Neighborhood-Level Socioeconomic Disadvantage Predicts Outcomes in Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy for Peritoneal Malignancy.

BACKGROUND: Cytoreductive surgery and hyperthermic intraperitoneal chemotherapy (CRS/HIPEC) improves survival in select patients with peritoneal metastases (PM), but the impact of social determinants of health on CRS/HIPEC outcomes remains unclear. PATIENTS AND METHODS: A retrospective review was conducted of a multi-institutional database of patients with PM who underwent CRS/HIPEC in the USA between 2000 and 2017. The area deprivation index (ADI) was linked to the patient's residential address. Patients were categorized as living in low (1-49) or high (50-100) ADI residences, with increasing scores indicating higher socioeconomic disadvantage. The primary outcome was overall survival (OS). Secondary outcomes included perioperative complications, hospital/intensive care unit (ICU) length of stay (LOS), and disease-free survival (DFS). RESULTS: Among 1675 patients 1061 (63.3%) resided in low ADI areas and 614 (36.7%) high ADI areas. Appendiceal tumors (n = 1102, 65.8%) and colon cancer (n = 322, 19.2%) were the most common histologies. On multivariate analysis, high ADI was not associated with increased perioperative complications, hospital/ICU LOS, or DFS. High ADI was associated with worse OS (median not reached versus 49 months; 5 year OS 61.0% versus 28.2%, P < 0.0001). On multivariate Cox-regression analysis, high ADI (HR, 2.26; 95% CI 1.13-4.50; P < 0.001), cancer recurrence (HR, 2.26; 95% CI 1.61-3.20; P < 0.0001), increases in peritoneal carcinomatosis index (HR, 1.03; 95% CI 1.01-1.05; P < 0.001), and incomplete cytoreduction (HR, 4.48; 95% CI 3.01-6.53; P < 0.0001) were associated with worse OS. CONCLUSIONS: Even after controlling for cancer-specific variables, adverse outcomes persisted in association with neighborhood-level socioeconomic disadvantage. The individual and structural-level factors leading to these cancer disparities warrant further investigation to improve outcomes for all patients with peritoneal malignancies.

Validation of the AJCC 8th Edition Staging System for Disseminated Appendiceal Cancer Patients Treated with Cytoreductive Surgery and Hyperthermic Intraperitoneal Chemotherapy: A Multi-institutional Analysis.

BACKGROUND: The AJCC 8th edition stratifies stage IV disseminated appendiceal cancer (dAC) patients based on grade and pathology. This study was designed to externally validate the staging system and to identify predictors of long-term survival. METHODS: A 12-institution cohort of dAC patients treated with CRS ± HIPEC was retrospectively analyzed. Overall survival (OS) and recurrence-free survival (RFS) were analyzed by using Kaplan-Meier and log-rank tests. Univariate and multivariate cox-regression was performed to assess factors associated with OS and RFS. RESULTS: Among 1009 patients, 708 had stage IVA and 301 had stage IVB disease. Median OS (120.4 mo vs. 47.2 mo) and RFS (79.3 mo vs. 19.8 mo) was significantly higher in stage IVA compared with IVB patients (p < 0.0001). RFS was greater among IVA-M1a (acellular mucin only) than IV M1b/G1 (well-differentiated cellular dissemination) patients (NR vs. 64 mo, p = 0.0004). Survival significantly differed between mucinous and nonmucinous tumors (OS 106.1 mo vs. 41.0 mo; RFS 46.7 mo vs. 21.2 mo, p < 0.05), and OS differed between well, moderate, and poorly differentiated (120.4 mo vs. 56.3 mo vs. 32.9 mo, p < 0.05). Both stage and grade were independent predictors of OS and RFS on multivariate analysis. Acellular mucin and mucinous histology were associated with better OS and RFS on univariate analysis only. CONCLUSIONS: AJCC 8th edition performed well in predicting outcomes in this large cohort of dAC patients treated with CRS ± HIPEC. Separation of stage IVA patients based on the presence of acellular mucin improved prognostication, which may inform treatment and long-term, follow-up strategies.

Patterns of Unnecessary Insurer Prior Authorization Denials in a Complex Surgical Oncology Practice.

INTRODUCTION: Insurance prior authorization (PA) is a determination of need, required by a health insurer for an ordered test/procedure. If the test/procedure is denied, a peer-to-peer (P2P) discussion between ordering provider and payer is used to appeal the decision. The objective of this study was to measure the number and patterns of unnecessary PA denials. METHODS: This was a retrospective review at a quaternary cancer center from October 2021 to March 2022. Included were all patients with outpatient imaging orders for surgical planning or surveillance of gastrointestinal, endocrine, or skin cancer. Primary outcome was unnecessary initial denial (UID) defined as an order that required preauthorization, was initially denied by the insurer, and subsequently overturned by P2P. RESULTS: Nine hundred fifty seven orders were placed and 419 required PA (44%). Of tests requiring authorization, 55/419 (13.1%) were denied. Variability in the likelihood of initial denial was seen across insurers, ranging from 0% to 57%. Following P2P, 32/55 were overturned (58.2% UID). The insurers most likely to have a UID were Aetna (100%), Anthem (77.8%), and Cigna (50.0%). UID was most common for gastrointestinal (58.9%) and endocrine (58.3%) cancers. Average P2P was 33.5 min (interquartile range 28-40). CONCLUSIONS: The majority of imaging studies initially denied were overturned after P2P. If all UIDs were eliminated, this would represent 108 less P2P discussions with an estimated time-savings of 60.3 h annually within a high-volume surgical oncology practice. Combined personnel costs to the health systems and stress on patients with cancer due to image-associated PAs and P2P appear hard to justify.

Prediction of Early Recurrence Following CRS/HIPEC in Patients With Disseminated Appendiceal Cancer.

INTRODUCTION: In patients with disseminated appendiceal cancer (dAC) who underwent cytoreductive surgery (CRS) with hyperthermic intraperitoneal chemotherapy (HIPEC), characterizing and predicting those who will develop early recurrence could provide a framework for personalizing follow-up. This study aims to: (1) characterize patients with dAC that are at risk for recurrence within 2 y following of CRS ± HIPEC (early recurrence; ER), (2) utilize automated machine learning (AutoML) to predict at-risk patients, and (3) identifying factors that are influential for prediction. METHODS: A 12-institution cohort of patients with dAC treated with CRS ± HIPEC between 2000 and 2017 was used to train predictive models using H2O.ai's AutoML. Patients with early recurrence (ER) were compared to those who did not have recurrence or presented with recurrence after 2 y (control; C). However, 75% of the data was used for training and 25% for validation, and models were 5-fold cross-validated. RESULTS: A total of 949 patients were included, with 337 ER patients (35.5%). Patients with ER had higher markers of inflammation, worse disease burden with poor response, and received greater intraoperative fluids/blood products. The highest performing AutoML model was a Stacked Ensemble (area under the curve = 0.78, area under the curve precision recall = 0.66, positive predictive value = 85%, and negative predictive value = 63%). Prediction was influenced by blood markers, operative course, and factors associated with worse disease burden. CONCLUSIONS: In this multi-institutional cohort of dAC patients that underwent CRS ± HIPEC, AutoML performed well in predicting patients with ER. Variables suggestive of poor tumor biology were the most influential for prediction. Our work provides a framework for identifying patients with ER that might benefit from shorter interval surveillance early after surgery.

Identifying patient profiles of disparate care in resectable pancreas cancer using latent class analysis.

BACKGROUND AND OBJECTIVES: Disparities in pancreas cancer care are multifactorial, but factors are often examined in isolation. Research that integrates these factors in a single conceptual framework is lacking. We use latent class analysis (LCA) to evaluate the association between intersectionality and patterns of care and survival in patients with resectable pancreas cancer. METHODS: LCA was used to identify demographic profiles in resectable pancreas cancer (n = 140 344) diagnosed from 2004 to 2019 in the National Cancer Database (NCDB). LCA-derived patient profiles were used to identify differences in receipt of minimum expected treatment (definitive surgery), optimal treatment (definitive surgery and chemotherapy), time to treatment, and overall survival. RESULTS: Minimum expected treatment (hazard ratio [HR] 0.69, 95% confidence interval [CI]: 0.65, 0.75) and optimal treatment (HR 0.58, 95% CI: 0.55, 0.62) were associated with improved overall survival. Seven latent classes were identified based on age, race/ethnicity, and socioeconomic status (SES) attributes (zip code-linked education and income, insurance, geography). Compared to the referent group (≥65 years + White + med/high SES), the ≥65 years + Black profile had the longest time-to-treatment (24 days vs. 28 days) and lowest odds of receiving minimum (odds ratio [OR] 0.67, 95% CI: 0.64, 0.71) or optimal treatment (OR 0.76, 95% CI: 0.72, 0.81). The Hispanic patient profile had the lowest median overall survival-55.3 months versus 67.5 months. CONCLUSIONS: Accounting for intersectionality in the NCDB resectable pancreatic cancer patient cohort identifies subgroups at higher risk for inequities in care. LCA demonstrates that older Black patients and Hispanic patients are at particular risk for being underserved and should be prioritiz for directed interventions.

Intraoperative Pancreatic Neck Margin Assessment During Pancreaticoduodenectomy for Pancreatic Adenocarcinoma in the Era of Neoadjuvant Therapy: A Multi-institutional Analysis from the Central Pancreatic Consortium.

BACKGROUND: Data regarding the survival impact of converting frozen-section (FS):R1 pancreatic neck margins to permanent section (PS):R0 by additional resection (i.e., converted-R0) during upfront pancreaticoduodenectomy for pancreatic ductal adenocarcinoma (PDAC) are conflicting. The impact of neoadjuvant therapy on this practice and its relationship with overall survival (OS) is incompletely understood. METHODS: We reviewed PDAC patients (80% borderline resectable/locally advanced [BR/LA]) undergoing pancreaticoduodenectomy after neoadjuvant therapy at seven, academic, high-volume centers (2010-2018). Multivariable models examined the association of PS:R0, PS:R1, and converted-R0 margins with OS. RESULTS: Of 272 patients receiving at least 2 (median 4) cycles of neoadjuvant chemotherapy (71% mFOLFIRINOX or gemcitabine/nab-paclitaxel) and undergoing pancreaticoduodenectomy with intraoperative frozen-section assessment of the transected pancreatic neck margin, PS:R0 (n = 220, 80.9%) was observed in a majority of patients; 18 patients (6.6%) had converted-R0 margins following additional resection, whereas 34 patients (12.5%) had persistently positive PS:R1 margins. At a median follow-up of 42 months, PS:R0 resection was associated with improved OS compared with either converted-R0 or PS:R1 resection (median 25 vs. 14 vs. 16 months, respectively; p = 0.023), with no survival difference between the converted-R0 and PS:R1 groups (p = 0.9). On Cox regression, SMA margin positivity (hazard ratio 2.2, p = 0.012), but not neck margin positivity (hazard ratio 1.2, p = 0.65), was associated with worse OS. CONCLUSIONS: In this multi-institutional cohort of predominantly BR/LA PDAC patients undergoing pancreaticoduodenectomy following modern neoadjuvant therapy, pursuing a negative neck margin intraoperatively if the initial margin is positive does not appear to be associated with improved survival.

Pathologic nodal staging for clinically node negative soft tissue sarcoma of the extremities.

BACKGROUND AND OBJECTIVES: Synovial, clear cell, angiosarcoma, rhabdomyosarcoma, and epithelioid (SCARE) soft tissue sarcoma are at risk for nodal involvement, although the nodal positivity rates and impact on prognostication in clinically node negative patients are not well described. METHODS: Patients with extremity SCARE sarcoma without clinical nodal involvement undergoing surgical resection in the National Cancer Database (2004-2017) were included. Logistic regression was used to evaluate the likelihood of nodal surgery and nodal positivity. Kaplan-Meier method and Cox regression were used to assess associations of nodal status to overall survival. RESULTS: We included 4158 patients, and 669 patients (16%) underwent regional lymph node surgery (RLNS). On multivariable logistic analysis, patients with epithelioid (odds ratio [OR]: 3.77; p < .001) and clear cell (OR: 6.38; p < .001) were most likely to undergo RLNS. Forty-five patients (7%) had positive nodes. Clear cell sarcoma (14%) and angiosarcoma (13%) had the highest rates of nodal positivity. Patients with positive nodes had reduced 5-year overall survival, and the stratification was largest in clear cell and angiosarcoma. CONCLUSION: Discordance exists between selection for pathologic nodal evaluation and factors associated with nodal positivity. Clinically node negative patients with clear cell and angiosarcoma should be considered for pathologic nodal evaluation.

Breast cancer treatment delays by socioeconomic and health care access latent classes in Black and White women.

BACKGROUND: Breast cancer mortality is higher for Black and younger women. This study evaluated 2 possible contributors to disparities-time to treatment and treatment duration-by race and age. METHODS: Among 2841 participants with stage I-III disease in the Carolina Breast Cancer Study, we identified groups of women with similar patterns of socioeconomic status (SES), access to care, and tumor characteristics using latent class analysis. We then evaluated latent classes in association with treatment delay (initiation >60 days after diagnosis) and treatment duration (in quartiles by treatment modality). RESULTS: Thirty-two percent of younger Black women were in the highest quartile of treatment duration (versus 22% of younger White women). Black women experienced a higher frequency of delayed treatment (adjusted relative frequency difference [RFD], 5.5% [95% CI, 3.2%-7.8%]) and prolonged treatment duration (RFD, 8.8% [95% CI, 5.7%-12.0%]). Low SES was significantly associated with treatment delay among White women (RFD, 3.5% [95% CI, 1.1%-5.9%]), but treatment delay was high at all levels of SES in Black women (eg, 11.7% in high SES Black women compared with 10.6% and 6.7% among low and high SES White women, respectively). Neither SES nor access to care classes were significantly associated with delayed initiation among Black women, but both low SES and more barriers were associated with treatment duration across both groups. CONCLUSIONS: Factors that influence treatment timeliness persist throughout the care continuum, with prolonged treatment duration being a sensitive indicator of differences by race, SES, and care barriers.

Disparities of Management of the Axilla in Women With Clinically Node Negative Breast Cancer.

BACKGROUND: In women with clinically node-negative breast cancer, sentinel lymph node biopsy is the first step in axillary staging. A randomized trial published in 2013 concluded that patients with sentinel lymph node micrometastases (N1mi) do not benefit from axillary lymph node dissection (ALND). We hypothesized that disparities exist in management of the axilla in node-negative patients. METHODS: We included women aged >40 years with nonmetastatic, clinically node-negative breast cancer from 2014 to 2016 in the National Cancer Database. Women treated neoadjuvantly, with large tumors (cT4), or no tumor (cT0) were excluded. Multivariable logistic regression identified patient and facility characteristics associated with undergoing ALND as first axillary surgery and completion ALND in the setting of N1mi disease. RESULTS: Of 273,951 patients, 22,898 (8%) underwent ALND first. These patients were more likely to be Hispanic (OR: 1.21, 95% CI: 1.10, 1.32), have Medicare (OR: 1.13, 95% CI: 1.03, 1.24), be uninsured (OR: 1.28, 95% CI: 1.08, 1.53), have lower educational attainment (OR: 1.24, 95% CI: 1.17, 1.32), be treated at a community hospital (OR: 1.62, 95% CI: 1.52, 1.74), or reside in the South (OR: 1.19, 95% CI: 1.12, 1.26). In the sentinel lymph node biopsy first group, 8,882 (4%) were classified as N1mi and 1,872 (21%) underwent subsequent ALND. These patients were more likely to be Hispanic (OR: 1.70, 95% CI: 1.19, 2.42) and have the lowest income (OR: 1.62, 95% CI: 1.15, 2.27). CONCLUSION: Disparities persist in implementation of evidence-based management of the axilla in women with clinically node-negative breast cancer.

Survival Outcomes Associated With Clinical and Pathological Response Following Neoadjuvant FOLFIRINOX or Gemcitabine/Nab-Paclitaxel Chemotherapy in Resected Pancreatic Cancer.

OBJECTIVE: To compare the survival outcomes associated with clinical and pathological response in pancreatic ductal adenocarcinoma (PDAC) patients receiving neoadjuvant chemotherapy (NAC) with FOLFIRINOX (FLX) or gemcitabine/nab-paclitaxel (GNP) followed by curative-intent pancreatectomy. BACKGROUND: Newer multiagent NAC regimens have resulted in improved clinical and pathological responses in PDAC; however, the effects of these responses on survival outcomes remain unknown. METHODS: Clinicopathological and survival data of PDAC patients treated at 7 academic medical centers were analyzed. Primary outcomes were overall survival (OS), local recurrence-free survival (L-RFS), and metastasis-free survival (MFS) associated with biochemical (CA 19-9 decrease ≥50% vs <50%) and pathological response (complete, pCR; partial, pPR or limited, pLR) following NAC. RESULTS: Of 274 included patients, 46.4% were borderline resectable, 25.5% locally advanced, and 83.2% had pancreatic head/neck tumors. Vein resection was performed in 34.7% and 30-day mortality was 2.2%. R0 and pCR rates were 82.5% and 6%, respectively. Median, 3-year, and 5-year OS were 32 months, 46.3%, and 30.3%, respectively. OS, L-RFS, and MFS were superior in patients with marked biochemical response (CA 19-9 decrease ≥50% vs <50%; OS: 42.3 vs 24.3 months, P < 0.001; L-RFS-27.3 vs 14.1 months, P = 0.042; MFS-29.3 vs 13 months, P = 0.047) and pathological response [pCR vs pPR vs pLR: OS- not reached (NR) vs 40.3 vs 26.1 months, P < 0.001; L-RFS-NR vs 24.5 vs 21.4 months, P = 0.044; MFS-NR vs 23.7 vs 20.2 months, P = 0.017]. There was no difference in L-RFS, MFS, or OS between patients who received FLX or GNP. CONCLUSION: This large, multicenter study shows that improved biochemical, pathological, and clinical responses associated with NAC FLX or GNP result in improved OS, L-RFS, and MFS in PDAC. NAC with FLX or GNP has similar survival outcomes.

Extensive cytoreductive surgery for appendiceal carcinomatosis: morbidity, mortality, and survival.

BACKGROUND: Cytoreductive surgery (CRS) and hyperthermic intraperitoneal chemoperfusion (HIPEC) are frequently used to treat appendiceal carcinomatosis. Some patients require multivisceral resection because of the volume of disease. It is unclear whether extent of CRS impacts survival in appendiceal carcinomatosis. METHODS: We analyzed 282 patients undergoing attempted CRS/HIPEC for appendiceal carcinomatosis. Patients were defined as having undergone Extensive CRS (n = 60) if they had >3 organ resections or >2 anastomoses; a subgroup of Extreme CRS patients (n = 10) had ≥5 organ resections and ≥3 anastomoses. Kaplan-Meier survival curves and multivariate Cox-regression models were used to identify prognostic factors affecting outcomes. RESULTS: Relative to the comparison group, patients undergoing Extensive CRS had a higher median peritoneal carcinomatosis index, operative duration, blood loss, and length of stay. No difference in completeness of cytoreduction, severe morbidity, or 60-day mortality was evident. Subgroup analysis of 10 patients undergoing extreme CRS likewise revealed no increase in severe morbidity or mortality. Median progression-free (PFS) and overall survival (OS) were 23.5 and 74 months in the comparison group; 18.5 (p = 0.086) and 51 (p = 0.85) months in the Extensive CRS group; and 40 months and not reached in the Extreme CRS subgroup. In a multivariable analysis, extent of CRS was not independently associated with PFS or OS. CONCLUSIONS: Extensive CRS is associated with greater OR time, blood loss, and length of stay, but is not associated with higher morbidity, mortality, or inferior oncologic outcomes in patients with appendiceal carcinomatosis.

Significance of serum tumor marker levels in peritoneal carcinomatosis of appendiceal origin.

BACKGROUND: The significance of tumor markers in patients with appendiceal carcinomatosis is poorly defined. We determined preoperative and postoperative tumor marker levels in patients undergoing cytoreductive surgery (CRS) and heated intraperitoneal chemoperfusion (HIPEC) and examined their association with clinicopathologic features and survival. METHODS: A total of 176 patients undergoing attempted CRS/HIPEC for appendiceal carcinomatosis had at least 1 tumor marker measured. Marker levels were correlated with tumor characteristics and oncologic outcomes. Kaplan-Meier curves and multivariate Cox regression models were used to identify prognostic factors affecting progression and survival. RESULTS: At least 1 marker was elevated prior to CRS/HIPEC in 70 % of patients (CEA, 54.1 %; CA19-9, 47.7 %; CA-125, 47.2 %). Among patients with elevated preoperative marker levels, normalization occurred postoperatively in 79.4 % for CEA, 92.3 % for CA19-9, and 60 % for CA-125. Absolute preoperative tumor marker levels correlated with peritoneal carcinomatosis index (PCI) (p < .0002), and the number of elevated markers was associated with PCI and progression-free survival (PFS). Elevated postoperative CEA level was associated with decreased PFS (median, 13 vs 36 months, p = .0008). On multivariate Cox regression analysis, elevated preoperative CA19-9 was associated with shorter PFS (hazard ratio [HR] 2.9, 95 % confidence interval [95 % CI] 1.5-5.3, p = .0008), whereas elevated CA-125 was associated with shorter overall survival (HR 2.6, 95 % CI 1.3-5.4, p = .01). CONCLUSIONS: Most patients with appendiceal carcinomatosis will have at least 1 elevated tumor marker and will normalize following CRS/HIPEC, allowing for ongoing surveillance. CA19-9 is a promising biomarker for early progression following CRS/HIPEC, whereas CA-125 is associated with shorter survival.

D2 lymphadenectomy with surgical ex vivo dissection into node stations for gastric adenocarcinoma can be performed safely in Western patients and ensures optimal staging.

BACKGROUND: The AJCC recommends examination of >16 nodes to stage gastric adenocarcinoma. D2 lymphadenectomy (LAD) followed by surgical ex vivo dissection (SEVD) into nodal stations is standard at many high-volume Asian centers, but potential increases in morbidity and mortality have slowed adoption of D2 LAD in some Western centers. METHODS: A total of 331 patients with gastric adenocarcinoma who underwent surgical resection at one Western institution from 1995 to 2010 were examined. RESULTS: Median age of patients was 69 years old, 65% were male, and 84% were white. D1 LAD was performed in 285 patients (86%) and D2 LAD in 46 patients (14%), with SEVD being performed in 17 patients (37%) in the D2 group. D2 LAD with or without SEVD was performed much more commonly between 2006 and 2010. For the D1, D2 without SEVD, and D2 with SEVD groups, the median number of examined nodes and percentage with >16 examined nodes were 16 and 51%, 27 and 93%, and 40 and 100%, respectively. Major complications occurred in 16% of the D1 group and 17% of the D2 group (p>0.05), and 30-day mortality was 3% for the D1 group and 0% for the D2 group. D2 LAD was a positive prognostic factor for overall survival on univariate (p=0.027) and multivariate analyses (p=0.005), but there were several possible confounding variables. CONCLUSIONS: D2 LAD at our Western institution was performed with low morbidity and no mortality. Optimal staging occurred after D2 LAD combined with SEVD, where a median of 40 nodes were examined and all patients had >16 examined nodes.

Noncurative gastrectomy for gastric adenocarcinoma should only be performed in highly selected patients.

BACKGROUND: The benefit of surgical resection in patients with incurable gastric adenocarcinoma is controversial. METHODS: A total of 289 patients who presented with advanced or metastatic gastric cancer from 1995 to 2010 were retrospectively reviewed. RESULTS: Ten patients (3.5 %) required emergent surgery at presentation and were excluded from further analyses. Patients who underwent nonemergent surgery at presentation (n = 110, 38.1 %) received either gastric resection (group A, n = 46, 42 %) or surgery without resection (group B, n = 64, 58 %). Procedures in group A included distal gastrectomy (n = 25, 54 %), total gastrectomy (n = 17, 37 %), and proximal/esophagogastrectomy (n = 4, 9 %). Procedures in group B included laparoscopy (n = 17, 27 %), open exploration (n = 25, 39 %), gastrostomy and/or jejunostomy tube (n = 12, 19 %), and gastrojejunostomy (n = 10, 16 %). Group A required a stay in the intensive care unit or additional invasive procedure significantly more often than group B (15 vs. 2 %, p = 0.009). Four patients in group A (8.7 %) and three patients in group B (4.7 %) died within 30 days of surgery (p = 0.45). When the 110 patients who underwent nonemergent surgery (groups A and B) were compared to nonoperatively managed patients (group C, n = 169, 58 %), median overall survival did not significantly differ (8.6 vs. 9.2 vs. 7.7 months; p > 0.05). Three patients in group B (4.7 %) and three in group C (1.8 %) ultimately required an operation for their primary tumor. CONCLUSIONS: Patients with gastric adenocarcinoma who present with advanced or metastatic disease not amenable to curative resection infrequently require emergent surgery. Noncurative resection is associated with significant perioperative morbidity and mortality as well as limited overall survival, and should therefore be performed judiciously.

Automated machine learning (AutoML) can predict 90-day mortality after gastrectomy for cancer.

Early postoperative mortality risk prediction is crucial for clinical management of gastric cancer. This study aims to predict 90-day mortality in gastric cancer patients undergoing gastrectomy using automated machine learning (AutoML), optimize models for preoperative prediction, and identify factors influential in prediction. National Cancer Database was used to identify stage I-III gastric cancer patients undergoing gastrectomy between 2004 and 2016. 26 features were used to train predictive models using H2O.ai AutoML. Performance on validation cohort was measured. In 39,108 patients, 90-day mortality rate was 8.8%. The highest performing model was an ensemble (AUC = 0.77); older age, nodal ratio, and length of inpatient stay (LOS) following surgery were most influential for prediction. Removing the latter two parameters decreased model performance (AUC 0.71). For optimizing models for preoperative use, models were developed to first predict node ratio or LOS, and these predicted values were inputted for 90-day mortality prediction (AUC of 0.73-0.74). AutoML performed well in predicting 90-day mortality in a larger cohort of gastric cancer patients that underwent gastrectomy. These models can be implemented preoperatively to inform prognostication and patient selection for surgery. Our study supports broader evaluation and application of AutoML to guide surgical oncologic care.

Comparison of a lymph node ratio-based staging system with the 7th AJCC system for gastric cancer: analysis of 18,043 patients from the SEER database.

OBJECTIVES: The American Joint Committee on Cancer (AJCC) staging system for gastric cancer bases N status on absolute number of metastatic nodes, regardless of the number of examined nodes. We examined a modified staging system utilizing node ratio (Nr), the ratio of metastatic to examined nodes. METHODS: A total of 18,043 gastric cancer patients who underwent gastrectomy were identified from the US Surveillance, Epidemiology, and End Results (SEER) database. A training set was divided into 5 Nr groups, and a TNrM staging system was constructed. Median survival and overall survival, based on 7th edition AJCC and TNrM staging systems, were compared, and the analysis was repeated in a validation set. RESULTS: Median examined nodes were 10 to 11. For the training set, overall survival for all 5 AJCC N categories was significantly different when subgrouped into 15 or fewer versus more than 15 examined nodes, but overall survival was similar regardless of the number of examined nodes in 4 of 5 Nr categories. Seven AJCC stages had statistically different overall survival between subgroups, whereas only 1 TNrM stage had statistically different overall survival between subgroups. When misclassification was defined as any subgroup in which median survival fell outside the 95% confidence interval of the group's overall median survival, AJCC staging misclassified 57% of patients and TNrM staging misclassified only 12%. Similar results were found in the validation set. CONCLUSIONS: The AJCC system classifies SEER gastric cancer patients into stages in which subgroups often have wide variations in survival. For patients undergoing limited lymph node analysis, the proposed TNrM system may predict survival more accurately.