Portomesenteric venous thrombosis following major colon and rectal surgery: incidence and risk factors.

BACKGROUND: Portomesenteric venous thrombosis (PMVT) is an uncommon complication of abdominal surgery. The objective of this study was to assess PMVT risk factor profiles and patient outcomes after colorectal surgery. METHODS: A single center retrospective review of patients undergoing colorectal surgery was performed (2007-2012). PMVT was defined as thrombus within the portal, splenic, or superior mesenteric vein on computed tomography (CT). Inferior mesenteric vein thrombosis was excluded. Independent samples t test was used to compare data variables between PMVT and non-PMVT patients. Univariate and multivariate logistic regression analyses were used to assess PMVT risk factors. RESULTS: There were 1,224 patients included (mean age 62 years, male = 566). Elective bowel resection was performed for colon carcinoma (n = 302), rectal carcinoma (n = 112), ulcerative colitis (n = 125), Crohn's disease (n = 78), polyps (n = 117), and diverticulitis (n = 215). Patients undergoing gynecological resections and emergent laparotomies were included (n = 275). Thirty-six patients (3%) were diagnosed with PMVT by CT: 17/36 on initial presentation and 19/36 by expert radiologist review. Patients with PMVT were younger (53 vs. 62 years, p = 0.001) with higher BMI (30.5 vs. 26.7, p < 0.001) and thrombocytosis (464 vs. 306, p < 0.001) compared to patients without PMVT. Univariate logistic regression identified younger age (p < 0.001), obesity (p < 0.001), ulcerative colitis (p < 0.001), thrombocytosis, (p < 0.001) and proctocolectomy as significant predictors of PMVT. Stepwise multivariate logistic regression identified that obesity (p < 0.001), thrombocytosis, (p < 0.001) and restorative proctocolectomy (p = 0.001) were still significant predictors. No patients in the PMVT group suffered bowel infarction and no related mortalities occurred. Thirty-day readmission rates were higher in the PMVT group (53% vs. 17%, p < 0.01). CONCLUSION: BMI ≥ 30 kg/m(2), thrombocytosis, and restorative proctocolectomy were significant predictors of PMVT. Initial diagnostic studies showed a PMVT rate of 1.4%; however, after expert focused radiologic review, the actual rate was 3%. Thus, the diagnosis of PMVT is difficult and readmission after colorectal surgery should prompt its consideration.

Comparison of tumor markers for predicting outcomes after resection of nonfunctioning pancreatic neuroendocrine tumors.

BACKGROUND: This study compares the predictability of 5 tumor markers for distant metastasis and mortality in pancreatic neuroendocrine tumors (PNETs). METHODS: A total of 128 patients who underwent pancreatectomy for nonfunctioning PNETs between 1998 and 2011 were evaluated. Tumor specimens were stained via immunochemistry for cytoplasmic and nuclear survivin, cytokeratin 19 (CK19), c-KIT, and Ki67. Univariate and multivariate regression analyses and receiver operating characteristics curve were used to evaluate the predictive value of these markers. RESULTS: A total of 116 tumors (91%) were positive for cytoplasmic survivin, 95 (74%) for nuclear survivin, 85 (66.4%) for CK19, 3 for c-KIT, and 41 (32%) for Ki67 >3%. Twelve (9%) tumors expressed none of the markers. Survivin, CK19, and c-KIT had no substantial effect on distant metastasis or mortality. Age >55 years, grade 3 histology, distant metastasis, and Ki67 >3% were associated with mortality (P < .05). A cut-off of Ki67 >3% was the best predictor (83%) of mortality with an area under the curve of 0.85. Ki67 >3% also predicted occurrence of distant metastases with odds ratio of 9.22 and 95% confidence interval of 1.55-54.55 (P < .015). CONCLUSION: Of the 5 markers studied, only Ki67 >3% was greatly associated with distant metastasis and death. Survivin, CK19, and c-KIT had no prognostic value in nonfunctioning PNETs.

Predicting aggressive behavior in nonfunctioning pancreatic neuroendocrine tumors.

PURPOSE: The biologic potential of nonfunctioning pancreatic neuroendocrine tumors (PNETs) is highly variable and difficult to predict before resection. This study was conducted to identify clinical and pathologic factors associated with malignant behavior and death in patients diagnosed with PNETs. METHODS: We used International Classification of Diseases 9th edition codes to identify patients who underwent pancreatectomy for PNETs from 1998 to 2011 in the databases of 4 institutions. Functioning PNETs were excluded. Multivariate regression Cox proportional models were constructed to identify clinical and pathologic factors associated with distant metastasis and survival. RESULTS: The study included 128 patients-57 females and 71 males. The age (mean ± standard deviation) was 55 ± 14 years. The body mass index was 28 ± 5 kg/m(2). Eighty-nine (70%) patients presented with symptoms, and 39 (30%) had tumors discovered incidentally. The tumor size was 3.3 ± 2 cm with 56 (44%) of the tumors measuring ≤2 cm. Seventy-three (57%) patients had grade 1 histology tumors, 37 (29%) had grade 2, and 18 (14%) had grade 3. Peripancreatic lymph node involvement was present in 31 patients (24%), absent in 75 (59%), and unknown in 22 (17%). Distant metastasis occurred in 18 patients (14%). There were 12 deaths, including 1 perioperative, 8 disease related, and 3 of unknown cause. With a median follow-up of 33 months, the overall 5-year survival was 75%. Multivariate Cox regression analysis identified age >55 (hazard ratio [HR], 5.89; 95% confidence interval [CI], 1.64-20.58), grade 3 histology (HR, 6.08; 95% CI, 1.32-30.2), and distant metastasis (HR, 8.79; 95% CI, 2.67-28.9) as risk factors associated with death (P < .05). Gender, race, body mass index, clinical symptoms, lymphovascular and perineural invasion, and tumor size were not related to metastasis or survival (P > .05). Three patients with tumors ≤2 cm developed distant metastasis resulting in 2 disease-related deaths. CONCLUSION: Age >55 years, grade 3 histology, and distant metastasis predict a greater risk of death from nonfunctioning PNETs. Resection or short-term surveillance should be considered regardless of tumor size.