one patients completed at least four cycles of TPC chemotherapy (Supplementary Table. S2). Of the other four patients, three progressed and one drop out. Twelve patients (48%) finished all six cycles of the TPC treatment. Dose reduction occurred in 11 (44%) patients, with the reasons of gastrointestinal reaction (n=5), hand-foot syndrome (n=3), and hematologic AEs (n=3). One patient required interruption of the treatment because of fall-induce femoral fracture (Supplementary Table. S2). Nineteen patients received the maintenance treatment, in which two patients (8%) required capecitabine dose reduction and cadonilimab interruption, respectively (Supplementary Table. S2). Fourteen patients are still on the study treatment at the date cutoff.