Published papers

Purpose: Hepatic veno-occlusive disease (HVOD) after liver transplantation (LT) is almost always a fatal complication. We assessed the outcomes of HVOD in a single institute and analyzed a literature-based cohort. Methods: We reviewed the medical records of recipients of LT performed between 1995 and 2020 at our institute and the literature on HVOD after LT. We then analyzed the clinical features based on a “pooled” cohort of cases identified in our institute and reported in the literature. Results: HVOD was diagnosed in 3 of 331 LT recipients, all of whom died in hospital, on days 164, 12, and 13, respectively. Our comprehensive review of the literature, as well as our cases, identified eight cases of HVOD that developed within 14 days after LT (early-onset type). Early-onset HVOD had a significantly worse prognosis than HVOD that developed beyond 2 weeks after LT (non-early-onset type), which was identified in 22 cases (25.0% vs. 86.1% of the 3-month graft survival rate). The most common causes of early-onset and non-early-onset types were acute cellular rejection (50%) and drug-induced disease (50%), respectively. Conclusion: Early-onset HVOD developing within 14 days after LT has a poor prognosis.

Objective: While the prevailing practice for treating chronic limb threatening ischaemia (CLTI) is revascularisation, there is a limited understanding of disease progression in patients with CLTI who do not undergo this intervention. This study compared the prognosis of CLTI patients with tissue loss who underwent revascularisation with the prognosis of those who were managed with non-revascularisation treatment, investigating the efficacy of revascularisation in improving the outcome of patients with CLTI. Methods: Utilising the Shizuoka Kokuho Database, a regionally representative longitudinal cohort in Japan, this retrospective analysis examined CLTI patients with tissue loss (ulcer and gangrene), between April 2012 and September 2020, identifying two distinct cohorts: those receiving non-revascularisation treatment and those undergoing revascularisation. The primary endpoint was amputation free survival, evaluated in both cohorts after applying propensity score matching. Results: Of the 3 160 CLTI patients with tissue loss, 559 received non-revascularisation treatment and 589 underwent revascularisation. The non-revascularisation treatment predominantly involved prostaglandin E1 injections (99.1%). After one to one propensity score matching, the revascularisation cohort (372 patients) exhibited a markedly enhanced amputation free survival rate compared with their non-revascularisation counterparts (amputation free survival rate at one year: 58.0% vs. 45.7%, p = .014). Subgroup analyses underscored the greater efficacy of revascularisation across multiple categories, including males (hazard ratio [HR] 0.59, 95% confidence interval [CI] 0.47 – 0.75), patients with diabetes mellitus (HR 0.67, 95% CI 0.50 – 0.89), and patients with severe wounds (HR 0.76, 95% CI 0.61 – 0.93). However, the analysis did not find differences between the revascularisation and non-revascularisation treatment groups among female patients (HR 1.06, 95% CI 0.78 – 1.45) and patients with non-severe wounds (HR 0.87, 95% CI 0.58 – 1.30). Conclusion: This regional Japanese cohort study revealed a potential enhancement in the amputation free survival rate post-revascularisation for CLTI patients with tissue loss compared with non-revascularisation treatment, except in female patients and patients with non-severe wounds. Additional studies are needed to further elucidate these results.

Introduction Surgical site infections (SSIs) are among the most common nosocomial infections in surgery patients. Two types of preparations, povidone-iodine and chlorhexidine-alcohol, are commonly used in preoperative antiseptic procedures worldwide. However, there are inconsistencies among international guideline recommendations concerning skin antiseptics. This trial aimed to evaluate the superiority of olanexidine, which reduced SSI rates more than povidone-iodine in our previous randomised trial, over chlorhexidine-alcohol in clean-contaminated surgery. Methods and analysis This multicentre randomised controlled clinical trial will compare two antiseptics (1.5% olanexidine and 1.0% chlorhexidine-alcohol) to prevent SSI in clean-contaminated gastrointestinal surgeries with surgical wounds. On providing consent, patients aged <18 years will be included. The primary outcome will be the postoperative 30-day overall SSI rate, while the secondary outcomes will be the postoperative 30-day superficial incisional SSI rate, deep incisional SSI rate, organ/space SSI rate, positive bacterial wound culture rate, cultured bacterial strains, rates of intervention-related toxicity and allergic events (eg, erythema, pruritus, dermatitis and other symptoms of allergy around the region disinfected by the antiseptic during surgery), rate of reoperations due to SSI, medical economic effect indicators (based on health insurance claims) and hospital duration. The Mantel-Haenszel method will be used to estimate the adjusted risk ratio and its 95% CI for the primary analysis, which will compare the treatment effects. Ethics and dissemination The protocol was approved by the Institutional Review Board of Keio University School of Medicine and subsequently by the board of each participating site. Participant recruitment began in January 2023. The final results will be published in medical journals after international peer review. Trial registration number UMIN000049712.

Background: Surgical-site antisepsis is used to prevent surgical-site infections (SSIs). Although several guidelines have indicated the efficacy of antiseptics, such as chlorhexidine, povidone-iodine, and olanexidine, in reducing the SSI rate, an optimal recommendation is still not established. The aim of this study was to evaluate the efficacy of aqueous olanexidine compared with chlorhexidine-alcohol as the optimal antiseptic for preventing SSI in clean-contaminated surgery. Methods: This multicentre randomized trial for surgical skin antisepsis in clean-contaminated gastrointestinal and hepatobiliary-pancreatic surgeries in five hospitals evaluated the efficacy of olanexidine and chlorhexidine-alcohol. The primary endpoint was 30-day SSI. Secondary outcomes included the occurrence of SSI types, intervention-related toxicity, and reoperation caused by SSI. Results: Overall, 700 patients from five institutions underwent randomization; 347 received olanexidine and 345 received chlorhexidine-alcohol in the full analysis set. The 30-day SSI rate was 12.4% (43 of 347) in the olanexidine group and 13.6% (47 of 345) in the chlorhexidine-alcohol group (adjusted risk ratio (aRR) 0.911 (95% c.i. 0.625 to 1.327); P = 0.626). No significant differences were observed between the groups regarding the secondary outcomes, including the occurrence of superficial incisional SSI, deep incisional SSI, organ/space SSI, and reoperation caused by SSI. Overall adverse effects were seen in two patients (0.58%) in the olanexidine group and in three patients (0.87%) in the chlorhexidine-alcohol group (aRR 0.663 (95% c.i. 0.111 to 3.951)). Conclusion: Olanexidine did not significantly reduce the occurrence of overall SSI compared with chlorhexidine-alcohol. Nevertheless, these findings provide valuable insights for developing novel surgical SSI management protocols. Registration number: UMIN 000049712 (University Hospital Medical Information Network Clinical Trials Registry).

Background: Juxtarenal aortic occlusion (JRAO), in which the occlusion of the aorta extends to just below the renal artery, is often treated by bypass surgery because of concerns about the risk of procedural failure and fatal embolization to abdominal organs when treated with endovascular treatment (EVT). This study assessed the outcome of EVT for JRAO compared with aorto-biiliac /aorto-bifemoral (AOB) or axillo-bifemoral (AXB) bypass. Methods: A retrospective review of an international database created by 30 centers in Asia (CHronic Abdominal Aortic Occlusion, ASian Multicenter registry) was performed for patients who underwent revascularization for chronic total occlusion of the infrarenal aorta from 2007 to 2017. Of the 436 patients, 130 with JRAO (Forty-seven AOBs, 32 AXBs, and 51 EVTs) from 25 institutions were included in this study. Results: Patients were significantly older in the AXB and EVT groups and more malnourished in the EVT group than the AOB group. EVT was attempted but failed in 1 patient. Seven patients (1 [2.1%] in the AOB group, 1 [3.1%] in the AXB group, and 5 [9.8%] in the EVT group) died during hospitalization, but most of the causes in the EVT group were not related to the revascularization procedure. No visceral embolism was observed, which had been concerned, even though protection was performed only in 2 cases of the EVT group. At the latest follow-up (median duration 3.0 years), the ankle-brachial pressure index was significantly higher in the order of AOB, EVT, and AXB. At 4 years, the estimated primary and secondary patency rates of the AOB group (87.5% and 90.3%, respectively) were significantly higher than the AXB group (66.7% and 68.6%, respectively). Conclusions: AOB remains the gold standard and should be the first choice for acceptable risk patients. For frail patients, EVT is a good option and likely preferable as a first-line treatment compared to AXB.

Purpose: To evaluate the 5-year performance of a bioresorbable vascular scaffold, the Remedy stent, for the treatment of iliac artery atherosclerotic disease. Materials and Methods: This prospective, multicenter, single-arm clinical study evaluated 97 patients (97 lesions) with symptomatic TransAtlantic Inter-Society Consensus II A/B iliac artery lesions for 5 years after stent placement. The primary efficacy end point was 12-month primary patency compared with the prespecified standard derived from historical data with metallic stents, and the primary safety end point was the occurrence of major adverse clinical events within 5 years. All angiographic and computed tomography angiographic findings were evaluated by an independent core laboratory for quantitative vessel analysis. Results: The 12-month primary patency rate was 88.6% (95% CI, 80.1%–94.4%), which was lower than the prespecified standard. No significant difference was noted between the diameter stenosis at 9 and 12 months. There were no device- or procedure-related deaths, major amputation, or distal embolization during the follow-up period. The ankle-brachial index maintained significant improvement through the 5-year period compared with that at baseline. The Kaplan–Meier estimates of freedom from target lesion revascularization (TLR), major adverse cardiovascular events, and major adverse cardiovascular and limb events were 95.8%, 91.7%, and 87.5% at 12 months and 85.4%, 72.1%, and 62.5% at 5 years, respectively. Conclusions: The 5-year follow-up outcomes, including freedom from TLR, of the Remedy stent in iliac artery lesions were satisfactory, with a good safety profile. Nevertheless, the 12-month primary patency did not meet the expected standard compared with that of contemporary metallic stents.

Objective: There is no established consensus or guidelines for wound management after revascularization for patients with chronic limb-threatening ischemia (CLTI) without severe infection. This study is designed to evaluate the clinical effect of the wound management strategy on toe wounds after revascularization for CLTI. Methods: This retrospective cohort study was performed at eight institutions affiliated with Keio University School of Medicine in Japan and included 261 patients who underwent revascularization for CLTI between April 2019 and July 2021. We identified 132 patients with toe wounds from the database who had restored in-line blood flow to the foot. Patients were divided into two groups by the timing of toe resection after revascularization, which dictated the wound management policy. Group A (62 patients) underwent early toe amputation for suspected osteomyelitis, whereas group B (70 patients) underwent watchful waiting. The primary outcome was wound healing after revascularization; the secondary outcome was major amputation. We compared outcomes between groups A and B after propensity score matching. Results: Using propensity score matching, each patient in group A (33 patients) was matched with a patient in group B (33 patients). Wound healing in matched group A was significantly better than that in matched group B (respectively: 1-year wound healing rate: 90.0% vs 68.2%, P < .001; median wound healing time: 65 days vs 258 days, P < .01). Although five major amputations were necessary in matched group B, none were required in matched group A (P = .05). The high rate of major amputations in group B was attributed to the sudden exacerbation of infection. Limb salvage rate in matched group A exceeded matched group B (100.0% vs 90.5%: 1-year limb salvage rate, P = .02). Conclusions: Early toe amputation for highly suspected osteomyelitis in patients with CLTI with toe wounds may expedite wound healing compared with watchful waiting, potentially avoiding unnecessary major amputation. Considering the wound management strategy is crucial when evaluating wound healing outcomes in patients with CLTI with revascularization.

Background: In Japan, there has never been a national analysis of pediatric deceased donor liver transplantation (pDDLT) based on donor and recipient factors. We constructed a Japanese nationwide database and assessed outcomes of pDDLT focusing on the pediatric prioritization system introduced in 2018. Methods: We collected data on pDDLTs (<18 years) performed between 1999 and 2021 from the Japan Organ Transplant Network and Japanese Liver Transplantation Society, identified risk factors for graft survival and compared the characteristics and graft survival in pDDLTs conducted before and after the introduction of the pediatric prioritization system. Results: Overall, 112 cases of pDDLT were included, with a 1-year graft survival rate of 86.6%. Four poor prognostic factors were identified: recipient intensive care unit stay, model for end-stage liver disease/pediatric end-stage liver disease score, donor cause of death, and donor total bilirubin. After the introduction of the system, allografts from pediatric donors were more reliably allocated to pediatric recipients and the annual number of pDDLTs increased. The 1-year graft survival rate improved significantly as did pDDLT conditions indicated by the risk factors. Conclusions: Under the revised allocation system, opportunities for pDDLT increased, resulting in favorable recipient and donor conditions and improved survival.

Objective: The optimal revascularization modality following complete resection of aortic graft infection (AGI) without enteric involvement remains unclear. The purpose of this investigation is to determine the revascularization approach associated with the lowest morbidity and mortality using real-world data in patients undergoing complete excision of AGI. Methods: A retrospective, multi-institutional study of AGI from 2002 to 2014 was performed using a standardized database. Baseline demographics, comorbidities, and perioperative variables were recorded. The primary outcome was infection-free survival. Descriptive statistics, Kaplan-Meier survival analysis, and univariate and multivariable analyses were performed. Results: A total of 241 patients at 34 institutions from seven countries presented with AGI during the study period (median age, 68 years; 75% male). The initial aortic procedures that resulted in AGI were 172 surgical grafts (71%), 66 endografts (27%), and three unknown (2%). Of the patients, 172 (71%) underwent complete excision of infected aortic graft material followed by in situ (in-line) bypass (ISB), including antibiotic-treated prosthetic graft (35%), autogenous femoral vein (neo-aortoiliac surgery) (24%), and cryopreserved allograft (41%). Sixty-nine patients (29%) underwent extra-anatomic bypass (EAB). Overall median Kaplan-Meier estimated survival was 5.8 years. Perioperative mortality was 16%. When stratified by ISB vs EAB, there was a significant difference in Kaplan-Meier estimated infection-free survival (2910 days; interquartile range, 391-3771 days vs 180 days; interquartile range, 27-3750 days; P < .001). There were otherwise no significant differences in presentation, comorbidities, or perioperative variables. Multivariable Cox regression showed lower infection-free survival among patients with EAB (hazard ratio [HR], 2.4; 95% confidence interval [CI], 1.6-3.6; P < .001), polymicrobial infection (HR, 2.2; 95% CI, 1.4-3.5; P = .001), methicillin-resistant Staphylococcus aureus infection (HR, 1.7; 95% CI, 1.1-2.7; P = .02), as well as the protective effect of omental/muscle flap coverage (HR, 0.59; 95% CI, 0.37-0.92; P = .02). Conclusions: After complete resection of AGI, perioperative mortality is 16% and median overall survival is 5.8 years. EAB is associated with nearly a two and one-half-fold higher reinfection/mortality compared with ISB. Omental and/or muscle flap coverage of the repair appear protective.

Aim: Liver allografts from brain-dead donors, which were declined and were eventually not transplanted due to accompanying marginal factors, have never been surveyed in Japan. We surveyed the declined allografts and discussed the graft potential focusing on various marginal factors. Methods: We collected data on brain-dead donors between 1999 and 2019 from the Japan Organ Transplant Network. We divided their liver allografts into declined (nontransplanted) and transplanted ones, and then characterized declined ones focusing on their timepoints of decline and accompanying marginal factors. For each marginal factor, we calculated the decline rate from the number of declined and transplanted allografts, and assessed the 1-year graft survival rate from transplanted allografts. Results: A total of 571 liver allografts were divided into 84 (14.7%) declined and 487 (85.3%) transplanted ones. In the declined allografts, a majority was declined after laparotomy (n = 55, 65.5%), most of which had steatosis and/or fibrosis (n = 52). Out of the moderate steatotic (without F ≥ 2 fibrosis) allografts (n = 33), 21 were declined and 12 were transplanted, leading to a 63.6% decline rate. The latter 12 achieved a 92.9% 1-year graft survival rate after transplantation. Comparison of donor background showed no significant difference between the declined and transplanted allografts. Conclusion: Pathological abnormalities of steatosis/fibrosis seem to be the most common donor factor leading to graft decline in Japan. Allografts with moderate steatosis were highly declined; however, transplanted ones achieved promising outcomes. This national survey highlights the potential utility of liver allografts with moderate steatosis.

Uterus transplantation (UTx) is now a potential option to allow women with uterine factor infertility to give birth. However, UTx is still at an experimental stage, and basic animal studies, including in nonhuman primates, are needed for the accumulation of data prior to clinical application. Considering that UTx may provide new hope to Japanese women, we launched UTx research in 2009 and have since accumulated a large archive of results in the UTx research field. Furthermore, we have carried out various activities aimed at the implementation of clinical applications of UTx in Japan while clarifying the ethical and social issues involved. Currently, the clinical application of UTx in Japan is just around the corner, and it is expected that UTx research will develop further in the future. Herein, we summarize our basic experiences using non-human primates and our activities with the goal of future clinical applications.

Introduction: Pure laparoscopy for living donor hepatectomy is gaining popularity due to its advantages. However, despite the long-standing application of laparoscopic donor left lateral sectionectomy, the dissection of the Glisson branch, portal vein, and biliary ducts, particularly those of the caudate lobe, remains insufficiently described. Although the Glissonean approach offers easy standardization for hilar dissection, clear landmarks for parenchymal transection, and reduces postoperative bile leakage, it has not been widely adopted in laparoscopic donor hepatectomy. Here, we introduce a modified Glissonean pedicle approach to address the movement restrictions in laparoscopic surgery. Surgical Technique: After liver mobilization, the Glisson of Spiegel lobe (G1L) was divided, followed by encircling the left Glissonean pedicle. A tape for the liver hanging maneuver was placed from the right edge of the left Glissonean pedicle, along the Arantius plate, to the left edge of the left hepatic vein. When the parenchymal transection was completed, the left hepatic vein was automatically taped. The left hepatic artery and left portal vein were exposed, and some branches of P1 were divided to lengthen for anastomosis. The left hepatic duct was taped by removing the left hepatic artery and left hepatic vein from the left Glissonean pedicle. The left hepatic duct was divided under intraoperative cholangiography. Next, the left hepatic artery, left portal vein, and left hepatic vein were sequentially divided, and the graft liver was retrieved. Discussion: Our Glissonean approach can help standardize donor left lateral sectionectomy, minimize the exposure of the left hepatic duct, and clarify B1 branch dissection.

Purpose: Acute liver failure is a life-threatening condition for which ABO-incompatible living donor liver transplantation (ABOi-LDLT) is sometimes the only life-saving treatment option. We reviewed a single-center experience of adult ABOi-LDLT treatment for acute liver failure (ALF). Methods: Preoperative treatment, immune indices (B cell marker, anti-donor blood-type antibody), and postoperative outcomes were compared between ALF and non-ALF groups. Results: There were 5 and 33 patients in the ALF and non-ALF groups, respectively. The ALF group received higher doses of steroids, underwent more rounds of plasma exchange (PE), and underwent transplantation for ALF with a shorter interval following preoperative rituximab (RTx) administration (median: 2 vs 13 days; P < 0.05) than the non-ALF group. Preoperatively, CD19-positive lymphocytes in the peripheral blood were sufficiently depleted in all of the non-ALF group patients, whereas they were poorly depleted in the ALF group. Postoperatively, neither group suffered anti-donor blood-type antibody titer rebound or antibody-mediated rejection. The ALF group had a comparable 5-year survival rate to the non-ALF group (80.0% vs 77.9%). Conclusions: Despite the delayed preoperative administration of RTx, the ALF group showed an uneventful immunological response and acceptable long-term survival rate. Thus, ABOi-LDLT seems a viable treatment option for ALF.

Purpose: To compare the long-term outcomes of the most widely used third-generation stent grafts, the Endurant and Excluder stent grafts, in Japanese patients using a multicenter registry. Materials and Methods: A retrospective analysis of endovascular aneurysm repairs for abdominal aortic and iliac artery aneurysms using either the Endurant or the Excluder stent grafts from January 2012 to July 2019 at 10 Japanese hospitals was performed. Results: A total of 332 and 378 repairs using the Endurant and Excluder stent grafts, respectively, were analyzed. Although the patients’ characteristics were generally similar in the two groups, the Endurant group exhibited significantly shorter (Endurant: 31.5±18.6 mm, Excluder: 37.4±21.0 mm; p<0.001), larger (Endurant: 22.4±4.2 mm, Excluder: 21.7±3.8 mm; p=0.029), and more reversed tapered (Endurant: 12.1%, Excluder: 5.8%; p=0.003) proximal necks. The incidence of instructions for use (IFU) violations was similar between the two groups (Endurant: 59.0%, Excluder: 54.5%; p=0.223). However, the Endurant group had significantly more proximal neck-related IFU violations (54.1% and 46.3%, respectively; p=0.039), more access-related IFU violations (8.1% and 4.0%, respectively; p=0.019), and fewer bilateral hypogastric artery embolizations (5.1% and 9.3%, respectively; p=0.035) compared with the Excluder group. The incidence of intraoperative (Endurant: 3.6%, Excluder: 3.7%; p=0.950) and perioperative complications (Endurant: 3.6%, Excluder: 3.4%, p=0.899) was equivalent in the two groups. However, there was a significantly higher incidence of postoperative type II endoleaks in the Excluder group (Endurant: 28%, Excluder: 46.0%, p<0.001). Aneurysm sac regression was more frequent in the Endurant group (Endurant: 40.7%, Excluder: 31.7%, p=0.013). The Endurant group also had significantly higher rates of sac increase (Endurant: 13.0%, Excluder: 7.7%, p=0.020). Kaplan-Meier curve and log-rank analyses revealed no statistical differences in late complications (p=0.868) and overall survival (p=0.926). Conclusions: There were no statistically significant differences between the Endurant and the Excluder stent grafts in terms of intraoperative, perioperative, and late complication rates; however, the anatomical characteristics of the patients were significantly different.

BACKGROUND Evidence has been published on the successful applications of the anti-tumor necrosis factor alpha antibody infliximab, such as induction therapy, salvage treatment for acute cellular rejection, and treatment for chronic ulcerative inflammation, in intestinal transplant recipients. However, the optimal protocol for the effective use of infliximab remains largely undetermined due to scarcity of available clinical data. We report a continuative application of infliximab as maintenance therapy for recurrent chronic ulcerative ileitis in a recipient of isolated intestinal transplantation (ITx). CASE SUMMARY The patient was a 11-year-old boy with intestinal motility disorder classified as a hypogenic type of intestinal dysganglionosis. The patient underwent living-donor related intestinal transplant. His immunosuppression regimen consisted of daclizumab, tacrolimus, and steroids. Although he did not show rejection while on tacrolimus monotherapy, routine screening endoscopy showed several ulcerative lesions in the distal end of the graft 2 years after the intestinal transplant. Endoscopic work up to evaluate the progression of anemia revealed stenosis with ulcerative inflammatory changes and multiple longitudinal ulcers in the graft. Since the endoscopic findings suggested ulcerative lesions in Crohn's disease, infliximab treatment was considered. Treatment with infliximab and a small dose of oral prednisolone afforded successful withdrawal of total parenteral nutrition and maintenance of a well-functioning graft without infectious complications for 5 years since the administration of the first dose of infliximab. CONCLUSION Infliximab is effective as maintenance therapy for recurrent chronic ulcerative ileitis in an isolated ITx patient.

Here, we describe a rare case of malignant lymphoma after liver transplantation for liver cirrhosis caused by human immunodeficiency virus (HIV) and hepatitis C virus (HCV) co-infection. A male patient was diagnosed with hemophilia A at 8 months of age. Since then, he had been receiving blood products, which led to HIV and HCV co‐infection. His HIV viral load was suppressed with antiretroviral therapy, and a sustained virologic response was achieved for HCV using direct-acting antivirals. However, his decompensated liver cirrhosis progressed, and deceased donor liver transplantation was performed. A post-transplant lymphoproliferative disorder (PTLD) developed 105 days after liver transplantation, with enlarged para-aortic and hilar lymph nodes, a right renal mass, and masses in the small and large intestines. Histopathological examination confirmed monomorphic PTLD (diffuse large B-cell lymphoma). Against the treatment (reduction of immunosuppression, rituximab, and chemotherapy), the response was poor, and the patient died 94 days after the outbreak of PTLD. Both transplantation and HIV infection are risk factors for lymphoproliferative diseases. To the best of our knowledge, this is one of the very few reports of PTLD in a patient with HIV/HCV co-infection.

Background: Primary infected aneurysms of the abdominal aorta and iliac arteries are potentially life-threatening. However, because of the rarity of the disease, its pathogenesis and optimal treatment strategy remain poorly defined. Methods: A nationwide retrospective cohort study investigated patients who underwent surgical treatment for a primary infected abdominal aortic and/or common iliac artery (CIA) aneurysm between 2011 and 2017 using a Japanese clinical registry. The study evaluated the relationships between preoperative factors and postoperative outcomes including 90-day and 3-year mortality, and persistent or recurrent aneurysm-related infection. Propensity score matching was used to compare survival between patients who underwent in situ prosthetic grafting and those who had endovascular aneurysm repair (EVAR). Results: Some 862 patients were included in the analysis. Preceding infection was identified in 30.2 per cent of the patients. The median duration of postoperative follow-up was 639 days. Cumulative overall survival rates at 30 days, 90 days, 1 year, 3 years and 5 years were 94.0, 89.7, 82.6, 74.9 and 68.5 per cent respectively. Age, preoperative shock and hypoalbuminaemia were independently associated with short-term and late mortality. Compared with open repair, EVAR was more closely associated with persistent or recurrent aneurysm-related infection (odds ratio 2.76, 95 per cent c.i. 1.67 to 4.58; P<0.001). Propensity score-matched analyses demonstrated no significant differences between EVAR and in situ graft replacement in terms of 3-year all-cause and aorta-related mortality rates (P=0.093 and P =0.472 respectively). Conclusion: In patients undergoing surgical intervention for primary infected abdominal aortic and CIA aneursyms, postoperative survival rates were encouraging. Eradication of infection following EVAR appeared less likely than with open repair, but survival rates were similar in matched patients between EVAR and in situ graft replacement.