A machine learning approach for predicting textbook outcome after cytoreductive surgery and hyperthermic intraperitoneal chemotherapy.

INTRODUCTION: Peritoneal carcinomatosis is considered a late-stage manifestation of neoplastic diseases. Cytoreductive surgery with hyperthermic intraperitoneal chemotherapy (CRS-HIPEC) can be an effective treatment for these patients. However, the procedure is associated with significant morbidity. Our aim was to develop a machine learning model to predict the probability of achieving textbook outcome (TO) after CRS-HIPEC using only preoperatively known variables. METHODS: Adult patients with peritoneal carcinomatosis and who underwent CRS-HIPEC were included from a large, single-center, prospectively maintained dataset (2001-2020). TO was defined as a hospital length of stay ≤14 days and no postoperative adverse events including any complications, reoperation, readmission, and mortality within 90 days. Four models (logistic regression, neural network, random forest, and XGBoost) were trained, validated, and a user-friendly risk calculator was then developed. RESULTS: A total of 1954 CRS-HIPEC procedures for peritoneal carcinomatosis were included. Overall, 13% (n = 258) achieved TO following CRS-HIPEC procedure. XGBoost and logistic regression had the highest area under the curve (AUC) (0.76) after model optimization, followed by random forest (AUC 0.75) and neural network (AUC 0.74). The top preoperative variables associated with achieving a TO were lower peritoneal cancer index scores, not undergoing proctectomy, splenectomy, or partial colectomy and being asymptomatic from peritoneal metastases prior to surgery. CONCLUSION: This is a data-driven study to predict the probability of achieving TO after CRS-HIPEC. The proposed pipeline has the potential to not only identify patients for whom surgery is not associated with prohibitive risk, but also aid surgeons in communicating this risk to patients.

Robotic-assisted endoluminal gastric leiomyoma resection: a novel surgical technique for benign gastroesophageal junction tumors.

Gastric leiomyomas are rare, benign smooth muscle tumors that arise from the muscularis propria and can be found in any part of the stomach. The American College of Gastroenterologists recommends resection only for symptomatic leiomyomas, which can often present with bleeding, abdominal pain, or dyspepsia. Notably, symptomatic leiomyomas that arise at the gastroesophageal (GE) junction, especially those that are large, pose unique challenges. Specifically, total gastrectomy with esophagojejunostomy is often necessary, which can be associated with a compromised quality of life and possible complications such as anastomotic stricture or reflux esophagitis. In this context, we present the case of a young, male patient with a large symptomatic leiomyoma at the GE junction who was offered a robotic-assisted endoluminal leiomyoma resection. By placing endoluminal trocars and utilizing the Da Vinci® robot, we were able to carefully excise the tumor without perforating the stomach or causing GE junction stenosis. This allowed the patient to preserve his stomach and avoid a high-risk anastomosis. Another notable highlight of the case included the use of the endoscope as both a bougie and a source of insufflation. The patient had an uncomplicated postoperative course and a rapid recovery, highlighting the feasibility of this approach for patients with benign GE junction tumors.

A Novel Machine Learning Approach to Predict Textbook Outcome in Colectomy.

BACKGROUND: Several calculators exist to predict risk of postoperative complications. However, in low-risk procedures such as colectomy, a tool to determine the probability of achieving the ideal outcome could better aid clinical decision-making, especially for high-risk patients. A textbook outcome is a composite measure that serves as a surrogate for the ideal surgical outcome. OBJECTIVE: To identify the most important factors for predicting textbook outcomes in patients with nonmetastatic colon cancer undergoing colectomy and to create a textbook outcome decision support tool using machine learning algorithms. DESIGN: This was a retrospective analysis study. SETTINGS: Data were collected from the American College of Surgeons National Surgical Quality Improvement Program database. PATIENTS: Adult patients undergoing elective colectomy for nonmetastatic colon cancer (2014-2020) were included. MAIN OUTCOME MEASURES: Textbook outcome was the main outcome, defined as no mortality, no 30-day readmission, no postoperative complications, no 30-day reinterventions, and a hospital length of stay of ≤5 days. Four models (logistic regression, decision tree, random forest, and eXtreme Gradient Boosting) were trained and validated. Ultimately, a web-based calculator was developed as proof of concept for clinical application. RESULTS: A total of 20,498 patients who underwent colectomy for nonmetastatic colon cancer were included. Overall, textbook outcome was achieved in 66% of patients. Textbook outcome was more frequently achieved after robotic colectomy (77%), followed by laparoscopic colectomy (68%) and open colectomy (39%, p < 0.001). eXtreme Gradient Boosting was the best performing model (area under the curve = 0.72). The top 5 preoperative variables to predict textbook outcome were surgical approach, patient age, preoperative hematocrit, preoperative oral antibiotic bowel preparation, and patient sex. LIMITATIONS: This study was limited by its retrospective nature of the analysis. CONCLUSIONS: Using textbook outcome as the preferred outcome may be a useful tool in relatively low-risk procedures such as colectomy, and the proposed web-based calculator may aid surgeons in preoperative evaluation and counseling, especially for high-risk patients. See Video Abstract . UN NUEVO ENFOQUE DE APRENDIZAJE AUTOMTICO PARA PREDECIR EL RESULTADO DE LOS LIBROS DE TEXTO EN COLECTOMA: ANTECEDENTES:Existen varias calculadoras para predecir el riesgo de complicaciones posoperatorias. Sin embargo, en procedimientos de bajo riesgo como la colectomía, una herramienta para determinar la probabilidad de lograr el resultado ideal podría ayudar mejor a la toma de decisiones clínicas, especialmente para pacientes de alto riesgo. Un resultado de libro de texto es una medida compuesta que sirve como sustituto del resultado quirúrgico ideal.OBJETIVO:Identificar los factores más importantes para predecir el resultado de los libros de texto en pacientes con cáncer de colon no metastásico sometidos a colectomía y crear una herramienta de apoyo a la toma de decisiones sobre los resultados de los libros de texto utilizando algoritmos de aprendizaje automático.DISEÑO:Este fue un estudio de análisis retrospectivo.AJUSTES:Los datos se obtuvieron de la base de datos del Programa Nacional de Mejora de la Calidad del Colegio Americano de Cirujanos.PACIENTES:Se incluyeron pacientes adultos sometidos a colectomía electiva por cáncer de colon no metastásico (2014-2020).MEDIDAS PRINCIPALES DE RESULTADO:El resultado de los libros de texto fue el resultado principal, definido como ausencia de mortalidad, reingreso a los 30 días, complicaciones posoperatorias, reintervenciones a los 30 días y una estancia hospitalaria ≤5 días. Se entrenaron y validaron cuatro modelos (regresión logística, árbol de decisión, bosque aleatorio y XGBoost). Finalmente, se desarrolló una calculadora basada en la web como prueba de concepto para su aplicación clínica.RESULTADOS:Se incluyeron un total de 20.498 pacientes sometidos a colectomía por cáncer de colon no metastásico. En general, el resultado de los libros de texto se logró en el 66% de los pacientes. Los resultados de los libros de texto se lograron con mayor frecuencia después de la colectomía robótica (77%), seguida de la colectomía laparoscópica (68%) y la colectomía abierta (39%) (p<0,001). XGBoost fue el modelo con mejor rendimiento (AUC=0,72). Los cinco principales variables preoperatorias para predecir el resultado en los libros de texto fueron el abordaje quirúrgico, la edad del paciente, el hematocrito preoperatorio, la preparación intestinal con antibióticos orales preoperatorios y el sexo femenino.LIMITACIONES:Este estudio estuvo limitado por la naturaleza retrospectiva del análisis.CONCLUSIONES:El uso de los resultados de los libros de texto como resultado preferido puede ser una herramienta útil en procedimientos de riesgo relativamente bajo, como la colectomía, y la calculadora basada en la web propuesta puede ayudar a los cirujanos en la evaluación y el asesoramiento preoperatorios, especialmente para pacientes de alto riesgo. (Traducción-Yesenia Rojas-Khalil ).

Instrument and Supply Variability: An Opportunity to Reduce the Carbon Footprint of the Operating Room.

INTRODUCTION: Reducing costs and carbon footprints are important, parallel priorities for the US health-care system. Within surgery, reducing the number of instruments that are sterilized and disposable supplies that are used for each operation may help achieve both goals. We wanted to measure the existing variability in surgical instrument and supply choices and assess whether standardization could have a meaningful cost and environmental impact. METHODS: We analyzed surgeon preference cards for common general surgery operations at our hospital to measure the number of sterilizable instrument trays and supplies used by each surgeon for each operation. From this data, we calculated supply costs, carbon footprint, and median operative time and studied the variability in each of these metrics. RESULTS: Among the ten operations studied, variability in sterilizable instrument trays requested on surgeon preference cards ranged from one to eight. Variability in disposable supplies requested ranged from 17 to 45. Variability in open supply costs ranged from $104 to $4184. Variability in carbon footprint ranged from 17 to 708 kg CO2e. If the highest-cost surgeon for each operation switched their preference card to that of the median-cost surgeon, $245,343 in open supply costs and 41,708 kg CO2e could be saved. CONCLUSIONS: There is significant variability in the instrument and supply choices of surgeons performing common general surgery operations. Standardizing this variability may lead to meaningful cost savings and carbon footprint reduction, especially if scaled across the entire health system.

A Machine Learning Approach to Predict Postoperative Pancreatic Fistula After Pancreaticoduodenectomy Using Only Preoperatively Known Data.

BACKGROUND: Clinically-relevant postoperative pancreatic fistula (CR-POPF) following pancreaticoduodenectomy (PD) is a major postoperative complication and the primary determinant of surgical outcomes. However, the majority of current risk calculators utilize intraoperative and postoperative variables, limiting their utility in the preoperative setting. Therefore, we aimed to develop a user-friendly risk calculator to predict CR-POPF following PD using state-of-the-art machine learning (ML) algorithms and only preoperatively known variables. METHODS: Adult patients undergoing elective PD for non-metastatic pancreatic cancer were identified from the ACS-NSQIP targeted pancreatectomy dataset (2014-2019). The primary endpoint was development of CR-POPF (grade B or C). Secondary endpoints included discharge to facility, 30-day mortality, and a composite of overall and significant complications. Four models (logistic regression, neural network, random forest, and XGBoost) were trained, validated and a user-friendly risk calculator was then developed. RESULTS: Of the 8666 patients who underwent elective PD, 13% (n = 1160) developed CR-POPF. XGBoost was the best performing model (AUC = 0.72), and the top five preoperative variables associated with CR-POPF were non-adenocarcinoma histology, lack of neoadjuvant chemotherapy, pancreatic duct size less than 3 mm, higher BMI, and higher preoperative serum creatinine. Model performance for 30-day mortality, discharge to a facility, and overall and significant complications ranged from AUC 0.62-0.78. CONCLUSIONS: In this study, we developed and validated an ML model using only preoperatively known variables to predict CR-POPF following PD. The risk calculator can be used in the preoperative setting to inform clinical decision-making and patient counseling.

Lymph Node Dissection in Intrahepatic Cholangiocarcinoma: a Critical and Updated Review of the Literature.

BACKGROUND: Lymphatic spread of intrahepatic cholangiocarcinoma (iCCA) is common and negatively impacts survival. However, the precise role of lymph node dissection (LND) in oncologic outcomes for patients with intrahepatic cholangiocarcinoma remains to be established. METHODS: Updated evidence on the preoperative diagnosis and prognostic value of lymph node metastasis is reviewed, as well as the potential benefit of LND in patients with iCCA. RESULTS: The ability to accurately determine nodal status for iCCA with current imaging modalities is equivocal. LND has prognostic value for both survival and disease recurrence. However, execution rates of LND are highly varied in the literature, ranging from 26.9 to 100%. At least 6 lymph nodes should be examined from nodal stations of the hepatoduodenal ligament and hepatic artery as well as based on the location of the primary tumor. Neoadjuvant therapies may be beneficial if lymph node metastases at diagnosis are suspected. Surgeons performing a minimally invasive approach should focus on increasing LND rates and harvesting ≥ 6 lymph nodes. Lymph node negativity is required in patients with iCCA being considered for liver transplantation under investigational protocols. CONCLUSION: Despite an upward trend in the LND rate, the reality is that only 10% of patients with iCCA receive an adequate LND. This review underscores the importance of routinely increasing the rate of adequate LND in these patients in order to achieve accurate staging, appropriately select patients for adjuvant therapy, and improve the prognosis of clinical outcomes. While prospective data is lacking, the therapeutic impact of LND remains unknown.

Virtual Interviews: Assessing How Expectations Meet Reality.

OBJECTIVE: This study aimed to determine the post-matriculation perceptions of interns and faculty who participated in the 2020-2021 virtual interview process and how their expectations of the program and the applicants, respectively, aligned with reality. DESIGN, SETTING, AND PARTICIPANTS: Published surveys on virtual interviewing were reviewed and modified to design two surveys, for interns and for faculty who interviewed. Interns and faculty members from the Departments of Surgery and Medicine at one institution who participated in the 2020-2021 virtual interview process completed the surveys four to six months after the start of the academic year. Following survey completion, surgical interns from the same application cycle participated in one in-person focus group nine months after the start of the academic year to clarify points raised in the surveys. RESULTS: Forty-six interns and faculty members responded to the survey (subgroup response rates ranging from 13-30%) and ten interns participated in the focus group (participation rate 100%). Most faculty and intern participants found that expectations formed during virtual interviews were accurate. However, our respondents noted limitations to virtual interviews that reduced their usefulness, including challenges with unnatural social interactions, understanding city fit, and getting a sense of resident and program culture. Participants provided possible solutions to address these challenges. CONCLUSIONS: Overall, this mixed-methods study at a single institution found that perceptions from virtual interviews were generally accurate but with some limitations. We describe several opportunities to improve the virtual interview process and optimize the application experience.

Impact and clinical usefulness of genetic data in the surgical management of colorectal cancer liver metastasis: a narrative review.

IMPORTANCE: In patients who undergo surgery for colorectal cancer liver metastases (CRLM), a number of somatic mutations have been associated with worse overall (OS) and recurrence-free survival (RFS). Although useful, an association with prognosis does not necessarily equate to an impact on surgical management. OBJECTIVE: The aim of this review was to investigate whether the best-studied somatic mutations impact surgical management of CRLM by informing: (I) post-hepatectomy surveillance; (II) selection of surgical technique; (III) selection of optimal margin width; and (IV) selection of patients for surgery. Lastly, we discuss the refinement of genetic data from overall mutation status to specific variants, as well as lesser studied somatic mutations. EVIDENCE REVIEW: We conducted a computerized search using PubMed and Google Scholar for reports published so far, using mesh headings and keywords related to genetic data and CRLM. FINDINGS: Genetic data may impact surgical management of CRLM in three ways. Firstly, KRAS mutations can predict lung recurrences. Secondly, KRAS mutations may help tailor margin width. Thirdly, KRAS mutations may help tailor surgical technique. CONCLUSIONS: Although genetic data may impact post-hepatectomy surveillance, selection of surgical technique and optimal margin width, their use to guide surgical selection remains elusive, as the data cannot support denying surgery to patients according to their somatic mutation profile.

PD-L1, PD-1, LAG-3, and TIM-3 in Melanoma: Expression in Brain Metastases Compared to Corresponding Extracranial Tumors.

Background Metastatic melanoma to the brain carries a particularly poor prognosis that may be associated with an attenuated antitumor response in the presence of central nervous system malignancies. Thus, the development of brain metastases could theoretically accelerate cancer progression both locally and systemically. Although dysregulation of checkpoint markers, such as programmed death-ligand 1 (PD-L1), programmed cell death receptor 1 (PD-1), lymphocyte activation gene 3 (LAG-3), and T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), have been implicated in immune dysfunction, the exact relationship between these markers and brain tumor-mediated immune suppression remains unclear. Thus, the objective of this study was to explore whether there exists a differential expression of the above checkpoint markers in the intracranial milieu as compared to tumors in the periphery, which may shed light on the mechanism behind the diminished antitumor response. Methods We identified nine patients with extracranial melanomas and matched intracranial metastases. Formalin-fixed, paraffin-embedded slides were stained for PD-L1, PD-1, LAG-3, and TIM-3 via immunohistochemistry. Qualitative analysis was performed to assess the staining of the markers in the neoplastic and lymphocytic cells, which were the two cell lineages in each biopsy.  Results Expression of PD-1 and TIM-3 between extracranial and intracranial tumoral sites was conserved. Specifically, in lymphocytes, PD-1 expression was observed in 100% of extracranial and 100% of intracranial slides, whereas TIM-3 expression was seen in 33.33% of extracranial and 33.33% of intracranial slides. Neither marker stained tumor cells, as expected. PD-L1 showed a slight variation in staining between sites, with lymphocyte staining in 100% of extracranial and 88.89% of intracranial slides, and the same percentages per site for tumor cells. The greatest variability was observed in LAG-3 lymphocyte staining, with staining in 77.78% of extracranial and 33.33% of intracranial slides. No LAG-3 staining of tumor cells was noted, as expected. Conclusion Preliminary analysis revealed the conservation of PD-L1, PD-1, LAG-3, and TIM-3 expression intra- and extracranially. This could suggest that these markers are important in maintaining an immunosuppressive phenotype at both sites. Another possibility is that this pattern of expression is associated with patients who develop brain metastasis, as this was the only subset of patients included in this study. Interestingly, LAG-3 staining of lymphocytes appeared more prominent in extracranial over intracranial tumors. Future studies should include more samples to draw out potential patterns masked by the small sample size, as well as to compare checkpoint expression in other patient groups, such as those with non-brain metastasis or those with no metastasis at all.