Comparison of retroperitoneal and transperitoneal surgical routes in laparoscopic nodal staging for locally advanced cervical cancers (FIGO IB3-IVA)

BackgroundThis study compares morbidity and mortality associated with retroperitoneal and transperitoneal para-aortic lymphadenectomy (PAAL) for pretherapeutic nodal staging of locally advanced cervical cancers (FIGO IB3–IVA).MethodsPre-, per- and postoperative data of patients treated for locally advanced stage cervical cancer between 1999 and 2018 in 12 French referral centers (FRANCOGYN Study Group) were retrospectively collected.ResultsThe study was conducted using a sample of 448 patients, of whom 223 (49,8%) underwent retroperitoneal (group 1) and 225 (50,2%) had transperitoneal PAAL (group 2). No differences were noted concerning clinical and histological characteristics between the two groups. Among these 448 patients, 23 (5,1%) had an intraoperative complication (9 (2,0%) in group 1 and 14 (3,1%) in group 2, p = 0.28) and 47 (10,5%) had a postoperative complication (22 (4,9%) in group 1 and 25 (5,6%) in group 2, p = 0.44), only one of which required revision surgery but the patient died. The length of hospital stay was significantly shorter in group 1 than in group 2 (3.97 versus 4.88 days, p < 0.001). There was no significant difference in mortality between the two groups; 34 of 223 patients in group 1 (15.3%) and 40 of 225 patients in group 2 (15.6%) died (HR = 0.968, 95% CI [0.591–1.585]). There was no significant difference in recurrence-free or overall survival between the two groups.ConclusionRetroperitoneal PAAL appears as a valuable and safety surgical route for nodal staging in locally advanced cervical cancer compared with standard transperitoneal PAAL.     

Hyperamylasemia grade versus drain fluid amylase: which better predicts pancreatectomy outcomes?

BackgroundThe clinical importance of postoperative hyperamylasemia (POHA) grade is unknown. Our objectives were to evaluate the association of POHA grade with clinically relevant postoperative pancreatic fistula (CR-POPF) and compare its prognostic utility against postoperative day 1 drain fluid amylase (DFA-1).MethodsPatients who underwent pancreatectomy from January 2019 through March 2020 were identified in the ACS NSQIP pancreatectomy-targeted dataset. POHA grade was assigned using post-operative serum amylase and clinical sequelae. The primary outcome was CR-POPF within 30 days. The association of POHA grade with CR-POPF was assessed using multivariable logistic regression, and c-statistics were used to compare POHA grade versus DFA-1.ResultsPOHA occurred in 520 patients at 98 hospitals, including 261 (50.2%) with grade A, 234 (45.0%) with grade B, and 25 (4.8%) with grade C POHA. CR-POPFs were increased among patients with grade B (66.2%, OR 9.28 [5.84–14.73]) and C (68.0%, OR 10.50 [3.77–29.26]) versus grade A POHA (19.2%). POHA-inclusive models better predicted CR-POPF than those with DFA-1 alone (p < 0.002) and models with both predictors outperformed POHA alone (p = 0.039).ConclusionPOHA grade represents a measure of post-pancreatectomy outcomes that predicts CR-POPF and outperforms DFA-1 but must be aligned with new international definitions.     

Skin3D: Detection and longitudinal tracking of pigmented skin lesions in 3D total-body textured meshes

We present an automated approach to detect and longitudinally track skin lesions on 3D total-body skin surface scans. The acquired 3D mesh of the subject is unwrapped to a 2D texture image, where a trained objected detection model, Faster R-CNN, localizes the lesions within the 2D domain. These detected skin lesions are mapped back to the 3D surface of the subject and, for subjects imaged multiple times, we construct a graph-based matching procedure to longitudinally track lesions that considers the anatomical correspondences among pairs of meshes and the geodesic proximity of corresponding lesions and the inter-lesion geodesic distances.We evaluated the proposed approach using 3DBodyTex, a publicly available dataset composed of 3D scans imaging the coloured skin (textured meshes) of 200 human subjects. We manually annotated locations that appeared to the human eye to contain a pigmented skin lesion as well as tracked a subset of lesions occurring on the same subject imaged in different poses. Our results, when compared to three human annotators, suggest that the trained Faster R-CNN detects lesions at a similar performance level as the human annotators. Our lesion tracking algorithm achieves an average matching accuracy of 88% on a set of detected corresponding pairs of prominent lesions of subjects imaged in different poses, and an average longitudinal accuracy of 71% when encompassing additional errors due to lesion detection. As there currently is no other large-scale publicly available dataset of 3D total-body skin lesions, we publicly release over 25,000 3DBodyTex manual annotations, which we hope will further research on total-body skin lesion analysis.