BARIATRIC SURGERY DECREASES THE RISK OF DEVELOPING CANCER: A MUTLICENTER POPULATION-BASED STUDY

Introduction. Microvascular invasion (MVI) is the main risk factor for overall mortality and recurrence after surgery for hepatocellular carcinoma (HCC). Its diagnosis can be made only postoperatively on the histological specimen. The aim of this preliminary study was to train machine-learning models to predict MVI on preoperative CT scan (fig.1).
Methods. Clinical data and 3-phases CT scans were retrospectively collected among 4 Italian centres. DICOM files were manually segmented to detect the liver and the tumor(s). An already available segmentation algorithm (Nnunet) was retrained to obtain automatic detection focused on HCC. An implementation was added to automatically extract radiomics features from the tumoral, peritumoral (among 5mm from the tumor margin) and healthy liver areas in each phase. Performance comparison between manual and algorithm segmentations was measured by intersection over union (Jaccard Index). Data obtained were explored and principal component analysis (PCA) was performed to reduce the dimensions of the dataset, keeping only the PC’s explaining 95% of the variability. After normalization, data were divided between training (70%) and test (30%) sets. Random-Forest (RF), fully connected MLP Artificial neural network (neuralnet) and extreme gradient boosting (XGB) models were fitted to predict MVI. Hyperparameters tuning was made per each model to reduce the out-of-bag error. Prediction accuracy was estimated in the test set and employed as the study end-point.
Results. Between 2008 and 2022, 218 consecutive preoperative CT scans of patients affected by HCC and submitted to surgery were collected with the relative clinical data. At the histological specimen, 72 (33.02%) patients had MVI. The Jaccard index between manual and algorithm segmentations was overall 90%. First and second order radiomics features were extracted, obtaining 672 variables per patient. After data exploration, PCA selected 58 dimensions explaining >95% of the variance. After standardization and normalization, RF, neuralnet and XGB were fitted to predict the presence of MVI. Tuning parameters were: 1) RF: n.tree=500, mtry=30; 2) Neuralnet: 2 hidden layer with 40 and 20 neurons, learning rate= 0.001, threshold for termination= 1%, activation function= sigmoid; 3) XGB: nrounds = 100, max_depth = 3, eta = 0.3. The models were then fitted in the testset to estimate prediction accuracy by confusion-matrix. RF was the best performer (Acc=98.4%, 95%CI: 0.91-0.99, Sens: 95.2%, Spec: 100%, PPV: 100% and NPV: 97.7%, fig.2).
Conclusion. Our model allowed an impressive prediction accuracy of the presence of MVI at the time of HCC diagnosis, never reached until now. This could lead to change the treatment allocation, the surgical extension and the follow-up strategy for those patients. The algorithm will be freely distributed online for medical purpose.

Introduction:
Obesity has a strong association with cancer; however, there is scarce literature of the protective effects of bariatric surgery (BSx) for de novo cancer. We aimed to assess whether BSx has a protective effect against de novo cancers in patients with obesity.

Methods:
Retrospective cohort study (2002-2022) using TriNetX (Cambridge, MA), a multi-institutional database from 47 US healthcare-organizations, containing >107 million patients. Adults with BMI >35 were identified by ICD-10-CM coding. History of BSx (sleeve gastrectomy, gastric bypass, and gastric band) was determined by ICD-10CM coding. Patients with history of prior malignancy, peptic ulcer disease, or ascites were excluded. Propensity score matching was performed for demographics, Elixhauser comorbidity-index, hormone therapy, and cancer screening modalities. We utilized the International Agency for Research on Cancer to determine cancers with sufficient evidence to be considered associated with obesity: esophageal adenocarcinoma, multiple myeloma, and cancers of the kidney, colon, rectum, stomach, liver, gallbladder, pancreas, ovary, endometrium, breast, and thyroid. Assessment for de novo cancer diagnosis started 1 year after index date: time of BSx (study group) or obesity diagnosis (nonsurgical control group) to minimize impact of preexisting cancers. All patients required at least 12 months of follow up designated by at least one visit after one year from index date, and patients followed up to 10 years from index date.

Results:
We initially identified 60,285 patients in the BSx group and 1,570,440 patients in nonsurgical control group (Table 1). After propensity score matching, we included 55,789 patients in each cohort. The cumulative incidence of obesity associated de novo cancers at 10 years was 4.0% (2,206 patients) in the BSx group and 8.9% (4,960 patients) in the nonsurgical control group (HR 0.482 [95% CI 0.459-0.507]; p <0.001). BSx group had lower prevalence rates for de novo breast cancer (HR 0.753 [CI 0.678-0.836]), colon cancer (HR 0.638 [CI 0.541-0.752]), liver cancer (HR 0.370 [CI 0.345-0.396]), pancreatic cancer (HR 0.84 [CI 0.569-1.080]), ovarian cancer (HR 0.654 [CI 0.531-0.806]), and endometrial cancer (HR 0.448 [CI 0.362-0.556]) when compared to nonsurgical control group (all p-values <0.05) (Table 2).

Conclusion:
Bariatric surgery significantly decreases the risk of de novo cancers in patients with obesity. We noted a protective effect of BSx with a lower cumulative incidence of all obesity associated cancers at 10 year follow up in addition to lower prevalence rate of de novo breast, colon, liver, pancreatic, ovarian, and endometrial cancers. This study supports the importance of treating obesity with bariatric surgery to reduce societal and economic burden of cancer.