Neuroendocrine Tumors of Unknown Primary in the Setting of Cytoreductive Hepatectomy.

BACKGROUND: Surgical cytoreduction for neuroendocrine tumor liver metastasis (NETLM) consistently shows positive long-term outcomes. Despite reservations in guidelines for surgery when the primary tumor is unidentified (UP-NET), this study compared the surgical and oncologic long-term outcomes between patients with these rare cases undergoing cytoreductive surgery and patients who had liver resection for known primaries. METHODS: The study identified 32 unknown primary liver metastases (UP-NETLM) in 522 retrospectively evaluated patients who underwent resection of well-differentiated NETLM between January 2000 and December 2020. Tumor and patient characteristics were compared with those in 490 cases of liver metastasis from small intestinal (SI-NETLM) or pancreatic (pNETLM) primaries. Survival analysis was performed to highlight long-term outcome differences. Surgical outcomes were compared between liver resections alone and simultaneous primary resections to assess surgical risk distinctions. RESULTS: The UP-NET patients had fewer NETLMs (p = 0.004), which on the average were larger than SI-NETLMs or pNETLMs (p = 0.002). Expression of Ki-67 was balanced among the groups. Major hepatectomy was performed more often in the UP-NETLM group (p = 0.017). The 10-year survival rate of 53% for UP-NETLM was comparable with that for SI-NETML (58%; p = 0.463) and pNETLMs (47%; p = 0.497). The median hepatic progression-free survival was 26 months for the UP-NETLM patients and 25 months for the SI-NETLM patients compared to 12 months for the pNETLM patients (p < 0.001). Perioperative mortality was lower than 2%, and severe postoperative morbidity occurred in 21%, similarly distributed among all the groups. CONCLUSION: The surgical risk and long-term outcomes for the UP-NETLM patients were comparable with those for other NETLM cases, affirming the validity of equally aggressive surgical cytoreduction as a therapeutic option in carefully selected cases.

Failed recovery after pancreatoduodenectomy: A significant problem even without surgical complications.

BACKGROUND: The absence of surgical complications has traditionally been used to define successful recovery after pancreas surgery. However, patient-reported outcome measures as metrics of a challenging recovery may be superior to objective morbidity. This study aims to evaluate the use of patient-reported outcomes in assessing recovery after pancreas surgery. METHODS: Patients scheduled for pancreatoduodenectomy were prospectively enrolled between 2016 to 2018. Patient-reported outcomes were collected using the linear analog self-assessment questionnaire preoperatively and on postoperative days 2, 7, 14, 30, and monthly until 6 months. Patients were also asked if they felt fully recovered at 30 days and 6 months. Thirty-day surgical morbidity was prospectively assessed, and the comprehensive complication index at 30 days was used to categorize morbidity as major or multiple minor complications (comprehensive complication index ≥26.2) vs uncomplicated (comprehensive complication index <26.2). Clinically significant International Study Group Pancreas Surgery Grade B and C pancreatic fistulas and delayed gastric emptying were reported. χ2 and Kruskal-Wallis tests were used to assess associations with recovery by 6 months and quality of life throughout the postoperative period. RESULTS: Of 116 patients who met inclusion criteria and were enrolled, 32 (28%) had major or multiple minor complications (comprehensive complication index ≥26.2). Overall, fewer than 1 in 10 patients (7%) reported feeling fully recovered at 30 days postoperatively, whereas 55% reported feeling fully recovered at 6 months. Of patients suffering major morbidity, 62% did not recover by 6 months, whereas 38% of those in the uncomplicated group reported not being recovered at 6 months (P = .03). Patients who experienced delayed gastric emptying reported low quality-of-life scores at 1 month (P = .04) compared to those with no delayed gastric emptying, but this did not persist at 6 months (P = .80). Postoperative pancreatic fistula was not associated with quality of life at 1 or 6 months (both P > .05). In the uncomplicated patients, age, sex, surgical approach, and cancer status were not associated with failed recovery at 6 months (all P > .05), and healthier patients (American Society of Anesthesiologists 1-2) were less likely to report complete recovery (42% vs 69% American Society of Anesthesiologists 3-4, P = .04). With the exception of higher preoperative pain scores (mean 2.3 [standard deviation 2.4] among patients not fully recovered at 6 months vs 1.6 [2.2] among those fully recovered, P = .04), preoperative patient-reported outcomes were not associated with failed recovery at 6 months (all P > .05). However, lower 30-day quality of life, social activity, pain, and fatigue scores were associated with incomplete recovery at 6 months. CONCLUSION: More than 1 in 3 patients with an uncomplicated course do not feel fully recovered from pancreas surgery at 6 months; the presence of surgical complications did not universally correspond with recovery failure. In patients with complications, delayed gastric emptying appears to drive quality of life more significantly than postoperative pancreatic fistula. In patients with uncomplicated recovery, healthier patients were less likely to report full recovery at 6 months. Thirty-day patient-reported outcomes may be able to identify patients who are at risk of incomplete long-term recovery.

Revolutionizing Postoperative Ileus Monitoring: Exploring GRU-D's Real-Time Capabilities and Cross-Hospital Transferability.

Background: Postoperative ileus (POI) after colorectal surgery leads to increased morbidity, costs, and hospital stays. Identifying POI risk for early intervention is important for improving surgical outcomes especially given the increasing trend towards early discharge after surgery. While existing studies have assessed POI risk with regression models, the role of deep learning's remains unexplored. Methods: We assessed the performance and transferability (brutal force/instance/parameter transfer) of Gated Recurrent Unit with Decay (GRU-D), a longitudinal deep learning architecture, for real-time risk assessment of POI among 7,349 colorectal surgeries performed across three hospital sites operated by Mayo Clinic with two electronic health records (EHR) systems. The results were compared with atemporal models on a panel of benchmark metrics. Results: GRU-D exhibits robust transferability across different EHR systems and hospital sites, showing enhanced performance by integrating new measurements, even amid the extreme sparsity of real-world longitudinal data. On average, for labs, vitals, and assisted living status, 72.2%, 26.9%, and 49.3% respectively lack measurements within 24 hours after surgery. Over the follow-up period with 4-hour intervals, 98.7%, 84%, and 95.8% of data points are missing, respectively. A maximum of 5% decrease in AUROC was observed in brutal-force transfer between different EHR systems with non-overlapping surgery date frames. Multi-source instance transfer witnessed the best performance, with a maximum of 2.6% improvement in AUROC over local learning. The significant benefit, however, lies in the reduction of variance (a maximum of 86% decrease). The GRU-D model's performance mainly depends on the prediction task's difficulty, especially the case prevalence rate. Whereas the impact of training data and transfer strategy is less crucial, underscoring the challenge of effectively leveraging transfer learning for rare outcomes. While atemporal Logit models show notably superior performance at certain pre-surgical points, their performance fluctuate significantly and generally underperform GRU-D in post-surgical hours. Conclusion: GRU-D demonstrated robust transferability across EHR systems and hospital sites with highly sparse real-world EHR data. Further research on built-in explainability for meaningful intervention would be highly valuable for its integration into clinical practice.

Deep learning predicts postoperative opioids refills in a multi-institutional cohort of surgical patients.

BACKGROUND: To combat the opioid epidemic, several strategies were implemented to limit the unnecessary prescription of opioids in the postoperative period. However, this leaves a subset of patients who genuinely require additional opioids with inadequate pain control. Deep learning models are powerful tools with great potential of optimizing health care delivery through a patient-centered focus. We sought to investigate whether deep learning models can be used to predict patients who would require additional opioid prescription refills in the postoperative period after elective surgery. METHODS: This is a retrospective study of patients who received elective surgical intervention at the Mayo Clinic. Adult English-speaking patients ≥18 years old, who underwent an elective surgical procedure between 2013 and 2019, were eligible for inclusion. Machine learning models, including deep learning, random forest, and eXtreme Gradient Boosting, were designed to predict patients who require opioid refills after discharge from hospital. RESULTS: A total of 9,731 patients with mean age of 62.1 years (51.4% female) were included in the study. Deep learning and random forest models predicted patients who required opioid refills with high accuracy, 0.79 ± 0.07 and 0.78 ± 0.08, respectively. Procedure performed, highest pain score recorded during hospitalization, and total oral morphine milligram equivalents prescribed at discharge were the top 3 predictors for requiring opioid refills after discharge. CONCLUSION: Deep learning models can be used to predict patients who require postoperative opioid prescription refills with high accuracy. Other machine learning models, such as random forest, can perform equal to deep learning, increasing the applicability of machine learning for combating the opioid epidemic.

Value of Surgical Cytoreduction in Patients with Small Intestinal Neuroendocrine Tumors Metastatic to the Liver and Peritoneum.

BACKGROUND: Cytoreductive hepatectomy can improve survival and symptoms of hormonal excess in patients with small intestinal neuroendocrine tumor (siNET) liver metastases, but whether to proceed when peritoneal metastases are encountered at the time of planned cytoreductive hepatectomy is controversial. METHODS: This was a retrospective review of patients who underwent surgical management of metastatic siNETs at Mayo Clinic between 2000 and 2020. Patients who underwent cytoreductive operation for isolated liver metastases or both liver and peritoneal metastases were compared. RESULTS: Of 261 patients who underwent cytoreductive operation for siNETs, 211 had isolated liver metastases and 50 had liver and peritoneal metastases. Complete cytoreduction was achieved in 78% of patients with isolated liver metastases and 56% of those with liver and peritoneal metastases (p = 0.002). After complete cytoreduction, median overall survival (OS) was 11.5 years for isolated liver metastases and 11.2 years for liver and peritoneal metastases (p = 0.10), and relief of carcinoid syndrome was ≥ 97% in both groups. After incomplete cytoreduction with debulking of > 90% of hepatic disease and/or closing Lyon score of 1-2, median OS was 6.4 years for isolated liver metastases and 7.1 years for liver and peritoneal metastases (p = 0.12). CONCLUSIONS: Patients with siNETs metastatic to both the liver and peritoneum have favorable outcomes after aggressive surgical cytoreduction, with the best outcomes observed after complete cytoreduction. Therefore, the presence of peritoneal metastases should not by itself preclude surgical cytoreduction in this population.

Academic Surgery in the Era of Large Language Models: A Review.

Importance: This review aims to assess the benefits and risks of implementing large language model (LLM) solutions in an academic surgical setting. Observations: The integration of LLMs and artificial intelligence (AI) into surgical practice has generated international attention with the emergence of OpenAI's ChatGPT and Google's Bard. From an administrative standpoint, LLMs have the potential to revolutionize academic practices by reducing administrative burdens and improving efficiency. LLMs have the potential to facilitate surgical research by increasing writing efficiency, building predictive models, and aiding in large dataset analysis. From a clinical standpoint, LLMs can enhance efficiency by triaging patient concerns and generating automated responses. However, challenges exist, such as the need for improved LLM generalization performance, validating content, and addressing ethical concerns. In addition, patient privacy, potential bias in training, and legal responsibility are important considerations that require attention. Research and precautionary measures are necessary to ensure safe and unbiased use of LLMs in surgery. Conclusions and Relevance: Although limitations exist, LLMs hold promise for enhancing surgical efficiency while still prioritizing patient care. The authors recommend that the academic surgical community further investigate the potential applications of LLMs while being cautious about potential harms.

Outcomes of Neoadjuvant Chemotherapy for Invasive Intraductal Papillary Mucinous Neoplasm Compared with de Novo Pancreatic Adenocarcinoma.

BACKGROUND: The management of invasive intraductal papillary mucinous cystic neoplasm (I-IPMN) does not differ from de novo pancreatic ductal adenocarcinoma (PDAC); however, I-IPMNs are debated to have better prognosis. Despite being managed similarly to PDAC, no data are available on the response of I-IPMN to neoadjuvant chemotherapy. METHODS: All patients undergoing pancreatic resection for a pancreatic adenocarcinoma from 2011 to 2022 were included. The PDAC and I-IPMN cohorts were compared to evaluate response to neoadjuvant therapy (NAT) and overall survival (OS). RESULTS: This study included 1052 PDAC patients and 105 I-IPMN patients. NAT was performed in 25% of I-IPMN patients and 65% of PDAC patients. I-IPMN showed a similar pattern of pathological response to NAT compared with PDAC (p = 0.231). Furthermore, positron emission tomography (PET) response (71% vs. 61%; p = 0.447), CA19.9 normalization (85% vs. 76%, p = 0.290), and radiological response (32% vs. 37%, p = 0.628) were comparable between I-IPMN and PDAC. A significantly higher OS and disease-free survival (DFS) of I-IPMN was denoted by Kaplan-Meier analysis, with a p-value of < 0.001 in both plots. In a multivariate analysis, I-IPMN histology was independently associated with lower risk of recurrence and death. CONCLUSIONS: I-IPMN patients have a longer OS and DFS after surgical treatment when compared with PDAC patients. The more favorable oncologic outcome of I-IPMNs does not seem to be related to early detection, as I-IPMN histological subclass is independently associated with a lower risk of disease recurrence. Moreover, neoadjuvant effect on I-IPMN was non-inferior to PDAC in terms of pathological, CA19.9, PET, and radiological response and thus can be considered in selected patients.

Supply and demand of hepatopancreatobiliary surgeons in the United States.

BACKGROUND: Hepatopancreatobiliary (HPB) surgery requires specialized training and adequate case volumes to maintain procedural proficiency and optimal outcomes. Studies of individual HPB surgeon supply related to annual HPB case demand are sparse. This study assesses the supply and demand of the HPB surgical workforce in the United States (US). METHODS: The National Inpatient Sample (NIS) was queried from 1998-2019 to estimate the number of HPB procedures performed. To approximate the number of HPB surgeons, models based on previous HPB workforce publications were employed. We then calculated the number of HPB surgeons needed to maintain volume-outcome thresholds at current reported levels of centralization. RESULTS: In 2019, approximately 37,335 patients underwent inpatient HPB procedures in the US, while an estimated 905-1191 HPB surgeons were practicing. Assuming 50% centralization and an optimal volume-outcome threshold of 24 HPB cases-per-year, only 778 HPB surgeons were needed. Without adjustment in centralization, by 2030 there will be a demand of fewer than 12 annual cases per HPB surgeon. CONCLUSION: The current supply of HPB surgeons may exceed demand in the United States. Without alteration in training pathways or improved care centralization, by 2030, there will be insufficient HPB case volume per surgeon to maintain published volume-outcome standards.

Early vs conventional onset pancreatic ductal adenocarcinoma: analysis of surgical and oncologic outcomes in patients undergoing curative intent resection.

BACKGROUND: Pancreatic ductal adenocarcinoma (PDAC) impacts patients in their 60s, but its incidence in younger patients is increasing. We hypothesize that younger patients may have worse oncologic outcomes. METHODS: Patients who underwent curative pancreatic resection for PDAC between January 2011 and December 2021 at a single institution were analyzed. Early-onset pancreatic cancer (EOPC) was defined as pancreatic cancer diagnosed in patients ≤50 years. Clinical and survival outcomes were compared between EOPC and Conventional Onset Pancreas Cancer (COPC). RESULTS: A total of 1133 patients were identified, 65 (5.7%) were EOPC. Preoperative patient characteristics including sex, smoking status, alcohol habitus, diabetes mellitus, CA 19-9, and neoadjuvant therapy were similar between EOPC and COPC (p > 0.05). EOPC patients were more likely non-white (p = 0.03), had lower ASA scores (p = 0.02) and larger median tumor size (33 vs 28 mm, p = 0.04), but had similar pathological stages and rate of R0 resections (p > 0.05). Postoperative outcomes were similar (p > 0.05). There was no statistically significant difference in overall (HR 0.93, CI 0.64, 1.33; p = 0.68) or recurrence free (HR 1.05, CI 0.75, 1.48; p = 0.77) survival between the EOPC and COPC after adjusting for significant factors. CONCLUSION: Patients with EOPC who underwent surgical resection had similar oncological outcomes compared to patients with COPC.

Cytoreduction and Hyperthermic Intraperitoneal Paclitaxel and Cisplatin for Gastric Cancer with Peritoneal Metastasis.

BACKGROUND: Peritoneal metastasis (PM) is the most common site of dissemination of gastric cancer (GC) and is associated with a poor prognosis. Cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) for GC with PM remains controversial due to modest survival and significant morbidity. METHODS: We conducted a retrospective analysis of patients with GC and PM treated with CRS and HIPEC with cisplatin and paclitaxel for 90 min from June 2019 to December 2022. RESULTS: Twenty-two patients were included and received a median of 7 (interquartile range [IQR] 4-8) cycles of neoadjuvant systemic therapy. Seventeen patients (77%) underwent a single neoadjuvant laparoscopic HIPEC, and six (27%) patients received chemoradiation. The median Peritoneal Carcinomatosis Index at the time of CRS was 1 (IQR 0-4), and 21 patients (95%) underwent complete cytoreduction (CC-0). An R0 resection was achieved in 20 (91%) patients, and the median length of stay was 5.5 (IQR 4-7.5) days. There were six (27%) 90-day major complications (Clavien-Dindo grade ≥ 3), one (4%) Common Terminology Classification for Adverse Events (CTCAE) grade 4 cytopenia, and one (4%) acute kidney injury. The rate of anastomotic leak (all grades) was 14%, the 30-day readmission rate was 18%, and the 90-day mortality rate was 0%. At a median follow-up of 24 months, the median progression-free survival (PFS) and overall survival (OS) were not reached. The 1-, 2-, and 3-year PFS rates were 65%, 56%, and 40%, respectively, and the 1-, 2-, and 3-year OS rates were 96%, 78%, and 55%, respectively. CONCLUSIONS: CRS and HIPEC with paclitaxel and cisplatin is well tolerated and is associated with favorable oncologic and perioperative outcomes.

An APRI+ALBI Based Multivariable Model as Preoperative Predictor for Posthepatectomy Liver Failure.

OBJECTIVE AND BACKGROUND: Clinically significant posthepatectomy liver failure (PHLF B+C) remains the main cause of mortality after major hepatic resection. This study aimed to establish an APRI+ALBI, aspartate aminotransferase to platelet ratio (APRI) combined with albumin-bilirubin grade (ALBI), based multivariable model (MVM) to predict PHLF and compare its performance to indocyanine green clearance (ICG-R15 or ICG-PDR) and albumin-ICG evaluation (ALICE). METHODS: 12,056 patients from the National Surgical Quality Improvement Program (NSQIP) database were used to generate a MVM to predict PHLF B+C. The model was determined using stepwise backwards elimination. Performance of the model was tested using receiver operating characteristic curve analysis and validated in an international cohort of 2,525 patients. In 620 patients, the APRI+ALBI MVM, trained in the NSQIP cohort, was compared with MVM's based on other liver function tests (ICG clearance, ALICE) by comparing the areas under the curve (AUC). RESULTS: A MVM including APRI+ALBI, age, sex, tumor type and extent of resection was found to predict PHLF B+C with an AUC of 0.77, with comparable performance in the validation cohort (AUC 0.74). In direct comparison with other MVM's based on more expensive and time-consuming liver function tests (ICG clearance, ALICE), the APRI+ALBI MVM demonstrated equal predictive potential for PHLF B+C. A smartphone application for calculation of the APRI+ALBI MVM was designed. CONCLUSION: Risk assessment via the APRI+ALBI MVM for PHLF B+C increases preoperative predictive accuracy and represents an universally available and cost-effective risk assessment prior to hepatectomy, facilitated by a freely available smartphone app.

Oncologic Outcomes of Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy for Highly Selected Patients with Metastatic Pancreatic Ductal Adenocarcinoma.

BACKGROUND: Peritoneal metastases (PM) from pancreatic ductal adenocarcinoma (PDAC) are currently treated with palliative systemic chemotherapy alone, with unsatisfactory results. Cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemotherapy (HIPEC) may provide an oncologic benefit for highly selected patients. PATIENTS AND METHODS: Patients with PDAC and isolated PM who completed ≥ 6 months of systemic chemotherapy with objective response between 2017 and 2022 were retrospectively reviewed. All patients met the inclusion/exclusion criteria as per our previously published PDAC CRS/HIPEC protocol. Patients who underwent CRS/HIPEC were compared with matched patients who underwent systemic therapy alone. Overall survival (OS) from diagnosis of PM and progression-free survival (PFS) from CRS/HIPEC was evaluated. RESULTS: In total, 61 patients met the inclusion criteria: 38 underwent systemic therapy alone and 23 CRS/HIPEC. There were no differences in baseline prognostic factors, including age, sex, tumor size, tumor location, anatomic resectability, or serum cancer antigen (CA) 19-9 (p > 0.05). Median OS from PM diagnosis in patients who underwent systemic therapy alone was 19 months with 1, 2, and 3 year OS of 81%, 31%, and 8%, respectively. In contrast, median OS from PM diagnosis in patients who underwent CRS/HIPEC was 41 months with improved 1, 2, and 3 year OS of 91%, 66%, and 59%, respectively (p = 0.002). In the 21 patients who achieved complete cytoreduction (CC-0), no adjuvant therapy was administered and the median PFS was 17 months. CONCLUSIONS: CRS/HIPEC in highly selected patients with PDAC and PM results in promising oncologic outcomes that are unlikely to be achieved with systemic chemotherapy alone. Further investigation is warranted and ongoing (NCT04858009).