Machine learning-based prediction of microsatellite instability and high tumor mutation burden from contrast-enhanced computed tomography in endometrial cancers.

To evaluate whether radiomic features from contrast-enhanced computed tomography (CE-CT) can identify DNA mismatch repair deficient (MMR-D) and/or tumor mutational burden-high (TMB-H) endometrial cancers (ECs). Patients who underwent targeted massively parallel sequencing of primary ECs between 2014 and 2018 and preoperative CE-CT were included (n = 150). Molecular subtypes of EC were assigned using DNA polymerase epsilon (POLE) hotspot mutations and immunohistochemistry-based p53 and MMR protein expression. TMB was derived from sequencing, with > 15.5 mutations-per-megabase as a cut-point to define TMB-H tumors. After radiomic feature extraction and selection, radiomic features and clinical variables were processed with the recursive feature elimination random forest classifier. Classification models constructed using the training dataset (n = 105) were then validated on the holdout test dataset (n = 45). Integrated radiomic-clinical classification distinguished MMR-D from copy number (CN)-low-like and CN-high-like ECs with an area under the receiver operating characteristic curve (AUROC) of 0.78 (95% CI 0.58-0.91). The model further differentiated TMB-H from TMB-low (TMB-L) tumors with an AUROC of 0.87 (95% CI 0.73-0.95). Peritumoral-rim radiomic features were most relevant to both classifications (p ≤ 0.044). Radiomic analysis achieved moderate accuracy in identifying MMR-D and TMB-H ECs directly from CE-CT. Radiomics may provide an adjunct tool to molecular profiling, especially given its potential advantage in the setting of intratumor heterogeneity.

Predicting breast cancer response to neoadjuvant chemotherapy using pretreatment diffuse optical spectroscopic texture analysis.

BACKGROUND: Diffuse optical spectroscopy (DOS) has been demonstrated capable of monitoring response to neoadjuvant chemotherapy (NAC) in locally advanced breast cancer (LABC) patients. In this study, we evaluate texture features of pretreatment DOS functional maps for predicting LABC response to NAC. METHODS: Locally advanced breast cancer patients (n=37) underwent DOS breast imaging before starting NAC. Breast tissue parametric maps were constructed and texture analyses were performed based on grey-level co-occurrence matrices for feature extraction. Ground truth labels as responders (R) or non-responders (NR) were assigned to patients based on Miller-Payne pathological response criteria. The capability of DOS textural features computed on volumetric tumour data before the start of treatment (i.e., 'pretreatment') to predict patient responses to NAC was evaluated using a leave-one-out validation scheme at subject level. Data were analysed using a logistic regression, naive Bayes, and k-nearest neighbour classifiers. RESULTS: Data indicated that textural characteristics of pretreatment DOS parametric maps can differentiate between treatment response outcomes. The HbO2 homogeneity resulted in the highest accuracy among univariate parameters in predicting response to chemotherapy: sensitivity (%Sn) and specificity (%Sp) were 86.5% and 89.0%, respectively, and accuracy was 87.8%. The highest predictors using multivariate (binary) combination features were the Hb-contrast+HbO2-homogeneity, which resulted in a %Sn/%Sp=78.0/81.0% and an accuracy of 79.5%. CONCLUSIONS: This study demonstrated that the pretreatment DOS texture features can predict breast cancer response to NAC and potentially guide treatments.

Liver Transplantation is a Preferable Alternative to Palliative Therapy for Selected Patients with Advanced Hepatocellular Carcinoma.

BACKGROUND: Patients with hepatocellular carcinoma (HCC) beyond the traditional criteria (advanced HCC) are typically offered palliation, which is associated with a 3-year survival rate lower than 30%. This study aimed to describe the outcomes for a subset of patients with advanced HCC who satisfied the Extended Toronto Criteria (ETC) and were listed for liver transplantation (LT). METHODS: All patients listed in the Toronto liver transplantation program with HCC beyond both the Milan and University of California, San Francisco criteria were included in this study. Data were extracted from the prospectively collected electronic database. All radiologic images were reviewed by two independent radiologists. The primary end point was patient survival. RESULTS: Between January 1999 and August 2014, 96 patients with advanced HCC were listed for LT, and 62 (65%) of these patients received bridging therapy while on the waiting list. Bridging therapy led to a significant reduction in tumor progression (p = 0.02) and tumor burden (p < 0.001). The majority of those listed underwent LT (n = 69, 72%). Both tumor progression on waiting list (hazard ratio [HR] 4.973; range1.599-15.464; p = 0.006) and peak alpha-fetoprotein (AFP) at 400 ng/ml or higher (HR, 4.604; range 1.660-12.768; p = 0.003) were independently associated with waiting list dropout. Post-LT HCC recurrence occurred in 35% of the patients (n = 24). Among those with HCC recurrence, survival was significantly better for those who received curative treatment (p = 0.004). The overall actuarial survival rates from the listing were 76% at 1 year, 56% at 3 years, and 47% at 5 years, and the corresponding rates from LT were 93, 71, and 66%. CONCLUSION: Liver transplantation provides significantly better survival rates than palliation for patients with selected advanced HCC.

Differential Diagnosis of Colonic Strictures: Pictorial Review With Illustrations from Computed Tomography Colonography.

Strictures of the colon can lead to significant morbidity requiring surgical management. The etiology of strictures is broad and generally categorized as benign, malignant, or pseudostrictures. Computed tomography (CT) is a crucial imaging modality in the assessment and characterization of colonic pathologies but colonoscopy remains the diagnostic gold standard. However, in the setting of incomplete colonoscopy due to strictures, the imaging features of CT will be relied on. This review will focus on the CT features of different colon pathologies leading to strictures and will be illustrated with images from 10 years of experience with CT colonography at our institutions from 2002-2012 (Hotel Dieu Hospital, Queen's University and Mount Sinai Hospital, University of Toronto).

The assessment of anemia from attenuation values of cranial venous drainage on unenhanced computed tomography of the head.

OBJECTIVE: To determine if prediction of anemia is possible from quantitative analyses of unenhanced cranial computed tomography (CT) images. METHODS: A retrospective chart review revealed 101 patients who had hemoglobin and hematocrit levels obtained within 24 hours of an unenhanced cranial CT. Regions of interest were the place on the torcular Herophili (confluence of sinuses) and the left and right transverse sinuses. Attenuation values were correlated with hemoglobin and hematocrit to investigate any possible relationship. RESULTS: Hemoglobin levels were correlated with torcular and transverse sinus attenuation levels. For every 10 Hounsfield unit (HU) increase in torcular attenuation, hemoglobin levels increased by 16.3 g/L (P < .001). When subgroup analysis was performed, by sex, for every 10 HU increase in torcular attenuation, hemoglobin levels increased by 22.6 g/L (P < .001) in men and 8.96 g/L (P < .05) in women. CONCLUSION: The attenuation values for venous drainage on unenhanced cranial CT constitute a practical adjunct in the assessment of anemia. Given the number of individuals undergoing CT without blood work, this rapid assessment technique for anemia is a convenient means for narrowing possible diagnoses.

Factors involved in discrepant preliminary radiology resident interpretations of neuroradiological imaging studies: a retrospective analysis.

OBJECTIVE: The purpose of this article is to determine discrepancy rates in radiology residents' interpretations of CT and MRI studies of the head, neck, and spine and to attempt to quantify its potential or realized effect on patient care. Secondarily, we attempted to determine the effect of resident, study, and patient variables on discrepancy rate. MATERIALS AND METHODS: A retrospective review of all relevant studies from March 1, 2009, through December 31, 2009, revealed 5695 cases with preliminary interpretations made by on-call residents. Preliminary and final attending radiologists' reports were compared for concordance, and for all discrepant reports, electronic patient records were investigated for patient impact. Discrepancies were deemed minor or major depending on the presence of a negative impact on care. Variables were investigated for their effect on discrepancy rate by univariate logistic regression analysis, with significant regressors included in subsequent multivariate modeling. RESULTS: The overall discrepancy rate for all included studies was 8.4% (7.2% minor and 1.2% major). Head CT showed the highest concordance (92.6%), whereas MRI studies had discrepancy rates when compared with equivalent CT studies. By multivariate binomial logistic regression, several variables positively correlated with concordance, including resident training level, outpatient status, and concluding hours of call shift. CONCLUSION: The observed rates of discrepancy for all study types were comparable to those in the current literature. Furthermore, a nearly linear downward trend in discrepancy rate with increasing resident training level was found. Similarly, we uniquely showed that discrepancy rate varies with patient status, study indication, and hour of call. These factors should be appropriately considered during evaluation or remediation of radiology residents.

Delayed recovery in patients with whiplash-associated disorders.

BACKGROUND: Prognosis of whiplash injury has been found to be related to a number of sociodemographic, treatment and clinical factors. In the current study, we attempt to identify several novel prognostic factors for delayed recovery in whiplash-associated disorders (WADs), using a validated and reliable measure of recovery. PATIENTS AND METHODS: Retrospective review of a large database of a national network of physiotherapy and rehabilitation service providers in Canada yielded 5581 individuals injured in motor vehicle collisions. Cases were grouped into 3 cohorts based on time lapsed between injury date and initial presentation. Acute (n=3075), early chronic (n=958) and chronic (n=1548) patient cohorts were compared regarding treatment outcome and relative distribution of 29 prognostic factors. Outcome was defined by a minimally important clinical change (10%) on a previously validated disability questionnaire between initial and discharge rehabilitation visits. RESULTS: Analysis demonstrated positive outcomes to be proportionally fewer in the chronic cohort (52.1%) relative to the early chronic (61.4%), which was in turn lower than the acute cohort (72.3%). Furthermore, individuals presenting with chronic pain were more likely to: (1) be female; (2) present with lower limb pain or nonorganic signs; (3) have returned to work; (4) have retained a lawyer; or (5) have undergone previous spinal surgery, and were less likely to: (1) present with neck or midback pain; (2) live in Ontario or Nova Scotia; or (3) have modified duties upon return to work. Acute, early chronic and chronic cohorts were also found to differ in the distribution of several other prognostic factors at initial clinical visit. CONCLUSIONS: Recovery in whiplash-associated disorder appears to be multifactorial with both medical and non-injury related factors influencing outcome. Further characterisation of these factors may prove invaluable in guiding future clinical treatment and referral practices.