A convolutional neural network for total tumor segmentation in [64Cu]Cu-DOTATATE PET/CT of patients with neuroendocrine neoplasms.

BACKGROUND: Segmentation of neuroendocrine neoplasms (NENs) in [64Cu]Cu-DOTATATE positron emission tomography makes it possible to extract quantitative measures useable for prognostication of patients. However, manual tumor segmentation is cumbersome and time-consuming. Therefore, we aimed to implement and test an artificial intelligence (AI) network for tumor segmentation. Patients with gastroenteropancreatic or lung NEN with [64Cu]Cu-DOTATATE PET/CT performed were included in our training (n = 117) and test cohort (n = 41). Further, 10 patients with no signs of NEN were included as negative controls. Ground truth segmentations were obtained by a standardized semiautomatic method for tumor segmentation by a physician. The nnU-Net framework was used to set up a deep learning U-net architecture. Dice score, sensitivity and precision were used for selection of the final model. AI segmentations were implemented in a clinical imaging viewer where a physician evaluated performance and performed manual adjustments. RESULTS: Cross-validation training was used to generate models and an ensemble model. The ensemble model performed best overall with a lesion-wise dice of 0.850 and pixel-wise dice, precision and sensitivity of 0.801, 0.786 and 0.872, respectively. Performance of the ensemble model was acceptable with some degree of manual adjustment in 35/41 (85%) patients. Final tumor segmentation could be obtained from the AI model with manual adjustments in 5 min versus 17 min for ground truth method, p < 0.01. CONCLUSION: We implemented and validated an AI model that achieved a high similarity with ground truth segmentation and resulted in faster tumor segmentation. With AI, total tumor segmentation may become feasible in the clinical routine.

High-dose versus standard-dose twice-daily thoracic radiotherapy for patients with limited stage small-cell lung cancer: an open-label, randomised, phase 2 trial.

BACKGROUND: Concurrent chemoradiotherapy is standard treatment for limited stage small-cell lung cancer (SCLC). Twice-daily thoracic radiotherapy of 45 Gy in 30 fractions is considered to be the most effective schedule. The aim of this study was to investigate whether high-dose, twice-daily thoracic radiotherapy of 60 Gy in 40 fractions improves survival. METHODS: This open-label, randomised, phase 2 trial was done at 22 public hospitals in Norway, Denmark, and Sweden. Patients aged 18 years and older with treatment-naive confirmed limited stage SCLC, Eastern Cooperative Oncology Group (ECOG) performance status 0-2, and measurable disease according to the Response Evaluation Criteria in Solid Tumors version 1.1 were eligible. All participants received four courses of intravenous cisplatin 75 mg/m2 or carboplatin (area under the curve 5-6 mg/mL × min, Calvert's formula) on day 1 and intravenous etoposide 100 mg/m2 on days 1-3 every 3 weeks. Participants were randomly assigned (1:1) in permuted blocks (sized between 4 and 10) stratifying for ECOG performance status, disease stage, and presence of pleural effusion to receive thoracic radiotherapy of 45 Gy in 30 fractions or 60 Gy in 40 fractions to the primary lung tumour and PET-CT positive lymph node metastases starting 20-28 days after the first chemotherapy course. Patients in both groups received two fractions per day, ten fractions per week. Responders were offered prophylactic cranial irradiation of 25-30 Gy. The primary endpoint, 2-year overall survival, was assessed after all patients had been followed up for a minimum of 2 years. All randomly assigned patients were included in the efficacy analyses, patients commencing thoracic radiotherapy were included in the safety analyses. Follow-up is ongoing. This trial is registered at ClinicalTrials.gov, NCT02041845. FINDINGS: Between July 8, 2014, and June 6, 2018, 176 patients were enrolled, 170 of whom were randomly assigned to 60 Gy (n=89) or 45 Gy (n=81). Median follow-up for the primary analysis was 49 months (IQR 38-56). At 2 years, 66 (74·2% [95% CI 63·8-82·9]) patients in the 60 Gy group were alive, compared with 39 (48·1% [36·9-59·5]) patients in the 45 Gy group (odds ratio 3·09 [95% CI 1·62-5·89]; p=0·0005). The most common grade 3-4 adverse events were neutropenia (72 [81%] of 89 patients in the 60 Gy group vs 62 [81%] of 77 patients in the 45 Gy group), neutropenic infections (24 [27%] vs 30 [39%]), thrombocytopenia (21 [24%] vs 19 [25%]), anaemia (14 [16%] vs 15 [20%]), and oesophagitis (19 [21%] vs 14 [18%]). There were 55 serious adverse events in 38 patients in the 60 Gy group and 56 serious adverse events in 44 patients in the 45 Gy group. There were three treatment-related deaths in each group (one neutropenic fever, one aortic dissection, and one pneumonitis in the 60 Gy group; one thrombocytic bleeding, one cerebral infarction, and one myocardial infarction in the 45 Gy group). INTERPRETATION: The higher radiotherapy dose of 60 Gy resulted in a substantial survival improvement compared with 45 Gy, without increased toxicity, suggesting that twice-daily thoracic radiotherapy of 60 Gy is an alternative to existing schedules. FUNDING: The Norwegian Cancer Society, The Liaison Committee for Education, Research and Innovation in Central Norway, the Nordic Cancer Union, and the Norwegian University of Science and Technology.

Phase II study of a 3-day schedule with topotecan and cisplatin in patients with previously untreated small cell lung cancer and extensive disease.

INTRODUCTION: Treatment with a topoisomerase I inhibitor in combination with a platinum results in superior or equal survival compared with etoposide-based treatment in extensive disease small cell lung cancer (SCLC). Five-day topotecan is inconvenient and therefore shorter schedules of topotecan and cisplatin are needed. The aim of this phase II study was to establish the response rate and response duration in chemo-naive patients with SCLC receiving a 3-day topotecan and cisplatin schedule. METHODS: Simons optimal two-stage design was used. Patients with previously untreated extensive disease SCLC, adequate organ functions and performance status less than 3 were eligible. Topotecan (2.0 mg/m, intravenously) was administered on days 1 to 3 with cisplatin (50 mg/m, intravenously) on day 3 every 3 weeks for a total of six cycles. RESULTS: Forty-three patients received 219 cycles of chemotherapy. Median age was 59 (range 44-74), 79% had performance status 0 or 1. Thirty-one patients completed all six cycles. Grade 3/4 anemia, neutrocytopenia, and thrombocytopenia were recorded in 9.5%, 66.7%, and 21.4% of patients, respectively. Fourteen percent of patients experienced neutropenic fever. No episodes of fatal sepsis occurred. Non-hematologic toxicity was mild and manageable. Overall and complete response rates were 72.1% and 9.3%, respectively. The median overall survival and response duration were 10.3 months (95% confidence interval: 8.6-12.0) and 7.0 months (95% confidence interval: 6.3-7.7), respectively. CONCLUSION: Three-day topotecan with cisplatin on day 3 is active and safe in extensive disease SCLC. An ongoing phase III randomized trial compares this combination to standard treatment.