Najaarsvergadering, 26 November 2010

Download Najaarsvergadering, 26 November 2010

Post on 06-Jun-2016




1 download




    Association of Surgeons from the Netherlands (ASN): Najaarsvergadering, 26 November 2010, Ede, the Netherlands

    Surgical treatment for colorectal carcinoma in the Netherlands; whichfactors are most predictive for postoperative mortality?

    N. E. Kolfschoten1, P. J. Marang van de Mheen2, G. A. Gooiker1,E. H. Eddes3, J. Kievit1,2, R. A. E. M. Tollenaar1, M Wouters1, onbehalf of the Dutch Surgical Colorectal Audit group1Leiden University Medical Centre, Department of Surgery 2Leiden UniversityMedical Centre, Department of Medical Decision Making 3Deventer Hospital,Department of Surgery

    Outcome after the resection of a colorectal carcinoma differs between hospitalsin the Netherlands. These differences are partially explained by variation incase-mix. The purpose of this study was to determine which case-mix factors arepredictive for postoperative mortality after resection of colon (CC) and rectumcarcinomas (RC).Patients that underwent a resection of a rst presentation of a colorectalcarcinoma in 2009 that were included in the Dutch Surgical Colorectal Auditwere analyzed. Case-mix factors that were predictive for postoperative mortalityin the univariate model were entered in the multivariate logistic regressionmodel.We included 4517 patients with CC and 1899 with RC. For CC, Age, ASAclassication, Charlson score, disseminated disease, a preoperative tumourcomplication and urgency of the resection were independent predictors forpostoperative mortality. For RC Age, gender, ASA class and a preoperativetumour complication were independent predictors. Previous abdominal surgery,Body mass index (BMI), surgical procedure, tumour stage, preoperativeradiotherapy and extended resection for a locally advanced tumour or metastasiswere not independently predictive.Older age, comorbidity and a preoperative tumor complication were the mostimportant case-mix factors predicting mortality after resection of a colorectalcarcinoma. Although extended resections and procedures in patients with a highBMI are considered complex procedures, they were not associated with a higherprobability for mortality. As the population is getting older, and the incidenceof colorectal carcinoma is rising, further research should focus on improvingcare for elderly patients with comorbidities, and patients with a preoperativetumour complication.

    Peri-operative care after esophageal resection in the Netherlands

    D. Faraj1, J. J. Bonenkamp2, F. J. H. vd Wildenberg 1, J. H. W. de Wilt2,C. Rosman1

    1Department of Surgery, Canisius-Wilhelmina Hospital, PO BOX 9015, 6500 GS,Nijmegen, The Netherlands 2Department of Surgery, Radboud University NijmegenMedical Center, Nijmegen, The Netherlands

    Introduction: Surgery for esophageal cancer causes signicant morbidity andmortality, but in spite of this, a national guideline for optimal peri-operativecare is lacking in TheNetherlands. The objective of this study was to investigatethe existing peri-operative care in Dutch hospitals where esophageal resectionsare performed.Methods: A questionnaire dealing with all aspects of peri-operative care wassend to all Dutch hospitals where esophageal resections are performed.

    Results: 86% (24/28) of the centers responded. Most hospitals (40%) performbetween 11 and 20 resections each year. 83% of the hospitals have a peri-operative care protocol. 54% of the hospitals use pre-operative physiotherapy.6 centers perform a minimal invasive approach for esophageal resection. After atranshiatal resection an intraabdominal drain is placed in 25% of the centers. In25% of the centers, a routine thoracic drain was placed; in 42% of the centersthis was only done when the pleura parietalis have been opened. The proximalanastomosis is drained in 50% of the centers. 96% of the centers use nasogastricdecompression of the gastric tube. Radiological contrast evaluation of thecervical anastomosis prior to oral intake is done in 63% of the centers. 75% ofthe hospitals intend to discharge patients without feeding via jejunostomy. 38%of the hospital do not have discharge guidelines.Conclusion: In the Netherlands, there is little uniformity in peri-operativecare in patients undergoing resection for esophageal cancer. In order toimprove patient outcome, evidence based peri-operative care protocols shouldbe developed and put it into practise.

    Long-term outcome of patients with locally advanced gastrointestinalstromal tumor after neoadjuvant treatment with imatinib mesylatefollowed by surgery: A Dutch multicentric perspective

    R. Tielen1, C. Verhoef3, F. van Coevorden5, H. Gelderblom7,S. Sleijfer4, H. H. Hartgrink6, J. J. Bonenkamp1,W. T. A. van der Graaf2, J. H. W. de Wilt1

    Department of 1Surgical Oncology and 2Medical Oncology, Radboud UniversityNijmegen Medical Centre, Nijmegen, The Netherlands; Department of 3Surgicaland 4Medical Oncology, Erasmus Medical Centre, Daniel Den Hoed Cancer Centre,Rotterdam, The Netherlands; Department of 5Surgical Oncology, The NetherlandsCancer Institute, Antoni van Leeuwenhoek Hospital, Amsterdam, The Netherlands;Department of 6Surgical and 7Medical Oncology, Leiden University Medical Centre,The Netherlands

    Background: Patients with locally advanced gastrointestinal stromal tumors(GISTs) have a high risk of tumor perforation, incomplete tumor resections,and often require multivisceral resections. Long-term disease-free and overallsurvival is usually impaired in this group of patients. Multimodality treatmentwith neoadjuvant imatinib followed by surgery seems to have a benecial effectin terms of improved surgical results and long-term outcome. We report on alarge cohort of locally advanced GIST patients that have been treated in fourcenters in the Netherlands specialized in the treatment of soft-tissue sarcoma.Methods: Between August 2001 and February 2010, fty-three patientsunderwent surgery for locally advanced GIST after neoadjuvant imatinibtreatment. Data of all patients were retrospectively collected from four Dutchcenters. Primary endpoints were PFS and OS.Results: After a median of 9 (range 155) months of neoadjuvant imatinibtreatment, all patients underwent surgery. No tumor perforation occurred, anda complete surgical (R0) resection was achieved in 45 patients (85%). Five-yearPFS was 65%, and 5-year OS was 87%. Eight patients had recurrent/metastaticdisease. At recent follow-up, 41 patients had no evidence of disease, 4 patientswere alive with disease, 3 patients died of disease, 3 patients had died of otherreasons, and 2 patients were lost to follow-up.Conclusions: Neoadjuvant imatinib in locally advanced GIST is feasible andenables a high rate of R0 resections without tumor rupture. The combination of

    While every effort is made by the Editorial Team and the Publishers to avoid inaccurate or misleading information appearing in BJS, data within theindividual abstracts are the responsibility of the Authors. The BJS society, the Publishers and members of the Editorial Team accept no liability whatsoeverfor the consequences of any such inaccurate or misleading data, opinion or statement.

    2011 British Journal of Surgery Society Ltd British Journal of Surgery 2011; 98(S6): 14Published by John Wiley & Sons Ltd

  • S2 ASN abstracts: Najaarsvergadering

    imatinib and surgery in patients with locally advanced GIST seems to improvePFS and OS.

    The standardized mortality ratio is not reliable for the assessmentof quality of care for patients with pancreatic cancer

    Y. R. B. M. van Gestel1, V. E. P. P. Lemmens1,2, J. W. Coebergh1,2,H. J. T. Rutten3, S. W. Nienhuijs3, I. H. J. T. de Hingh3

    1Eindhoven Cancer Registry, Comprehensive Cancer Centre South (IKZ), Eindhoven,the Netherlands 2Department of Public Health, Erasmus University Medical Centre,Rotterdam, the Netherlands 3Department of Surgery, Catharina Hospital, Eindhoven,the Netherlands

    Background: In the Netherlands, Hospital Standardized Mortality Ratios(HSMRs) will be disclosed to the general public next year. The SMR forpancreatic cancer appeared to be increased in our hospital. This is alarmingsince our hospital serves as a regional referral center for these patients toimprove quality of care. The aim of this study was to retrospectively analyze thepancreatic cancer patients who died in our hospital.Patients and Methods: During 2006 and 2008, 178 patients were admittedto our hospital with pancreatic cancer as the primary diagnosis. The observedin-hospital mortality rate was 15 patients while the predicted mortality ratewas 12 patients. The 3 excess deaths resulted in an increased SMR of 128.The medical records of the 15 patients who died during the hospital stay werereviewed.Results: In total, 84 patients underwent surgery for pancreatic cancer of whom3 patients (36%) died in the postoperative period. Furthermore, 1 cardiaccompromised patient died due to complications after endoscopic retrogradecholangiopancreatography.The remaining 11 patients were all admitted with end-stage disease and severesymptoms to receive palliative care only.Conclusions: The in-hospital mortality rate after pancreatic surgery isacceptable and in accordance with the (international) literature. The increasedSMR appeared to be mainly caused by the admission of palliative patients. Asthe SMR model does not adjust for reason of admission, use of this measure isnot appropriate to make a reliable comparison between hospitals for quality ofcare of pancreatic cancer.

    Screening for AAA in men with peripheral arterial disease is effective

    R. R. Smeets, J. A. Ten Bosch, D. E. J. G. J. Dolmans, A. G. Krasznai,R. J. Th. J. WeltenDepartment of Surgery, Atrium Medical Centre, Heerlen, The Netherlands

    Introduction: Despite of recent evidence there is currently no screeningprogram for abdominal aortic aneurysm (AAA) in the Netherlands. The presentstudy aims to identify a subgroup in which screening might be effective.Patients and Methods: Between 2004 and 2009, all men older than 55 whowere referred to the outpatient clinic of a large teaching hospital under suspicionof peripheral arterial disease (PAD)were screened for presence of an AAA.Thesepatients received an ankle-brachial index measurement and an ultrasound of theabdominal aorta and iliac arteries. Patient history, co-morbidity and risk factorswere registered.Results: Of all referred men older than 55 (n = 1061), 1032 (973%) receiveda screening ultrasound and were included in the study. In this population 112AAAs (> 3 cm) were detected (109%), 16 measured > 5 cm. Peripheral arterialdisease was present in 764 of the 1032 screened patients (741%). An AAA waspresent in 75% of patients without PAD and in 12% of patients with PAD.Except PAD (p = 0038) our study showed other risk factors, namely: increaseof age (P < 0001) a prior cerebrovascular accident (p = 0020), smoking (p =0035), cardiac history and hypertension. In our study period (median follow-up37 years) 32 patients (286%) with a proven AAA reached operation criteria.Conclusion: Screening for AAA in men older than 55 years of age withperipheral arterial disease is effective. In addition, this data suggests thatscreening for AAA in all patients, in suspicion of PAD, that have been referredto the vascular surgery outpatient clinic is effective.

    Patients with a post-Pfannenstiel pain syndrome: consider abdominalwall endometriosis

    T. Verhagen, M. J. A. Loos, L. Mulders, M. R. Scheltinga,R. M. H. RoumenDepartments of Surgery and Gynaecology/Obstetrics Maxima Medical CentreVeldhoven, the Netherlands

    Introduction:Chronic pain after Pfannenstiel incisionsmay be caused by nerveentrapment. However, pain is occasionally due to abdominal wall endometriosis(AWE). Aim of the study is to describe the clinical picture of AWE and discussresults of surgery.Materials and Methods: Women with a Pfannenstiel incision performedbetween January 2000 and August 2010 with a histological diagnosis of AWEunderwent an interview and physical examination. Pain was measured at varioustime points using VAS (0100).Results: All Pfannenstiel incisions (n = 21) were performed for cesareandeliveries. Pain had started at a median of 38 months after the initial procedure.It was moderate to severe (mean VAS = 44 [95%CI: 3255]) and appearedrelated to the menstrual cycle in three-quarters of the women. Nine reportedcyclic swelling. More than half of the population (n = 11) suffered fromfunctional impairments due to the pain (severe, n = 5). Surgical removal wasperformed some 15 months after onset of pain. All women were pain free3 months after AWE removal (VAS = 44 [95%CI: 3255] versus 5 [95%CI:19], p = 0002). However, at follow up (26 mo), six initially pain free womenexperienced renewed pain which was based on recurrent AWE in four.Conclusions: Pain in a Pfannenstiel incision several years after a cesareandelivery may be caused by abdominal wall endometriosis. AWE pain is oftencyclic and a mass may be palpated. The pain is substantial and may lead tosevere functional impairment. Surgical removal is successful on the long termbut recurrent disease is not uncommon.

    High variability between systems that predict non-sentinel lymphnode involvement in the individual sentinel node positive breastcancer patient

    I. van den Hoven, G. Kuijt, A. Voogd, R. RoumenDepartment of Surgery, Maxima Medical Centre, Veldhoven, The Netherlands.

    Introduction: Several systems have been developed to predict the risk of non-sentinel lymph node (non-SLN) involvement in the sentinel lymph node (SLN)positive breast cancer patient. Aim of the study was to compare the predictedprobabilities and scores of the available models and to evaluate their variability,within these patients.Patients and Methods: Data were collected on SLN positive breast cancerpatients with a completion axillary lymph node dissection (ALND). Predictedprobabilities and scores for non-SLN metastasis were calculated using 9predictive systems. The number of patients that were considered low-riskor with a probability of 10%, by at least one of the different systems wascalculated. For each nomogram a box plot was constructed and the inter quartilerange (IQR) calculated. Ten patients were randomly selected from the rst (lowrisk) percentile to compare the various predicted probabilities and scores forthis selected group.Results: Almost two-thirds (642%, n = 77) of patients (n = 120) wereassigned to the low-risk/low-probability group according at least one model.None of the patients carried a low risk according to all nine systems. CompletionALND demonstrated non-SLN metastasis in more than one thirds (35%, n =27) of these patients. At group level, considerable variation in distribution ofthe predicted probabilities was observed. At individual level, the variability inpredicted probabilities for 10 randomly selected patients was huge.Conclusions:There is an unacceptable high variability in individual predictedprobabilities and scores, according to the various available systems that predictnon-SLN metastasis in SLN-positive breast cancer patients.

    2011 British Journal of Surgery Society Ltd British Journal of Surgery 2011; 98(S6): 14Published by John Wiley & Sons Ltd

  • ASN abstracts: Najaarsvergadering S3

    Reduced quality of life and body image in patients with incisionalhernia: a prospective cohort study

    Gabrielle H. van Ramshorst1, M. D., Hasan H. Eker2, M. D., WimC. J. Hop3, PhD, Johannes Jeekel4, M.D., PhD, Professor of Surgery,Johan F. Lange2, M.D., PhD, Professor of Surgery1Erasmus University Medical Center, Department of Surgery, Rotterdam,Netherlands. Currently: Red Cross Hospital, Beverwijk, the Netherlands 2ErasmusUniversity Medical Center, Department of Surgery, Rotterdam, the Netherlands3Erasmus University Medical Center, Department of Biostatistics, Rotterdam, theNetherlands 4Erasmus University Medical Center, Department of Neuroscience,Rotterdam, the Netherlands

    Introduction: Incisional hernia (IH) is a frequent complication of openabdominal surgery. We investigated risk factors for IH and the impact of IH onquality of life and body image.Patients and Methods: Between 20072009, a prospective cohort studywas performed in which 967 eligible open abdominal surgery patients wereincluded. Patients were invited for clinical evaluation after 12 months and ShortForm 36 (SF-36) and body image questionnaires were compared for patientswith and without IH. Putative relevant variables were evaluated in univariateanalysis and subsequently entered in multivariate stepwise logistic regressionmodels to delineate major independent predictors of IH.Results: 374 patients were examined after a median follow-up period of16 months (range 1024 months). Seventy-ve patients had developed IH(20%); 63 (84%) were symptomatic. Complaints included bulging (n =50), pain/discomfort (n = 45) and cosmesis (n = 8). Adjusted for age andgender, patients with IH reported signicantly lower mean scores for physicalfunctioning (p = 0045), role physical (p = 0002), general health (p = 0034)and physical component summary (p = 0012). No signicant differences werefound for other SF-36 components. Adjusted for age and gender, patients withIH reported signicantly worse cosmetic scores (p = 0002), body image andtotal body image scores (both p < 0001). Major independent risk factors forIH were male gender (p = 0002), chronic pulmonary disease (p = 0025), bodymass index (p = 0006) and surgical site infection (p < 0001).Conclusions: Patients with IH have lower quality of life and body image.Major independent risk factors for IH were male gender, chronic pulmonarydisease, body mass index and surgical site infection.

    Increased levels of non-invasively measured advanced glycationend products (AGEs) in patients with abdominal aortic aneurysmand aorto-iliac occlusive disease

    K. A. Vakalopoulos1, J. J. Harlaar1, H. H. Eker1, E. B. Deerenberg1,W. W. Vrijland2, J. F. Lange1, A. C. van der Ham2

    1Department of Surgery, Erasmus MC, University Medical Centre Rotterdam.Rotterdam, the Netherlands. 2Department of Surgery, Sint Franciscus Gasthuis.Rotterdam, the Netherlands.

    Background: In patients with aneurysm of the abdominal aorta (AAA), aortictissue shows local connective tissue destruction with accumulation of AGEs atthe aneurysm site. This may be linked to a systemic connective tissue disorderresulting in an increased risk of incisional hernia in these patients. Patients withaorto-iliac occlusive disease (AIOD) show little to no relation to connectivetissue disorders. Skin autouorescence (AF) is a validated, non-invasive way tomeasure AGE levels in skin. We hypothesized that AAA patients might havehigher skin AGE levels when compared to patients with AIOD and to a healthycontrol group.Patients & Methods: Patients with asymptomatic AAA (n = 28) and patientswith AIOD (n = 20) were included preoperatively, as well as a healthy controlgroup (n = 31). Skin AGE levels were measured with a non-invasive AGEreader (DiagnOptics BV. Groningen, the Netherlands).Results: Mean skin AGE levels were signicantly higher in patients withAAA (30; SD 087) and AIOD (313; SD 056) compared to healthy controls(213; SD058) (p < 0001), also after allowing for the effect of calendar age(p < 0001). No signicant differences were found when AAA patients werecompared to AIOD patients (p = 047).Conclusions: AGE levels are increased in both AAA patients and AIODpatients, compared to healthy controls. Contrary to our hypothesis AAA patients

    did not have higher AGE levels than AIOD patients. AGE accumulation seemsa promising marker for vascular disease in general.

    Site recurrence after RFA of CRLM therapy

    M. O. Mackintosh, B. M. Zonderhuis, M. R. Meijerink,A. J. M. van Tilborg, E. F. I. Comans, S. Meijer, M. P. van den TolDepartment of Surgical Oncology, Department of Radiology, Department of NuclearMedicine, Vumc, Amsterdam, The Netherlands

    Introduction: Patients with colorectal liver metastases (CRLM), who arenot suitable for resection, can be treated by Radiofrequency ablation (RFA).Literature does mention a RFA site recurrence rate of 3660%. PositronEmission Tomography (PET) CT is most accurate to detect recurrences.Without extrahepatic disease, a solitary site recurrence can be treated by RFA,surgical resection or stereotactic body radiation therapy (SBRT).The aim of this study was to evaluate the effectiveness of these treatments ofRFA site recurrences.Patients and Methods: From July 20002010, 518 CRLM in 132consecutive patients were treated by RFA. Follow up was conducted by4-monthly FDG-PET/CT. A site recurrence is dened as a FDG avide localrecurrence in or adjacent to the treated lesion. Data on treatment, morbidity,mortality and follow-up were analyzed.Results:43 patientswith 49CRLMdeveloped a site recurrence: 95% (49/518).92% of site recurrences were detected by PET/CT within 9 months. 27 CRLM(23 patients) underwent repeat treatment (55%,(27/49)). After a re-treatment (22re-RFA, 2 resection, 3 SBRT), 17 CRLM were successfully treated. 10 CRLMdeveloped a second site-recurrence; 6 were eligible for re-retreatment with a100% success rate. A total of 23 (85%,23/27)) site recurrences were treatedsuccessfully (follow-up 29 months). There was no peri-operative mortality;morbidity was seen in 6%.Conclusions: Over 90% of CRLM were successfully treated by RFA; therate of site recurrences was 95%. 92% of site recurrences were detected byPET/CT within 9 months. 55% of site recurrences were eligible for renewedlocal treatment of which 85% was successful.

    Fistuloscopy: a promising tool for diagnosis and treatment of analfistula

    W. E. Hueting, B. B. Nieuwkamer, I. A. M. J. Broeders, E. C. J. Consten

    surgeon, Diaconessenhuis Leiden, Leiden, the Netherlands resident surgery, MeanderMedical Centre, Amersfoort, the Netherlands surgeon, Meander Medical Centre,Amersfoort surgeon, Meander Medical Centre, Amersfoort

    Background: Surgical treatment of anal stula is characterized by itsrecurrence. Unnoticed stula can lead to recurrence. Fistuloscopy could beof added-value: anal stulas trajectory is viewed live and more complete,therefore superior therapy can be applied.Patients and Methods: From January to May 2010, patients with MRI-proven recurrence of stula were included in this pilot-study. Exclusion criteriawere malignancy and inammatory bowel disease. After obtaining informedconsent, stuloscopy was performed with a rigid 3 mm scope (Storz) underregional or general anesthesia. With continuous ow of glucose-saline solution,the trajectory of the stula was documented using video-imaging and comparedwith theMRI. Low transsncterical stula were treated with stulectomy. Hightransncterical stula, intitially were treated with seton drainage, followed bymucosa advancement plasty (MAP) six weeks later.Results: Twelve patients with recurrent stula (seven male, mean age 468yrs, range 2175) underwent stuloscopy. Due to narrowness of the stula,conversion to conventional exploration was performed in two patients. In sevenpatients, the stula trajectory did not correspond with the MRI. Two patientshad no stula at all, two patients had additional stula, in one the stula wasinter- instead of transsncterical, one patient was diagnosed with submucosalstula and one stula came to a dead end. This resulted in therapeutical changesin ve patients: stulotomy/stulectomy in 3 patients and seton placement intwo more.

    2011 British Journal of Surgery Society Ltd British Journal of Surgery 2011; 98(S6): 14Published by John Wiley & Sons Ltd

  • S4 ASN abstracts: Najaarsvergadering

    Conclusions: Discrepancy in the stula trajectory between stuloscopy andMRI was found. Fistuloscopy is a promising tool in diagnosis and treatment ofanal stula.

    A simple risk score to predict the presence of NSN metastasesin breast cancer patients with a positive sentinel node

    R. F. D. van la Parra1, P. G. M. Peer2, W. K. de Roos1, M. F. Ernst3,K. Bosscha3

    1Department of Surgery, Gelderse Vallei Hospital, Ede, The Netherlands 2Departmentof Epidemiology, Biostatistics and Health Technology Assessment, Radboud UniversityNijmegen Medical Center, Nijmegen, The Netherlands 3Department of Surgery,Jeroen Bosch Hospital, s-Hertogenbosch, The Netherlands

    Introduction: Completion axillary lymph node dissection (ALND) remainsthe standard of care for patients with a positive sentinel lymph node (SLN).However, in 4060% of patients the sentinel node is the only positive node. Arisk score was developed to identify the patients individual risk for non-SLNmetastases.Materials and Methods: The risk score was developed based on data of 182breast cancer patients who underwent successful SLN biopsy and completionaxillary lymph node dissection for SLN metastases from one hospital. Based onthe predictive factors of NSN metastases, identied in a previous meta analysis,a risk score was developed using pathological tumour size ( 20 mm/> 20 mm),lymphovascular invasion (yes/no), extracapsular extension (yes/no), size of theSLN metastases ( 2 mm/> 2 mm) and number of positive SLNs (1/> 1). Therisk score was subsequently validated on an external population (n = 180). Areceiver operating characteristic (ROC) curve was drawn and the area under thecurve was calculated to assess the discriminative ability of the nomogram.Results: The area under the ROC curve was 078 (range 071085) in theoriginal population and 078 (range 070085) in the validation population.Conclusions: This simple risk score integrates 5 clinicopathological variablesto provide an individualized risk estimate of the likelihood of NSNmetastases inbreast cancer patients with a positive sentinel node. This risk score may assist inindividual decision making regarding axillary lymph node dissection in sentinelnode positive patients.

    C-reactive protein concentration is associated with prognosis inpatients suffering from peritoneal carcinomatosis of colorectal origin

    Marcel C. G. van de Poll1, Yvonne L. B. Klaver1,Valery E. P. P Lemmens2, Bart Leenders1, Simon W. Nienhuijs1,Ignace H. J. T. de Hingh1

    Background:Only a limited number of patientswith peritoneal carcinomatosis(PC) of colorectal origin benet from palliative chemotherapy. Identicationof prognostic factors may aid in patient selection. The plasma concentration ofC-reactive protein is increasingly recognized as prognostic factor in a varietyof malignancies. However, its value in PC of colorectal origin is currently

    unknown. Aim of the present study was to investigate the association of plasmaCRP concentrations with survival in patients suffering from PC of colorectalorigin who receive palliative chemotherapy.Methods: Fifty patients with colorectal PC were identied from our regionalCancer Registration. Relevant data were retrieved from their clinical records.The most discriminatory CRP concentration was identied and patients werestratied accordingly, resulting in a group with low and a group with high CRPconcentrations. Further comparisons were made between these groups.Results: A CRP concentration < 35 mg/L was associated with a betterprognosis (median survival 224 months) than a CRP concentration 35 mg/L(79 months) (p = 00002). CRP concentrations were inversely related toalbumin concentrations which could predict survival at a cut-off value of 35 g/L(median survival 72 versus 129 months, p = 001). High CRP concentrationswere related to a decreased resectability rate of the primary tumor.Conclusions: Elevated CRP plasma concentrations are associated withdecreased survival in patients with colorectal PC. This reects the importanceof inammation in cancer survival. Further research is warranted to assess theclinical applicability of the current ndings.

    The use of the COX-2 inhibitor carprofen as an analgesic yieldsa model for anastomotic leakage in the rat small intestine

    R. J. van der Vijver, C. J. H. M. van Laarhoven, B. M. de Man,R. M. L. M. Lomme, T. HendriksDepartment of Surgery, Radboud University Nijmegen Medical Centre, Nijmegen,The Netherlands

    Introduction: Dehiscence of intestinal anastomoses remains a disastrous andrelatively frequent surgical complication. This study aims to provide a longneeded animal model for leakage developing after construction of a patentintestinal anastomosis.Methods: Twenty male Wistar rats underwent resection and anasto-mosis of both ileum and colon. In 10 rats (controls) buprenorphine(002 mg/kg/12 hours) was given as an analgesic and 10 rats received theCOX-2 inhibitor carprofen (5 mg/kg/24 hours) for pain relief during 48 hourspostoperatively. The clinical condition of the animals was followed closely andall animals were sacriced at the third day after operation. The abdomen wasinspected carefully for signs of anastomotic leakage. Thereafter wound strengthwas measured.Results: All rats in the control group showed an uneventful postoperativecourse. At autopsy, no signs of anastomotic dehiscence were observed. In thecarprofen group one animal died on day 2 from ileal dehiscence. Altogether 6 outof 10 (p = 00011) ileal dehiscences were found, while all colonic anastomosesremained intact. Mean anastomotic bursting pressure in the ileum was 67 mmHg in the controls and 25 mmHg (p < 0001) in the carprofen group.Conclusions:Treatment with carprofen results in a high percentage of failuresof anastomoses in the ileum within three postoperative days. This phenomenonshould be further studied mechanistically since it implies that COX-2 inhibitionmight interfere with healing. However, it also yields a model which offers anunique possibility to test methods for controlling leakage and contain its effects.

    2011 British Journal of Surgery Society Ltd British Journal of Surgery 2011; 98(S6): 14Published by John Wiley & Sons Ltd