Clinical and radiological pattern of olaparib-induced interstitial lung disease.

BACKGROUND: PARP inhibitors (PARPi) are used in the treatment of ovarian, breast, pancreatic, and prostate cancers. Pneumonitis has been identified as a potential side effect, with a higher meta-analysis-assessed risk for olaparib versus other PARPi. Olaparib-induced interstitial lung disease (O-ILD) was first described within the Japanese population, with few information available for Caucasian patients. METHODS: We performed a retrospective study by pooling data from the French and Belgian pharmacovigilance databases from 2018 to 2022. Patients with O-ILD were included following a central review by: 1) pharmacologists using the French drug causality assessment method; 2) senior pneumologists or radiologists, using the Fleischner Society's recommendations. RESULTS: Five patients were identified and analysed. All were females, with ovarian or breast cancer. Median age at O-ILD diagnosis was 71 (38-72) years old, with no smoking history. Median delay between treatment initiation and symptom occurrence was 12 (6-33) weeks. Pneumonitis severity assessed using the Common Terminology Criteria for Adverse Events V5 was Grade 3 (n = 4) or 2 (n = 1). CT-scan review (n = 3) described hypersensitivity pneumonitis reaction as a common pattern. Bronchioalveolar lavage (n = 4) revealed lymphocytic alveolitis. Treatments relied on olaparib discontinuation (n = 5) and glucocorticoid intake (n = 4), with no fatal issue. Safe re-challenge with PARPi occurred in two patients. Forty additional O-ILD cases were identified in the WHO VigiBase database, including one fatal case. CONCLUSIONS: PARPi-ILD is a rare but potentially life-threatening disease, presenting as a hypersensitivity pneumonitis pattern within 3 months of PARPi initiation. Treatment primarily relies on medication discontinuation. Re-challenging with another PARPi could be considered. CLINICAL TRIAL NUMBER: CEPRO #2023-010.

Size-matching in lung transplantation by computed tomography for end-stage interstitial disease: a retrospective analysis.

Background: Size matching between donors and recipients is a major issue in lung transplantation (LTx), especially in patients with restrictive lung disease (RLD). This study aims to evaluate computed tomography (CT) as an additional method for defining the total lung capacity (TLC) in patients with end-stage interstitial disease awaiting LTx. Methods: Clinical data and CT scans from patients who underwent a first LTx from January 2014 to July 2018 in Bichat Hospital, Paris, were prospectively included in a database. CT TLC (ctTLC) was retrospectively calculated after semi-automatic contouring of the parenchyma and compared with measured TLC (mTLC) and predicted TLC (pTLC) values. Results: The study group included 89 patients (male:female =68:21; mean age, 59.5±10.0 years). The time between pulmonary function tests (PFTs) and CT scan was 162±270 days [median, 67 days; interquartile range (IQR), 0-233 days]. ctTLC was inferior to mTLC and pTLC (respectively 2,979±1,001 mL, 3,530±1,077 and 6,381±955 mL, P<0.001). The relative difference between CT lung volume (ctLV) and measured lung volume (mLV) was higher on the left than on the right side (25.4% vs. 16.3%, respectively, P=0.11). After exclusion of two outliers, we found a significant correlation between ctTLC and mTLC (r=0.762, P<0.001). Conclusions: CT volume is a feasible method to assess TLC in patients with end-stage interstitial disease awaiting LTx. This study highlights potential size-mismatch for graft selection before LTx and opens the perspective of a prospective trial evaluating impact of size-matching by donor-recipient (D-R) ctTLC ratio on postoperative outcomes.

A YAP-centered mechanotransduction loop drives collective breast cancer cell invasion.

Dense and aligned Collagen I fibers are associated with collective cancer invasion led by protrusive tumor cells, leader cells. In some breast tumors, a population of cancer cells (basal-like cells) maintain several epithelial characteristics and express the myoepithelial/basal cell marker Keratin 14 (K14). Emergence of leader cells and K14 expression are regarded as interconnected events triggered by Collagen I, however the underlying mechanisms remain unknown. Using breast carcinoma organoids, we show that Collagen I drives a force-dependent loop, specifically in basal-like cancer cells. The feed-forward loop is centered around the mechanotransducer Yap and independent of K14 expression. Yap promotes a transcriptional program that enhances Collagen I alignment and tension, which further activates Yap. Active Yap is detected in invading breast cancer cells in patients and required for collective invasion in 3D Collagen I and in the mammary fat pad of mice. Our work uncovers an essential function for Yap in leader cell selection during collective cancer invasion.

Clinical characteristics and outcomes of respiratory syncytial virus-associated ARF in immunocompetent patients: A seven-year experience at a tertiary hospital in France.

BACKGROUND: Respiratory syncytial virus (RSV) is widely recognized as a cause of acute respiratory failure in infants and immunocompromised patients. However, RSV can also contribute to acute respiratory failure in adults, particularly among the elderly population. The objective of this study was to analyze the clinical characteristics and outcomes of immunocompetent adults hospitalized for RSV infection. METHODS: This retrospective study included all immunocompetent adult patients consecutively admitted to a tertiary care hospital with RSV-related acute respiratory failure over a seven-year period (2016-2023). Diagnosis of RSV infection was made through nasal swabs or pulmonary samples, with multiplex reverse transcription polymerase chain reaction (RT-PCR). Patients were eligible for inclusion if they required supplemental oxygen therapy for at least 48 h. RESULTS: One hundred and four patients met the inclusion criteria. Median age [IQR] was 77 years [67-85]. Ninety-seven patients had at least one comorbidity (97/104, 93%). At the time of RSV diagnosis, 67 patients (67/104, 64%) experienced acute decompensation of a pre-existing chronic comorbidity. Antibiotics were started in 80% (77/104) of patients; however, only 16 patients had a confirmed diagnosis of bacterial superinfection. Twenty-six patients needed ventilatory support (26/104, 25%) and 21 were admitted to the intensive care unit (21/104, 20%). The median duration of oxygen therapy [IQR] was 6 days [3-9], while the median hospital length of stay [IQR] was 11 days [6-15]. The overall mortality rate within 1 month of hospital admission was 13% (14/104). The sole variables associated with one-month mortality were age and maximum oxygen flow during hospitalization. CONCLUSION: RSV-associated acute respiratory failure affected elderly individuals with multiple comorbidities and was associated with prolonged hospitalization and a high mortality rate.

Detection and severity quantification of pulmonary embolism with 3D CT data using an automated deep learning-based artificial solution.

PURPOSE: The purpose of this study was to propose a deep learning-based approach to detect pulmonary embolism and quantify its severity using the Qanadli score and the right-to-left ventricle diameter (RV/LV) ratio on three-dimensional (3D) computed tomography pulmonary angiography (CTPA) examinations with limited annotations. MATERIALS AND METHODS: Using a database of 3D CTPA examinations of 1268 patients with image-level annotations, and two other public datasets of CTPA examinations from 91 (CAD-PE) and 35 (FUME-PE) patients with pixel-level annotations, a pipeline consisting of: (i), detecting blood clots; (ii), performing PE-positive versus negative classification; (iii), estimating the Qanadli score; and (iv), predicting RV/LV diameter ratio was followed. The method was evaluated on a test set including 378 patients. The performance of PE classification and severity quantification was quantitatively assessed using an area under the curve (AUC) analysis for PE classification and a coefficient of determination (R²) for the Qanadli score and the RV/LV diameter ratio. RESULTS: Quantitative evaluation led to an overall AUC of 0.870 (95% confidence interval [CI]: 0.850-0.900) for PE classification task on the training set and an AUC of 0.852 (95% CI: 0.810-0.890) on the test set. Regression analysis yielded R² value of 0.717 (95% CI: 0.668-0.760) and of 0.723 (95% CI: 0.668-0.766) for the Qanadli score and the RV/LV diameter ratio estimation, respectively on the test set. CONCLUSION: This study shows the feasibility of utilizing AI-based assistance tools in detecting blood clots and estimating PE severity scores with 3D CTPA examinations. This is achieved by leveraging blood clots and cardiac segmentations. Further studies are needed to assess the effectiveness of these tools in clinical practice.

The extracellular matrix as hallmark of cancer and metastasis: From biomechanics to therapeutic targets.

The extracellular matrix (ECM) is essential for cell support during homeostasis and plays a critical role in cancer. Although research often concentrates on the tumor's cellular aspect, attention is growing for the importance of the cancer-associated ECM. Biochemical and physical ECM signals affect tumor formation, invasion, metastasis, and therapy resistance. Examining the tumor microenvironment uncovers intricate ECM dysregulation and interactions with cancer and stromal cells. Anticancer therapies targeting ECM sensors and remodelers, including integrins and matrix metalloproteinases, and ECM-remodeling cells, have seen limited success. This review explores the ECM's role in cancer and discusses potential therapeutic strategies for cell-ECM interactions.

Splenic Artery Infectious Aneurysms in Infective Endocarditis - An Observational Study and Comprehensive Literature Review.

BACKGROUND: To determine the prevalence, the clinical and radiological features, associated factors, treatment, and outcome of splenic artery aneurysms (SAAs) in infective endocarditis (IE). METHODS: We retrospectively reviewed 474 consecutive patients admitted to our institution with definite IE (2005-2020). RESULTS: Six patients had SAAs (1.3%; 3 women; mean age: 50 years). In all cases, the diagnosis was obtained by abdominal computed tomography angiography (CTA). SAAs-IE were solitary and saccular with a mean diameter of 30 mm (range: 10-90 mm). SAAs-IE were intrasplenic (n = 4) or hilar (n = 2). Streptococcus spp. were the predominant organisms (n = 4). In all cases, a left-sided native valve was involved (aortic, n = 3; mitral, n = 2; mitral-aortic, n = 1). SAAs were silent in half patients and were revealed by abdominal pain (n = 2) and by the resurgence of fever after cardiac surgery (n = 1). All patients underwent emergent valve replacement. One patient died within 24 hr from multiorgan failure. For the others, uneventful coil embolization was performed in 4 patients after valve replacement (3 diagnosed early and 1 at 8 weeks). In the remaining patient, SAA-IE diagnosed at abdominal CTA at day 16, with complete resolution under appropriate antibiotherapy alone. CONCLUSIONS: SAAs-IE are a rare occurrence that may be clinically silent. SAAs-IE can be intrasplenic or hilar in location. Endovascular treatment in this context was safe. According to current guidelines, radiologic screening by abdominal CTA allowed the detection of silent SAAs which could be managed by endovascular treatment to prevent rupture. The delayed formation of these SAAs could justify a CTA control at the end of antibiotherapy.

Hemoptysis: From Diagnosis to Treatment.

Management of hemoptysis begins with an angio-CT to identify the location, the bleeding vessel, mapping of systemic arteries and the cause of the hemoptysis. Endovascular treatment is the first-line therapy, in 90% of cases by embolization of the systemic arteries and in 10% of cases by occlusion of the pulmonary arteries.

Ultrasonography in thoracic and abdominal stab wound injury: results from the FETTHA study.

BACKGROUND: While the role of Extended Focused Assessment with Sonography in Trauma (eFAST) is well defined in the management of severe blunt trauma, its performance in injuries caused by stab wounds has been poorly assessed. METHODS: Prospective single centre study which included all patients with stab wounds to the thorax or abdomen between December 2016 and December 2018. All patients underwent initial investigation with both eFAST and CT scan, except in cases of haemodynamic or respiratory instability, and in cases with a positive diagnosis by eFAST in which case surgery without CT scan was performed. RESULTS: Of the 200 consecutive patients included, 14 unstable patients underwent surgery immediately after eFAST. In these 14 patients, 9 had cardiac tamponade identified by eFAST and all were confirmed by surgery. In the remaining 186 patients, the median time between eFAST and CT scan was 30 min (IQR 20-49 min). Test characteristics (including 95% CI) for eFAST compared with reference standard of CT scan for detecting pneumothorax were as follows: sensitivity 77% (54%-92%), specificity 93% (90%-97%), positive predictive value (PPV) 60% (49%-83%), negative predictive value (NPV) 97% (93%-99%). Test characteristics (including 95% CI) for eFAST compared with CT scan for detecting haemothorax were as follows: sensitivity 97% (74%-99%), specificity 96% (92%-98%), PPV 83% (63%-93%) and NPV 99% (96%-100%). Finally, test characteristics (including 95% CI) for eFAST compared with CT scan for detecting haemoperitoneum were as follows: sensitivity 75% (35%-97%), specificity 97% (93%-99%), PPV 55% (23%-83%) and NPV 99% (96%-99%). CONCLUSIONS: In patients admitted with stab wounds to the torso, eFAST was not sensitive enough to diagnose pneumothorax and haemoperitoneum, but performed better in the detection of cardiac tamponade and haemothorax than the other injuries. More robust multicentre studies are needed to better define the role of eFAST in this specific population.

Detection and quantification of pulmonary embolism with artificial intelligence: The SFR 2022 artificial intelligence data challenge.

PURPOSE: In 2022, the French Society of Radiology together with the French Society of Thoracic Imaging and CentraleSupelec organized their 13th data challenge. The aim was to aid in the diagnosis of pulmonary embolism, by identifying the presence of pulmonary embolism and by estimating the ratio between right and left ventricular (RV/LV) diameters, and an arterial obstruction index (Qanadli's score) using artificial intelligence. MATERIALS AND METHODS: The data challenge was composed of three tasks: the detection of pulmonary embolism, the RV/LV diameter ratio, and Qanadli's score. Sixteen centers all over France participated in the inclusion of the cases. A health data hosting certified web platform was established to facilitate the inclusion process of the anonymized CT examinations in compliance with general data protection regulation. CT pulmonary angiography images were collected. Each center provided the CT examinations with their annotations. A randomization process was established to pool the scans from different centers. Each team was required to have at least a radiologist, a data scientist, and an engineer. Data were provided in three batches to the teams, two for training and one for evaluation. The evaluation of the results was determined to rank the participants on the three tasks. RESULTS: A total of 1268 CT examinations were collected from the 16 centers following the inclusion criteria. The dataset was split into three batches of 310, 580 and 378 C T examinations provided to the participants respectively on September 5, 2022, October 7, 2022 and October 9, 2022. Seventy percent of the data from each center were used for training, and 30% for the evaluation. Seven teams with a total of 48 participants including data scientists, researchers, radiologists and engineering students were registered for participation. The metrics chosen for evaluation included areas under receiver operating characteristic curves, specificity and sensitivity for the classification task, and the coefficient of determination r2 for the regression tasks. The winning team achieved an overall score of 0.784. CONCLUSION: This multicenter study suggests that the use of artificial intelligence for the diagnosis of pulmonary embolism is possible on real data. Moreover, providing quantitative measures is mandatory for the interpretability of the results, and is of great aid to the radiologists especially in emergency settings.

Strategies to safely rule out pulmonary embolism in COVID-19 outpatients: a multicenter retrospective study.

OBJECTIVES: The objective was to define a safe strategy to exclude pulmonary embolism (PE) in COVID-19 outpatients, without performing CT pulmonary angiogram (CTPA). METHODS: COVID-19 outpatients from 15 university hospitals who underwent a CTPA were retrospectively evaluated. D-Dimers, variables of the revised Geneva and Wells scores, as well as laboratory findings and clinical characteristics related to COVID-19 pneumonia, were collected. CTPA reports were reviewed for the presence of PE and the extent of COVID-19 disease. PE rule-out strategies were based solely on D-Dimer tests using different thresholds, the revised Geneva and Wells scores, and a COVID-19 PE prediction model built on our dataset were compared. The area under the receiver operating characteristics curve (AUC), failure rate, and efficiency were calculated. RESULTS: In total, 1369 patients were included of whom 124 were PE positive (9.1%). Failure rate and efficiency of D-Dimer > 500 µg/l were 0.9% (95%CI, 0.2-4.8%) and 10.1% (8.5-11.9%), respectively, increasing to 1.0% (0.2-5.3%) and 16.4% (14.4-18.7%), respectively, for an age-adjusted D-Dimer level. D-dimer > 1000 µg/l led to an unacceptable failure rate to 8.1% (4.4-14.5%). The best performances of the revised Geneva and Wells scores were obtained using the age-adjusted D-Dimer level. They had the same failure rate of 1.0% (0.2-5.3%) for efficiency of 16.8% (14.7-19.1%), and 16.9% (14.8-19.2%) respectively. The developed COVID-19 PE prediction model had an AUC of 0.609 (0.594-0.623) with an efficiency of 20.5% (18.4-22.8%) when its failure was set to 0.8%. CONCLUSIONS: The strategy to safely exclude PE in COVID-19 outpatients should not differ from that used in non-COVID-19 patients. The added value of the COVID-19 PE prediction model is minor. KEY POINTS: • D-dimer level remains the most important predictor of pulmonary embolism in COVID-19 patients. • The AUCs of the revised Geneva and Wells scores using an age-adjusted D-dimer threshold were 0.587 (95%CI, 0.572 to 0.603) and 0.588 (95%CI, 0.572 to 0.603). • The AUC of COVID-19-specific strategy to rule out pulmonary embolism ranged from 0.513 (95%CI: 0.503 to 0.522) to 0.609 (95%CI: 0.594 to 0.623).

Multimodal Techniques to Study Tumor Growth, Basement Membrane Breaching, and Invasion in 3D Matrices.

Cancer-derived organoids and three-dimensional (3D) extracellular matrix (ECM) are taking center stage as in vitro models to study neoplastic cell behavior, since they recapitulate the heterogeneous cellular composition of tumors and their extracellular environment. In combination with imaging and molecular/biochemical techniques, 3D organoid models have contributed substantially to our knowledge about the cellular and molecular mechanisms that regulate the growth of tumors and invasion into the surrounding tissue. We here outline a set of protocols that describe culturing of cancer-derived organoids in 3D matrices and various strategies that allow modeling of tumor growth, tumor cell penetration into basement membranes, and invasion into Collagen I-rich ECM. Furthermore, we specify protocols for subsequent handling of organoids cultured in 3D ECM for confocal microscopy and analysis of gene expression at the protein and mRNA level. Although we here use breast cancer-derived organoids, these protocols can be directly applied or adapted for organoids derived from other cancer types or healthy tissues. Thus, in addition to investigating cell behavior of multiple cancer types, the combination of protocols described here may be used to study processes such as cell differentiation and migration during homeostasis and normal development.

Cerebral perfusion using ASL in patients with COVID-19 and neurological manifestations: A retrospective multicenter observational study.

BACKGROUND AND PURPOSE: Cerebral hypoperfusion has been reported in patients with COVID-19 and neurological manifestations in small cohorts. We aimed to systematically assess changes in cerebral perfusion in a cohort of 59 of these patients, with or without abnormalities on morphological MRI sequences. METHODS: Patients with biologically-confirmed COVID-19 and neurological manifestations undergoing a brain MRI with technically adequate arterial spin labeling (ASL) perfusion were included in this retrospective multicenter study. ASL maps were jointly reviewed by two readers blinded to clinical data. They assessed abnormal perfusion in four regions of interest in each brain hemisphere: frontal lobe, parietal lobe, posterior temporal lobe, and temporal pole extended to the amygdalo-hippocampal complex. RESULTS: Fifty-nine patients (44 men (75%), mean age 61.2 years) were included. Most patients had a severe COVID-19, 57 (97%) needed oxygen therapy and 43 (73%) were hospitalized in intensive care unit at the time of MRI. Morphological brain MRI was abnormal in 44 (75%) patients. ASL perfusion was abnormal in 53 (90%) patients, and particularly in all patients with normal morphological MRI. Hypoperfusion occurred in 48 (81%) patients, mostly in temporal poles (52 (44%)) and frontal lobes (40 (34%)). Hyperperfusion occurred in 9 (15%) patients and was closely associated with post-contrast FLAIR leptomeningeal enhancement (100% [66.4%-100%] of hyperperfusion with enhancement versus 28.6% [16.6%-43.2%] without, p = 0.002). Studied clinical parameters (especially sedation) and other morphological MRI anomalies had no significant impact on perfusion anomalies. CONCLUSION: Brain ASL perfusion showed hypoperfusion in more than 80% of patients with severe COVID-19, with or without visible lesion on conventional MRI abnormalities.

Effect of diagnosis level of certainty on adherence to antibiotics' guidelines in ED patients with pneumonia: a post-hoc analysis of an interventional trial.

BACKGROUND AND IMPORTANCE: Clinical diagnosis of community-acquired pneumonia (CAP) is difficult to establish with certainty. Adherence to antibiotic guidelines independently affects the prognosis of CAP patients. OBJECTIVE: We aimed to determine whether guidelines' adherence was related to CAP diagnosis level of certainty and could be reinforced accordingly to diagnosis improvement. DESIGN: Secondary analysis of a prospective, multicenter study, which evaluated the impact of early thoracic CT scan on diagnosis and therapeutic plan in patients with clinically suspected CAP visiting emergency departments. SETTING AND PARTICIPANTS: In total 319 patients with clinically suspected CAP were enrolled in four emergency departments, Paris, France, between Nov 2011 and Jan 2013. OUTCOME MEASURES AND ANALYSIS: We evaluated guidelines' adherence before and after CT scan and its relationship with CAP diagnosis level of certainty. Antibiotics were categorized as adherent according to 2010 French guidelines. CAP diagnosis level of certainty was prospectively classified by the emergency physicians based on a Likert scale as excluded, possible, probable or definite before and immediately after the CT scan. These classifications and therapeutic plans were also completed by an independent adjudication committee. Determinants of adherence were assessed using Poisson regression with robust variance. MAIN RESULTS: Adherence to guidelines increased from 34.2% before CT scan to 51.3% after CT scan [difference 17.1% (95% CI, 9.5-24.7)], meanwhile CAP diagnosis with high level of certainty (definite and excluded CAP) increased from 46.1 to 79.6% [difference 33.5% (95% CI, 26.5-40.5)]. Diagnosis level of certainty before CT scan was the strongest determinant of adherence in multivariate analysis (RR, 2.63; 95% CI, 1.89-3.67). CONCLUSION: Antibiotic guidelines' adherence was poor and positively related to CAP diagnosis level of certainty. The results suggest that improvements in CAP diagnosis may increase adherence to antibiotic guidelines. Clinical trial registered with www.clinicaltrials.gov (NCT01574066).

Breast metastasis of a lung carcinoma.

Breast metastasis is a rare phenomenon (0.2-1.3%)1 compared to primary breast lesions. Several neoplasms have been reported to metastasize to the breast such as melanoma, lymphoma and lung cancer. In this article, we report a case of breast metastasis of lung cancer confirmed by biopsy and immunohistochemistry with CT and ultrasound imaging.

Clinical assessment of scannographic markers for sarcopenia in lung transplant candidates.

INTRODUCTION: The selection of patients for lung transplantation is difficult. An aspect of the patient's general condition and frailty can be assessed by measuring the surface area of certain muscles on CT. Indeed, sarcopenia, assessed by measuring the area of psoas muscles on scannographic sections has already been shown to be associated with poor outcomes in lung transplant and other major surgeries and could thus be helpful to evaluate candidates to lung transplant. However, it is not routinely performed by radiologists. As a pilot study, we compared the reliability of computerized tomography scan assessment for sarcopenia by clinicians with that of radiologists. MATERIALS AND METHODS: We conducted a retrospective single-centre study in which preoperative abdominal CT scans of lung transplant patients from 2014 to 2018 were analysed to assess sarcopenia by measuring the surface areas (mm2) of the psoas, paraspinal muscles at the L4 vertebra and diaphragm pillars at the celiac trunk bilaterally. The surfaces were blindly assessed by clinicians and radiologists and the intraclass correlation coefficients between the two measurements were estimated by a generalized linear model using SAS 9.4 software. RESULTS: We performed a double reading of 200 scans. The clinicians 'measurements were comparable to those of the radiologists for the psoas and paraspinal muscles but not for the diaphragm pillars. CONCLUSION: CT measurement of psoas and paraspinal muscle areas by clinicians appears reliable and feasible in routine practice and could be used in the evaluation of lung transplant candidates.

Grade 3-4 Immune-Related Adverse Events Induced by Immune Checkpoint Inhibitors in Non-Small-Cell Lung Cancer (NSCLC) Patients Are Correlated with Better Outcome: A Real-Life Observational Study.

Background: Immune checkpoint inhibitors (ICIs) have been a major advance in treating non-small-cell lung cancer (NSCLC). Programmed cell death protein-1/programmed death-ligand 1 blockade enhances immune function, mediating anti-tumor activity, yet causing immune-related adverse events (irAEs). We investigated the prognostic role of Grade 3−4 irAEs on overall survival (OS). Methods: This observational study recruited advanced NSCLC patients who received ICIs at Bichat-Claude Bernard University Hospital and in a community hospital, Saint-Joseph Foundation (Paris), between 1 January 2016 and 31 December 2019. Immunotherapy as a single-agent or double-drug combination was applied in the first and later lines. Univariable and multivariable analyses were instrumental in evaluating the prognostic impact of irAEs. Results: Overall, 201 consecutive ICI-treated patients were enrolled. High-grade irAEs (Grades 3−4) occurred in 36 patients (17.9%), including 11 (30.5%) cases of pneumonitis, 8 (22.2%) of colitis, 4 (11.1%) hepatic, 3 (8.3%) dermatological, 2 (5.5%) neurological events, and 2 cases (5.5%) of poly-arthralgia. The median OS was 10.4 ± 1.36 months (95% CI:7.7−13.1), being significantly higher in patients with high-grade irAEs than those without, 27.8 months vs. 8.1 months, respectively (HR = 2.5; p < 0.0001). Multivariable analysis revealed an independent association between high-grade irAEs and longer OS (HR = 0.29, 95% CI: 0.2−0.6, p < 0.0001). Conclusions: Our real-life study confirms that high-grade irAEs predict longer OS in advanced NSCLC.