A modified Rives–Stoppa technique with composite mesh (FLaPp) in large incisional hernia: a multicentric retrospective cohort study

The preoperatively recorded variables were sex, age, body mass index (BMI), coexisting diseases, preoperative ASA assessment score, type of incisional hernia, operative time, and length of hospital stay. All patients underwent preoperative computed tomography (CT) of the abdomen with functional maneuvers that increase intraabdominal pressure (Valsalva, lateral decubitus) and, particularly, multidetector CT, which allows three-dimensional reconstruction to assess the presence of loss of domain.

The FLaPp is composed of two layers: an upper light polypropylene mesh (PM) and a thin polypropylene transparent film (PPF). The two layers possess different properties in order to best perform their functions. The mesh for the parietal side is macroporous, with 91% porosity and 48 gr/m², and is made of polypropylene monofilament of 120 μm diameter. The woven polypropylene is designed for tissue ingrowth for adequate prosthesis integration, due to its 2D scaffold feature which improves cell proliferation [11, 12]. The film for the visceral side is composed of nonporous, smooth, and transparent polypropylene with a thickness of 50 μm (type IV) [13] to minimize adhesion formation at the intestinal side due to its plane surface. In fact, the fibroblasts on the film do not proliferate, and, as a consequence, undergo apoptosis, a physiological death [13, 14]. The transparency of the polypropylene film enables easy identification of blood vessels, nerves, and underlying tissues during prosthesis placement and allows the risk of bowel injuries to be limited. The PM and the PPF are sewn together in the central oval area of the composite prosthesis: the two free peripheral flaps of the mesh have a dimension adaptable to the defect size. Moreover, as regards to the tensile strength of the polypropylene film, mechanical tests have demonstrated that its resistance exceeds that of the thread, and the passage of the needle does not cause any tearing of the film itself. De Maria et al. analyzed the influence of the topology of polypropylene for abdominal wall repair, evaluating its ability to prevent and minimize adhesion formation and promote tissue ingrowth. Finally, they found that the mechanical behavior of mesh presents an anisotropy index similar to that of natural tissue as well as a high safety index [14].

All patients received prophylactic preoperative chemotherapy with cefazolin (2 g) i.v. and all procedures were performed under general anesthesia. Once the peritoneum has been incised and the contents of the sack have been freed from adhesions, the medial edge of the rectus muscle fascia is incised on the posterior side to ensure continuity of the anterior fascial plane, which is essential to allow easy supraprosthetic closure [6]. The lateral extent of this dissection is the linea semilunaris, confirmed by visualizing the junction between the posterior and anterior rectus sheaths. Inferiorly, the space of Retzius is entered to expose the pubis symphysis and both Cooper’s ligaments. This dissection is blunt, in what is typically a bloodless plane. Since this area is below the arcuate line, the posterior layer includes peritoneum and transversalis fascia only. Proximally, the posterior fascia is transversally incised, bringing the dissection back to a preperitoneal plane, allowing it to be extended beyond the retroxiphoid region and the costal margins. Once this surgical time was over, in all the treated cases, only a posterior separation of the components would have allowed the closure of the posterior fascial due to the tissue deficit. The FLaPp prosthesis allows this necessity to be overcome, permitting realization of a simpler and faster intervention (Fig. 1, [6]). The PPF flap was sutured to the margins of the residual peritoneum and posterior rectus sheath to replace the tissue defect [15, 16]; short running sutures were used, adapting this flap to the shape and size of the defect, and the excess film was eliminated. This lower layer of the prosthesis had to be large enough to cover the defect and overlap the tissues for at least 1 cm. At this point, the prosthesis already appears well fixed and the PM flap is positioned on the large retromuscular dissection area according to the principles of Rives–Stoppa (Fig. 2). Superiorly, the PM flap can be placed beyond the costal margin and in the retroxiphoid space. It is secured with one stich to the xiphoid process; this suture is placed 4–5 cm off the edge of the PM flap to allow for large overlap, especially for subxiphoid defects. Finally, the macroporous mesh is anchored at the lateral extremities with two transparietal transfixed sutures or with full-thickness transabdominal stitches using the Reverdin needle and can be further fixed with fibrin sealant (TISSEEL; Baxter, Deerfield, IL, USA) [15, 17]. In all patients we used a FLaPp prosthesis with dimensions of 20 × 30 cm, 26 × 34 cm, or 30 × 40 cm, according to the defect size. Closure of the anterior plane with reinsertion of the rectus is performed easily, because access to the posterior fascia of the rectum is achieved while strictly maintaining continuity of the anterior fascial plane. Running sutures with an absorbable monofilament (polydioxanone) are used [6]. Two closed-suction drains were usually positioned directly over the mesh. The drains were removed when the output was < 25–30 mL of drainage over a 24-hour period. All patients used an elastic waistband for at least 2 months.

Primary outcome was the evaluation of feasibility and safety of the FLaPp prosthesis in patients affected by LIH by the assessment of early events such as postoperative surgical site occurrence (SSO), morbidity, mortality, and pain. Secondary outcome was the assessment of patients’ satisfaction with the procedure and the onset of recurrence and of other late complications (i.e., chronic mesh inflammation, mesh bowel erosion, mesh migration, sinus tract formation, onset of occasional pain or stiff abdomen).

Follow-up data were obtained from analysis of the surgeons’ office records, the patients’ hospital and outpatient electronic medical records, and a survey conducted via telephone. Follow-up appointments were scheduled at 2 weeks, 1 month, 6 months, and at the end of the follow-up (3 years). In the first 30 days, the incidence of SSO such as seroma, hematoma, surgical site infection (SSI), intensity of pain, and perioperative morbidity and mortality were evaluated [18]. Moreover, the incidence of chronic mesh inflammation, mesh bowel erosion, mesh migration, sinus tract, stiff abdomen, chronic pain, and recurrence rate were also recorded. The degree of pain was determined using a 10 cm visual analog scale (VAS) [19] (0 = no pain and 10 = worst possible pain) on postoperative day 15 and day 30, and by assessing the number of oral analgesic drugs (ketorolac 30 mg) taken by each patient. In fact, all patients received a prescription for analgesic drugs to be taken as required and they duly recorded analgesic use on a paper to be given back to the surgeon. Hernia recurrence was diagnosed by physical examination, which was performed serially in the outpatient setting.

To assess patients’ satisfaction, a quality of life (QoL) questionnaire (the so-called 36-Items Short-Form Health Survey, SF-36) was administered by an external observer (i.e., a surgical fellow) 7–10 days before surgery, and at the third year of follow-up. [20, 21]. The SF-36 defines eight domains of health status: physical functioning (PF), physical role limitations (PR), bodily pain (BP), general health perceptions (GH), energy/vitality (EV), social functioning (SF), emotional role limitations (ER), and mental health (MH). The number of questions contributing to each domain varies from 2–10. Response values for each question range from 1–6. All domain scales are standardized from 0–100, with higher scores corresponding to better health status.

The database of this study was completed in January 2020 in order to achieve a minimum follow-up time of 3 years for each patient, including the last one operated in 2016.

Continuous variables were expressed by mean and standard deviation (SD), categorical variables were reported in terms of median and interquartile range (IQR; 25th–75th percentile). The two-tailed Mann–Whitney U test or the chi-square test was applied to analyze the relationship between each preoperative characteristic and recurrence, according to variable type. The one-tailed paired Student t-test was applied to evaluate the increase in SF-36 score recorded preoperatively and at long-term follow-up for each of the eight domains separately. A p-value < 0.05 was considered statistically significant. Statistical analysis was carried out using MATLAB® 2021a (MathWorks; Natick, MA, USA) software. Analysis of SF-36 scores was performed according to guidelines developed by Ware and Sherbourne and Apolone and Mosconi [20, 21].

No major complications related to the operation or to anesthesia were observed, and no mortality occurred. SSO included 7 seromas (16.3%), 5 hematomas (11.6%), and 3 wound infections (7.0%; Table 2). Three of the 7 patients with seroma were treated by 3 or 4 needle aspirations, while the other 4 patients healed spontaneously; of the 5 patients with hematoma, only one was treated with a small skin incision under local anesthesia and with surgical drainage. All patients with wound infection were treated with antibiotic drugs after microbiological and antibiogram examinations. Two of them also required surgical drainage of the subcutaneous space and negative pressure wound therapy (VAC therapy). Early postoperative pain was controlled for the first 72 h with infusion therapy using an elastomeric pump prepared as follows: 50 mg of morphine (5 ampoules) + 50 mg of metoclopramide (5 ampoules) + 210 mg of ketorolac (7 ampoules) + 300 mg of clonidine (2 ampoules), at the infusion rate of 2 ml/hour. At postoperative day 15, 39 patients (90.7%) had a VAS score lower than 4 (range 1–3.5) and only 6 patients took one analgesic pill (ketorolac 30 mg) per day for another six days. At 30 days, the patient’s VAS score was less than 3 (range 0–2) and no patient required analgesic drugs parenterally or orally. Medical complications occurred in 6/43 patients (14.0%): particularly prolonged postoperative ileus occurred in 3 subjects (7.0%), pneumonia in one (2.3%), and urinary tract infection in 2 patients (4.7%). Moreover, all three patients with chronic renal failure (7.0%) suffered from exacerbation of renal disease and were treated with intravenous infusions of hydroelectrolyte solutions.

No patients presented chronic mesh inflammation, mesh bowel erosions, or mesh migration. No patients suffered from chronic pain: only one (2.3%) described pain that occasionally limited activities such as gardening or light gym work. However, we documented two cases (4.7%) of sinus tract and 4 patients (9.3%) complained about a stiff abdomen. Recurrence rate was 9.3% (4/43; Table 2). Of the abovementioned cases, one elderly patient required reoperation for total mesh excision 48 h after surgery for biliary peritonitis due to ischemic bowel perforation; this patient relapsed after 3 months. Two patients underwent partial mesh excision for development of a chronic sinus tract and one of these two developed a recurrence 8 months after primary surgery. Finally, the other two patients relapsed at 15 and 18 months after surgery, respectively. None of the preoperative patient characteristics were significant for the development of hernia recurrence (p > 0.2); however, two patients were obese and two patients were diabetic, and all four cases were irreducible IHs.

QoL results are summarized in Table 3. In terms of the preoperative functional status, significant deficiencies (mean score < 50) were reported in four SF-36 domains: physical function (34.9 ± 1.7), physical role limitations (35.1 ± 2.1), bodily pain (27.8 ± 1.5), and emotional role limitations (39.4 ± 1.6). At 3‑year follow-up, we documented a significant improvement in all domains (p < 0.0001) except for energy/vitality (p = 0.45).