Extremity Soft Tissue Sarcoma: A Multi-Institutional Validation of Prognostic Nomograms

BACKGROUND

Prognostic nomograms for patients with resected extremity soft tissue sarcoma (STS) include the Sarculator and Memorial Sloan Kettering (MSKCC) nomograms. We sought to validate these two nomograms within a large, modern, multi-institutional cohort of resected primary extremity STS patients.

METHODS

Resected primary extremity STS patients from 2000 to 2017 were identified across nine high-volume U.S. institutions. Predicted 5- and 10-year overall survival (OS) and distant metastases cumulative incidence (DMCI), and 4-, 8-, and 12-year disease-specific survival (DSS) were calculated with Sarculator and MSKCC nomograms, respectively. Predicted survival probabilities stratified in quintiles were compared in calibration plots to observed survival assessed by Kaplan-Meier estimates. Cumulative incidence was estimated for DMCI. Harrell's concordance index (C-index) assessed discriminative ability of nomograms.

RESULTS

A total of 1326 patients underwent resection of primary extremity STS. Common histologies included: undifferentiated pleomorphic sarcoma (35%), fibrosarcoma (13%), and leiomyosarcoma (9%). Median tumor size was 8.0 cm (IQR 4.5-13.0). Tumor grade distribution was: Grade 1 (13%), Grade 2 (9%), Grade 3 (78%). Median OS was 172 months, with estimated 5- and 10-year OS of 70% and 58%. C-indices for 5- and 10-year OS (Sarculator) were 0.72 (95% CI 0.70-0.75) and 0.73 (95% CI 0.70-0.75), and 0.72 (95% CI 0.69-0.75) for 5- and 10-year DMCI. C-indices for 4-, 8-, and 12-year DSS (MSKCC) were 0.71 (95% CI 0.68-0.75). Calibration plots showed good prognostication across all outcomes.

CONCLUSIONS

Sarculator and MSKCC nomograms demonstrated good prognostic ability for survival and recurrence outcomes in a modern, multi-institutional validation cohort of resected primary extremity STS patients. External validation of these nomograms supports their ongoing incorporation into clinical practice.

Weight loss after minimally-invasive esophagectomy for esophageal cancer

e16565

Background: Nutritional deficiencies are common in esophageal cancer patients after esophagectomy. Patients frequently present with weight loss due to obstruction. Cancer therapies (chemoradiation and esophagectomy) further compromise nutrient intake through alterations in appetite, gastrointestinal function, and the catabolic effects of therapy. We aimed to understand and identify risk factors associated with post-esophagectomy weight loss. Methods: Patients undergoing minimally-invasive esophagectomy via laparoscopic and thoracoscopic techniques with intra-thoracic anastomosis between January 2015 and July 2019 were identified. Post-operative weight loss at 3- and 6-months post-surgery was calculated as percent change from preoperative weight. Univariate and multivariable generalized linear method (GLM) analysis was performed. Results: 176 patients (145 male, 31 female) underwent esophagectomy, with a median age of 64 [IQR 57-71] and median 3-month postoperative weight loss of 7.9% [IQR 1.5-12.3%]. The majority of patients were Caucasian (89%), histologically diagnosed adenocarcinoma (90%), and received neoadjuvant chemoradiation (71%). Median preoperative BMI was 28.3 [IQR 25.5-32.5]. Preoperative enteral feeding tubes were placed in 86 patients (49%). Five patients (2.8%) were converted to open surgery (laparotomy) and anastomotic leaks occurred in 12 (6.8%) patients. The median length of stay was 9 days [IQR 8-13]. Among a subset of 147 patients with data available, the median 6-month postoperative weight loss was 10% [IQR 5.3-15.0%]. On multivariable analysis, greater preoperative BMI and anastomotic leak were associated with increased weight loss at 3 and 6 months post-esophagectomy. Conclusions: Significant postoperative weight loss is common among minimally-invasive esophagectomy patients. Higher preoperative BMI and postoperative anastomotic complications are significantly associated with increased postoperative weight loss at 3 and 6 months. Need for preoperative feeding tube and conversion to open surgery were further associated with postoperative weight loss. [Table: see text]

Interaction of Postoperative Morbidity and Receipt of Adjuvant Therapy on Long-Term Survival After Resection for Gastric Adenocarcinoma: Results From the U.S. Gastric Cancer Collaborative

BACKGROUND

Postoperative complications (POCs) can negatively impact survival after oncologic resection. POCs may also decrease the rate of adjuvant therapy completion. We evaluated the impact of complications on gastric cancer survival and analyzed the combined effect of complications and adjuvant therapy on survival.

METHODS

We analyzed 824 patients from 7 institutions of the U.S. Gastric Cancer Collaborative who underwent curative resection for gastric adenocarcinoma between 2000 and 2012. POC were graded using the modified Clavien-Dindo system. Survival probabilities were estimated using the method of Kaplan and Meier and analyzed using multivariate Cox regression.

RESULTS

Median follow-up was 35 months. The overall complication rate was 41 %. The 5-year overall survival (OS) and recurrence-free survival (RFS) of patients who experienced complications were 27 and 23 %, respectively, compared with 43 and 40 % in patients who did not have complications (p < 0.0001 for OS and RFS). On multivariate analysis, POC remained an independent predictor for decreased OS and RFS (HR 1.3, 95 % CI 1.1-1.6, p = 0.03 for OS; HR 1.3, 95 % CI 1.01-1.6, p = 0.03 for RFS). Patients who experienced POC were less likely to receive adjuvant therapy (OR 0.5, 95 % CI 0.3-0.7, p < 0.001). The interaction of complications and failure to receive adjuvant therapy significantly increased the hazard of death compared with patients who had neither complications nor adjuvant therapy (HR 2.3, 95 % CI 1.6-3.2, p < 0.001).

CONCLUSIONS

Postoperative complications adversely affect long-term outcomes after gastrectomy for gastric cancer. Not receiving adjuvant therapy in the face of POC portends an especially poor prognosis following gastrectomy for gastric cancer.

Gastric remnant cancer: A distinct entity or simply another proximal gastric cancer?

BACKGROUND

The purpose of this study was to compare outcomes following resection of gastric remnant (GRC) and conventional gastric cancer.

METHODS

Patients who underwent resection for gastric cancer in 8 academic institutions from 2000-2012 were evaluated to compare morbidity, mortality, and survival based on history of prior gastrectomy.

RESULTS

Of the 979 patients who underwent gastrectomy with curative-intent during the 12-year study period, 55 patients (5.8%) presented with GRC and 924 patients (94.4%) presented with conventional gastric cancer. Patients with GRC were slightly older (median 69 vs. 66 years). GRC was associated with higher rates of complication (56% vs. 41%, P = 0.028), longer operative times (301 vs. 237 min, P < 0.001), higher intraoperative blood loss (300 vs. 200 ml, P = 0.012), and greater need for blood transfusion (43% vs. 23%, P = 0.001). There were no significant differences in 30-day (3.6% vs. 4%) or 90-day mortality (9% vs. 8%) between the two groups. Overall survival rates were similar between GRC and conventional gastric cancer (5-year 20.3% vs. 38.6%, P = 0.446). Multivariate analysis revealed that history of gastrectomy was not predictive of survival while established predictors (older age, advanced T-stage, nodal involvement, blood transfusion, multivisceral resection, and any complication) were associated with poor survival.

CONCLUSIONS

Despite higher morbidity, prognosis after resection of gastric remnant cancer is similar to conventional gastric cancer.

Multivisceral Resection for Gastric Cancer: Results from the US Gastric Cancer Collaborative

BACKGROUND

Resection of an adjacent organ during gastrectomy for gastric cancer is occasionally necessary to achieve margin clearance. The short- and long-term outcomes of this approach remain unclear.

METHODS

Patients who underwent gastric cancer resection in seven U.S. academic institutions from 2000 to 2012 were evaluated to compare perioperative morbidity, mortality, and survival outcomes, stratified by the need for and type of multivisceral resection (MVR).

RESULTS

Of 835 patients undergoing curative-intent gastrectomy, 159 (19 %) had MVR. The most common adjacent organs resected were the spleen (48 %), pancreas (27 %), liver segments 2/3 (14 %), and colon (13 %). As extent of resection increased (gastrectomy only, n = 676; MVR without pancreatectomy, n = 116; and MVR with pancreatectomy, n = 43), perioperative morbidity was higher: any complication (45, 60, 59 %, p = 0.012), major complication (17, 31, 33 %, p = 0.001), anastomotic leak (5, 11, 19 %, p < 0.001), and respiratory failure (9, 15, 22 %, p = 0.012). However, perioperative mortality did not significantly increase (30-day: 3, 4, 2 %, p = 0.74; 90-day: 6, 8, 9 %, p = 0.61). Overall survival after resection decreased as extent of resection increased (5-year: 42, 28, 6 %). After controlling for age, race, T stage, N stage, grade, margin status, perineural invasion, adjuvant therapy, and blood transfusion, MVR with pancreatectomy (HR 1.67, p = 0.044), but not MVR without pancreatectomy (HR 0.97, p = 0.759), remained an independent predictor of poor survival.

CONCLUSION

In this modern, multi-institutional cohort of gastric cancer patients, multivisceral resection was associated with higher perioperative morbidity but not significantly higher perioperative mortality. If concomitant pancreatectomy is anticipated, patients should be selected with extreme caution because long-term survival remains poor.

Number of Lymph Nodes Removed and Survival after Gastric Cancer Resection: An Analysis from the US Gastric Cancer Collaborative

BACKGROUND

Examination of at least 16 lymph nodes (LNs) has been traditionally recommended during gastric adenocarcinoma resection to optimize staging, but the impact of this strategy on survival is uncertain. Because recent randomized trials have demonstrated a therapeutic benefit from extended lymphadenectomy, we sought to investigate the impact of the number of LNs removed on prognosis after gastric adenocarcinoma resection.

STUDY DESIGN

We analyzed patients who underwent gastrectomy for gastric adenocarcinoma from 2000 to 2012, at 7 US academic institutions. Patients with M1 disease or R2 resections were excluded. Disease-specific survival (DSS) was calculated using the Kaplan-Meier method and compared using log-rank and Cox regression analyses.

RESULTS

Of 742 patients, 257 (35%) had 7 to 15 LNs removed and 485 (65%) had ≥16 LNs removed. Disease-specific survival was not significantly longer after removal of ≥16 vs 7 to 15 LNs (10-year survival, 55% vs 47%, respectively; p = 0.53) for the entire cohort, but was significantly improved in the subset of patients with stage IA to IIIA (10-year survival, 74% vs 57%, respectively; p = 0.018) or N0-2 disease (72% vs 55%, respectively; p = 0.023). Similarly, for patients who were classified to more likely be "true N0-2," based on frequentist analysis incorporating both the number of positive and of total LNs removed, the hazard ratio for disease-related death (adjusted for T stage, R status, grade, receipt of neoadjuvant and adjuvant therapy, and institution) significantly decreased as the number of LNs removed increased.

CONCLUSIONS

The number of LNs removed during gastrectomy for adenocarcinoma appears itself to have prognostic implications for long-term survival.

Optimal extent of lymphadenectomy in gastric adenocarcinoma: A seven-institution study of the U.S. Gastric Cancer Collaborative

115

Background: The optimal extent of lymphadenectomy in the treatment of gastric adenocarcinoma continues to be a subject of intense debate. We aimed to compare gastrectomy outcomes following limited (D1) or extended (D2) lymphadenectomy. Methods: Using the multi-institutional U.S. Gastric Cancer Collaborative database, we reviewed the morbidity, mortality, recurrence, and overall survival (OS) of 727 patients receiving D1 or D2 lymphadenectomies. Patients with stage IV disease, prior gastrectomy, and age 85 or greater were excluded. Multivariate analyses included variables with p values less than 0.1. Results: Between 2000 and 2014, 266 (36.6%) and 461 (63.4%) patients received a D1 and D2 lymphadenectomy, respectively. ASA class, mean number of comorbidities, grade, stage, and signet ring cell subtypes were similar between groups. Neoadjuvant and adjuvant chemotherapy was more common in the D2 group (p<0.001). The mean number of lymph nodes recovered was significantly higher in patients receiving a D2 lymphadenectomy (21.5 for D2 vs. 17.1 for D1, p<0.001). Median follow up was 1.3 years. While Clavien III/IV major morbidity was similar (15.0% for D1 vs. 14.5% for D2, p=0.85), mortality was worse for those receiving a D1 lymphadenectomy (4.9% vs. 1.3%, p=0.004). Recurrence rates for patients receiving D1 and D2 lymphadenectomies were 25.8% and 27.0%, respectively (p=0.74). D2 lymphadenectomy was associated with improved median OS in stage I (4.7 years for D1 vs. not reached for D2, p=0.003) stage II (3.6 years for D1 vs. 6.3 for D2, p=0.42), and stage III patients (1.3 years for D1 vs. 2.1 for D2, p=0.01). After adjusting for significant predictors of OS which included ASA, stage, grade, neoadjuvant chemotherapy, and adjuvant radiation, D2 lymphadenectomy remained a significant predictor of improved survival when compared with D1 lymphadenectomy (HR 1.5, 95% CI 1.1-2.0, p=0.008). Conclusions: D2 lymphadenectomy is associated with improved survival that is more prominent in early stages of disease. It can be performed safely without increased risk of morbidity and perioperative mortality and should be the preferred lymphadenectomy technique for the treatment of gastric adenocarcinoma.

The prognostic value of preoperative helicobacter pylori infection in resected gastric cancer

137

Background: Limited data exist on the prognostic implication of pre-operative Helicobacter pylori (H. pylori) infection in gastric adenocarcinoma (GAC). Our aim was to assess the association of H. pyloriwith recurrence and survival in patients undergoing resection of GAC. Methods: All patients who underwent curative intent resection for GAC from 2000 to 2012 at seven academic institutions comprising the US Gastric Cancer Collaborative were included. 30-day mortalities were excluded. Survival analyses were conducted with Kaplan Meier log rank and multivariate Cox regression. Primary endpoints were recurrence-free survival (RFS) and overall survival (OS). Results: Of 965 patients, 559 met inclusion criteria and had documented pre-operative H. pylori testing. 18.6% (n=104) of patients tested positive for H. pylori pre-operatively. Data regarding treatment of H. pylori was not available. H. pylori infection was associated with younger age (62.1 vs 65.1 years; p=0.041), distal tumor location (82.7% vs 71.9%; p=0.033), and receipt of adjuvant radiation therapy (47.0% vs 34.9%; p=0.032). There were no significant differences in ASA class, margin status, Grade, PNI, LVI, or nodal metastases. The distribution of TNM stage I-III was similar between the two groups. H. pylori status was not associated with tumor recurrence. However, pre-operative H. pylori infection was associated with longer OS (84.3 mo vs 44.2 mo; p=0.008). When accounting for differences in age, tumor location, and delivery of radiation therapy, H. pylori infection persisted as a positive prognostic factor for OS (HR 0.60; CI 0.40-0.91; p = 0.016). Conclusions: Patients with and without preoperative H. pylori infection had no significant differences in adverse pathologic factors including positive margin, high grade, lymph node metastases, or advanced TNM stage. Despite similar disease presentation, pre-operative H. pylori infection was independently associated with improved overall survival. Further studies examining the interaction between H. pylori and tumor immunology and genetics are needed to better understand the relationship between H. pylori and survival in gastric cancer.

The optimal length of the proximal resection margin in patients with proximal gastric adenocarcinoma: A multi-institutional study of the U.S. Gastric Cancer Collaborative

108

Background: A 5cm margin is advocated for distal gastric cancers. The optimal length of the proximal resection margin (PM) for proximal (GEJ Siewert II and III, cardia, and fundus) gastric adenocarcinoma (GAC) is not established. Methods: Patients who underwent curative intent abdominal-approach resection for proximal GAC from 2000-2012 at 7 academic US institutions were included. Patients with positive distal margins were excluded. PM length was analyzed by 0.5cm increments and was also dichotomized at the mean and median value. Primary endpoints were local recurrence (LR), recurrence-free survival (RFS) and overall survival (OS). Results: Out of 965 patients, 211 had proximal GAC, and 162 had data available on PM length. 151 patients had negative microscopic margins with a mean value of 2.6cm and a median of 1.7cm (range 0.1-15cm). When PM length was sequentially dichotomized and analyzed at 0.5cm increments (0.5-6.5cm), a greater margin distance for each analysis was not associated with LR, RFS, or OS. Similarly, a PM distance greater than the mean or median value was not associated with LR, RFS, or OS. 11 patients had a positive PM (R1), which was associated with higher N-stage (N3: 73% vs 26%; p=0.007) and increased LR (HR6.1; p=0.009). When accounting for other adverse prognostic factors (grade, lymphovascular invasion, tumor size, T-stage, and N-stage), a positive PM was not independently associated with LR. A positive PM was also not associated with decreased RFS or OS. Conclusions: For an abdominal-approach resection of proximal gastric adenocarcinoma, the length of the proximal margin is not associated with local recurrence, recurrence-free survival, or overall survival. A positive microscopic margin is associated with advanced N-stage but is not independently associated with recurrence or survival. When performing an abdominal-approach resection of proximal gastric adenocarcinoma, a grossly negative proximal margin is sufficient. Efforts to achieve a specific margin distance, especially if it necessitates an esophagectomy, should be abandoned.

Is linitis plastica a contraindication for surgical resection? A 7-institution study of the U.S. Gastric Cancer Collaborative

118

Background: Linitis plastica (LP) describes a diffusively infiltrative gastric adenocarcinoma that portends poor prognosis. Current treatment guidelines do not differentiate between LP and non-LP cancers and it is not known if the same staging system should be applied to both situations. Methods: Using the multi-institutional U.S. Gastric Cancer Collaborative database, 869 patients with gastric adenocarcinoma who underwent resection between 2000-2012 were identified. Clinicopathologic, perioperative and survival outcomes of the 58 patients with LP were compared to the 811 patients without LP. Results: Advanced disease (stage III/IV) at presentation was more common in patients with LP compared to non-LP patients (90 vs 44%, p<0.01). Despite the fact that most LP patients underwent total gastrectomy (88% vs 57%, p<0.01), positive margins were more common in LP patients (33 vs 7%, p<0.01). There was no difference in perioperative complications (48 vs 43%, p=0.45) or mortality (7 vs 3%, p=0.12) between LP and non-LP patients. While survival correlated with stage in non-LP patients, there was no difference in median overall survival (OS) of LP patients based on stage (I/II, 17.3 mos; III, 10.6 mos; IV, 12.0 mos; p=0.46). Median OS was significantly worse in patients with LP (11.6 vs 37.8 months, p<0.01) when margin status and extent of lymphadenectomy were not factored in the analysis. However, when analyzing only patients with optimal resections (R0, D2 lymphadenectomy), the median OS for stage III LP (n=22) and non-LP (n=185) patients was nearly identical (26.8 vs 25.3 mos, p=0.69). There were no independent prognostic factors identified to predict survival in LP patients undergoing curative resection. Conclusions: The poor prognosis of LP gastric cancer is due primarily to its advanced stage at diagnosis. However, LP patients who undergo optimal resections can expect similar long term survival compared to optimally resected non-LP patients with advanced stage disease. Patient selection and multidisciplinary management are paramount when considering surgical resection in patients with gastric LP.

The prognostic value of signet ring cell histology in resected gastric cancer

128

Background: Conflicting data exist on the prognostic implication of signet ring cell (SRC) histology in gastric adenocarcinoma (GAC). Our aim was to assess the association of SRC with recurrence and survival in patients undergoing resection of GAC. Methods: All pts who underwent curative intent resection for GAC from 2000 to 2012 at 7 academic institutions comprising the US Gastric Cancer Collaborative were included. 30-day mortalities were excluded. Survival analyses included Kaplan Meier log rank and multivariate Cox regression. Primary endpoints were recurrence-free survival (RFS) and overall survival (OS). Stage-specific analysis was performed. Results: Of 965 pts, 768 met inclusion criteria. SRC was present in 39.5% and was associated with female gender (52.9% vs 38.6%; p<0.001), younger age (61 vs 67 yrs; p<0.001), poor differentiation (94.8% vs 50.3%; p<0.001), perineural invasion (PNI: 41.4% vs 23%; p<0.001), distal location (82.2% vs 70.1%; p<0.001), receipt of adjuvant therapy (63% vs 51.2%; p=0.002), and more advanced stage (Stage 3: 55.2% vs 36.5%; p<0.001). SRC was associated with earlier recurrence (56.7mo vs median not reached (MNR); p=0.009) and decreased OS (33.7mo vs 46.6mo; p=0.011). When accounting for other adverse pathologic features, PNI (HR 1.57; p=0.016) and higher TNM stage (HR 2.63; p<0.001) were associated with decreased RFS, but SRC was not. PNI (HR 1.53; p=0.006), higher TNM Stage (HR 2.10; p<0.001), greater size (HR 1.05; p=0.014), and adjuvant therapy (HR 0.50; p<0.001) were associated with OS. SRC was not an independent predictor of OS. Stage-specific analysis showed no association between SRC and RFS or OS in Stage 1 or 3. In Stage 2, SRC was associated with earlier recurrence (38.1mo vs MNR; p=0.005) but not OS. The negative association of SRC with decreased RFS persisted in multivariate analysis (HR 3.11; p=0.015). Conclusions: Signet ring histology is associated with other adverse pathologic features including higher grade and higher TNM stage but is not independently associated with reduced RFS or OS. Identification of signet ring histology during preoperative evaluation should not, in isolation, dictate treatment strategy.

An assessment of feeding jejunostomy tube placement at the time of resection for gastric adenocarcinoma: A seven-institution analysis of 837 patients from the U.S. Gastric Cancer Collaborative

120

Background: A recent single institutional study demonstrated that jejunostomy feeding tubes (J-tubes) placed during resection of gastric adenocarcinoma (GAC) are associated with increased complications and no change in receipt of adjuvant therapy. Our aim was to validate these findings in a large multi-institutional cohort. Methods: All patients who underwent resection for GAC at 7 institutions participating in the U.S. Gastric Cancer Collaborative from 2000-2012 were identified. Patients with metastatic disease were excluded. Univariate and multivariate logistic regression were performed to assess the association of J-tubes with postoperative complications and receipt of adjuvant therapy. Subset analysis of patients who underwent total vs subtotal gastrectomy was also performed. Results: Of 965 patients, 837 were included for analysis, of whom 265 (32%) received a J-tube. Patients receiving J-tubes demonstrated greater incidence of preoperative weight loss, lower BMI, greater extent of resection, and more advanced TNM stage. J-tube placement was associated with increased infectious complications (36% vs 19%;p<0.001), including surgical site infections (14% vs 6%;p<0.001) and deep intra-abdominal infections (11% vs 4%;p<0.001). On multivariate analysis, J-tubes remained independently associated with increased risk of infectious complications (HR=1.93;p=0.001), surgical site infections (HR=2.85;p=0.001), and deep intra-abdominal infections (HR=2.13;p=0.04). J-tubes were not associated with increased receipt of adjuvant therapy (HR=0.82;p=0.34). Subset analysis of patients undergoing total and subtotal gastrectomy similarly demonstrated an association of J-tubes with increased risk of infectious outcomes andno association with increased receipt of adjuvant therapy. Conclusions: J-tubes placed during resection of gastric adenocarcinoma are independently associated with increased postoperative infections and are not associated with increased receipt of adjuvant therapy, despite being placed in patients with advanced TNM stage tumors. Selective use of J-tubes is recommended.

Value of peritoneal drain placement after total gastrectomy for gastric adenocarcinoma: A multi-institutional analysis from the U.S. Gastric Cancer Collaborative

131

Background: A recent randomized trial of peritoneal drain (PD) placement after pancreaticoduodenectomy concluded that placement of PDs decreased the frequency and severity of complications. The role of PD placement after total gastrectomy for gastric adenocarcinoma (GAC) is not well-established. Methods: Patients who underwent total gastrectomy for GAC at 7 institutions from the U.S. Gastric Cancer Collaborative from 2000-2012 were identified. Univariate and multivariate analyses were performed to evaluate the association of PD placement with postoperative outcomes. Results: 344 patients were identified and anastomotic leak rate was 9%.253 (74%) patients received a PD. Those with PD placed had similar ASA class, tumor size, TNM stage, and need for additional organ resection when compared to their counterparts with no PD. No difference was observed in the rate of any complication (54% vs. 48%;p=0.45), major complication (25% vs. 24%;p=0.90), or 30-day mortality (7% vs. 4%;p=0.51) between the two groups. In addition, no difference in anastomotic leak (9% vs. 10%;p=0.90), need for secondary drainage (10% vs. 9%;p=0.92), or reoperation (13% vs. 8%;p=0.28) was identified. On multivariate analysis, PD placement was not associated with a decrease in frequency or severity of postoperative complications. Subset analysis of patients stratified by whether they underwent concomitant pancreatectomy similarly demonstrated no association of PD placement with reduced complications or mortality. In patients who experienced an anastomotic leak (n=31), placement of PD was similarly not associated with a decrease in complications, need for secondary drainage, or mortality. Conclusions: Peritoneal drain placement after total gastrectomy for adenocarcinoma, regardless of concomitant pancreatectomy, is not associated with a decrease in the frequency and severity of adverse postoperative outcomes, including anastomotic leak and mortality, or decrease in the need for secondary drainage procedures or reoperation. Routine use of peritoneal drains is not warranted.

Impact of external-beam radiation therapy on outcomes among patients with resected gastric cancer: A multi-institutional analysis

84

Background: Use of perioperative chemotherapy (CTx) alone versus chemo-radiation therapy (cXRT) in the treatment of resectable gastric cancer remains varied. We sought to define the utilization and effect of CTx alone versus cXRT on patients having undergone curative-intent resection for gastric cancer. Methods: Using the multi-institutional U.S. Gastric Cancer Collaborative database, we identified 505 patients between 2000 and 2012 with gastric cancer who received perioperative therapy in addition to curative-intent resection. The impact of perioperative therapy on survival was analyzed by the use of propensity-score matching of clinicopathologic factors among patients who received CTx alone versus cXRT. Results: Median patient age was 62 years and the majority of patients were male (58%). Surgical resection involved either partial gastrectomy (54%) or total gastrectomy (46%). On pathology, median tumor size was 5.0 cm; most patients had a T3 (37%) or T4 (36%) lesion and lymph node metastasis (74%). Margin status was R0 in most patients (89%). 211 (42%) patients received perioperative CTx alone whereas the remaining 294 (58%) patients received 5-FU based cXRT. Factors associated with receipt of cXRT were younger age (OR 0.98), T3 tumors (OR 1.52), and lymph node metastasis (OR 2.03) (all P < .05). Recurrence occurred in 214 (39%) patients. At a median follow-up of 28 months, median overall survival (OS) was 33.4 months and 5-year survival was 36.7%. Factors associated with worse OS included tumor size (HR 1.1), T-stage (HR 1.5), and lymph node metastasis (HR 1.58) (all P<0.05). In contrast, receipt of cXRT was associated with improved long-term OS (CTx alone: 21 months vs. cXRT 45 months; p<0.001). In the propensity-matched multivariate model that adjusted for tumor size, T-stage, and nodal status, cXRT remained associated with an improved long-term disease-free (HR 0.43) and overall (HR 0.41) survival (both P<0.001). Conclusions: XRT was utilized in 58% of patients undergoing curative-intent resection for gastric cancer. Using propensity-matched analysis, cXRT was an independent factor associated with improved recurrence-free and overall survival.

Factors associated with recurrence in lymph node-negative gastric adenocarcinoma: Results from the U.S. Gastric Cancer Collaborative

80

Background: Lymph node (LN) status is a predictor of recurrence after gastrectomy for gastric adenocarcinoma. Clinicopathologic predictors of recurrence in patients with node-negative disease are less well established. Methods: Patients who underwent surgery with curative intent for gastric adenocarcinoma from between 2000-2012 from participating institutions of the U.S. Gastric Cancer Collaborative were analyzed. Patients who died within 30 days of surgery were excluded. Univariate (UV) and multivariate (MV) analysis of clinicopathologic factors was associated with recurrence was performed. Results: Nine-hundred sixty-five patients from seven institutions were included in the analysis. Three-hundred forty-five (36%) had LN- disease, of whom 63 (18%) had disease recurrence after a median follow-up of 24 months. The most common patterns of recurrence were: peritoneal alone (44%), liver (22%), or combined liver/peritoneal (9%). This distribution did not differ significantly from LN+ disease. UV analysis identified tumor size, linitis plastica, diffuse histology, poor differentiation, signet ring histology, T stage ≥3, perineural invasion, and lymphvascular invasion as risk factors for recurrence (Table). On MV analysis, T stage≥3 (OR 3.6, 95% CI=1.7-7.5) and poorly differentiated histology (OR 2.4, 95% CI=1.2-4.9) were independent predictors of recurrence. Conclusions: Despite the presence of negative lymph nodes, patients with T stage ≥3 and poorly differentiated histology are at high risk of recurrence after gastrectomy for adenocarcinoma of the stomach. These factors, along with other patient and treatment-related variables, may be used to select patients who may benefit from more aggressive adjuvant therapy and to guide subsequent monitoring for disease recurrence. [Table: see text]

The effect of perioperative transfusion on recurrence and survival following gastric cancer resection: A seven-institution analysis of 765 patients from the U.S. Gastric Cancer Collaborative

100

Background: Whether perioperative blood transfusion has a negative prognostic effect on recurrence and survival in patients undergoing resection of gastric adenocarcinoma (GAC) is unknown. Methods: All patients who underwent resection for GAC from 2000-2012 at 7 institutions were identified. The effect of transfusion on recurrence-free (RFS) and overall survival (OS) in the context of adverse clinicopathologic variables was examined by univariate (UV) and multivariate (MV) regression analyses. Results: Out of 965 pts, 765 underwent curative intent, R0 resection. Median FU for survivors was 44 mos; 30-day deaths were excluded. Median estimated blood loss (EBL) was 250cc and 166 pts (22%) received perioperative RBC transfusions. 5-yr RFS was 51% in transfused and 61% in non-transfused patients (p=0.01). Median OS was decreased in patients receiving transfusions (19 vs 50 mos, p<0.001). On MV analysis, transfusion remained an independent risk factor for decreased RFS (HR 2.8; 95% CI: 1.2-6.5; p=0.01) and decreased OS (Table), regardless of EBL or need for splenectomy. Timing (intraop vs postop) and volume (# of units) did not alter the effect of transfusion on survival. Non-transfused pts were more likely to receive adjuvant therapy (56% vs 44%; p=0.01). Conclusions: Perioperative blood transfusion is associated with decreased recurrence-free and overall survival following resection of gastric cancer, independent of adverse clinicopathologic factors. This supports the judicious use of perioperative transfusion during resection of gastric cancer. [Table: see text]

The effect of postoperative morbidity on survival after resection for gastric adenocarcinoma: Results from the U.S. Gastric Cancer Collaborative

5

Background: The negative impact of postoperative complications (POCs) on survival is well documented for many cancer types, but has not been well described in gastric cancer. Here, we evaluated the effect of POCs on survival after surgery for gastric cancer in a cohort of patients from a multi-institutional database. Methods: Patients who underwent surgery with curative intent for gastric adenocarcinoma between 2000-2012 from participating institutions of the U.S. Gastric Cancer Collaborative were analyzed. Patients who died within 30 days of surgery were excluded. Ninety-day postoperative complication data were collected. Survival probabilities were estimated by Kaplan-Meier analysis and compared using the log-rank test. Results: A total of 853 patients from seven institutions met inclusion criteria. Median follow-up was 32 months. The overall complication rate was 40% (n=344). The most frequent complications were: infectious (25%, including surgical site infection [8%]), and anastomotic leak (6%). 7% of patients underwent reoperation during the same hospitalization. Five-year overall survival (OS) for patients without perioperative complications was 54%, compared with 39% for patients with POCs (p=0.001). Disease free survival (DFS) at five years was 61% for patients without POCs compared to 49% in patients with POCs (p=0.002). Patients without POCs were significantly more likely to receive adjuvant therapy (55% vs 42%; p<0.001). Conclusions: In a large, multi-institutional cohort, POCs were associated with decreased survival in patients undergoing surgery for gastric adenocarcinoma. This may be due, in part, to the negative impact of complications on the receipt of adjuvant therapy. Efforts aimed at reducing perioperative morbidity are important not only for short-term surgical outcomes, but also for enhancing long-term oncologic outcomes in patients with gastric cancer. [Table: see text]

Utility of the proximal margin frozen section for resection of gastric adenocarcinoma: A 7-institution study of the U.S. gastric cancer collaborative

103

Background: The proximal gastric margin dictates the extent of resection for gastric adenocarcinoma (GAC). The value of achieving negative margins by additional gastric resection after a positive proximal margin frozen section (FS) is unknown. Methods: The US Gastric Cancer Collaborative includes all patients who had resection of GAC at 7 institutions by oncologic surgeons from 2000-2012. Intraoperative proximal margin FS data were classified as R0 or R1 based on final permanent section (PS); positive distal margins were excluded. Primary aim was to evaluate the impact on local recurrence (LR) of converting a positive proximal margin FS to an R0 final margin by additional resection. Secondary endpoints were recurrence-free (RFS) and overall survival (OS). Results: Of 860 pts, 520 had a proximal margin FS; 67 were positive. Of these 67, 48 were converted to R0 on PS by additional resection. R0 proximal margin was achieved in 447 pts (86%), R1 in 25 (5%), and R1 converted to R0 in 48 (9%). Median FU was 44 mos. Although LR was decreased in the converted R1 to R0 group compared to the R1 group (10% vs 32%, p=0.01), when accounting for other pathologic variables on multivariate (MV) analysis, R1 to R0 conversion was not associated with decreased LR. Median RFS was similar between the R1 to R0 and R1 cohort (37 vs 31 mos; p=0.6) compared to 110 mos for the R0 group. Median OS was similar between the R1 to R0 conversion and R1 groups (36 vs 26 mos; p=0.14) compared to 50 mos for the R0 group. On MV analysis, increasing T-stage and positive lymph nodes were associated with worse OS; R1 to R0 conversion of the proximal margin was not associated with improved OS (p=0.5; Table). Conclusions: Conversion of a positive intraoperative proximal margin frozen section during gastric cancer resection does not decrease local recurrence or improve recurrence-free or overall survival. This may guide decisions regarding the extent of resection. [Table: see text]

Differential expression and prognostic value of ERCC1 and thymidylate synthase in resected gastric adenocarcinoma

38

Background: Excision repair cross complementing gene-1 (ERCC1) and thymidylate synthase (TS) are key regulatory enzymes whose expression patterns are variably associated with overall survival (OS) in several malignancies. For example, in lung cancer, high ERCC1 expression is inversely associated with OS depending on whether or not patients receive perioperative chemotherapy with surgery. The expression pattern and prognostic value of ERCC1 and TS in resected gastric adenocarcinoma (GAC) are not known. Methods: 109 patients who underwent resection of GAC between 1/00-6/11 had tissue available for analysis. Primary objective was to assess for differential expression of ERCC1 and TS using immunohistochemistry. Secondary objective was to assess for association of ERCC1 and TS expression with OS. Results: Median age was 64yrs. Median FU was 21.2mos and median OS was 28.8mos. Resected GAC exhibited differential expression of ERCC1 (23% high, n=25) and TS (43% high, n=47). ERCC1 and TS expression were not associated with OS. In a planned subset analysis, however, of patients who received chemotherapy (n=73), high ERCC1 expression was associated with decreased OS (16.7 vs. 53.8mos; p=0.03). After controlling for tumor size, margin, grade, T-stage, lymph node involvement, and presence of lymphovascular or perineural invasion, the negative prognostic value of high ERCC1 expression persisted on multivariate Cox regression analysis (HR 2.5; 95%CI: 1.03-6.0; p=0.04). By contrast, in patients who underwent resection only (n=35), high ERCC1 expression was associated with improved OS (40.4 vs. 12.7mos; p=0.10). This finding persisted on multivariate analysis (HR 0.20; 95%CI: 0.04-.86; p=0.03). Conclusions: Resected gastric adenocarcinoma exhibits differential expression of TS and ERCC1. TS expression is not associated with OS. However, similar to what is reported in lung cancer, high ERCC1 tumor expression is associated with decreased OS in patients receiving chemotherapy, but is associated with increased OS in those treated with surgery alone. ERCC1 expression has prognostic value in resected gastric cancer and further investigation is warranted.

A novel simplified approach to incorporating lymph node ratio into gastric cancer staging

24

Background: Previous investigators have reported on the value of lymph node ratio (LNR, defined as the number of positive nodes divided by the total number of nodes assessed) in gastric adenocarcinoma (GAC) staging. Given the complexity of previously proposed staging systems, it has not gained widespread acceptance. The aim of our study was to offer a novel simplified approach to incorporating LNR into gastric cancer staging. Methods: 131 patients who underwent curative intent resection with lymphadenectomy for GAC between 1/00-6/11 were identified. Clinicopathologic factors were assessed. Primary outcome was overall survival (OS). Results: Median age was 64 yrs, 51% were male. Median tumor size was 3.5 cm, 67% were poorly differentiated, 20% had perineural invasion, 31% had lymphovascular invasion, and 6% had a positive margin. Locoregional nodal metastases were present in 59% (n=77, N0: 41%, N1: 18%, N2: 22%, N3a: 14%, N3b: 5%). Median number of lymph nodes (LN) assessed was 15.5. Mean FU was 27.3 mos, median OS was 29.3 mos. Median LNR was 0.4 (.04-1). Patients with LNR ≥0.4 had decreased OS as compared to patients with LNR <0.4 (15.1 vs 41.5 mos, p<0.0001); the survival of patients with LNR <0.4 was similar to that of node negative pts (48 mos, p=0.882). On Cox regression analysis, LNR ≥0.4 was more strongly associated with decreased OS (HR 3.09, 95%CI: 1.81-5.26; p<0.0001) compared to the AJCC 7th edition N stage (HR 1.36, 95%CI: 1.11-1.68; p=0.004). In the subset of patients who were inadequately staged and had <16 nodes examined, a LNR ≥0.4 was associated with reduced survival compared to a LNR <0.4 (17.3 vs 41.5 mos, p=.04). Conclusions: Compared to the current lymph node staging system, a lymph node ratio using 0.4 as the cutoff may more accurately predict survival outcomes. It seems to be particularly useful in patients who have inadequate nodal assessment. This simplified approach to lymph node ratio may be a more valuable staging tool than the current AJCC nodal staging system for gastric cancer and needs to be validated.

Differential HER2 expression in resected gastric cancer: Is there prognostic value?

54

Background: For advanced gastric cancer, the ToGA trial established HER2 as an important therapeutic target in the 20% of patients whose tumors exhibited HER2 overexpression or gene amplification. Others have reported that HER2-positive tumors are associated with poor survival in advanced disease. The expression profile and prognostic value of HER2 in resectable gastric cancer are unknown. Methods: 111 pts underwent curative intent resection of gastric adenocarcinoma between 1/00-6/11 and had tissue available for analysis. Immunohistochemistry (IHC) for HER2 was performed on banked tumor specimens and graded by two pathologists blinded to outcomes utilizing ToGA trial criteria. An IHC score of 0+ or 1+ was regarded as negative, 3+ as positive. Fluorescence in-situ hybridization (FISH) for HER2 was performed on equivocal (2+) IHC samples. Primary outcome was differential expression, secondary outcome was overall survival (OS). Results: Median age was 64 years, 54% were male. Median tumor size was 4 cm, 7.2% had a positive margin, 67.6% were poorly differentiated, 23.4% had perineural invasion, 35.1% had lymphovascular invasion, and 61.3% had nodal metastases. 24 patients had stage I disease (21.6%), 32 stage II (28.8%), and 55 stage III (49.6%). Mean follow-up was 28.9 months, median OS was 27.2 months. HER2 expression by IHC was negative in 61 (55%), equivocal in 37 (33.3%), and positive in 13 (11.7%). Of the 37 equivocal cases, FISH was positive in 8, for a total of 21 HER2-positive cases (18.9%) and 90 HER2-negative cases (81.1%). HER2 status did not correlate with T or N stage, tumor size or location, tumor grade, or perineural or lymphovascular invasion. HER2 status was not associated with OS (p=0.36). Conclusions: Resectable gastric cancer exhibits differential expression of HER2, similar to that of advanced disease. Despite reports suggesting HER2 positive status is associated with aggressive disease and worse outcomes in the advanced setting, HER2 status is not associated with adverse pathologic factors or survival in resectable disease. Although not prognostic, the predictive value of HER2 status for response to trastuzumab in the adjuvant setting requires further investigation.