Introduction

In certain scenarios, abdominal wall surgery constitutes a real challenge for surgeons. The results of umbilical hernia surgery may be affected in patients with cirrhosis and secondary ascites.

In such patients, ascites is a common complication, and elevated intraabdominal pressure as a consequence of ascites leads to the development of this type of hernias in patients with latent umbilical defects.

In our experience, in patients hospitalized in gastroenterology wards due to decompensated cirrhosis, the presence of umbilical hernia is underestimated, and surgeons are only called upon in case of complicated hernias.

Material and methods

Between January 2017 and January 2019, 15 patients with liver cirrhosis and symptomatic umbilical hernia underwent surgery in a public hospital. The clinical records were reviewed. All the patients were men. The diagnosis of umbilical hernia was made after physical examination and no further tests were necessary. There was no evidence of obstruction or strangulation; however, all hernias were irreducible, and in 8 patients (53.3%) there were signs of vascular involvement at the level of the skin covering the hernia sac.

The preoperative risk was estimated using the Child-Pugh score (Table 1).

Thirteen patients (86.6%) were classified as Child class C. After specific preoperative treatment, these patients were reclassified as Child class B.

The remaining patients (13.3%) were categorized as Child class B.

All patients had preanesthetic evaluation to assess the preoperative risk and were classified as American Society of Anesthesiologists (ASA) grade 3.

In 100% of the patients, the procedure was performed under epidural regional anesthesia. Some patients required anxiolytics to relief the anxiety caused by surgery.

Patients were hospitalized 48 hours before surgery and routine laboratory tests were carried out with measurement of white blood cell count, platelet count, prothrombin time, total bilirubin, AST, ALT, albumin levels, BUN and glycemia.

Evacuation of ascites was an essential preoperative step; occasionally, it was necessary to perform an ultrasound scan to detect the fluid in the perihepatic space and Douglas’ pouch.

A catheter was placed in the abdomen to allow for intermittent paracentesis. Neutrophil count in the abdominal fluid was done routinely to rule out bacterial peritonitis.

Vitamin K, fresh plasma and recombinant factor VII were administered to improve the coagulation status.

Surgical technique

Once in the operating room, the skin was prepared with povidone-iodine scrub before establishing the surgical field. The umbilical region was approached using a concave infraumbilical incision when it was not necessary to resect the involved skin. Incisions of the skin overlying the peritoneal sac were avoided, since at this level the skin is very thin, poorly vascularized and prone to poor healing. In case of vascular involvement, a lozenge-shaped skin incision was performed, and the umbilical scar was removed. These patients require careful hemostasis using electrocautery until reaching the aponeurotic plane.

The hernia sac and the fibrous umbilical ring were resected in all the patients. Mean defect size was 3 cm (2- 4.5). Muscle atrophy due to malnutrition became evident when the borders of the umbilical ring were raised. The defect was closed in two layers with non-absorbable suture and a 10 × 6 cm preaponeurotic lightweight polypropylene mesh was placed in 14 patients. In a patient with a 2-cm umbilical ring, a cone with a truncated beak made of heavyweight polypropylene mesh was prepared and placed in the preperitoneal plane and was fixed with four cardinal points.

When necessary, the subcutaneous cellular tissue was sutured to the aponeurotic plane with separate stitches of absorbable suture to minimize dead space and the development of seromas. The skin was closed with locked running suture to prevent leakage of ascitic fluid. We did not leave drains in any of the cases.

Patients were discharged within the first 3 days and were followed-up monthly during the first 6 months by gastroenterologists and surgeons. Thereafter, visits were scheduled twice a year.

Results

The clinical records of patients with cirrhosis and ascites undergoing elective umbilical hernia repair in a public hospital were retrospectively reviewed.

The analysis corresponds to 15 patients; 100% were men. Thirteen patients were classified as Child class C of the Child-Pugh score (Table 2).

Mean age was 60 years (range 28-72). The intraoperative variables analyzed included the ring size (between 2 and 4.5 cm), the presence of ascites and the type of procedure used (Table 3).

Mean operative time was 60 minutes and the patients were discharged within 72 hours. There were no long-term recurrences. Two patients were lost to follow-up 12 months after surgery and 1 patient died 6 months after the procedure due to progression of cirrhosis; thus, 86.6% of the patients completed follow-up.

Discussion

Evidence-based medicine (EBM) and its favorite daughter, evidence-based surgery, are based on a triad: external clinical evidence, clinical experience and patients’ preferences or needs. An inadequate sample size, in some special situations, is not an argument to disregard an outcome¹.

In this sense, the literature available on the management of umbilical hernias in cirrhotic patients is not definite -in view of the current knowledge- about the best treatment based on the existing evidence, as most of the publications are case series with limited follow-up and few patients.

Umbilical hernia repair is a safe technique in the absence of ascites; however, the presence of postoperative complications in cirrhotic patients has led gastroenterologists and particularly surgeons to prefer conservative treatment rather than elective surgery².

When complications develop, morbidity and mortality increase significantly, mostly in case of rupture of the hernia sac with spontaneous paracentesis and eviscerated bowel loops facilitated by a phenomenon similar to lubrication (Figures 1 and 2).

The rupture of the umbilicus constitutes the most serious complication which, along with trophic ulcers of the skin overlying the hernia, is always associated with coagulopathy and lack of response of ascites to treatment. In these cases, hernia repair and concomitant insertion of a peritoneojugular shunt could be indicated despite the high rate of infections associated with this treatment³.

Even so, there are contraindications to surgery for patients with chronic liver disease, especially for those classified as Child-Pugh class C⁴.

According to Chapman, 42% of cirrhotic patients with ascites will develop an umbilical hernia and require urgent repair during the disease⁵.

In our environment, umbilical hernia occurs in 20% of patients admitted with chronic liver disease and ascites.

For many years, we have been performing elective surgery in patients with cirrhosis and umbilical hernia, and we have been satisfied with the results in terms of patients’ well-being, quality of life and hernia recurrence.

This approach made us visit internal medicine and gastroenterology wards to look for these patients and convince their primary care physicians of the importance of timely treatment.

Another dilemma is to choose the appropriate surgical technique for these patients, since recurrences range between 5 and 30%⁶.

In a prospective randomized trial, Arroyo et al. compared mesh repair with primary suture and reported that hernia recurrence rate was 11% after suture repair and 1% after mesh repair⁷.

Once mesh repair has been decided, the site of mesh placement must be defined. The preperitoneal space is adequate, but considering the pathological characteristics of the tissues in these patients and the presence of ascites (refractory in many cases), we considered the supra-aponeurotic space as the ideal site for mesh location.

The technique is easy and reproducible, and the rate of complications was not different compared with that of patients without cirrhosis.

None of our patients presented complicated hernias, but the external characteristics of the hernia sacs were complex. Strangulated hernias associated with upper gastrointestinal bleeding due to variceal bleeding may rarely occur⁸.

We do not consider minimally invasive surgery as an adequate approach in this type of patients. In endoscopic extended totally extraperitoneal (eTEP) repair, a mesh is placed in the preperitoneal-retromuscular space and involves significant dissection with greater incidence of bleeding. In intraperitoneal onlay mesh (IPOM) placement, recurrent ascites has been described as a cause of recurrence and as a conditioning factor in mesh tissue integration and host-tissue response.

Some authors as Saric have reported good results with the laparoscopic approach for complicated umbilical hernias in cirrhotic patients with ascites⁹.

When the data were grouped in a meta-analysis, the recurrence rate was 45% in uncontrolled ascites and 4% in controlled ascites. The authors concluded that uncontrolled ascites strongly correlated with umbilical hernia recurrence¹⁰. In our experience, we did not find hernia recurrence despite most patients presented refractory ascites during postoperative follow-up (Figure 3).

Surgery has not been considered an indication for this type of hernia in cirrhotic patients with ascites because spontaneous reduction is achieved after paracentesis. Nevertheless, complicated hernias increase the likelihood of postoperative adverse events and can lead to fluid leakage when an ostomy is used after bowel resection¹¹.

However, when liver transplantation is planned in the short term, concurrent repair of an umbilical hernia present at the time of liver transplantation is feasible¹². A recent recommendation issued by European and American experts suggests that minimally invasive umbilical hernia repair seems safe in patients without ascites.

In case of ascites, open repair with onlay or preaponeurotic mesh seems to be a good option in patients with Child class A or B. A MELD score above 15 (based on bilirubin, INR and creatinine levels) is a risk factor for poor outcome, particularly in end-stage disease¹³.

Conclusions

The incidence of umbilical hernia in cirrhotic patients with ascites is 20%. Watchful waiting can lead to serious complications. Once the general status has been compensated, elective surgery provides benefit by increasing patients’ quality of life. Placement of a polypropylene mesh in the preaponeurotic plane after omphalectomy is a reproducible technique. Management of ascites is essential to avoid recurrence.

Large prospective randomized series with long-term follow-up are needed to provide strong evidence on the ideal technique.