Background Persistent smoking after cancer diagnosis is associated with increased overall mortality (OM) and cancer mortality (CM). According to the 2020 Surgeon General's report, smoking cessation may reduce CM but supporting evidence is not wide. Use of deep learning-based modeling that enables universal natural language processing of medical narratives to acquire population-based real-life smoking data may help overcome the challenge. We assessed the effect of smoking status and within-1-year smoking cessation on CM by an in-house adapted freely available language processing algorithm. Materials and methods This cross-sectional real-world study included 29 823 patients diagnosed with cancer in 2009-2018 in Southwest Finland. The medical narrative, International Classification of Diseases-10th edition codes, histology, cancer treatment records, and death certificates were combined. Over 162 000 sentences describing tobacco smoking behavior were analyzed with ULMFiT and BERT algorithms. Results The language model classified the smoking status of 23 031 patients. Recent quitters had reduced CM [hazard ratio (HR) 0.80 (0.74-0.87)] and OM [HR 0.78 (0.72-0.84)] compared to persistent smokers. Compared to never smokers, persistent smokers had increased CM in head and neck, gastro-esophageal, pancreatic, lung, prostate, and breast cancer and Hodgkin's lymphoma, irrespective of age, comorbidities, performance status, or presence of metastatic disease. Increased CM was also observed in smokers with colorectal cancer, men with melanoma or bladder cancer, and lymphoid and myeloid leukemia, but no longer independently of the abovementioned covariates. Specificity and sensitivity were 96%/96%, 98%/68%, and 88%/99% for never, former, and current smokers, respectively, being essentially the same with both models. Conclusions Deep learning can be used to classify large amounts of smoking data from the medical narrative with good accuracy. The results highlight the detrimental effects of persistent smoking in oncologic patients and emphasize that smoking cessation should always be an essential element of patient counseling.

Background: Locoregional recurrence remains a major cause of failure in head and neck squamous cell carcinoma (HNSCC). Human papilloma virus (HPV)-associated HNSCCs generally have a good prognosis but may recur even after standard photon radiotherapy (RT). Another incentive in observing patterns of recurrence is increased use of highly conformal techniques such as proton therapy. We therefore studied geographic distribution of recurrent tumors in relation to the high-risk treatment volume in a cohort of patients with HNSCC receiving combined modality therapy. Methods: Medical records of 508 patients diagnosed with HNSCC in 2010-2015 were reviewed. We identified a subgroup that had local and/or regional recurrence at hybrid positron emission tomography (PET)/computed tomography (CT) and/or magnetic resonance imaging (MRI). We adapted p16 as a surrogate marker for HPV-positivity and only patients with known p16 status were eligible for a detailed analysis where recurrent tumor was copied on the planning CT and the dose received by the recurrent tumor volume was determined using dose-volume histograms. Results: Twenty-five patients who had received either cisplatin (n = 23) or cetuximab-enhanced (n = 2) RT were identified. 31 locoregional recurrent tumors were detected among 18 p16 negative and 7 p16 positive patients. Of recurrent tumors 14 (45%) were classified as in-field, 5 (16%) as marginal miss, and 12 (39%) as true miss. p16 positive patients had 4 in-field, 2 marginal, and 1 true miss. By contrast, p16 negative patients had 10 in-field, 3 marginal, and 11 true miss recurrences. Conclusions: Both p16 positive and negative HNSCC recur in high-risk treatment volume despite the common view of high radiosensitivity of the former. Biomarkers predicting radioresistance should be characterized in p16 positive tumors before widely embarking on de-escalated CRT protocols. Another concern is how to decrease the number of true or marginal misses in p16 negative cases despite multimodality imaging-based target delineation.

ObjectiveHead and neck cancer follow-up length, interval and content are controversial. Therefore, this study aimed to evaluate the efficacy of the follow-up protocol after curative treatment in head and neck cancer patients.MethodClinical data of 456 patients with new malignancy of the head and neck from a tertiary care centre district from 1999 to 2008 were analysed. Time from treatment, symptoms and second-line treatment outcomes of patients with recurrent disease were evaluated.ResultsA total of 94 (22 per cent) patients relapsed during the 5-year follow-up period; 90 per cent of recurrences were found within 3 years. Fifty-six per cent of the patients had subjective symptoms indicating a recurrence of the tumour. All recurrent tumours found during routine follow-up visits without symptoms were found within 34 months after completion of treatment.ConclusionRoutine follow up after three years is questionable; recurrent disease beyond this point was detected in only 2 per cent of patients. In this study, all late tumour recurrences had symptoms of the disease. Easy access to extra follow-up visits when symptoms occur could cover the need for late follow up.