Single centre observational study to evaluate the safety and efficacy of the Proceed TM Ventral Patch to repair small ventral hernias

Transcription

1 Hernia (2014) 18: DOI /s ORIGINAL ARTICLE Single centre observational study to evaluate the safety and efficacy of the Proceed TM Ventral Patch to repair small ventral hernias J. Bontinck I. Kyle-Leinhase P. Pletinckx V. Vergucht R. Beckers F. Muysoms Received: 12 April 2013 / Accepted: 14 July 2013 / Published online: 24 July 2013 Ó Springer-Verlag France 2013 Abstract Background There is evidence that mesh repair for primary umbilical hernias results in less recurrences and similar wound complication rates compared to tissue repair. In recent years, several mesh devices for the repair of small ventral hernias have been developed, but some reports have been published on serious complications and adverse effects encountered with those mesh devices. Methods The Proceed TM Ventral Patch (PVP TM ) is a partially absorbable lightweight polypropylene mesh. We introduced PVP TM in our department in April 2009 and collected patient data and outcome in an observational study of 101 consecutive patients until December 2011 (Clinical.Trials.gov: NCT ). In addition to the routine control 3 weeks postoperative, prospective followup included a questionnaire, clinical investigation and ultrasound after 12 months. Results The study included 91 primary (76 umbilical/15 epigastric) and 10 incisional ventral hernias (including 6 trocar hernias). In all patients a PVP TM with a diameter of 6.4 cm was used. Wound problems were the most frequent complication (n = 18). Follow-up of at least 12 months was achieved in 98 patients (97 %) and the mean follow-up time was 15.9 months. Follow-up by clinical examination diagnosed a recurrence in 11/92 patients (12.0 %). Only four patients were aware of their recurrent hernia, the seven J. Bontinck I. Kyle-Leinhase P. Pletinckx V. Vergucht F. Muysoms (&) Department of Surgery, AZ Maria Middelares, Kortrijksesteenweg 1026, Ghent 9000, Belgium filip.muysoms@azmmsj.be R. Beckers Department of Radiology, AZ Maria Middelares, Kortrijksesteenweg 1026, Ghent 9000, Belgium others reported no problems in the questionnaire. The additional ultrasound performed did not reveal recurrences that were not already diagnosed by clinical examination. In five patients a reoperation for repair of the recurrence was performed (reoperation rate 5/98 = 5.1 %). Hernia defect size (p = 0.032) and type of hernia (p = 0.029) were found to be a significant risk factors for development of a recurrent hernia (Fisher s exact test). Hernia size was a significant risk factor both in a univariate (p = 0.005) and in a multivariate Cox model (p = 0.017). Incisional hernia was of borderline significance in a univariate (p = 0.047) and in a multivariate Cox model (p = 0.08). Conclusion Intensive clinical follow-up yields a high percentage (12.0 %) of clinically proven, but often asymptomatic recurrences after repair of small ventral hernias with the PVP TM. Reoperation rate for recurrence was 5.1 %. Hernia defect size is a significant risk factor for recurrence. Therefore, we recommend using the PVP TM only for primary ventral hernias smaller than 2 cm. For larger or incisional hernias other techniques allowing the use of larger meshes is advocated. Keywords Umbilical hernias Epigastric hernias Ventral hernias Mesh devices Complications Recurrences Introduction Background Although some studies report favorable outcome after a non-mesh repair of umbilical hernias smaller than 2 cm [1, 2], there is evidence that mesh repair for primary umbilical hernias results in reduced recurrence and similar wound complication rates compared to tissue repair [3, 4]. The

2 672 Hernia (2014) 18: largest population based study on the treatment of ventral hernias is from the Danish Ventral Hernia database [5]. The cumulative risk for reoperation after repair of small (\2 cm) primary umbilical and epigastric hernias was found to be significantly lower for mesh repair compared to suture repair (Kaplan Meier analysis; 2.2 vs. 5.6 %; p = 0.001) [6]. In another study the authors showed that the cumulative reoperation rate underestimated the clinical recurrence rate by a fourfold (Kaplan Meier analysis; 4 vs. 15 %; p \ 0.001) [7]. Moreover, a high readmission rate for wound complications was seen after elective primary ventral hernia repair [8] and many patients complained about pain and discomfort after 3 years [9]. This illustrates that although umbilical hernia repair is considered a minor surgical procedure, its outcome is often not satisfactory. In recent years, several mesh devices for the repair of small ventral hernias have been developed, including the Proceed TM Ventral Patch (PVP TM, Ethicon Inc, Johnson & Johnson, Sommerville, NJ, USA) [10 18]. The technique proved user friendly and initial patient satisfaction was high. Nevertheless, published patient series on PVP TM report only small cohorts and short follow-up periods [12, 18]. Previously, we published a word of caution after encountering some serious complications and adverse effects with those mesh devices [15]. Berrevoet et al. [16] performed a systematic laparoscopic control and found that the intra-peritoneal deployment of the mesh device was often unsatisfactory, and that the recurrence rate was higher than open retro-muscular mesh repair in hernias smaller than 3 cm [17]. Objectives In line with the conclusions of our previous publication [15] we started a registry of all patients treated in our department for small ventral hernias with the PVP TM.In this observational study we wanted to investigate its safety, by monitoring serious adverse events, and efficacy at 12 months follow-up. Primary endpoint of the study was freedom of recurrence proven by clinical examination at a minimum follow-up of 12 months. Methods This study was written in accordance with the STROBE statement: Strengthening the reporting of observational studies in epidemiology [19]. Study design All patients treated with a PVP TM in our surgical department were included in the cohort, regardless of the indication. The patients were enrolled after the operation and were prospectively evaluated after a minimum followup of 12 months. Demographic and indication data, and short-term complications were documented in the medical charts and operative reports. All patients were invited for a clinical follow-up at 12 months after surgery. During the control visit patients were first questioned on the presence of chronic pain, foreign body feeling, wound problems and other complications. Subsequently, a clinical examination was performed with the patient standing and performing Valsalva maneuver. Then patients were invited to undergo an ultrasound by a radiologist, who was unaware of the findings during the clinical examination. Patients that had not shown up for their 12 months control visit were contacted by phone between November 2012 and January Those not willing to come for a clinical control were interviewed by phone on the occurrence of reoperation, chronic pain or other complications. Patients seen at our department after their 12 months visit and before January 31st 2013 for another reason were also included in the follow-up data if a clinical examination of the hernia repair site was performed. Setting and participants The study was conducted at the surgical department of the AZ Maria Middelares Hospital in Ghent, Belgium. Patients were operated by one of three abdominal surgeons. Postoperative ultrasound at 12 months was performed at the department of radiology. Prior to the start of the study a surgeon made the radiologists acquainted with the PVP TM mesh and the placement technique during a lecture. The first patient of the cohort was the first patient receiving a PVP TM in our department. The study cohort was operated between April 2009 and December All consecutive patients receiving a PVP TM repair were included in the cohort. The ethical committee of the hospital approved the study protocol and the trial was registered at ClinicalTrials.gov (NCT ). The patients gave their informed consent during the postoperative control visit. Surgical technique Surgery was planned either as an outpatient procedure or with a 24 h hospitalization, depending on the general condition or age of the patient and on the patients preference. Operations were performed under general anesthesia. A single dose of antibiotic prophylaxis (Cefacidal Ò, 2 g) was administered at the induction of anesthesia. For umbilical and epigastric hernias in the vicinity of the umbilicus, a curvilinear incision was made in the umbilicus either caudally or cranially. The hernia sac and herniated

3 Hernia (2014) 18: omentum were dissected and separated from the skin. The hernia sac was either reduced through the hernia defect or resected if the sac was too voluminous. Through the hernia defect the pre-peritoneal plane was widely developed around the defect. In case the hernia defect was too small to allow a finger, it was slightly enlarged. Cranially, the plane between the round ligament of the liver and the abdominal wall was developed. Caudally, the plane between the umbilical fold and the abdominal wall was developed. Care was taken to achieve a good hemostasis. If a hole in the peritoneum was made and this was easily accessible, it was closed. If not it was accepted and left open. Only if the development of the pre-peritoneal plane was not at all possible the mesh was positioned intraperitoneally. A PVP Ò medium size mesh (diameter 6.4 cm) was used in all cases and the mesh was inserted through the hernia defect. The mesh was deployed, centered round the defect and well extended without wrinkles. Care was taken not to have pre-peritoneal fat between the mesh and the posterior rectus fascia, which was checked with the index finger. The mesh was fixed to the edge of the hernia defect with two U-stitches through its central leaflets cranially and caudally. The leaflets were cut at the level of the fascia. The hernia defect was closed above the mesh with 2 to 4 sutures 2/0, either slowly absorbable or non-absorbable, with the vest over pants technique. The skin was closed with separate stitches 4/0. Postoperative wound care consisted of daily dry dressings and the general practitioner removed the skin stitches after 10 days. Patients were advised to have unrestricted activity except for heavy weight lifting or intense exercise during 3 weeks. Variables Primary endpoint of the study was recurrent ventral hernia. This was defined as a protruding bulge felt by palpation of the hernia repair site during straining at clinical examination or if the radiologist found a recurrent hernia defect. The secondary endpoint, wound infection, was defined as the need for antibiotic treatment or drainage of pus. Infection was classified as superficial when confined to the skin and/or subcutaneous tissue, and as deep when fascia or mesh was involved. Descriptive data at baseline included age, gender, body mass index, ASA-score, presence of diabetes or other comorbidities, history of abdominal wall or groin hernia operation, type of hernia (primary umbilical, epigastric, incisional), emergency operation, recurrent hernia, size of the hernia (cm), mesh positioning (pre-peritoneal, intraperitoneal, retro-muscular), closure of the hernia defect, intra-operative complications, early postoperative complications (graded according to the Clavien-Dindo classification [20]) and hospital stay. Outcome data at postoperative follow-up after at least 12 months were type of follow-up (telephone questioning, clinical examination w/wo ultrasound), abdominal wall pain at the operation site (classified according to Cunningham et al. [21]), foreign body feeling, symptoms of small bowel obstruction, reoperation for recurrent hernia and repair technique, subsequent unrelated abdominal operations, cosmetic result as perceived by the patient (very good, good, acceptable or bad), clinically proven recurrent hernia, recurrence seen with ultrasound and date of recurrence diagnosis. Bias The cohort consisted of consecutive operations using a PVP TM mesh, thus eligibility depended on the decision of the surgeon to use this device, allowing some inclusion bias. In contaminated or clean-contaminated operations the use of a mesh is avoided and for hernias larger than 3 cm we generally use an open retro-muscular mesh repair or a laparoscopic technique. No other mesh devices were used during the inclusion period of the trial. For our primary endpoint we tried to diminish the investigator bias to diagnose a recurrence by proposing an ultrasound investigation. In 18 patients no ultrasound was performed because no radiologist was available at the timing of the control visit or the patient refused because of lack of time. At the beginning of our experience the hernia defect was closed above the mesh with non-absorbable sutures. We changed to slowly absorbable sutures because in some skinny patients the knots of the non-absorbable sutures remained palpable and caused irritation of the skin. Study size In the study protocol a sample size of 100 patients was empirically chosen as a population large enough to investigate our objectives of safety and efficiency of the new mesh, and small enough to complete the study in a reasonable time-frame. By the end of 2011, the study size was reached and a total of 101 patients had been included. Quantitative variables Patients were grouped in five age categories. Body Mass Index was grouped in three categories (\25/25 to 30/ C30 kg/m 2 ). Hernias were grouped according to their size using cut-off values at 1 and 2 cm. Size measurements performed intraoperatively were used. If intra-operative measurement was missing, the preoperative measurement was used.

4 674 Hernia (2014) 18: Statistical methods Statistical significance was assumed at p \ The association between baseline characteristics and the recurrence of clinically proven hernia was statistically evaluated according to Fisher s exact test. Estimation of recurrence-free survival curves was done according to the Kaplan Meier method. Time-to-event curves according to hernia size were compared using the log-rank test. Univariate and multivariate Cox proportional hazard modeling was used to study risk factors for hernia recurrence. The model assumption of proportionality of hazards was checked by plotting log [ log(s(t))] against time in the different subgroups, where S(t) represents the Kaplan Meier survival estimate. P values were obtained through Wald Chi square statistics. All statistical analyses were performed using SAS software (version 6.12) (The SAS system, Cary, NC, USA: SAS Institute Inc). Results Participants A total of 101 eligible patients were included in the study and analyzed for the descriptive and early outcome data. A flow diagram of the study cohort is shown in Fig. 1. Three patients were lost to follow-up: two had died, one could not be contacted. Follow-up data were obtained in 98 patients. Six patients could only be interviewed by phone. Clinical follow-up was thus obtained in 92 patients and an additional ultrasound was performed in 74 patients. Descriptive data Patient data at baseline are described in Table 1. Indication and operative data are described in Table 2. The PVP TM was placed in the pre-peritoneal position in all but seven patients. It was placed intraperitoneal if the pre-peritoneal plane could not be developed, most often in incisional hernias. A retro-muscular placement between the muscle layers was done in one patient with a trocar hernia in the flank and one patient with a primary umbilical hernia in whom a catheter for peritoneal dialysis was placed during the same operation. Outcome data Early postoperative data are listed in Table 3. In 33 patients a postoperative complication occurred (32.7 %), mostly minor complications. We saw one case of deep wound infection. This patient was operated for an incarcerated umbilical hernia. At the operation the small bowel seemed viable. This proved to be wrong at the second postoperative day, when a bowel perforation was diagnosed. We performed a laparotomy with small bowel resection and removal of the mesh. The wound infection was treated with topical negative pressure therapy (VAC TM ) and the patient was discharged after 25 days. Main results The results of 98 patients for whom follow-up of at least 12 months was available are listed in Table 4. Reoperation Fig. 1 Study flow chart of 101 patients treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia and aged over 18 years

5 Hernia (2014) 18: Table 1 Patient data at baseline of 101 patients treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia Patient data % (n/n) a or Mean (SD) b Age (at the time of surgery) \40 years 13.9 % (14/101) years 20.8 % (21/101) years 33.7 % (34/101) years 16.8 % (17/101) C70 years 14.9 % (15/101) Mean (SD) b 54.0 years (13.3) % Women 32.7 % (33/101) Body Mass Index \25 kg/m % (26/101) kg/m % (35/101) C30 kg/m % (40/101) Mean (SD) b 28.5 kg/m 2 (5.0) ASA-score Normal health 58.4 % (59/101) Mild to moderate systemic disease 28.7 % (29/101) Serious systemic disease 12.9 % (13/101) Life threatening systemic disease 0 % (0/101) Diabetes 9.9 % (10/101) Type I 2.0 % (2/101) Type II 7.9 % (8/101) History of previous abdominal wall or groin 18.8 % (19/101) hernia operation a %(n/n) n number of patients/n total number of patients b Mean (SD) mean value of variable (standard deviation) Table 2 Indication and operative data at baseline of 101 patients treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia Indication and operative data % (n/n) a or Mean (SD) b Type of the hernia Primary Umbilical 75.3 % (76/101) Primary Epigastric 14.9 % (15/101) Incisional c 9.9 % (10/101) Midline Umbilical 6.9 % (7/101) Midline Infraumbilical 1.0 % (1/101) Lateral Flank 2.0 % (2/101) Emergency repair 6.9 % (7/101) Recurrent hernia 3.0 % (3/101) Size of the hernia B1 cm 27.6 % (24/87) cm 52.9 % (46/87) [2 cm 19.5 % (17/87) Mean (SD) b 1.66 cm (0.67) Placement of mesh Preperitoneal 93.1 % (94/101) Intraperitoneal 5.0 % (5/101) Retromuscular 2.0 % (2/101) Closure of hernia defect above the mesh 98.0 % (99/101) Type of sutures Absorbable 48.5 % (48/99) Non-absorbable 51.5 % (51/99) a %(n/n) n number of patients/n total number of patients b Mean (SD) mean value of variable (standard deviation) c Trocar hernias N = 6 rate for recurrent hernia was 5.1 % (5/98) with a mean observation period in event free patients of 15.9 months (SD 4.4; median 14.3; range ). Chronic pain at the operation site was reported by 11.4 % of patients. This pain was moderate or severe in three of them (3.1 %). Cosmetic result of the operation was perceived as very good or good by 92.0 % of the patients. Foreign body feeling was absent in 88.4 %. Clinically proven recurrent hernia was diagnosed in 12.0 % (11/92) of patients with a mean observation period in event free patients of 15.8 months (SD 4.4; median 14.2; range ). The clinically proven recurrence rate was 9.3 per 100 person-years. Only four patients were aware of their recurrent hernia. In the seven others, recurrence was asymptomatic and not noticed previously by the patients. The estimated recurrence-free survival curve (Kaplan Meier) of clinically proven recurrent hernia is shown in Fig. 2. This graph shows that recurrences are diagnosed also beyond 12 months indicating longer follow-up studies are needed to determine the real longterm efficiency. Ultrasound was performed as a control of the clinical examination in 74 patients. The radiologist was able to visualize the mesh in 63.5 % (47/74) of cases. In ten of those patients the radiologist noticed contraction of the mesh. No additional recurrences that were not diagnosed by clinical examination were found. In five patients the recurrent hernia was repaired, which was performed by laparoscopy in three patients and by open retro-muscular mesh repair in two. The mesh was removed during the operation in four patients and important centripetal contraction of the mesh, diminishing the surface area, was observed in all cases. In the fifth patient the mesh had already been removed during an early reoperation, mentioned previously.

6 676 Hernia (2014) 18: Table 3 Early complications data of 101 patients treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia Early postoperative data % (n/n) a Early postoperative complications b No complications 67.3 % (68/101) Grade I 24.8 % (25/101) Grade II 6.9 % (7/101) Grade IIIa 0.0 % (0/101) Grade IIIb 1.0 % (1/101) Grade IV 0.0 % (0/101) Grade V 0.0 % (0/101) Type of complication Wound complications 17.8 % (18/101) Superficial wound infection 11.8 % (12/101) Deep wound infection 1.0 % (1/101) Seroma 3.0 % (3/101) Hematoma 2.0 % (2/101) Prolonged postoperative pain needing 15.8 % (16/101) pain medication ([ 3 weeks) Miscellaneous Urinary retention 1.0 % (1/101) Urinary tract infection 1.0 % (1/101) Postoperative hoarseness 1.0 % (1/101) Hospital stay Ambulatory surgery 31.7 % (32/101) 24 h admission 53.5 % (54/101) 2 days 7.9 % (8/101) [2 days 6.9 % (7/101) a %(n/n) n number of patients/n total number of patients b Using the Clavien-Dindo classification for postoperative complications [20] Table 4 Follow-up data (C12 months postoperative) of an observational cohort study on 101 patients treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia Follow-up data (C12 months postoperative) % (n/n) a Questionnaire (N = 98) Reoperation for recurrent hernia 5.1 % (5/98) Chronic pain at the operation site b No pain 88.5 % (85/96) Mild pain 8.3 % (8/96) Moderate pain 2.1 % (2/96) Severe pain 1.0 % (1/96) Cosmetic result as perceived by the patient Very good 62.1 % (54/87) Good 29.9 % (26/87) Acceptable 6.9 % (6/87) Bad 1.1 % (1/87) Foreign body feeling at the operation site No foreign body feeling 88.4 % (84/95) Sporadic (once a month) 9.5 % (9/95) Frequent (once a week) 0.0 % (0/95) Constant (almost daily) 1.1 % (1/95) Clinical examination (N = 92) Clinical recurrence 12.0 % (11/92) Known by the patient 4.4 % (4/92) Asymptomatic and not noticed by the patient 7.6 % (7/92) Ultrasound investigation (N = 74) Mesh could be visualized 63.5 % (47/74) Contraction of the mesh mentioned by radiologist 21.3 % (10/47) Additional recurrences found with ultrasound 0.0 % (0/74) a %(n/n) n number of patients/n total number of patients b Using the Cunningham classification for chronic pain [21] Other analyses Among those variables where sufficient data were present, no significant risk factors for wound complication at 3 weeks could be found. Table 5 shows the analysis of risk factors for clinically proven recurrent hernia. The type of hernia was a significant risk factor, with more recurrences for primary epigastric hernias and incisional hernias compared to primary umbilical hernias (p = 0.029). Size of the hernia defect was also a significant risk factor, with larger diameters leading to more recurrences (p = 0.032). The estimated recurrence-free survival curve according to the size of the hernia defect is shown in Fig. 3. An additional proportional hazard modeling was performed for risk factors of hernia recurrence with a univariate and a multivariate Cox model. These results are shown in Table 6. Using the univariate Cox proportional hazards model, larger hernia size was found to be associated with an increased risk of hernia recurrence (p = 0.005). In the Fig. 2 Estimated recurrence-free survival curve (Kaplan Meier) of 101 patients treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia same model, incisional hernia proved to be a significant risk factor for recurrence, albeit with borderline significance (p = 0.047). Using the multivariate Cox model,

7 Hernia (2014) 18: Table 5 Analysis of risk factors for recurrence in 92 patients that were clinically evaluated at least 12 months after implantation of a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia Risk factors for (clinically proven) recurrent hernias % Recurrent hernia (n/n)* Significance** Age p = 0.72 \50 years 15.6 % (5/32) years 9.1 % (3/33) C60 years 11.1 % (3/27) Gender p = 0.81 Men 12.5 % (8/64) Women 10.7 % (3/28) Body Mass Index p = 0.39 \25 kg/m % (2/24) kg/m % (2/29) C30 kg/m % (7/39) ASA-score p = 0.44 ASA I 8.9 % (5/56) ASA II 15.4 % (4/26) ASA III 20.0 % (2/10) Diabetes p = 0.59 No 13.1 % (11/84) Yes 0.0 % (0/8) History of abdominal wall or groin hernia operation p = 0.42 No 10.7 % (8/75) Yes 17.6 % (3/17) Type of the hernia p = Primary Umbilical 8.3 % (6/72) Primary Epigastric 16.7 % (2/12) Incisional 37.5 % (3/8) Emergency repair p = 0.84 No 11.8 % (10/85) Yes 14.3 % (1/7) Size of the hernia p = B1 cm 0 % (0/24) cm 15.0 % (6/40) [2 cm 25.0 % (4/16) Placement of mesh p = 0.40 Pre-peritoneal 11.4 % (10/88) Intra-peritoneal 33.3 % (1/3) Retro-muscular 0.0 % (0/1) Type of sutures p = 0.84 Absorbable 11.4 % (5/44) Non-absorbable 12.8 % (6/47) Bold values indicate statistical significant (p \ 0.05) *%(n/n) n number of patients/n total number of patients ** According to Fisher s exact test hernia size remained a significant risk factor (p = 0.017), but hernia type did not (p = 0.08). Discussion Key results Of 101 eligible patients, 92 were examined clinically at least 1 year after operation with a recurrence rate of 12 % and a reoperation rate of 5 %. Seventy-four patients had an additional ultrasound of the abdominal wall. This examination showed shrinkage of the mesh in ten cases, but could not detect additional recurrences. Size of the hernia defect and incisional hernia (versus primary ventral hernia) proved to be significant risk factors for recurrence. Limitations Although data at 12 months follow-up were gathered prospectively, data at baseline and follow-up at 3 weeks were gathered retrospectively. Consequently, the recording of patient characteristics especially comorbidities, smoking and occupation was incomplete in a significant part of our cohort. Besides inferior accuracy of retrospectively collected data, this also reduces power when analyzing these variables for risk factors. Most recurrences after ventral hernia repair occur within 2 years after the operation [22]. Since our study had a mean follow-up of 16 months, it is likely that a longer follow-up would yield a higher recurrence rate. All small ventral hernias of our department during the study period were repaired with the PVP TM and therefore the inclusion bias mentioned above seems to be of limited magnitude. Investigator bias was reduced by the ultrasound investigation, performed by the radiologist. Interpretation Recurrence is the most frequent long-term complication in ventral hernia repair. We know from the Danish Ventral Hernia Database that even for small ventral hernias the reoperation rate for recurrence is high (4 %) and clinically proven recurrence rate is even higher (15 %) [7]. Our study, with a high follow-up percentage of 97 %, shows very similar rates. Other series on the use of mesh for repair of small ventral hernias, ranging from 28 to 152 patients, report very variable recurrence rates between 0 and 14.8 % [10 18, 23]. In studies without prospective follow-up data [10, 11, 18, 23] recurrence rates reported are

8 678 Hernia (2014) 18: Fig. 3 Estimated recurrence-free survival curve (Kaplan Meier) of 101 patients, according to the size of the hernia defect, treated with a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia 0 to 1.7 %. We consider these data irrelevant because too many patients were not evaluated. In one study a recurrence rate of 2.6 % was reported after follow-up with a questionnaire by mail and selective clinical investigation [13]. In studies with a better follow-up, both in methodology and percentage of follow-up, systematically higher recurrence rates are reported [12, 16, 17]. Our data show the importance of correct evaluation and reporting the recurrences. Follow-up consisted of a standardized schedule, with all interviews and clinical examinations conducted by the same surgeon (FM). These data were gathered prospectively. The majority of the patients with a recurrent hernia had no symptoms and were not aware they had a recurrent hernia (7/11). It took a physical exam to detect their recurrence. The fact that ultrasound has no added value in detecting recurrent hernias, confirms the importance of a thorough clinical examination. Any follow-up based on less than a clinical examination of the patient will result in an important underestimation of the true recurrence rate. Moreover, recurrence rate should be calculated based only on those patients that were followed. Because no information is available on the lost to followup patients, they should not be used in the equation to calculate the recurrence rate. Our study found that hernia defect size and the type of hernia are significant risk factors for hernia recurrence. Venclauskas has stated that for umbilical hernias with a size [2 cm and higher BMI can be considered risk factors for recurrence [24]. Our study supports this conclusion for hernia size but not for BMI. It appears that with the PVP TM good results are achieved only in hernias smaller than 2 cm. This finding indicates the importance of the amount of overlap of mesh around the hernia defect. Polypropylene meshes, like the PVP TM, have demonstrated an in vivo centripetal shrinkage percentage of up to 77 % in some patients [15]. This finding of mesh contraction was confirmed in those patients that were reoperated for recurrence and in 21 % of the patients where the radiologist was able to visualize the mesh. The overlap obtained with a mesh of 6.4 cm in diameter is insufficient with hernias larger than Table 6 Analysis with Cox proportional hazards modeling of risk factors for recurrence in 92 patients that were clinically evaluated at least 12 months after implantation of a Proceed TM Ventral Patch for a primary umbilical hernia or another abdominal wall hernia Bold values indicate statistical significant (p \ 0.05) * Estimated regression coefficient (standard error) in Cox PH model ** History of abdominal wall or groin hernia operation Risk factors for recurrent hernia: Cox Proportional Hazards modeling b (SE)* Wald v 2 Significance Univariate Cox model Age (per year)?0.016 (0.024) 0.42 p = 0.51 Gender (Female vs. Male) (0.678) 0.11 p = 0.74 Body Mass Index (per kg/m 2 )?0.031 (0.056) 0.31 p = 0.58 ASA-score (vs. normal health) Mild to moderate systemic disease?0.612 (0.672) 0.83 p = 0.36 Serious systemic disease?0.877 (0.840) 1.09 p = 0.30 History of a.w. or g.h. operation (yes vs. no)**?0.504 (0.679) 0.55 p = 0.46 Type of the hernia (vs. Primary Umbilical) Primary Epigastric?0.548 (0.819) 0.45 p = 0.50 Incisional?1.409 (0.711) 3.93 p = Emergency repair (vs. elective)?0.656 (1.066) 0.38 p = 0.54 Size of the hernia (per cm)?1.403 (0.499) 7.89 p = Type of sutures (non-absorbable vs absorbable)?0.204 (0.608) 0.11 p = 0.74 Multivariate Cox model Type of the hernia (vs. Primary Umbilical) Primary Epigastric?0.604 (0.859) 0.49 p = 0.48 Incisional?1.354 (0.778) 3.03 p = 0.08 Size of the hernia (per cm)?1.194 (0.499) 5.72 p = 0.017

9 Hernia (2014) 18: cm. Therefore, we recommend not to use PVP TM in hernias of 2 cm or more. In these patients, another technique allowing the use of a larger mesh should be preferred. Berrevoet et al. [16] observed a recurrence rate of 14.8 % with the Ventralex TM patch and stated that, to achieve an adequate overlap, a patch of 8 cm diameter is necessary for almost all ventral hernias. They conclude that a retro-muscular repair is still the gold standard mesh repair for umbilical hernias and all new devices should be evaluated against this. Although PVP TM and similar mesh devices are developed with an anti-adhesive barrier to allow intraperitoneal placement of the mesh we prefer to place the mesh in a preperitoneal position. Exposure of the Ventralex TM patch to the bowel has resulted in some patients in serious complications like bowel erosions [15]. Therefore, we like to keep the peritoneum intact and place the mesh directly underneath the fascia. Sometimes small peritoneal tears are unavoidable during the dissection which still gives a benefit to the anti-adhesive barrier, although strictly spoken this would not be necessary in a peritoneal position. Our study shows that the repair of small ventral hernias with a PVP TM mesh device is a safe procedure. One patient had a serious complication (Grade IIIb) not directly related to the use of the device, but leading to removal of the mesh in contaminated operative conditions during early reoperation. All other complications, including the need for pain medication after 3 weeks, were minor complications. In 11.8 % a superficial wound infection was present, but this did not lead to mesh infection. The superficial wound infection in more than 10 % of our patients seems high. However, Farrow et al. [25] found in their analysis of umbilical hernias treated in a Veterans Administration Medical Centre a surgical site infection rate of 19.7 %. Most studies report lower frequencies of infections, i.e. around 5 %. Similar to the comments made concerning recurrence rates, study methodology and accuracy might give an important bias towards underreporting of the number of surgical site infections. As umbilical hernia repair is mostly done in day surgery or during a short hospitalization, many postoperative superficial surgical site infections will occur at home and remain unknown to the surgeon. Generalisability We think that our cohort was a representative sample of adult patients with small ventral hernias. The number of patients with an incisional hernia, mainly trocar hernias and recurrent umbilical hernias, was low. Therefore, our conclusions are more relevant for primary ventral hernias than for incisional hernias. The study included only patients treated with the PVP TM mesh. This mesh is unique in its composition and structure and differs substantially from other existing mesh devices for small ventral hernias. Therefore, we cannot extrapolate our findings concerning the safety and efficacy to other mesh devices. Conclusion Intensive clinical follow-up yields a high number of asymptomatic recurrences after repair of small ventral hernias with the PVP TM. Reoperation rate for recurrence was 5.1 %. Hernia defect size is a significant risk factor for recurrence. Therefore, we recommend using the PVP TM only for primary ventral hernias smaller than 2 cm. For larger or incisional hernias other techniques allowing the use of larger meshes is advocated. Acknowledgments The Committee for Innovation of the AZ Maria Middelares Hospital funded a consultancy fee for the statistical analysis. The statistical analysis was performed by Prof. Dr. Dirk De Bacquer, Department of Public Health, Ghent University, Belgium. No other funding was requested for this study. Conflict of interest The authors report no conflict of interest in relation to this study. References 1. Dalenbäck J, Andersson C, Ribokas D, Rimbäck G (2012) Longterm follow-up after adult umbilical hernia repair: low recurrence rates also after non-mesh repairs. Hernia. doi: /s Halm JA, Heisterkamp J, Veen HF, Weidema WF (2005) Longterm follow-up after umbilical hernia repair: are there risk factors for recurrence after simple and mesh repair. Hernia 9: Arroyo A, Garcia P, Pérez F, Andreu J, Candela F, Calpena R (2001) Randomized clinical trial comparing suture and mesh repair of umbilical hernias in adults. BJS 88: Aslani N, Brown CJ (2010) Does mesh offer an advantage over tissue in open repair of umbilical hernias? A systematic review and meta-analysis. Hernia 14: Helgstrand F, Rosenberg J, Bay-Nielsen M, Friis-Andersen H, Wara P, Jorgensen LN, Kehlet H, Bisgaard T (2010) Establishment and initial experiences from the Danish Ventral Hernia Database. Hernia 14: Christoffersen MW, Helgstrand F, Rosenberg J, Kehlet H, Bisgaard T (2013) Lower reoperation rate for recurrence after mesh versus sutured repair in small umbilical and epigastric hernias. A nationwide register study. Hernia 17(suppl 2):S1 S26 7. Helgstrand F, Rosenberg J, Kehlet H, Strandfelt P, Bisgaard T (2012) Reoperation versus clinical recurrence rate after ventral hernia repair. Ann Surg 256: doi: /sla.0b013e f5b9 8. Bisgaard T, Kehlet H, Bay-Nielsen M, Iversen MG, Rosenberg J, Jørgensen LN (2011) A nationwide study on readmission, morbidity, and mortality after umbilical and epigastric hernia repair. Hernia 15: Erritzøe-Jervild L, Christoffersen MW, Helgstrand F, Bisgaard T (2013) Long-term complaints after elective repair for small umbilical or epigastric hernias. Hernia 17: doi: / s z

10 680 Hernia (2014) 18: Hadi HI, Maw A, Sarmah S, Kumar P (2006) Intraperitoneal tension-free repair of small midline ventral abdominal wall hernias with a Ventralex hernia patch: initial experience in 51 patients. Hernia 10: Martin DF, Williams RF, Mulrooney T, Voeller GR (2007) Ventralex mesh in umbilical/epigastric hernia repairs: clinical outcomes and complications. Hernia 12: Tollens T, Struyve D, Aelvoet C, Vanrijkel JP (2010) Introducing the Proceed Ventral Patch as a new device in surgical management of umbilical and small ventral hernias: preliminary results. Surg Technol Int 19: Iversen E, Lykke A, Hensler M, Jorgensen LN (2010) Abdominal wall hernia repair with a composite eptfe/polypropylene mesh: clinical outcome and quality of life in 152 patients. Hernia 14: Tollens T, Den Hondt M, Devroe K, Terry C, Speybroeck S, Aelvoet C, Vanrykel JP (2011) Retrospective analysis of umbilical, epigastric and small incisional hernia repair using the Ventralex TM hernia patch. Hernia 15: doi: / s y 15. Muysoms FE, Bontinck J, Pletinckx P (2011) Complications of mesh devices for intraperitoneal umbilical hernia repair: a word of caution. Hernia 15: doi: /s x 16. Berrevoet F, Van den Bossche B, de Baerdemaeker L, de Hemptinne B (2010) Laparoscopic evaluation shows deficiencies in memory ring deployment during small ventral hernia repair. World J Surg 34: Berrevoet F, D Hont F, Rogiers X, Troisi R, De Hemptinne B (2011) Open intraperitoneal versus retromuscular mesh repair for umbilical hernias less than 3 cm diameter. Am J Surg 201: Ambe P, Meyer A, Köhler L (2013) Repair of small and medium size ventral hernias with a Proceed Ventral Patch: a single center retrospective analysis. Surg Today 43: doi: / s von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP, STROBE Initiative (2007) Strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ 335: Dindo D, Demartines N, Clavien PA (2004) Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 240: Cunningham J, Temple WJ, Mitchell P, Nixon JA, Preshaw RM, Hagen NA (1996) Cooperative hernia study. Pain in the postrepair patient. Ann Surg 224: Singhal V, Szeto P, VanderMeer TJ, Cagir B (2012) Ventral hernia repair: outcomes change with long-term follow-up. JSLS 16: Kulacoglu H, Yazicioglu D, Ozyaylali I (2012) Prosthetic repair of umbilical hernias in adults with local anesthesia in a day-case setting: a comprehensive report from a specialized hernia center. Hernia 16: Venclauskas L, Silanskaite J, Kiudelis M (2008) Umbilical hernia: factors indicative of recurrence. Medicina (Kaunas) 44: Farrow B, Awad S, Berger DH, Albo D, Lee L, Subramanian A, Bellows CF (2008) More than 150 consecutive open umbilical hernia repairs in a major Veterans Administration Medical Center. Am J Surg 196: