Survival analysis for lung cancer patients: A comparison of Cox regression and machine learning models.

INTRODUCTION: Survival analysis based on cancer registry data is of paramount importance for monitoring the effectiveness of health care. As new methods arise, the compendium of statistical tools applicable to cancer registry data grows. In recent years, machine learning approaches for survival analysis were developed. The aim of this study is to compare the model performance of the well established Cox regression and novel machine learning approaches on a previously unused dataset. MATERIAL AND METHODS: The study is based on lung cancer data from the Schleswig-Holstein Cancer Registry. Four survival analysis models are compared: Cox Proportional Hazard Regression (CoxPH) as the most commonly used statistical model, as well as Random Survival Forests (RSF) and two neural network architectures based on the DeepSurv and TabNet approaches. The models are evaluated using the concordance index (C-I), the Brier score and the AUC-ROC score. In addition, to gain more insight in the decision process of the models, we identified the features that have an higher impact on patient survival using permutation feature importance scores and SHAP values. RESULTS: Using a dataset including the cancer stage established by the Union for International Cancer Control (UICC), the best performing model is the CoxPH (C-I: 0.698±0.005), while using a dataset which includes the tumor size, lymph node and metastasis status (TNM) leads to the RSF as best performing model (C-I: 0.703±0.004). The explainability metrics show that the models rely on the combined UICC stage and the metastasis status in the first place, which corresponds to other studies. DISCUSSION: The studied methods are highly relevant for epidemiological researchers to create more accurate survival models, which can help physicians make informed decisions about appropriate therapies and management of patients with lung cancer, ultimately improving survival and quality of life.

Anti-TL1A Antibody PF-06480605 Safety and Efficacy for Ulcerative Colitis: A Phase 2a Single-Arm Study.

BACKGROUND & AIMS: An immune component of inflammatory bowel disease is up-regulated tumor necrosis factor-like ligand 1A (TL1A). Anti-TL1A antibodies such as PF-06480605, a fully human immunoglobulin G1 monoclonal antibody, may have therapeutic potential. METHODS: This Phase 2a, multicenter, single-arm, open-label study (TUSCANY) evaluated safety, tolerability, efficacy, pharmacokinetics, and immunogenicity in PF-06480605-treated participants with moderate to severe ulcerative colitis (UC). Participants received 500 mg intravenous PF-06480605 every 2 weeks, 7 doses total, with a 3-month follow-up period. Primary safety and efficacy endpoints were the incidence of adverse events (AEs) and week 14 endoscopic improvement (EI) (Mayo endoscopic subscore = 0 or 1), respectively. Secondary endpoints included total soluble TL1A (free/drug-bound) (sTL1A), incidence of anti-drug and neutralizing antibodies, PF-06480605 concentrations, and changes in fecal calprotectin and high-sensitivity C-reactive protein. Histology was assessed at week 14. RESULTS: The study enrolled 50 participants; 42 completed. Of 109 treatment-emergent AEs, 18 were treatment-related. The most common AEs were UC disease exacerbation and arthralgia (6 participants each). Four serious AEs, no deaths, and no malignancies were reported. Week 14 EI was observed in a statistically significant proportion of participants (38.2% [uniformly minimum-variance unbiased estimator, per protocol population]). Minimal histologic disease was observed after treatment (Robarts Histopathology Index ≤5: 33.3%; Geboes Index ≤3.2: 47.6%). sTL1A increase over time from baseline indicated sustained target engagement. Forty-one participants (82%) tested positive for anti-drug antibodies and 5 (10%) for neutralizing antibodies. CONCLUSIONS: PF-06480605 demonstrated an acceptable safety profile and statistically significant EI in participants with moderate to severe UC, warranting further study in a larger participant cohort. Tissue histopathology analyses support this conclusion. TRIAL REGISTRATION NUMBER: https://clinicaltrials.gov/NCT02840721.

Randomised trial and open-label extension study of an anti-interleukin-6 antibody in Crohn's disease (ANDANTE I and II).

OBJECTIVE: Neutralising pro-inflammatory interleukin-6 (IL-6) may effectively treat Crohn's disease (CD). Effects of PF-04236921, an anti-IL-6 antibody, in adults with CD are reported. DESIGN: Parallel-group, dose-ranging, double-blind trial with 4-week screening and 12-week treatment periods. After induction, patients entered 28-week follow-up or 48-week open-label extension (OLE) with 28-week follow-up. Adults with confirmed CD and inadequate response to anti-tumour necrosis factor (TNF) therapy were included. Induction study: 249 patients randomised 1:1:1:1 to placebo, PF-04236921 10, 50 or 200 mg by subcutaneous injection on days 1 and 28. OLE study: PF-04236921 50 mg every 8 weeks up to six doses followed by 28-week follow-up. RESULTS: 247 patients were randomised and received treatment in the induction study. The 200 mg dose was discontinued due to safety findings in another study (NCT01405196) and was not included in the primary efficacy analysis. Crohn's Disease Activity Index (CDAI)-70 response rates with PF-04236921 50 mg were significantly greater than placebo at weeks 8 (49.3% vs 30.6%, P<0.05) and 12 (47.4% vs 28.6%, P<0.05) and met the primary end point. Week 12 CDAI remission rates with PF-04236921 50 mg and placebo were 27.4% and 10.9%, respectively (16.5% difference; P<0.05). 191 subjects received treatment in the OLE. Common treatment-emergent and serious adverse events in both studies included worsening CD, abdominal pain and nasopharyngitis. CONCLUSIONS: PF-04236921 50 mg induced clinical response and remission in refractory patients with moderate-to-severe CD following failure of anti-TNF therapy. GI abscess and perforation were observed, a specific focus of attention during future clinical development. TRIAL REGISTRATION NUMBER: NCT01287897 and NCT01345318.

Clinical results from a randomized, double-blind, dose-ranging study of pantoprazole in children aged 1 through 5 years with symptomatic histologic or erosive esophagitis.

In an 8-week, multicenter, randomized, double-blind study, we evaluated the efficacy and tolerability of pantoprazole (0.3mg/kg [low dose (LD)], 0.6 mg/kg [medium dose (MD)], and 1.2 mg/kg [high dose (HD)]) for delayed-release oral suspension (granules) in patients 1 to 5 years with documented symptoms of gastroesophageal reflux disease (GERD) and endoscopic evidence of reflux-related erosive esophagitis (EE) or histologic esophagitis (HE) consistent with GERD. Patients with HE were randomly assigned to LD, MD, or HD, and patients with EE, to MD or HD. A daily eDiary captured 5 individual GERD symptoms. A total of 60 patients (56 HE, 4 EE) were randomized. Mean weekly GERD symptom score (WGSS, sum of weekly mean frequency scores for 5 individual GERD symptoms) for the modified intention-to-treat HE population at the final week was improved with LD ( P < .001), MD (P = .063), and HD (P < 0.001) (paired t-tests). Patients with EE were healed at week 8. Adverse event incidences did not increase with dose.