Suture materials and subsequent wound strength

This issue contains two case reports of caesarean section wounds that ruptured. We have all been aware, at least in British practice, that if a patient has undergone a previous caesarean section for a non-recurring problem, that she should be considered for a ‘trial of scar’, and from time to time we discover at the subsequent section that the wound has dehisced or rarely ruptured. Therefore, what is remarkable about these two cases is that the ruptures were spontaneous and occurred in the second trimester before the lower segment had re-formed.

In the first report (Izzat 2008), the patient had previously had an emergency caesarean section with extension of the lower segment wound laterally into the broad ligament. The wound was repaired with polyglactin 910 sutures and the left uterine artery was ligated. In her next pregnancy, she presented at 16 weeks with a 4 l haemoperitoneum arising from an anterolateral uterine rupture. Subtotal hysterectomy was performed and the subsequent pathology demonstrated a posterior placenta without involvement of the previous scar.

In the second report (Bhattacharya et al. 2008), the patient had two normal vaginal deliveries, a section and a subsequent vaginal delivery and spontaneous complete abortion. We do not have any information on what suture material was used to repair her section wound. She presented with a twin pregnancy at 22 weeks and underwent laparotomy because of a half-litre haematoma, with the uterus ruptured along the previous section scar, one fetal sac expelled into the peritoneal cavity and the other still within the uterus, and a placenta separated from but still within the uterus.

We have published recent case reports (El-Matary et al. 2007; Ismail and Toon 2007) of caesarean scar pregnancy where implantation into the scar produced early pregnancy vaginal bleeding or ultrasound features of implantation into the caesarean scar. Caesarean scar pregnancy has been reviewed by Ash et al. (2007) and they describe that there were only 19 cases reported in the English medical literature until 2001, but since then there has been a ‘substantial increase’ in reporting. The reasons behind this increase could include repair technique and suture material, as well as the increasing section rate. The diagnostic features include implantation into the anterior uterine wall in the vicinity of the previous scar, with an empty uterine cavity. The details in the two cases we report in this issue do not suggest implantation into the scar, but of early scar rupture and raise the question of scar integrity.

In the bad ‘old days’ we used chromic catgut to close uterine incisions. Williams (1930) had described the use of ‘formol or chromicized’ catgut sutures to close the uterus during caesarean section, and Berkeley and Bonney (1935) advocated catgut in reconstructing the uterus after myomectomy, silk being reserved for when considerable tension in the suture was required. This use of chromic catgut led to a significant inflammatory response and subsequently good wound fibrosis, which was an advantage for uterine surgery. We speculate that non-absorbable or slowly absorbable suture material does not produce a uterine scar with the same strength from a vigorous fibrous response. Polyglycolic acid or polyglactin sutures are reported to have advantages over chromic catgut for perineal repair with less pain or need for analgesia (Kettle and Johanson 2000; Upton et al. 2002). These materials are said to cause less tissue reaction and irritation in the wound and these properties may be an advantage in the perineum but may be a serious disadvantage within the uterine wall.

Rather surprisingly, there appears to be little information about wound strength when ‘modern’ suture material is used for the uterus. We have recently published (Kelly et al. 2008) that laparoscopic myomectomy may have a greater risk of subsequent uterine rupture in labour, but it is not clear whether this followed the selection of suture material or the laparoscopic suturing technique.

On the other hand, surgeons have studied the role of polyglycolic acid and other synthetic suture material for more than 30 years. Breaking strength was assessed in rats by Craig et al. (1975) who compared polyglactin 910 and polyglycolic acid, and Sanz et al. (1988) who compared polyglyconate, polyglactin and polydioxanone with chromic catgut. Polydioxanone and chromic surgical gut were compared in urological surgery in dogs by Cohen et al. (1987) for both breaking strength and for stone formation. Outlaw et al. (1998) compared chromic gut with polydiaxone, Maxon, Monocryl, Vicryl, Dexon and Polysorb after implantation in rats, and showed that breaking strength was decreased in chromic gut and Monocryl after 1 day, and all the multifilament (i.e. Vicryl, Dexon, Polysorb) sutures by 7 days. However, suture diameter increased in chromic gut and the multifilament sutures and this was attributed to inflammatory tissue infiltration.

Rodeheaver et al. (1981) observed that the strength of both polyglactin and polyglycolic acid sutures were similarly diminished by the formation of the knot. Those of us who grew up with catgut learnt about its handling and knotting and quickly realised that similar knotting techniques were inadequate for these newer sutures. Rodeheaver et al. (1983) studied this when comparing coated sutures, and reported that coated polyglycolic acid sutures achieved knot security with one less throw than similar sizes of coated polyglactin 910 sutures. Williams et al. (2008) have used a tensile tester to assess arthroscopically positioned sutures – they showed that non-absorbable materials had better knot security following heat treatment with an electrosurgical unit and electrode. Several studies have looked at suture technique and wound strength. In a study by Zantop et al. (2006), it was found that meniscal repair with horizontal suture techniques can withstand elongation due to shear forces more effectively than can vertical mattress sutures. Poole et al. (1984) found that with laparotomy wounds, the simple interrupted technique was unaffected by suture tension but was generally inferior to the running stitch in terms of wound-bursting strength. Cengiz et al. (2000) found for laparotomy wounds that wound bursting strength is higher with a conventional running suture than with a continuous double loop closure when the effect of the suture length: wound length ratio is accounted for. Finally, the experience of the surgeon may be more important than the suture material – Irvin et al. (1976) showed that when polyglycolic acid and polyglactin sutures were compared with polypropylene in closing abdominal wounds, dehiscence and herniation were significantly more common in wounds closed by surgeons in training. It is not clear what roles knot insecurity or overly tight suturing leading to wound devitalisation were responsible for loss of wound strength.

Have we been right to abandon chromic catgut for uterine repair? Do we need to carefully select an alternative if there are ethical or other reasons for its discontinuation? The only study we could find that was specific to caesarean section was published in Croatian (Sestanovic et al. 2003) and which reported 1,059 vaginal deliveries that occurred after previous section. Plain catgut was compared with Dexon and Vicryl at the original section, and 15 women were found at subsequent delivery to have uterine rupture or dehiscence. Perhaps not surprisingly, 11 of the 15 uterine ruptures occurred in the plain catgut group, none in the Dexon group and four in the Vicryl group; the obvious differences between chromic gut and plain catgut would not encourage us to put plain catgut into such wounds. What was as surprising was that wound closure included the four options of one row of interrupted sutures, one row of continuous suture, two rows of interrupted and two rows of continuous and interrupted suture. We need a study where the uterus has been closed in two continuous layers with different suture materials. We believe that more information is necessary about the risks of uterine rupture and the use of newer suture materials.