DLP-201

Patients with melanoma progressing on immune checkpoint blockade may benefit from adoptive transfer of tumor-infiltrating lymphocytes (TIL).

We investigated the impact of a pegylated IFNα conditioning and support regimen on the safety and efficacy of TIL plus nivolumab (NCT03638375). Patients with immune checkpoint blockade–resistant stage III/IV melanoma were treated with TIL plus nivolumab without (n = 9) or with (n = 25) IFNα.

The treatment was safe, and side effects included IFNα-induced lymphopenia (16%) and neutropenia (12%). No febrile neutropenia or >grade 4 adverse events were observed. Disease control was obtained in 11.1% (95% confidence interval, −14.5%–36.7%) of the patients treated without and in 41.7% (95% confidence interval, 20.4%–62.9%) of the patients treated with IFNα, clearly suggesting the need for IFNα support. IFNα treatment strongly reduced the numbers of circulating leukocytes and neutrophils, more consistently in therapy responders. No differences were observed in the phenotype and dose of TIL administered.

Taken together, our low-toxicity therapy comprising TIL, nivolumab, and IFNα is safe, shows evidence of clinical activity, and may be particularly suitable for more frail patients who are less able to tolerate lymphodepletion and high-dose IL-2 regimens.

GOLDEN model is designed to predict hepatitis B surface antigen (HBsAg) loss based on patients with chronic hepatitis B (CHB) receiving nucleos(t)ide analogues therapy. We aimed to validate GOLDEN model's effectiveness in predicting HBsAg loss or decline in interferon-α-treated patients with CHB.

Interferon-α-treated patients were enrolled from EXCEL study, a randomized controlled trial that included hepatitis B e antigen-positive, noncirrhotic, treatment-naive patients with CHB, and Search-B cohort, a prospective real-world observational cohort of CHB. Using multiple quantitative HBsAg (qHBsAg) measurements, GOLDEN model was used to calculate a patient's probability of achieving HBsAg loss or decline.

Among 200 patients in the EXCEL study and 1041 patients from the Search-B cohort, the corresponding cumulative incidence of qHBsAg <100 IU/mL or HBsAg loss was 20.0% and 6.7%, after the median follow-up of 18.0 (interquartile range, 18.0-30.0) and 66.7 (interquartile range, 48.8-84.7) months, respectively. The GOLDEN model achieved an area under the curve of 0.820 (95% confidence interval, 0.737-0.902) for predicting qHBsAg <100 IU/mL in the EXCEL study and 0.964 (95% confidence interval, 0.953-0.974) for predicting HBsAg loss in the Search-B cohort, maintaining robust performance across subgroups. The favorable group showed higher cumulative incidences of qHBsAg <100 IU/mL (42.5% vs 4.6%; P < .001) or HBsAg loss (37.2% vs 0%; P < .001) than the unfavorable group, along with significantly lower qHBsAg levels and faster qHBsAg decline rates. Moreover, the favorable group defined by GOLDEN model and qHBsAg levels at enrollment were confirmed as independent predictors for HBsAg loss or decline.

GOLDEN model is a robust tool for predicting HBsAg loss or decline in interferon-α-treated patients with CHB, offering valuable support for clinicians in developing personalized, effective management strategies for patients with CHB.