Assessment of the current and emerging criteria for the histopathological classification of lung neuroendocrine tumours in the lungNENomics project.

BACKGROUND: Six thoracic pathologists reviewed 259 lung neuroendocrine tumours (LNETs) from the lungNENomics project, with 171 of them having associated survival data. This cohort presents a unique opportunity to assess the strengths and limitations of current World Health Organization (WHO) classification criteria and to evaluate the utility of emerging markers. PATIENTS AND METHODS: Patients were diagnosed based on the 2021 WHO criteria, with atypical carcinoids (ACs) defined by the presence of focal necrosis and/or 2-10 mitoses per 2 mm2. We investigated two markers of tumour proliferation: the Ki-67 index and phospho-histone H3 (PHH3) protein expression, quantified by pathologists and automatically via deep learning. Additionally, an unsupervised deep learning algorithm was trained to uncover previously unnoticed morphological features with diagnostic value. RESULTS: The accuracy in distinguishing typical from ACs is hampered by interobserver variability in mitotic counting and the limitations of morphological criteria in identifying aggressive cases. Our study reveals that different Ki-67 cut-offs can categorise LNETs similarly to current WHO criteria. Counting mitoses in PHH3+ areas does not improve diagnosis, while providing a similar prognostic value to the current criteria. With the advantage of being time efficient, automated assessment of these markers leads to similar conclusions. Lastly, state-of-the-art deep learning modelling does not uncover undisclosed morphological features with diagnostic value. CONCLUSIONS: This study suggests that the mitotic criteria can be complemented by manual or automated assessment of Ki-67 or PHH3 protein expression, but these markers do not significantly improve the prognostic value of the current classification, as the AC group remains highly unspecific for aggressive cases. Therefore, we may have exhausted the potential of morphological features in classifying and prognosticating LNETs. Our study suggests that it might be time to shift the research focus towards investigating molecular markers that could contribute to a more clinically relevant morpho-molecular classification.

Osimertinib treatment based on plasma T790M monitoring in patients with EGFR-mutant non-small-cell lung cancer (NSCLC): EORTC Lung Cancer Group 1613 APPLE phase II randomized clinical trial.

BACKGROUND: The APPLE trial aimed to evaluate the feasibility of longitudinal plasma epidermal growth factor receptor (EGFR) T790M monitoring for the best sequencing strategy of gefitinib and osimertinib. METHODS: APPLE is a randomized, non-comparative, phase II study in patients with common EGFR-mutant, treatment-naive non-small-cell lung cancer including three arms: arm A (osimertinib upfront until RECIST progression, PD), arm B [gefitinib until emergence of circulating tumor DNA (ctDNA) EGFR T790M mutation by cobas EGFR test v2 or RECIST PD], and arm C (gefitinib until RECIST PD), and then switch to osimertinib in both arms. The primary endpoint is the progression-free survival (PFS) rate 'on osimertinib' at 18 months (PFSR-OSI-18) after randomization in arm B (H0: PFSR-OSI-18 of ≤40%). Secondary endpoints include response rate, overall survival (OS), and brain PFS. We report the results of arms B and C. RESULTS: From November 2017 to February 2020, 52 and 51 patients were randomized into arms B and C, respectively. Most patients were females (70%) and had EGFR Del19 (65%); one-third had baseline brain metastases. In arm B, 17% of patients (8/47) switched to osimertinib based on the emergence of ctDNA T790M mutation before RECIST PD, with a median time to molecular PD of 266 days. The study met its primary endpoint of PFSR-OSI-18 of 67.2% (84% confidence interval 56.4% to 75.9%) in arm B versus 53.5% (84% confidence interval 42.3% to 63.5%) in arm C, with a median PFS of 22.0 months versus 20.2 months, respectively. The median OS was not reached in arm B versus 42.8 months in arm C. Median brain PFS in arms B and C was 24.4 and 21.4 months, respectively. CONCLUSIONS: The serial monitoring of ctDNA T790M status in advanced EGFR-mutant non-small-cell lung cancer during treatment with first-generation EGFR inhibitors was feasible, and a molecular progression before RECIST PD led to an earlier switch to osimertinib in 17% of patients with satisfactory PFS and OS outcomes.

ESMO expert consensus statements on the management of EGFR mutant non-small-cell lung cancer.

The European Society for Medical Oncology (ESMO) held a virtual consensus-building process on epidermal growth factor receptor (EGFR)-mutant non-small-cell lung cancer in 2021. The consensus included a multidisciplinary panel of 34 leading experts in the management of lung cancer. The aim of the consensus was to develop recommendations on topics that are not covered in detail in the current ESMO Clinical Practice Guideline and where the available evidence is either limited or conflicting. The main topics identified for discussion were: (i) tissue and biomarkers analyses; (ii) early and locally advanced disease; (iii) metastatic disease and (iv) clinical trial design, patient's perspective and miscellaneous. The expert panel was divided into four working groups to address questions relating to one of the four topics outlined above. Relevant scientific literature was reviewed in advance. Recommendations were developed by the working groups and then presented to the entire panel for further discussion and amendment before voting. This manuscript presents the recommendations developed, including findings from the expert panel discussions, consensus recommendations and a summary of evidence supporting each recommendation.

Early Loss of Fat Mass During Chemoradiotherapy Predicts Overall Survival in Locally Advanced Squamous Cell Carcinoma of the Lung, but Not in Locally Advanced Squamous Cell Carcinoma of the Head and Neck.

Background: Cancer cachexia is highly prevalent in advanced non-small cell lung cancer (NSCLC) and locally advanced head and neck squamous cell carcinoma (LAHNSCC), and compromises treatment tolerance and overall survival (OS). NSCLC and LAHNSCC patients share similar risk factors, and receive comparable anti-cancer treatment regimens. The aim of this study was to determine the predictive value of body composition assessed by bioelectrical impedance analysis (BIA) and handgrip strength (HGS) (baseline and early changes during therapy) on OS in NSCLC and LAHNSCC patients treated with platinum-based chemoradiotherapy (CRT) or cetuximab-based bioradiotherapy (BRT). To elucidate potential underlying determinants of early changes in body composition and HGS, specific (fat and fat free) mass loss patterns of squamous NSCLC (sNSCLC) were compared to human papilloma virus negative (HPV-) LAHNSCC patients treated with CRT. Methods: Between 2013 and 2016, BIA and HGS were performed at baseline and after 3 weeks of CRT/BRT in LAHNSCC and NSCLC patients treated with curative intent. Results: Two hundred thirty-three patients were included for baseline measurements. Fat free mass index (FFMI) and HGS<10th percentile of reference values at baseline were both prognostic for poor OS in NSCLC and LAHNSCC [HR 1.64 [95%CI 1.13-2.39], p = 0.01 and HR 2.30 [95%CI 1.33-3.97], p = 0.003, respectively], independent of Charlson Comorbidity Index, cancer site, and gross tumor volume. Early fat mass (FM) loss during CRT was predictive for poor OS in sNSCLC (n = 64) [HR 3.80 [95%CI 1.79-8.06] p ≤ 0.001] but not in HPV- LAHNSCC (n = 61). In patients with significant weight loss (>2%) in the first 3 weeks of CRT (sNSCLC n = 24, HPV- LAHNSCC n = 23), the FM change was -1.4 ± 14.5% and -8.7 ± 9.0% in sNSCLC and HPV- LAHNSCC patients, respectively (p < 0.05). Fat fee mass change was -5.6 ± 6.3% and -4.0 ± 4.3% for sNSCLC and HPV- LAHNSCC, respectively (p = 0.31). Conclusion: FFMI and HGS<10th percentile at baseline are independent prognostic factors for poor OS in NSCLC and LAHNSCC patients treated with CRT/BRT. The specific composition of mass loss during first 3 weeks of CRT significantly differs between sNSCLC and HPV- LAHNSCC patients. Early FM loss was prognostic in sNSCLC only.

Pulmonary neuroendocrine neoplasms with well differentiated morphology and high proliferative activity: illustrated by a case series and review of the literature.

OBJECTIVES: Pulmonary neuroendocrine neoplasms (NENs) are subdivided in carcinoids and neuroendocrine carcinomas (small cell lung carcinoma and large cell neuroendocrine carcinoma (LCNEC)), based on the presence of necrosis and mitotic index (MI). However, it is unclear if tumors with well differentiated morphology but high proliferation rate should be regarded as LCNEC or as high grade carcinoids. In previous case series, a longer overall survival then expected in LCNEC has been suggested. We describe 7 of those cases analyzed for pRb expression and overall survival. MATERIAL AND METHODS: Cases with well differentiated morphology, but MI > 10/2mm2 and/or Ki-67 proliferation index >20% were selected based on pathology reports of consecutive NENs in our university medical center (Maastricht UMC+, 2007-2018) and confirmed by pathological review. Immunohistochemistry was performed to assess pRb expression. RESULTS: Seven stage IV cases were included in this study. Median overall survival was 8 months (95% confidence interval 5-11 months). Cases with well differentiated morphology and preserved pRb expression (4/7) had a median overall survival of 45 months. CONCLUSION: A subgroup of pulmonary NENs with well differentiated morphology but high proliferation rate likely exists. pRb staining might be helpful to predict prognosis, but clinical relevance remains to be studied.

Immune checkpoint inhibitors versus second line chemotherapy for patients with lung cancer refractory to first line chemotherapy.

PURPOSE: Anti Programmed Death-ligand (PD1/PD-L1) directed immune-checkpoint-inhibitors (ICI) are widely used to treat patients with advanced non-small cell lung cancer (NSCLC) who progress after first line chemotherapy. The best strategy after early progression under first line has not been specifically studied. PATIENTS AND METHODS: We conducted a multicenter, retrospective study including all consecutive NSCLC patients progressing within the first 3 months following introduction of first-line chemotherapy and being treated with second line ICI monotherapy or chemotherapy between March 2010 and November 2017. We analysed the clinicopathological data and outcome under second line chemotherapy vs. second line ICI: objective response rate (ORR), progression-free survival (PFS), overall survival (OS. RESULTS: We identified 176 patients with refractory disease, 99 who received subsequent immunotherapy and 77 undergoing chemotherapy. The 2 populations were comparable regarding the main prognostic criteria, median age was 60, main histology was adenocarcimoma (68.2%). PFS was not significantly different between both treatments 1.9 [1.8-2.1] versus 1.6 month [1.4-2.0] (P=0.125). Compared to chemotherapy, ICI treated patients had a superior OS (P=0.03) (Median [95% CI] OS 4.6 [2.8-6.7] versus 4.2 months [3.4-5.9] and a non-significant improvement in ORR (17.2% versus 7.9%, respectively, P=0.072). Poor performance status (ECOG PS≥2) and a higher number of metastatic sites (≥3) were associated with poorer prognosis. KRAS-mutated patients did not seem to benefit more from ICI than chemotherapy. CONCLUSIONS: ICI appears to be the preferred second-line treatment for patients who are refractory to first line chemotherapy.

Exploring imaging features of molecular subtypes of large cell neuroendocrine carcinoma (LCNEC).

OBJECTIVES: Radiological characteristics and radiomics signatures can aid in differentiation between small cell lung carcinoma (SCLC) and non-small cell lung carcinoma (NSCLC). We investigated whether molecular subtypes of large cell neuroendocrine carcinoma (LCNEC), i.e. SCLC-like (with pRb loss) vs. NSCLC-like (with pRb expression), can be distinguished by imaging based on (1) imaging interpretation, (2) semantic features, and/or (3) a radiomics signature, designed to differentiate between SCLC and NSCLC. MATERIALS AND METHODS: Pulmonary oncologists and chest radiologists assessed chest CT-scans of 44 LCNEC patients for 'small cell-like' or 'non-small cell-like' appearance. The radiologists also scored semantic features of 50 LCNEC scans. Finally, a radiomics signature was trained on a dataset containing 48 SCLC and 76 NSCLC scans and validated on an external set of 58 SCLC and 40 NSCLC scans. This signature was applied on scans of 28 SCLC-like and 8 NSCLC-like LCNEC patients. RESULTS: Pulmonary oncologists and radiologists were unable to differentiate between molecular subtypes of LCNEC and no significant differences in semantic features were found. The area under the receiver operating characteristics curve of the radiomics signature in the validation set (SCLC vs. NSCLC) was 0.84 (95% confidence interval (CI) 0.77-0.92) and 0.58 (95% CI 0.29-0.86) in the LCNEC dataset (SCLC-like vs. NSCLC-like). CONCLUSION: LCNEC appears to have radiological characteristics of both SCLC and NSCLC, irrespective of pRb loss, compatible with the SCLC-like subtype. Imaging interpretation, semantic features and our radiomics signature designed to differentiate between SCLC and NSCLC were unable to separate molecular LCNEC subtypes, which underscores that LCNEC is a unique disease.

Effects of checkpoint inhibitors in advanced non-small cell lung cancer at population level from the National Immunotherapy Registry.

OBJECTIVE: Phase III studies of checkpoint inhibitors changed the therapeutic landscape for lung cancer. In 2015 the Dutch Society of Chest Physicians (NVALT) introduced a national immunotherapy registry for patients with lung cancer; quality standards for hospitals were implemented. At population level we studied clinical benefit in daily practice and in patients who are underrepresented in phase III trials. MATERIALS AND METHODS: From the initial introduction of checkpoint inhibitors in the Netherlands patients were centrally registered. Educational programs and quality control were provided under supervision of NVALT. The largest immunotherapy providing hospitals were compared to hospitals who provided less checkpoint inhibitors as marker of experience. Patients characteristics, treatment and side effects, response rate and survival were studied. RESULTS: A total of 2676 patients were registered, 2302 with follow up data were evaluated. Between October 2015 and December 2017 a gradual increase from 12 to 30 qualified hospitals showed no major toxicity differences. Toxicity led to a hospital admission rate of 9.1 with an average duration of 10.4 days. Overall tumor response was 21.8 % and median overall survival 12.6 months. Overall survival was not significantly different for patients aged ≥ 75 years, those having brain metastases or selected auto-immune diseases before start checkpoint inhibitors compared to younger patients or those without, respectively. Survival outcomes were worse in patients with PS 2+, non-smokers, and patients who received any palliative radiotherapy (HR 2.1, 95 % CI 1.7-2.7; 1.3, 95 % CI 1.0-1.6 and 1.2, 95 % CI 1.1-1.4, respectively). CONCLUSIONS: Changes in the therapeutic landscape did not lead to major differences in quality of care between hospitals. Elderly patients, those with brain metastases or selected auto-immune disease underrepresented in clinical trials did not do worse on checkpoint inhibitors, except for those with PS 2 + .

Highly accurate DNA-based detection and treatment results of MET exon 14 skipping mutations in lung cancer.

OBJECTIVES: The oncogenic MET exon 14 skipping mutation (METex14del) is described to drive 1.3 %-5.7 % of non-small-cell lung cancer (NSCLC) and multiple studies with cMET inhibitors show promising clinical responses. RNA-based analysis seems most optimal for METex14del detection, however, acquiring sufficient RNA material is often problematic. An alternative is DNA-based analysis, but commercially available DNA-based panels only detect up to 63 % of known METex14del alterations. The goal of this study is to describe an optimized DNA-based diagnostic test for METex14del in NSCLC, including clinical features and follow-up of patients treated with cMET-targeted therapy and consequent resistance mechanisms. MATERIAL AND METHODS: Routinely processed diagnostic pathology non-squamous NSCLC specimens were investigated by a custom-made DNA-based targeted amplicon-based next generation sequencing (NGS) panel, which includes 4 amplicons for METex14del detection. Retrospectively, histopathological characteristics and clinical follow up were investigated for advanced non-squamous NSCLC with METex14del. RESULTS: In silico analysis showed that our NGS panel is able to detect 96 % of reported METex14 alterations. METex14del was found in 2 % of patients with non-squamous NSCLC tested for therapeutic purposes. In total, from May 2015 - Sep 2018, METex14del was found in 46 patients. Thirty-six of these patients had advanced non-squamous NSCLC, they were predominantly elderly (76.5 years [53-90]), male (25/36) and (ex)-smokers (23/36). Five patients received treatment with crizotinib (Pfizer Oncology), in a named patient based program, disease control was achieved for 4/5 patients (3 partial responses, 1 stable disease) and one patient had a mixed response. Two patients developed a MET D1228N mutation during crizotinib treatment, inducing a resistance mechanism to crizotinib. CONCLUSIONS: This study shows that METex14del can be reliably detected by routine DNA NGS analysis. Although a small cohort, patients responded well to targeted treatment, underlining the need for routine testing of METex14del in advanced non-squamous NSCLC to guarantee optimal personalized treatment.

Health-related quality of life after prophylactic cranial irradiation for stage III non-small cell lung cancer patients: Results from the NVALT-11/DLCRG-02 phase III study.

BACKGROUND AND PURPOSE: The NVALT-11/DLCRG-02 phase III trial (clinicaltrials.gov identifier: NCT01282437) showed that, after standard curative intent treatment, prophylactic cranial irradiation (PCI) decreased the incidence of symptomatic brain metastases (BM) in stage III non-small cell lung cancer (NSCLC) patients compared to observation. In this study we assessed the impact of PCI on health-related quality of life (HRQoL). In addition, an exploratory analysis was performed to assess the impact of neurocognitive symptoms and symptomatic BM on HRQoL. MATERIALS AND METHODS: Stage III NSCLC patients were randomized between PCI and observation. HRQoL was measured using the EuroQol 5D (EQ-5D-3L), EORTC QLQ-C30 and QLQ-BN20 instruments at completion of standard curative intent treatment and 4 weeks, 3, 6, 12, 24 and 36 months thereafter. Generalized linear mixed effects (GLM) models were used to assess the impact of PCI compared to observation over time on three HRQoL metrics: the EORTC QLQ-C30 global health status and the EQ-5D-3L utility and visual analogue scale (EQ VAS) scores. RESULTS: In total, 86 and 88 patients were included in the PCI and observation arm, with a median follow-up of 48.5 months (95% CI 39-54 months). Baseline mean HRQoL scores were comparable between the PCI and observation arm for the three HRQoL metrics. In the GLM models, none of the HRQoL metrics were clinically relevant or statistically significantly different between the PCI and the observation arm (p-values ranged between 0.641 and 0.914). CONCLUSION: No statistically significant nor a clinically relevant impact of PCI on HRQoL was observed.

Large cell neuroendocrine carcinoma with a solitary brain metastasis and low Ki-67: a unique subtype.

INTRODUCTION: Stage IV large cell neuroendocrine carcinoma (LCNEC) of the lung generally presents as disseminated and aggressive disease with a Ki-67 proliferation index (PI) 40-80%. LCNEC can be subdivided in two main subtypes: the first harboring TP53/RB1 mutations (small-cell lung carcinoma (SCLC)-like), the second with mutations in TP53 and STK11/KEAP1 (non-small-cell lung carcinoma (NSCLC)-like). Here we evaluated 11 LCNEC patients with only a solitary brain metastasis and evaluate phenotype, genotype and follow-up. METHODS: Eleven LCNEC patients with solitary brain metastases were analyzed. Clinical characteristics and survival data were retrieved from medical records. Pathological analysis included histomorphological analysis, immunohistochemistry (pRB and Ki-67 PI) and next-generation sequencing (TP53, RB1, STK11, KEAP1 and MEN1). RESULTS: All patients had N0 or N1 disease. Median overall survival (OS) was 12 months (95% confidence interval (CI) 5.5-18.5 months). Mean Ki-67 PI was 59% (range 15-100%). In 6/11 LCNEC Ki-67 PI was ≤40%. OS was longer for Ki-67 ≤40% compared to >40% (17 months (95% CI 11-23 months) vs 5 months (95% CI 0.7-9 months), P = 0.007). Two patients were still alive at follow-up after 86 and 103 months, both had Ki-67 ≤40%. 8/11 patients could be subclassified, and both SCLC-like (n = 6) and NSCLC-like (n = 2) subtypes were present. No MEN1 mutation was found. CONCLUSION: Stage IV LCNEC with a solitary brain metastasis and N0/N1 disease show in the majority of cases Ki-67 PI ≤40% and prolonged survival, distinguishing them from general LCNEC. This unique subgroup can be both of the SCLC-like and NSCLC-like subtype.

DLL3 expression in large cell neuroendocrine carcinoma (LCNEC) and association with molecular subtypes and neuroendocrine profile.

OBJECTIVES: For stage IV pulmonary large cell neuroendocrine carcinoma (LCNEC), the only therapeutic option is palliative chemotherapy. DLL3 is a new therapeutic target, which seems to be often expressed in SCLC and LCNEC. It has recently been reported that DLL3 mRNA expression is particularly upregulated in the LCNEC subgroup with STK11/KEAP1 and TP53 co-mutations, in contrast to lower expression levels in RB1 and TP53 co-mutated LCNEC. Our aim was to investigate DLL3 protein expression in stage IV LCNEC and correlate data with mutational profiles (i.e.STK11/KEAP1/RB1), immunostaining results (pRb, NE markers) and clinical characteristics. MATERIALS AND METHODS: Immunohistochemical analysis for DLL3 (SC16.65) and ASCL1 (SC72.201) was performed on 94 and 51 FFPE tissue sections, respectively, of pathologically reviewed stage IV LCNEC. DLL3 and ASCL1 were scored positive if ≥1% of the tumor cells showed cytoplasmic/membranous or dotlike (DLL3) or nuclear (ASCL1) immunostaining. Data were correlated with available sequencing (TP53, RB1, STK11, KEAP1), immunostaining (pRb, NE markers) and clinical data. RESULTS: DLL3 was expressed in 70/94 (74%) LCNEC, 56 (80%) of which showed cytoplasmic/membranous staining. Median H-score was 55 (interquartile range 0-160). DLL3 staining was not different in pRb immunohistochemistry negative and positive patients (DLL3+ in 53/70 (76%) vs. 14/21 (67%), p = 0.409) or RB1 mutated and wildtype patients (DLL3+ in 27/34 (79%) vs. 23/33 (70%), p = 0.361). Nevertheless, 6/6 (100%) STK11 mutated, 10/11 (91%) KEAP1 mutated and 9/9 (100%) TP53 wildtype tumors were DLL3+ . Furthermore, DLL3 expression was associated with expression of ASCL1 and at least 2 out of 3 neuroendocrine markers. CONCLUSION: The high percentage (74%) of DLL3 expression in stage IV LCNEC denotes the potential of DLL3 targeted therapy in this patient group.

ALK immunohistochemistry positive, FISH negative NSCLC is infrequent, but associated with impaired survival following treatment with crizotinib.

OBJECTIVE: Metastasized non-small cell lung cancer (NSCLC) with an anaplastic lymphoma kinase (ALK) rearrangement is usually sensitive to a range of ALK-tyrosine kinase inhibitors. ALK-positive NSCLC have been identified in pivotal phase III trials with fluorescence in situ hybridization (ALK FISH+). These tumors are also expressing the fusion product (ALK immunohistochemistry (IHC)+). However, discrepant cases occur, including ALK IHC + FISH-. The aim of this study was to collect ALK IHC + cases and compare within this group response to crizotinib treatment of ALK FISH + cases with ALK FISH- cases. MATERIALS AND METHODS: In this European prospective multicenter research study patients with Stage IV ALK IHC + NSCLC treated with crizotinib were enrolled. Tumor slides were validated centrally for ALK IHC and ALK FISH. RESULTS: Registration of 3523 ALK IHC tests revealed a prevalence of 2.7% (n = 94) ALK IHC + cases. Local ALK FISH analysis resulted in 48 concordant (ALK IHC+/FISH+) and 16 discordant (ALK IHC+/FISH-) cases. Central validation revealed 37 concordant and 7 discordant cases, 5 of which had follow-up. Validation was hampered by limited amount of tissue in biopsy samples. The PFS at 1 year for ALK concordant and discordant was 58% and 20%, respectively (HR = 2.4; 95% CI: 0.78-7.3; p = 0.11). Overall survival was significantly better for concordant cases than discordant cases after central validation (HR=4.5; 95% CI= 1.2-15.9; p=0.010. CONCLUSION: ALK IHC + FISH- NSCLC is infrequent and associated with a worse outcome on personalized treatment. A suitable predictive testing strategy may be to screen first with IHC and then confirm with FISH instead of considering ALK IHC equivalent to ALK FISH according to the current guidelines.

Can radiomics help to predict skeletal muscle response to chemotherapy in stage IV non-small cell lung cancer?

BACKGROUND: Muscle depletion negatively impacts treatment efficacy and survival rates in cancer. Prevention and timely treatment of muscle loss require prediction of patients at risk. We aimed to investigate the potential of skeletal muscle radiomic features to predict future muscle loss. METHODS: A total of 116 patients with stage IV non-small cell lung cancer included in a randomised controlled trial (NCT01171170) studying the effect of nitroglycerin added to paclitaxel-carboplatin-bevacizumab were enrolled. In this post hoc analysis, muscle cross-sectional area and radiomic features were extracted from computed tomography images obtained before initiation of chemotherapy and shortly after administration of the second cycle. For internal cross-validation, the cohort was randomly split in a training set and validation set 100 times. We used least absolute shrinkage and selection operator method to select features that were most significantly associated with muscle loss and an area under the curve (AUC) for model performance. RESULTS: Sixty-nine patients (59%) exhibited loss of skeletal muscle. One hundred ninety-three features were used to construct a prediction model for muscle loss. The average AUC was 0.49 (95% confidence interval [CI]: 0.36, 0.62). Differences in intensity and texture radiomic features over time were seen between patients with and without muscle loss. CONCLUSIONS: The present study shows that skeletal muscle radiomics did not predict future muscle loss during chemotherapy in non-small cell lung cancer. Differences in radiomic features over time might reflect myosteatosis. Future imaging analysis combined with muscle tissue analysis in patients and in experimental models is needed to unravel the biological processes linked to the radiomic features.