Deep learning-assisted interactive contouring of lung cancer: Impact on contouring time and consistency.

BACKGROUND AND PURPOSE: To evaluate the impact of a deep learning (DL)-assisted interactive contouring tool on inter-observer variability and the time taken to complete tumour contouring. MATERIALS AND METHODS: Nine clinicians contoured the gross tumour volume (GTV) using the PET-CT scans of 10 non-small cell lung cancer (NSCLC) patients, either using DL-assisted or manual contouring tools. After contouring a case using one contouring method, the same case was contoured one week later using the other method. The contours and time taken were compared. RESULTS: Use of the DL-assisted tool led to a statistically significant decrease in active contouring time of 23 % relative to the standard manual segmentation method (p < 0.01). The mean observation time for all clinicians and cases made up nearly 60 % of interaction time for both contouring approaches. On average the time spent contouring per case was reduced from 22 min to 19 min when using the DL-assisted tool. Additionally, the DL-assisted tool reduced contour variability in the parts of tumour where clinicians tended to disagree the most, while the consensus contour was similar whichever of the two contouring approaches was used. CONCLUSIONS: A DL-assisted interactive contouring approach decreased active contouring time and local inter-observer variability when used to delineate lung cancer GTVs compared to a standard manual method. Integration of this tool into the clinical workflow could assist clinicians in contouring tasks and improve contouring efficiency.

Real-World Impact of SABR on Stage I Non-Small-Cell Lung Cancer Outcomes at a Scottish Cancer Centre.

INTRODUCTION: Stereotactic ablative body radiotherapy (SABR) offers patients with stage I non-small-cell lung cancer (NSCLC) a safe, effective radical therapy option. The impact of introducing SABR at a Scottish regional cancer centre was studied. METHODS: The Edinburgh Cancer Centre Lung Cancer Database was assessed. Treatment patterns and outcomes were compared across treatment groups (no radical therapy (NRT), conventional radical radiotherapy (CRRT), SABR and surgery) and across three time periods reflecting the availability of SABR (A, January 2012/2013 (pre-SABR); B, 2014/2016 (introduction of SABR); C, 2017/2019, (SABR established)). RESULTS: 1143 patients with stage I NSCLC were identified. Treatment was NRT in 361 (32%), CRRT in 182 (16%), SABR in 132 (12%) and surgery in 468 (41%) patients. Age, performance status, and comorbidities correlated with treatment choice. The median survival increased from 32.5 months in time period A to 38.8 months in period B to 48.8 months in time period C. The greatest improvement in survival was seen in patients treated with surgery between time periods A and C (HR 0.69 (95% CI 0.56-0.86), p < 0.001). The proportion of patients receiving a radical therapy rose between time periods A and C in younger (age ≤ 65, 65-74 and 75-84 years), fitter (PS 0 and 1), and less comorbid patients (CCI 0 and 1-2), but fell in other patient groups. CONCLUSIONS: The introduction and establishment of SABR for stage I NSCLC has improved survival outcomes in Southeast Scotland. Increasing SABR utilisation appears to have enhanced the selection of surgical patients and increased the proportion of patients receiving a radical therapy.

Hypoalbuminaemia as a Prognostic Biomarker of First-Line Treatment Resistance in Metastatic Non-small Cell Lung Cancer.

Introduction: Despite significant advances in systemic anticancer therapy (SACT) for non-small cell lung cancer (NSCLC), many patients still fail to respond to treatment or develop treatment resistance. Albumin, a biomarker of systemic inflammation and malnutrition, predicts survival in many cancers. We evaluated the prognostic significance of albumin in patients receiving first-line targeted therapy or immunotherapy-based SACT for metastatic NSCLC. Methods: All patients treated with first-line targeted therapy or immunotherapy-based SACT for metastatic NSCLC at a regional Scottish cancer centre were identified. Serum albumin at pre-treatment, after 12-weeks of treatment, and at the time of progressive disease were recorded. The relationship between albumin (≥ 35g/L v <35g/L) and overall survival (OS) was examined. Results: Data were available for 389 patients of both targeted therapy cohort (n = 159) and immunotherapy-based therapy cohort (n = 230). Pre-treatment albumin was predictive of OS in each cohort at HR1.82 (95%CI 1.23-2.7) (p =0.003) and HR2.55 (95%CI 1.78-3.65) (p < 0.001), respectively. Pre-treatment albumin <35 g/L was associated with a significantly higher relative risk of death within 12 weeks in each cohort at RR9.58 (95%CI 2.20-41.72, p = 0.003) and RR3.60 (95%CI 1.74-6.57, p < 0.001), respectively. The 12-week albumin was predictive of OS in each cohort at HR1.88 (95%CI 1.86-4.46) (p < 0.001) and HR2.67 (95%CI 1.74-4.08) (p < 0.001), respectively. 46 out of 133 (35%) evaluable patients treated with targeted therapy and 43 out of 169 (25%) treated with immunotherapy-based therapy crossed over albumin prognostic groups between pre-treatment and 12-week. The prognostic value of 12-week albumin was independent of pre-treatment albumin status. A majority of patients had albumin <35g/L at the time of progressive disease when it was also predictive of survival following progressive disease at HR2.48 (95%CI 1.61-3.82) (p < 0.001) and HR2.87 (95%CI 1.91-4.31) (p < 0.001) respectively). Conclusions: Albumin is a reliable prognostic factor in patients with metastatic NSCLC, predicting survival independent of the class of drug treatment at various time points during the patient journey. Tracking albumin concentrations during systemic therapy may indicate disease activity or treatment response over time.

Evaluation of microscopic disease in oral tongue cancer using whole-mount histopathologic techniques: implications for the management of head-and-neck cancers.

PURPOSE: To map the distribution of microscopic disease (MD) in head-and-neck cancer by analyzing digital images of whole-mounted serial sections of tongue cancer specimens. METHODS AND MATERIALS: Ten T1-3 oral tongue cancer specimens were evaluated. The specimens were sliced into 3-mm blocks from which one or more 4-µm slides were taken and digitized to create whole-mounted serial sections. Gross tumor and microscopic disease were digitally contoured on each slide. Lines perpendicular to the gross tumor volume (GTV) edge were created at 0.05-mm intervals and the distance between GTV and MD measured. RESULTS: Of 88 slides assessed, 44 (50%) had evidence of MD. Of the 63,809 perpendicular lines drawn along the GTV edges, 2320 (3.6%) encountered microscopic disease along their path. The majority of MD abutted the GTV, and only 26.7% was noncontiguous with the GTV edge. The maximum distance from the border was 7.8 mm. Ninety-nine percent of all MD was within 4.75 mm and 95% was within 3.95 mm of the GTV. CONCLUSION: In this study we were able to assess the distribution of MD more accurately than has been possible with routine pathologic techniques. The results indicate that when the GTV is correctly identified, there is very little MD to be found outside this volume. This has implications for the volume of tissue resected at surgery and the volume included in the clinical target volume in conformal radiotherapy planning.