The place of mesh

Pelvic organ prolapse (POP) is a multifactorial pathology. Pelvic floor traumas resulting from pregnancy and vaginal delivery, postmenopausal atrophy of the pelvic floor tissues, obesity and chronic bronchial obstruction are well known risk factors. However, the main risk factor for pelvic floor relaxation is the quality of the connective tissue in the pelvis and the perineum.

Many recent studies have shown significant modifications of smooth muscle cells, matrix metalloproteinase -9, collagen types I, III and IV, tenascin, elastin, vitronectin in the extracellular matrix of pelvic floor connective tissue in cases of POP. Fascia and ligament tissue are less elastic and more breakable.

Three studies have underlined the primary weakness of the collagen in POP. Pal et al. showed a positive correlation between low bone densitometry and pelvic organ prolapse. We know that osteoporosis is primarily disease of the collagen matrix of the bone, especially of its turnover, and we can imagine a similar mechanism for POP. Salter et al. established a strong association (OR 3.12, P < 0.05) between two connective tissue disorders: striae and POP. Handa et al. recently confirmed that small prolapses were likely to regress in time: regression rates for grade 1 prolapses after 2–8 years were 23.5% for cystocoele, 22% for rectocele and 48% for uterine prolapse. Progression rates were only 9.5%, 13.5% and 1.9%, respectively. This means that damaged tissues in many women who have vaginal distension and distortion at the time of delivery will repair themselves. However, if this does not happen, POP will develop.

Genetically defective connective tissue or premature ageing of this tissue may be the primary cause of POP. This means that correct repair should focus on the anatomic correction of the disruptions or distensions, and also the improvement of the quality of the supportive tissue.

When a mesh is inserted, it will trigger a foreign-body reaction. Fibrin resulting from the dissection will quickly surround the mesh, giving it a soft primary hold. Granulocytes and macrophages will colonize the mesh within 48 h during the inflammatory response process. After a week, fibroblasts form on the mesh, and start to produce the extracellular matrix of a new connective tissue around the mesh. Thus, the goal of the mesh is mainly to ‘stake’ the building of this fresh new connective tissue in the correct dissection plane, in the right direction. This tissue, is continuous with the pelvic fascias and ligaments around it, and will therefore be more resistant to tearing and traction than any suture alone. This new tissue growth is the key to resistance, not the mesh itself.

As connective tissue is a live tissue, it needs to be continuously remodelled to stay resistant and elastic. The proteic matrix needs to be renewed, especially in women with POP whose collagen is of poor quality. If an absorbable mesh is inserted, it will induce an inflammatory response followed by macrophages and fibroblasts, and a new fascia or ligament will be produced on this site. However, when the mesh is finally absorbed, the new collagen will age and undergo the natural evolution of the autologous one, which is poor. This is the experience that many surgeons have had with cross-linked porcine dermis implants. If a non-absorbable mesh is used, the foreign-body reaction will persist, the collagen will be renewed on a regular basis around the mesh, and the quality of the new ligament or fascia will remain constant.

In a study of the resistance of various pelvic fascias and ligaments, Rubod et al. reported poor quality of native fixation structures. Tendinous arches, uterosacral ligaments and even sacrospinous ligaments are weaker than the softest mesh or band currently being used. Thus, we have to understand that the efficiency of POP repair is not dependent on the resistance of the mesh nor on the nature of the sutures attaching it to the fragile structures of the pelvis, but on a constant soft fixation of the structures in the correct axis. The vagina is a sophisticated anatomical structure and its perfect reconstruction is a real challenge. The results of pelvic floor surgery depend on this anatomical correction.

The functional anatomy related to balanced forces has been largely described by Petros et al. in their ‘Integral Theory’. Pelvic floor repair should aim to restore these forces without any over-tensioning. Over-correction of one segment will always lead to pain, dysfunction or breakage involved by increased tension on another segment. This has led to the tissue-sparing attitude associated with mesh surgery of POP: avoid hysterectomy and colpectomy to avoid tension and reduce the rate of recurrence or complication.

Prolapse is not an illness of the uterus, and hysterectomy does not cure the prolapse. The apical defect has to be treated specifically or the POP will recur. Hysterectomy increases morbidity without evidence of increasing success rate, and shortens vaginal length. Colpectomy aims to trim the excessive vaginal epithelium in order to ‘tailor’ a ‘normal looking’ vagina. Vaginal epithelium, however, is a live tissue and is able to recover after incision and distension. For instance, after distension by vaginal delivery, it retracts back to its normal size after 2 h. After abdominal or laparoscopic sacrocolpopexy (almost never associated with a colpectomy), it retracts to a normal shape within a few days, keeping normal thickness, vascularisation and innervations. This is crucial if you use mesh to reinforce the fascia; a local necrosis of the overlying epithelium will expose the prosthesis and result in erosion.

Furthermore, keeping the whole vaginal wall avoids tension and dyspareunia. The main difference between traditional vaginal POP surgery and mesh surgery is that the first is used for treating anterior and posterior colpoceles and the latter is used for treating cystoceles and rectoceles. The use of mesh in POP repair, therefore, aims to remodel the defective collagen of the pelvic floor and its turnover, and to reconstruct the anatomy using a tissue-sparing, tension-free approach.

On the basis of these considerations, what are the ideal indications for this kind of repair, bearing in mind the extra cost of the graft, retraining to shorten the learning curve and specific complications of diagnoses and treatment? POP surgery is indicated in the following women: all women with an obvious genetic defect of their connective tissue; women who have a family history of genital prolapse (i.e. mother, grandmother or sister) or abdominal wall defects (hernias); women who have experienced POP before the age of 40 years; and women experiencing recurrent POP after accurate previous surgery.

The risk of failure of an autologous repair is high, probably above the reported average of 30% for cystocele repair. The indication of mesh reinforcement in POP-Q grade 3 and 4 prolapse looks reasonable, especially for cystocele, and has been used by many professional associations, such as the French college of Gynecology and Obstetrics and the French society of pelvic and gynecologic surgery. Comparative trials have shown the superiority of mesh repair in anterior compartment defect, and the current evolution of mesh weight and porosity is dramatically reducing the risk of complications such as erosion, exposition and infection.

Contraindications to the use of synthetic mesh include women with an infection or women who are at high risk of infection (i.e. unstable diabetes, spontaneous or induced immunodeficiency). Women undergoing radiotherapy for gynaecological cancer is also a relative contraindication because of poor vascular supply and the high risk of mesh exposition.

If high risk of recurrence in women is a good indication for mesh repair, the approach may be abdominal (or rather laparoscopic, now) or vaginal for the insertion. It seems that both approaches are complementary rather than conflicting. Sacrocolpopexy is most effective for dominant level 1 defect, which is more frequent among young women. Vaginal mesh repair is most effective every time a large cystocele with lateral defect is present.

Recurrence after sacrocolpopexy is best treated using the vaginal approach and vice versa. If laparoscopy is considered as dangerous in fragile women, vaginal surgery can be carried out under local or spinal anaesthesia. When hip stiffness prevents correct positioning of the woman for vaginal surgery, laparoscopy is the solution. In a modern urogynaecological unit, both approaches should be available and selected according to the patient’s features.

Inserting the mesh

Once the correct indication has been established, several types and shapes of mesh can be used. Light meshes are preferable (i.e. pore size of 1 mm; 30–40 g per square meter). The limit of decreasing the weight is not known yet. Reducing the quantity of implanted synthetics helps to reduce the risk of complications, such as erosion and infection. It also reduces the inflammatory foreign-body reaction and excessive fibrosis.

The dissection plane is between the vaginal fascia, the bladder anteriorly, and the rectum posteriorly. The fascia remains affixed to the mucosa to enable good vascularisation, and to avoid necrosis and mesh exposition. The mesh is secured to pelvic landmarks in a standardised way: the sacrospinous ligaments and the perineal body in the posterior mesh, the tendinous arc and the fascia obturatoris close to the ischial spine or the sacrospinous ligaments again for the anterior mesh. The mesh must be well spread without tension but also without folds to reduce the risk of fibrosis and retraction. Then, the vagina is closed by full thickness stitches bringing together fascia and mucosa without fascial plicature or colpectomy.

Revision and removal

Vaginal mesh surgery is a new technique, involving a new set of complications, that has to be learned by the surgeons inserting implants and by the practitioners that are likely to deal with the women receiving the implant. Unrecognised complications may lead to catastrophic events, such as cellulitic phlegmasia and death. The guidelines for treating complications are again specific to the type of mesh used. The main complications requiring revision or removal are erosion, symptomatic retraction, persistent pain, failure and dysfunction.