Deep learning and digital pathology powers prediction of HCC development in steatotic liver disease.

BACKGROUND AND AIMS: Identifying patients with steatotic liver disease who are at a high risk of developing HCC remains challenging. We present a deep learning (DL) model to predict HCC development using hematoxylin and eosin-stained whole-slide images of biopsy-proven steatotic liver disease. APPROACH AND RESULTS: We included 639 patients who did not develop HCC for ≥7 years after biopsy (non-HCC class) and 46 patients who developed HCC <7 years after biopsy (HCC class). Paired cases of the HCC and non-HCC classes matched by biopsy date and institution were used for training, and the remaining nonpaired cases were used for validation. The DL model was trained using deep convolutional neural networks with 28,000 image tiles cropped from whole-slide images of the paired cases, with an accuracy of 81.0% and an AUC of 0.80 for predicting HCC development. Validation using the nonpaired cases also demonstrated a good accuracy of 82.3% and an AUC of 0.84. These results were comparable to the predictive ability of logistic regression model using fibrosis stage. Notably, the DL model also detected the cases of HCC development in patients with mild fibrosis. The saliency maps generated by the DL model highlighted various pathological features associated with HCC development, including nuclear atypia, hepatocytes with a high nuclear-cytoplasmic ratio, immune cell infiltration, fibrosis, and a lack of large fat droplets. CONCLUSIONS: The ability of the DL model to capture subtle pathological features beyond fibrosis suggests its potential for identifying early signs of hepatocarcinogenesis in patients with steatotic liver disease.

Necitumumab plus gemcitabine and cisplatin versus gemcitabine and cisplatin alone as first-line treatment for stage IV squamous non-small cell lung cancer: A phase 1b and randomized, open-label, multicenter, phase 2 trial in Japan.

OBJECTIVES: This open-label, multicenter, phase 1b/2 study assessed necitumumab plus gemcitabine and cisplatin (GC + N) in patients with previously untreated squamous non-small cell lung cancer in Japan. MATERIALS AND METHODS: The phase 1b part determined the gemcitabine dose for the phase 2 part, in which patients were randomized 1:1 to GC + N or gemcitabine and cisplatin (GC) (gemcitabine 1250 mg/m2 on days 1 and 8; cisplatin 75 mg/m2 on day 1 of maximum four 3-week cycles; nectimumab 800 mg on days 1 and 8 of a 3-week cycle continued until progressive disease or unacceptable toxicity). The primary endpoint of the phase 2 part was overall survival. RESULTS: In the phase 2 part, 181 patients received GC + N (N = 90) or GC (N = 91). Overall survival was significantly improved with GC + N versus GC (median, 14.9 months vs 10.8 months; hazard ratio [HR] = 0.66, 95% CI: 0.47 - 0.93, p = 0.0161). Improvements were also observed in progression-free survival (median, 4.2 months vs 4.0 months; HR = 0.56; p = 0.0004) and objective response rate (51% vs 21%; p < 0.0001). Survival was also significantly prolonged with GC + N versus GC for patients with epidermal growth factor receptor-positive tumors. Grade ≥3 treatment-emergent adverse events at ≥5% higher incidence with GC + N than GC were neutrophil count decreased (42% vs 35%), febrile neutropenia (12% vs 3%), decreased appetite (11% vs 4%), and dermatitis acneiform (6% vs 0%). CONCLUSION: GC + N is well tolerated and has significant and clinically meaningful treatment benefit in the first-line treatment of patients with squamous non-small cell lung cancer in Japan. Clinicaltrials.gov identifier: NCT01763788.