Radiomics and Machine Learning for prediction of two-year disease-specific mortality and KRAS mutation status in metastatic colorectal cancer.

PURPOSE: Colorectal cancer is usually accompanied by liver metastases. The prediction of patient evolution is essential for the choice of the appropriate therapy. The aim of this study is to develop and evaluate machine learning models to predict KRAS gene mutations and 2-year disease-specific mortality from medical images. METHODS: Clinical and follow-up information was collected from patients with metastatic colorectal cancer who had undergone computed tomography prior to liver resection. The dominant liver lesion was segmented in each scan and radiomic features were extracted from the volumes of interest. The 65% of the cases were employed to perform feature selection and to train machine learning algorithms through cross-validation. The best performing models were assembled and evaluated in the remaining cases of the cohort. RESULTS: For the mortality model development, 101 cases were used as training set (64 alive, 37 deceased) and 35 as test set (22 alive, 13 deceased); while for KRAS mutation models, 55 cases were used for training (31 wild-type, 24 mutated) and 30 for testing (17 wild-type, 13 mutated). The ensemble of top performing models resulted in an area under the receiver operating characteristic curve of 0.878 for mortality and 0.905 for KRAS prediction. CONCLUSIONS: Predicting the prognosis of patients with metastatic colorectal cancer is useful for making timely decisions about the best treatment options. This study presents a noninvasive method based on quantitative analysis of baseline images to identify factors influencing patient outcomes, with the aim of incorporating these tools as support systems.

Long-term results in patients with lung resections for colorectal cancer metastases. Experience of a single center in Argentina / [Resultados a largo plazo en pacientes con resecciones pulmonares por metástasis de cáncer colorrectal. Experiencia de un solo centro de Argentina.]

Introduction: Between 5-10% of patients who undergo curative surgery for colorectal cancer (CRC) will present recurrence of their disease on the lungs. Surgical treatment of pulmonary metastases (PM) has gained popularity over the years, different publications report an overall survival rate at 5 years of between 30% and 60%. We present a review of patients with PM resections of CRC treated in a single center in Argentina. Materials and methods: A descriptive, observational, retrospective study was conducted between 2008 and 2018. All patients with pulmonary metastasectomy of colorectal cancer were included. The primary endpoint was to evaluate overall survival and disease-free survival. Possible prognostic factors were evaluated as a secondary endpoint. Results: A total of 99 surgeries were performed in the 68 patients, the resection was multiple in 48.5%, with more than 3 nodules in 13%. Overall and progression-free survival at 5 years was 55% and 27%, respectively. In the statistical analysis we found that the lesions in more than one lobe (p = 0.015) and the resection of more than 3 nodules (p = 0.011) presented a lower overall survival. Conclusions: In this retrospective analysis we evidenced comparable values to the world literature regarding morbidity, mortality, overall survival and progression-free. In our series, patients with disease in more than one lobe or more than three resected lesions had significantly lower overall survival.

Introducción: Entre el 5-10% de los pacientes que se someten a cirugía curativa por cáncer colorrectal (CCR) presentarán en algún momento recurrencia de su enfermedad a nivel pulmonar. El tratamiento quirúrgico de las metástasis pulmonares (MP) fue ganando popularidad a través de los años, diferentes publicaciones informan una tasa de supervivencia global a los 5 años de entre 30% a 60%. Presentamos una revisión de pacientes con resecciones por MP de CCR tratados en un único centro en Argentina. Materiales y métodos: Se realizó un estudio descriptivo, observacional, retrospectivo entre 2008 y 2018. Se incluyeron todos los pacientes con resecciones de MP de CCR.  El objetivo primario fue evaluar la supervivencia global y  supervivencia libre de enfermedad. Como objetivo secundario se evaluaron posibles factores pronósticos. Resultados: Se realizaron un total de 99 cirugías en los 68 pacientes, la resección fue múltiple en un 48,5%, siendo más de 3 nódulos en un 13%. La supervivencia global y libre de progresión a los 5 años fue del 55% y del 27% respectivamente. En el análisis estadístico encontramos que las lesiones en más de un lóbulo (p=0.015) y la resección de más de 3 nódulos (p=0.011) presentaron una menor supervivencia global. Conclusiones: En este análisis retrospectivo evidenciamos valores comparables a la literatura mundial con respecto a morbilidad, mortalidad, supervivencia global y libre de progresión. En nuestra serie los pacientes con enfermedad en más de un lóbulo o más de tres lesiones resecadas presentan significativamente menor supervivencia global.

Outcomes of patients with non-small cell lung cancer and poor performance status treated with immune checkpoint inhibitors in the real-world setting.

BACKGROUND: Immune-checkpoint inhibitors (ICIs) are standard treatments for metastatic non-small cell lung cancer (NSCLC). Patients with poor performance status (PS) are underrepresented in clinical trials. We evaluate the efficacy and safety of ICIs in a real-world setting. METHODS: We conducted a multi-institutional retrospective study to assess clinical outcomes of NSCLC treated with ICIs. We categorized pts within two groups (PS 0-1 vs 2) and assessed clinical outcomes and safety. RESULTS: Two hundred and sixty nine patients were included, 44 patients (16.4%) had baseline PS 2 and 223 patients (82.9%) PS 0-1. The overall response rate (ORR) was 30.4%, median PFS was 7.26 months (95% CI 5.1-9.4), and median OS was 15.18 months (95% CI 9.5-20.9). Patients with a PS 2 were most likely to received ICIs in the second or later line (84.1% vs 64.6%; p = 0.01), had baseline steroids (21.4% vs 8.2%; p 0.010), lower response rate (16.7% vs 34.5%; p 0.02) and clinical benefit (35.7% vs 71%; p 0.000) compared to PS 0-1 pts. Moreover, PS ≥ 2 patients had shorter PFS, median 2.2 months (95% CI 1.3-3.1) compared to 9.9 months (95% CI 6.7-13.1] and shorter OS, 3.3 months (95% CI 2.6-4.2) versus 24.1 months (95% CI 16.1-32.1), respectively. PS was significantly associated with PFS and OS in multivariate analysis. As it was expected, immunotherapy was well tolerated with a safety profile comparable to the previous published data. CONCLUSION: Based on these retrospective results, patients with poor baseline performance status seem to have poor clinical outcomes with ICIs in the real-world setting.