Prediction of bowel resection due to bowel necrosis in recent portal vein thrombosis based on CT presentations and clinical data Authors: Junyang Luo, Junwei Chen, Hua Tang, Jialin Wu, Xiaohong Wang, Zaibo Jiang, Jie Qin Published: 2025 ABSTRACT BackgroundThis study aims to identify predictors of bowel resection due to bowel necrosis in patients with recent portal vein thrombosis (PVT) based on computed tomography (CT) presentations and clinical data.MethodsThis retrospective study was conducted from January 2013 to June 2023 and included patients with noncirrhotic, nontumoral recent PVT. Patients were followed up from initial hospitalization until September 2023 or death. Presentations on CT images, laboratory tests, adverse events, and clinical outcomes were recorded. Independent predictors of bowel resection due to bowel necrosis were identified using univariate and multivariate logistic regression models.ResultsSeventy-six patients (mean age, 41.6 years ± 13.0; 57 males) were evaluated. All patients received low molecular weight heparin, fasting, and intravenous antibiotics during hospitalization, and long-term oral anticoagulation after discharge. CT imaging revealed completely occlusive main portal vein in 73.7% (56/76) of patients and partially occlusive in 26.3% (20/76). Thrombus extension into the superior mesenteric vein was present in 80.3% (61/76) of patients and into the splenic vein in 69.7% (53/76). Additionally, 44.7% (34/76) of patients were classified as high density (CT value of main portal vein ≥ 50 Hounsfield Units on unenhanced CT scans), while 55.3% (42/76) were classified as low density (CT value < 50 Hounsfield Units). Bowel resection was performed in 15.8% (12/76) of patients due to bowel necrosis. Independent predictors of bowel resection were low CT value of the portal vein thrombus and a high neutrophil count.ConclusionsIn patients with noncirrhotic, nontumoral recent PVT, a low CT value of the main portal vein thrombus and a high neutrophil count are important predictive factors for bowel resection due to bowel necrosis. Introduction Recent portal vein thrombosis (PVT), also known as acute PVT, refers to the recent development of a thrombus in the main portal vein and/or the left and right branches, potentially involving the splenic vein (SV) and superior mesenteric vein (SMV) [1]. PVT is a frequent complication in cirrhotic patients [2], occurring in approximately 5–20% of cases, while the incidence of PVT unrelated to cirrhosis or liver tumor ranges from 0.35 to 2.5 cases per 100,000 per year [3]. Patients with recent PVT might experience abdominal pain, vomiting, fever, or other symptoms [3]. Anticoagulation is the first-line treatment for recent PVT, although its effectiveness can vary depending on the extent of the thrombus and the patient’s coagulation status [1]. Other therapies, such as local thrombolysis, thrombectomy, and transjugular intrahepatic portosystemic shunt, may lead to a higher rate of portal vein recanalization, while being associated with a relatively higher risk of technical failure and bleeding complications. However, if the thrombus extends into the superior mesenteric vein (SMV), it can cause acute intestinal ischemia in 4–58% of patients and potentially result in bowel necrosis with a mortality rate of 10–45% [4]. Therefore, it is crucial to identify patients at high risk of bowel infarction at the time of diagnosis. Additionally, these patients require antibiotics and intensive monitoring to reduce the need for bowel resection. Currently, dynamic contrast-enhanced CT is recommended as the preferred diagnostic tool for recent PVT when ultrasound and clinical findings are suggestive [1]. The typical manifestation of recent PVT is the presence of a filling defect within the lumen in the venous phase. Although a prior study demonstrated that hyperdense thrombus on baseline unenhanced CT predicts better outcomes in acute ischemic stroke [5], research focusing on recent PVT remains limited. Thus, this retrospective single-center study aims to identify predictors of bowel resection due to bowel necrosis in patients with noncirrhotic, nontumoral recent PVT based on CT features and clinical data. Materials and methods Patients with recent PVT The study was reviewed and approved by the ethics committee of the Third Affiliated Hospital of Sun Yat-sen University (LL2024-296-01), and informed consent was waived. All procedures were performed in compliance with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. The records of consecutive patients diagnosed with recent PVT between January 2013 and June 2023 at our center were retrospectively reviewed. Patients were included in the study based on the following inclusion and exclusion criteria. The inclusion criterion was recent PVT diagnosed by contrast-enhanced CT. The exclusion criteria were: (a) evidence of abdominal malignancy; (b) evidence of chronic liver disease such as alcohol abuse or viral hepatitis; and (c) Budd-Chiari syndrome. A flow chart of patient selection is shown in Fig. 1. Fig. 1Flowchart shows the patient selection criteria Flowchart shows the patient selection criteria CT technique All patients presenting with abdominal pain underwent dynamic contrast-enhanced CT examinations. As our study was retrospective, the CT examinations were conducted using different scanners (Aquilion One, Toshiba, Japan; CT LightSpeed, GE Healthcare, Milwaukee, Wis, USA; Syngo.via, version VB20A; Siemens Healthineers) and varying helical CT techniques. In general, unenhanced examinations were performed with 5-mm collimation. The standard dynamic contrast-enhanced CT protocol involved the use of a total of 120–150 mL of nonionic intravenous contrast material (350 mg of iodine per milliliter) administered using a high-pressure syringe at 3 mL/sec during the study period. Additionally, venous phase CT images were acquired 60–80 s after contrast material injection with a collimation of 5 mm and a pitch of 1.0–1.4 for all patients. Image analysis The CT features of the main portal vein (MPV), SMV, SV, ascites, and spleen volume were evaluated. The CT images were analyzed by two experienced radiologists (Z.J. and J.L.), each with more than 10 years of experience in abdominal imaging. Both radiologists used the same standardized measurement technique, and all measurements were repeated twice. The site of thrombosis was identified from the venous phase of contrast-enhanced CT. Unenhanced axial CT images were analyzed to measure thrombosis density (in Hounsfield Units, HU) at three standardized locations in the MPV: the MPV bifurcation, the middle MPV segment, and the SMV-SV confluence. Unenhanced scans were displayed with a window width of 150 HU and a center level of 50 HU. The average of the measurements from the two experienced radiologists was then calculated. A mean CT value of the MPV thrombus ≥ 50 HU was defined as high density, while a CT value < 50 HU was defined as low density. The extension of portal vein thrombosis (PVT) was recorded, and the percentage of occlusion of the main portal vein was measured in accordance with the Recommended Standardized Nomenclature, using portal venous-phase CT images obtained after contrast material administration [1]. Treatment strategy and follow-up All patients were prescribed anticoagulation therapy with low molecular weight heparin, along with fasting and intravenous antibiotics. The anticoagulation protocol was determined in accordance with guidelines and consensus on the management of non-cirrhotic, non-malignant portal vein thrombosis [1, 3, 4, 6]. Interventional procedures, such as thrombectomy and local thrombolysis via percutaneous transhepatic or transjugular intrahepatic portosystemic shunt approach, were performed in some patients based on the decision of the same experienced team under fluoroscopic guidance [7, 8]. All patients with recent PVT were evaluated by a surgical team. The diagnosis of bowel necrosis, which was the primary outcome, was strictly reserved for patients who underwent laparotomy. The decision to operate was based on a comprehensive assessment including persistent severe abdominal pain and concerns about bowel necrosis. Bowel necrosis was confirmed intraoperatively by the surgeon and, in all resected cases, by pathological examination of the specimen. Long-term oral anticoagulation (warfarin targeting an international normalized ratio of 2.0–3.0, or rivaroxaban 10 mg twice daily) was prescribed for all patients upon discharge. Meanwhile, patients with thrombophilic conditions or hematological malignancies were advised to consult with a hematology specialist. Throughout the follow-up period, all patients underwent regular physical examinations, blood tests, and Doppler ultrasonography or contrast-enhanced CT scans every 3–6 months. Any adverse events related to anticoagulation or portal hypertension, as well as 30-day mortality, were documented. The follow-up period was defined as the time interval from the initial admission to the hospital for recent PVT to either death or the end of the study (September 30, 2024). Statistical analyses Quantitative variables were reported as mean ± standard deviation, and qualitative variables were presented as frequency (percentage). The Shapiro-Wilk test was employed to assess the normality of continuous data. For comparisons between qualitative data, the chi-square test or Fisher’s exact test were used, and Student’s t-test or Mann-Whitney U test were used for quantitative data. To identify independent predictors for bowel resection due to bowel necrosis, all baseline variables with a P value < 0.10 in univariate analysis were included in the multivariate analysis using a logistic regression model. Odds ratios were calculated based on simultaneous analysis of all pertinent variables, and 95% confidence intervals were calculated to assess the confidence of the relative risk estimates. All statistical tests were performed as two-sided tests, and P <.05 was considered statistically significant. Statistical analyses were conducted using R software (Version 4.2.1, http://www.r-project.org). Results Clinical characteristics The baseline clinical characteristics of the patients are summarized in Table 1. A total of 76 patients (57 males and 19 females) were included in the study, with a mean age of 41.6 years ± 13.0 (range: 18–77 years). These patients had noncirrhotic, nontumoral recent PVT. Table 1Baseline characteristics of patientsVariablePatients (n = 76)Age (years), mean ± SD41.6 ± 13.0Male sex, n (%)57 (75.0)Underlying diseases27 (36) Hematologic or prothrombotic diseases, n (%)14 (18.4) Postnatal, n (%)2 (2.6) Nephrotic syndrome, n (%)4 (5.3) Pancreatitis or cholangitis, n (%)7 (9.2)Melena, n (%)25 (38.9)Diabetes, n (%)3 (3.9)Cerebral venous sinus thrombosis, n (%)3 (3.9)Deep venous thrombosis, n (%)11 (11.5)D-Dimer, median (IQR)6.8 (3.1–16.5)Fibrinogen, median (IQR)4.6 (3.5–5.7)INR, median (IQR)1.17 (1.10–1.29)Hemoglobin (g/L), median (IQR)132 (110.5- 148.5)Leukocyte count(×109/L), median (IQR)9.74 (6.42–14.69)Neutrophil count(×109/L), median (IQR)6.47 (3.86–12.49)Lymphocyte count(×109/L), median (IQR)1.25 (0.98–1.77)Platelet count(×109/L), median (IQR)226.5 (165- 309.75)Serum albumin (g/L), median (IQR)36.8 (30.6–39.8)Serum creatinine (µmol/L), median (IQR)71.5 (58.1–82.3)Serum sodium (mmol/L), median (IQR)139.5 (136–141)Data presented as number of patients (percentage), mean ± SD, or median (IQR) where appropriate Table 1Baseline characteristics of patientsVariablePatients (n = 76)Age (years), mean ± SD41.6 ± 13.0Male sex, n (%)57 (75.0)Underlying diseases27 (36) Hematologic or prothrombotic diseases, n (%)14 (18.4) Postnatal, n (%)2 (2.6) Nephrotic syndrome, n (%)4 (5.3) Pancreatitis or cholangitis, n (%)7 (9.2)Melena, n (%)25 (38.9)Diabetes, n (%)3 (3.9)Cerebral venous sinus thrombosis, n (%)3 (3.9)Deep venous thrombosis, n (%)11 (11.5)D-Dimer, median (IQR)6.8 (3.1–16.5)Fibrinogen, median (IQR)4.6 (3.5–5.7)INR, median (IQR)1.17 (1.10–1.29)Hemoglobin (g/L), median (IQR)132 (110.5- 148.5)Leukocyte count(×109/L), median (IQR)9.74 (6.42–14.69)Neutrophil count(×109/L), median (IQR)6.47 (3.86–12.49)Lymphocyte count(×109/L), median (IQR)1.25 (0.98–1.77)Platelet count(×109/L), median (IQR)226.5 (165- 309.75)Serum albumin (g/L), median (IQR)36.8 (30.6–39.8)Serum creatinine (µmol/L), median (IQR)71.5 (58.1–82.3)Serum sodium (mmol/L), median (IQR)139.5 (136–141)Data presented as number of patients (percentage), mean ± SD, or median (IQR) where appropriate Baseline characteristics of patients Data presented as number of patients (percentage), mean ± SD, or median (IQR) where appropriate All 76 patients presented with abdominal pain, and 32.9% (25/76) of them also suffered from melena. Furthermore, 56.6% (43/76) of the patients were diagnosed with a high risk of thrombosis, which included hematologic or prothrombotic diseases, pancreatitis, cholangitis, postnatal state, and nephrotic syndrome. Three patients (3.9%) had a history of cerebral venous sinus thrombosis, while 11 patients (14.5%) had a history of deep venous thrombosis. Additionally, only three patients (3.9%) were diagnosed with diabetes. CT features (Table 2) (Figs. 2 and 3) Table 2CT featuresVariablePatients (n = 76)Percent occlusion of main portal vein, n (%) Completely occlusive, n (%)56 (73.7) Partially occlusive, n (%)20 (26.3)Intrahepatic portal vein involved, n (%)62 (81.6)Splenic vein involved, n (%)61 (80.3)Superior mesenteric vein involved, n (%)53 (69.7)CT value, Hounsfield Units, median (IQR)47.5 (35- 62.25) Hounsfield Units ≥ 50, n (%)34 (44.7) Hounsfield Units < 50, n (%)42 (55.3)Cavernous transformation, n (%)23 (30.3)Data presented as number of patients (percentage), or median (IQR) where appropriate Table 2CT featuresVariablePatients (n = 76)Percent occlusion of main portal vein, n (%) Completely occlusive, n (%)56 (73.7) Partially occlusive, n (%)20 (26.3)Intrahepatic portal vein involved, n (%)62 (81.6)Splenic vein involved, n (%)61 (80.3)Superior mesenteric vein involved, n (%)53 (69.7)CT value, Hounsfield Units, median (IQR)47.5 (35- 62.25) Hounsfield Units ≥ 50, n (%)34 (44.7) Hounsfield Units < 50, n (%)42 (55.3)Cavernous transformation, n (%)23 (30.3)Data presented as number of patients (percentage), or median (IQR) where appropriate CT features Data presented as number of patients (percentage), or median (IQR) where appropriate Fig. 2A 33-year old man with polycythemia vera and acute abdominal pain for 4 days, received percutaneous transhepatic portal vein thrombectomy and thrombolysis and long-term oral anticoagulation without bowel resection. He remains asymptomatic after discharged. A unenhanced CT shows high density of the main portal vein (black arrow). B and C portal venous-phase shows completely occlusive main portal vein (black arrow) and mild cavernous transformation of portal vein (white arrow). D portal venous-phase on the 26-month follow-up CT shows enlarged cavernous transformation of portal vein (white arrow) without patent main portal vein. CT = computed tomography A 33-year old man with polycythemia vera and acute abdominal pain for 4 days, received percutaneous transhepatic portal vein thrombectomy and thrombolysis and long-term oral anticoagulation without bowel resection. He remains asymptomatic after discharged. A unenhanced CT shows high density of the main portal vein (black arrow). B and C portal venous-phase shows completely occlusive main portal vein (black arrow) and mild cavernous transformation of portal vein (white arrow). D portal venous-phase on the 26-month follow-up CT shows enlarged cavernous transformation of portal vein (white arrow) without patent main portal vein. CT = computed tomography Fig. 3A 28-year old man with acute abdominal pain for 10 days, received bowel resection due to bowel necrosis and then long-term oral anticoagulation. He remains asymptomatic after discharged and the underlying disease remains cryptogenic. A unenhanced CT shows relative low density of the main portal vein (black arrow) and mild ascites (*). B portal venous-phase shows completely occlusive main portal vein (black arrow) and mild cavernous transformation of portal vein (white arrow). C portal venous-phase on the 3-month follow-up CT shows enlarged cavernous transformation of portal vein (white arrow) without patent main portal vein. D Small bowel were resected. CT = computed tomography A 28-year old man with acute abdominal pain for 10 days, received bowel resection due to bowel necrosis and then long-term oral anticoagulation. He remains asymptomatic after discharged and the underlying disease remains cryptogenic. A unenhanced CT shows relative low density of the main portal vein (black arrow) and mild ascites (*). B portal venous-phase shows completely occlusive main portal vein (black arrow) and mild cavernous transformation of portal vein (white arrow). C portal venous-phase on the 3-month follow-up CT shows enlarged cavernous transformation of portal vein (white arrow) without patent main portal vein. D Small bowel were resected. CT = computed tomography The median time from initial symptoms to receiving the first CT examination was 7 days, with a range of 1–21 days. On unenhanced CT scans, quantitative HU measurements classified PVT as high density in 44.7% (34/76) of cases and low density in 55.3% (42/76) of cases. Based on the CT images, completely occlusive MPV was observed in 73.7% (56/76) of patients, while partially occlusive MPV was observed in 26.3% (20/76) of patients. Thrombus extension into the SMV was present in 80.3% (61/76) of patients and into the SV in 69.7% (53/76). Additionally, minimal cavernous transformation was found in 30.3% (23/76) of patients, and 48.7% (37/76) had ascites. Treatment specifics and clinical outcomes (Table 3) Table 3Treatment specifics and clinical outcomesVariablePatients (n = 76)Thrombectomy, n (%)25 (32.9)Thrombolysis, n (%)34 (44.7)Transjugular intrahepatic portosystemic shunt, n (%)15 (19.7)Interventioal procedure-related adverse events, n (%)1 (1.3)Procedure related or 30-day mortality, n (%)0Duration of follow-up (months), median (range)55 (12–156)Major bleeding related to anticoagulation0Bowel resection due to bowel necrosis, n (%)12 (15.8) During the initial hospitalization, n (%)10 (13.2) Delayed resection within 6 months, n (%) 2 (2.6) Extension of thrombosis into the SMV, n (%)11 (14.5)Mortality, n (%)1 (1) Pulmonary embolism, n (%)1 (1)Data presented as number of patients (percentage), mean ± SD, or median (range) where appropriate Table 3Treatment specifics and clinical outcomesVariablePatients (n = 76)Thrombectomy, n (%)25 (32.9)Thrombolysis, n (%)34 (44.7)Transjugular intrahepatic portosystemic shunt, n (%)15 (19.7)Interventioal procedure-related adverse events, n (%)1 (1.3)Procedure related or 30-day mortality, n (%)0Duration of follow-up (months), median (range)55 (12–156)Major bleeding related to anticoagulation0Bowel resection due to bowel necrosis, n (%)12 (15.8) During the initial hospitalization, n (%)10 (13.2) Delayed resection within 6 months, n (%) 2 (2.6) Extension of thrombosis into the SMV, n (%)11 (14.5)Mortality, n (%)1 (1) Pulmonary embolism, n (%)1 (1)Data presented as number of patients (percentage), mean ± SD, or median (range) where appropriate Treatment specifics and clinical outcomes Data presented as number of patients (percentage), mean ± SD, or median (range) where appropriate Among these patients, 32.9% (25/76) received thrombectomy, with 19.7% (15/76) via the transjugular intrahepatic portosystemic shunt approach and 13.2% (10/76) via the transhepatic approach. In addition to thrombectomy, thrombolysis was performed in 44.7% (34/76) of patients. Local thrombolysis via the transarterial approach alone was performed in 13.2% (10/76) of patients, and catheter-directed thrombolysis in the portal vein was performed in 31.6% (24/76) of patients. An interventional procedure-related adverse event developed in one patient, who experienced hematemesis after receiving thrombolysis therapy and recovered after medical treatment without requiring a blood transfusion. Bowel resection was performed in 15.8% (12/76) of patients due to bowel necrosis. Among these patients, 10 underwent bowel resection during their initial hospitalization because their severe abdominal pain did not improve despite anticoagulation, bowel rest, medical treatment, or interventional procedures. However, 2 patients required delayed bowel resection upon readmission within 6 months due to complaints of chronic abdominal pain and recurrent vomiting. Importantly, 4 of the 12 resection patients (33.3%) had previously undergone interventional treatments before surgery. The diagnosis of transmural bowel necrosis was confirmed by postoperative pathological examination of the resected specimen. There were no cases in which surgery was performed but pathology failed to confirm necrosis. Among these 12 patients, extension of thrombosis into the SMV was present in 11 (91.7%). The remaining 64 patients (84.2%) who were evaluated by surgeons did not require surgery, as their symptoms improved with medical and/or interventional treatments. The 76 patients had a median follow-up time of 55 months (range: 12–132 months). Long-term (at least 2 years) oral warfarin or rivaroxaban were prescribed for 30.3% (23/76) and 69.7% (53/76) of the patients, respectively. Additionally, no major bleeding related to anticoagulation was observed in this study. One patient experienced recurrent variceal bleeding and was treated with endoscopic band ligation and medication, despite initially receiving a transjugular intrahepatic portosystemic shunt as a treatment. Shunt dysfunction occurred 15 months after the procedure, and shunt revision was unsuccessful. Another patient died of pulmonary embolism 44 months after being diagnosed with PVT. The remaining 72 patients remained asymptomatic, without experiencing abdominal pain or variceal bleeding. Routine endoscopic surveillance for varices was not performed. Univariate and multivariate analysis (Table 4) Table 4Factors associated with bowel resectionVariableUnivariate analysisMultivariate analysisOdds Ratio†P ValueOdds Ratio†P ValueCT value < 50 Hounsfield Units5.00 (1.01, 24.65)0.04838.05 (1.24, 1168.16)0.037Ascites (yes vs. no)16.08 (1.95, 132.21)0.010……Hemoglobin (per one g/L increase)0.97 (0.94, 0.99)0.005……Neutrophil (per one ×109/L increase)1.08 (1.00, 1.17)0.0371.23 (1.01, 1.49)0.039Serum sodium (per one mmol/L increase)1.20 (1.01, 1.40)0.033……Serum albumin (per one µmol/L increase)1.14 (1.03, 1.26)0.008……D-Dimer (per one ug/ml increase)1.10 (1.01, 1.21)0.031Only variables with a P value < 0.1 in the univariate analysis are shown. †Data in parentheses are 95% CIs Table 4Factors associated with bowel resectionVariableUnivariate analysisMultivariate analysisOdds Ratio†P ValueOdds Ratio†P ValueCT value < 50 Hounsfield Units5.00 (1.01, 24.65)0.04838.05 (1.24, 1168.16)0.037Ascites (yes vs. no)16.08 (1.95, 132.21)0.010……Hemoglobin (per one g/L increase)0.97 (0.94, 0.99)0.005……Neutrophil (per one ×109/L increase)1.08 (1.00, 1.17)0.0371.23 (1.01, 1.49)0.039Serum sodium (per one mmol/L increase)1.20 (1.01, 1.40)0.033……Serum albumin (per one µmol/L increase)1.14 (1.03, 1.26)0.008……D-Dimer (per one ug/ml increase)1.10 (1.01, 1.21)0.031Only variables with a P value < 0.1 in the univariate analysis are shown. †Data in parentheses are 95% CIs Factors associated with bowel resection Only variables with a P value < 0.1 in the univariate analysis are shown. †Data in parentheses are 95% CIs Among all radiological and clinical factors, seven predictors with a P value less than 0.10—including two CT features (ascites and CT value of MPV) and five clinical variables (hemoglobin, neutrophil count, D-dimer, serum sodium, and serum albumin)—were identified by univariate analysis. In the multivariable analysis, the CT value of MPV (odds ratio, 38.05; 95% confidence interval: 1.24, 1168.16) and neutrophil count (odds ratio, 1.23; 95% confidence interval: 1.01, 1.49) remained independent predictors of bowel resection. Discussion This retrospective single-center study focused on the clinical outcomes in patients with recent PVT and the predictors correlated with bowel resection due to bowel necrosis. These findings demonstrated that physicians could effectively stratify patients into subgroups with a high or low risk of bowel necrosis. Diagnosis of recent PVT can be sometimes challenging due to nonspecific symptoms and the possibility of mistaking it for other gastrointestinal disorders. Contrast-enhanced CT scans are valuable in accurately diagnosing recent PVT and assessing the extent of thrombosis [9]. These scans provide detailed imaging of the portal vein and surrounding structures, allowing for the detection of thrombosis and any associated complications. To the best of our knowledge, there is currently no CT analysis available for predicting bowel necrosis in recent PVT. Our study also revealed that a higher density of PVT on initial unenhanced CT scans is associated with a lower rate of bowel resection, likely due to the fact that older thrombus has a lower HU and patients experience longer periods of intestinal ischemia. This finding can be useful in guiding treatment decisions and assessing the response to therapy. Our study utilized an unadjusted HU threshold of 50 to differentiate high- and low-density thrombus. This simplified approach offers greater clinical practicality by providing a rapid, simple tool for risk stratification at the time of initial CT scanning without requiring additional laboratory standardization. Prior studies in cerebral venous thrombosis have established different methods for assessing thrombus density, such as a quantitative 62-HU threshold standardized by hematocrit [10] and a purely qualitative visual approach [5]. In contrast, our use of a 50-HU threshold provides a more conservative and clinically expedient benchmark. This threshold, which is also commonly referenced in other acute vascular conditions [5], was identified in our comparatively larger cohort and enhances utility in emergency clinical decision-making for acute abdominal pain. On the other hand, a higher neutrophil count, indicating more severe systemic inflammation or septic thrombophilia, could also be another predictor of bowel necrosis. Firstly, infections could cause inflammation and promote the production of pro-inflammatory proteins or substances, which can damage blood vessel walls and trigger the formation of a thrombus [11]. Additionally, septic thrombophilic conditions often involve abnormalities in the coagulation system, and these abnormalities can lead to a hypercoagulable state, worsening the severity of PVT and intestinal ischemia, ultimately leading to bowel necrosis. The incidence of bowel resection due to bowel necrosis in our study may be higher than that reported in previous studies. This could be attributed to the fact that all patients in our study have non-cirrhotic PVT and do not have chronic liver diseases, which can be caused by various factors such as genetic conditions, prothrombotic disorders, abdominal inflammation, and other reasons [12]. Despite being prescribed oral anticoagulation, one patient in this study died of pulmonary embolism. Therefore, it is important for physicians to pay closer attention to the underlying risk factors of the patients. Additionally, direct oral anticoagulants have been recommended for both deep venous thrombosis and non-cirrhotic PVT [13, 14]. However, the long-term use of anticoagulants (such as low molecular weight heparin, oral warfarin, and direct oral anticoagulants) in PVT should be individualized based on the patients’ underlying diseases and in consultation with a hematologist. Although our study did not show a significant difference in the rate of bowel resection between patients who underwent interventional treatments and those who did not, the use of such treatments, such as local thrombolysis, thrombectomy, or transjugular intrahepatic portosystemic shunt for recent PVT, has become increasingly popular in recent years [15, 16]. These interventions can facilitate recanalization of the occluded portal vein and mitigate the risk of chronic portal occlusion and portal hypertension. Nevertheless, interventional treatments remain technically challenging in cases of complete PVT occlusion, and abdominal hemorrhage represents a serious potential adverse event, especially among patients undergoing thrombolysis. Therefore, these treatments should only be performed in experienced centers. Our data reflect a pragmatic step-up approach to managing these complex cases. The observation that one-third of patients who ultimately required resection had previously undergone interventional treatments underscores that surgery is generally considered a last resort. The decision to operate was likely driven by the emergence of clinical signs of peritonitis or lack of clinical improvement, indicating progression to irreversible necrosis [6, 16]. This treatment sequence is clinically informative, suggesting that failure of minimally invasive salvage therapy serves as a critical indicator of compromised bowel viability. This study has several limitations. Firstly, its retrospective and single-center design introduces an inherent risk of selection and information bias. Secondly, the small number of patients who underwent bowel resection limits the statistical power of the multivariate model, as evidenced by the wide confidence intervals, underscoring the need for cautious interpretation and validation in larger cohorts. Thirdly, while we focused on the practical CT feature of thrombus density, we did not comprehensively evaluate specific CT signs of intestinal ischemia, such as bowel wall thickening, pneumatosis, decreased contrast enhancement, or portal venous gas [17], which could provide additional diagnostic information. Fourth, laboratory markers of ischemia and inflammation, such as lactate, C-reactive protein, and procalcitonin, were not routinely available in our cohort and could improve prognostic assessment. Fifth, the heterogeneity of CT scanners and protocols may have introduced variability in absolute HU measurements, despite our use of standardized techniques to minimize this effect. Finally, some patients were diagnosed with cryptogenic PVT, indicating that not all underlying prothrombotic conditions were fully elucidated; more extensive hematological workup would be beneficial in future studies. Nevertheless, our study provides valuable long-term follow-up data, which adds considerable clinical relevance, and highlights the need for further multi-center prospective studies to validate and refine our predictive model. In conclusion, our study determined that low CT value and high neutrophil count might be associated with bowel resection due to recent PVT and bowel necrosis. These findings can help physicians to identify patients with high risk, and the choice of treatment modality will primarily depend on risk stratification in future practice.