Prospective randomised controlled trial to compare tissue reactions and scarring with polypropylene versus polyglactin 910 sutures in periocular skin

Abstract

Background/aims

The aim of this study was to compare tissue reactions and scarring produced by polypropylene (Prolene) versus polyglactin 910 (Vicryl) sutures in eyelid skin.

Methods

47 patients with entropion were randomised to 6–0 Prolene or 6–0 Vicryl. Patients underwent everting sutures 1 month prior to definitive surgery involving wedge excision and retractor plication. Only the randomised suture type was used in both procedures. The study suture within the wedge was analysed histologically. Patients were assessed clinically 2 weeks and 6 months following definitive surgery.

The primary outcome was histological inflammatory reaction diameter around the suture assessed by a masked histopathologist. Secondary outcomes were clinical scarring assessed using a validated scar grading questionnaire completed by patients and masked assessors and clinical photographs assessed by two masked oculoplastic surgeons.

Results

Mean inflammatory reaction diameter was 0.39 mm for Prolene and 0.56 mm for Vicryl. Median histological grade sum was 32 (IQR 22–37) for Prolene and 40 (IQR 27–46) for Vicryl. The median number of foreign body giant cells around the suture was 1 (IQR 0–3) for Prolene and 4 (IQR 3–8) for Vicryl. Only this last result was statistically significant; the lack of significance for other results may be due to the small sample size. One patient developed a clinical granuloma following surgery with Vicryl. There was no significant difference between suture scarring assessed by questionnaire responses or clinical photography.

Conclusion

Prolene may cause less histological inflammation than Vicryl in eyelid skin but clinical scarring is similar.

Trial registration number

NCT04724512.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Different suture materials evoke varying reactions in the host tissue.

• Some studies have found more inflammatory-related complications with Vicryl than Prolene.

WHAT THIS STUDY ADDS

• Histological tissue inflammatory reaction may be less with Prolene than Vicryl in the short term, but long-term clinical scarring is very good with both sutures in eyelid skin. Some patients develop clinical ‘granuloma’ to Vicryl, which may require further surgery.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• In patients more prone to significant scarring, there may be less risk of a significant inflammatory response using a monofilament suture such as Prolene rather than an absorbable braided multifilament suture such as Vicryl.

Introduction

The choice of suture for surgery is dependent on multiple factors, including characteristics of the patient, wound, tissue and anatomic location as well as suture material characteristics, suture availability and surgeon preference.

One consideration may be the degree of suture-induced inflammatory tissue reaction. This is of particular interest to oculoplastic surgeons since inflammation within the eyelid may directly affect the ocular surface, and in the worst cases, secondary scarring can result in eyelid malposition and/or damage to the cornea or conjunctiva. In addition, given the cosmetic importance of eyelids in self-perception and social interactions, the ideal suture would cause minimal acute inflammation enabling rapid recovery as well as minimal or no long-term visible scarring.¹

Two widely used sutures in oculoplastic surgery are polyglactin 910 (Vicryl; Ethicon, Somerville, New Jersey) and polypropylene (Prolene; Ethicon, Somerville, New Jersey). Vicryl has the advantage of being absorbable, thus reducing the need for suture removal. However, a significant foreign body granulomatous reaction (referred to clinically and in most papers as a granuloma) has been reported with the use of Vicryl sutures in various anatomical locations.²,⁸ In addition, there have been anecdotal reports of granulomas in some patients following surgery with Vicryl by British Oculoplastic Surgery Society members. Most of the published literature on suture-induced tissue reactions has either been carried out in animal models or has assessed clinical rather than histological scarring in humans.²,⁸ Our group developed an ethical model to study the histological reaction to implanted suture material in normal human cutaneous tissue.⁹ Following informed consent, patients with symptomatic involutional entropion were offered temporary everting sutures including one laterally placed study suture while waiting for definitive surgery. A month later, a definitive entropion procedure was carried out involving a wedge excision and external retractor plication; the study suture was included with the eyelid wedge excision allowing histological examination of the tissue reaction around the suture.

The ethical justification for carrying out two procedures was that everting sutures could be carried out as a clinic procedure on the day of presentation or soon after, providing immediate relief of symptoms compared with being placed on a waiting list for definitive surgery. Previous randomised controlled trials have found that everting sutures alone have a lower long-term success rate than surgery involving lid tightening and shortening for entropion.^{10 11} Patients were advised that everting sutures alone may be an effective treatment for some patients, but long-term success is significantly higher with a lid tightening procedure.

Using this model, we previously found that Vicryl sutures elicit a significant foreign body inflammatory response, which is proportional to the suture gauge.⁹ However, as far as the authors are aware, the difference between the histopathological reactions induced by Prolene and Vicryl sutures has not been studied. Our aim was to conduct a prospective randomised study to compare the histopathological reaction and clinical scarring, resulting from Prolene and Vicryl sutures.

Methods

This prospective, randomised, triple-masked clinical study was carried out across two sites (University Hospitals Dorset and Imperial College Healthcare NHS Trust) between 2021 and 2024. The research adhered to the tenets of the Declaration of Helsinki, and the trial has been registered (Identifier NCT04724512, www.clinicaltrials.gov).

Consecutive patients aged over 18 years referred to oculoplastic clinics with involutional entropion and significant horizontal lid laxity were eligible for inclusion. Exclusion criteria included a history of previous eyelid surgery, a systemic illness that could influence tissue healing response (uncontrolled diabetes, concurrent steroid use or Addison’s disease) or the inability to provide informed consent.

Due to the higher bleeding risk with clopidogrel, warfarin and direct oral anticoagulants, patients were required to stop these before both procedures if medically safe to do so. Otherwise, these patients were excluded from the study. Aspirin or non-steroidal anti-inflammatory drugs were stopped prior to definitive surgery only.

Most patients had been identified during triage and were sent the participant information sheet in advance. A full detailed oral and written explanation of the study was given by the clinician or research nurse at the clinic appointment.

Once informed consent was obtained, randomisation to the Prolene or Vicryl groups was carried out with the electronic database Castor, with variable block sizes: https://www.castoredc.com (EU Headquarters, 1076 EE Amsterdam, The Netherlands).

The sutures chosen were 6–0 gauge MPV489 Vicryl and MPP8697 Prolene, which have identical single-armed, 11 mm 3/8 circle reverse cutting multipass needles.

Surgical technique

The surgical technique for placement of everting sutures and definitive entropion repair has been described previously.⁹ In brief, after administration of local anaesthesia, patients received three everting sutures spaced evenly along the eyelid, with the lateral study suture placed 10 mm from the lateral commissure. One month later, they underwent definitive surgery. After local anaesthesia, a subciliary incision was made below the level of the everting sutures up to the position of the medial edge of the planned wedge excision. The lower lid retractors were plicated to the inferior edge of the tarsal plate using two or three interrupted sutures of the study suture material. A full-thickness wedge encompassing the lateral suture was excised and sent for histological analysis. The lid margin and subciliary incision were closed using the study suture (continuous suture for skin closure). Patients were prescribed carbomer gel three times a day and paraffin ointment at night for 2 weeks.

Histological assessment

Specimens were fixed in formalin. Two or three horizontal ‘bread slices’ were taken and processed overnight. A mean of six H&E levels was taken. The diameter of tissue response (TRD) and the diameter of the suture (SD) were measured at the same place. The latter was subtracted from the former to assess the diameter of the inflammatory reaction (IRD), that is, TRD−SD=IRD. If the suture had fallen out but it was clear where it had been from the cavity left behind with a surrounding inflammatory reaction, the measurements could still be taken: the size of the cavity where the suture had been was taken as SD. In contrast, in specimens where the pathologists could not identify any suture or suture cavity, the specimen could not be analysed.

To minimise the risk of bias from oblique sectioning, measurements were taken at six levels and the smallest diameter recorded was used for mean diameter calculation. The nature of the inflammation and degree of fibrous scarring were assessed at four points immediately around the suture. This included semiquantitative assessment as to the number of neutrophils, lymphocytes, plasma cells, macrophages, giant cells, fibrosis and necrosis.

Study outcomes

The primary study outcome was the diameter of the histological inflammatory reaction (IRD) measured on horizontal sections where possible over six levels.

The secondary outcomes were subdivided into histological and clinical outcomes. The degree of inflammation was measured using the histological scale from the ISO standards 10933-6:2016, which was designed to characterise the history and evolution of the local tissue response after implantation of a biomaterial.

At the 2-week postoperative visit, clinical assessment was carried out by a masked clinician following suture removal by a specialist nurse. The clinical scarring was assessed by the patient and a masked clinician at 2 weeks and 6 months using the validated Patient and Observer Scaring Assessment Scale (POSAS) Linear scar V.3.0 (posas.nl) as well as with standardised clinical photographs. Each question in the POSAS contributed one point to the overall score giving a total sum score. Two further independent masked evaluators assessed the photographic images for the quality of scarring using the scale invisible (excellent), mildly visible (satisfactory), moderately visible (poor) and highly visible (very poor).

Statistical analysis

The sample size was calculated using the results of Cartmill et al’s study, on which this study design was based.⁹ Using their histological outcome measure of ‘granuloma diameter’, the effect size was calculated to be 1.3. Since we were expecting less inflammation with Prolene than Vicryl, the only suture material used in the original study, we reduced the effect size to a more conservative 1.0. With a two-tailed test, power of 80% and alpha of 0.05, the suggested sample size was 17 in each group. This was increased by 20% to allow for patients excluded or lost to follow-up resulting in a total of 44 patients (22 participants in each arm). Data were analysed using a t-test for the primary endpoint and Wilcoxon rank sum tests for the remainder.

Results

Figure 1 shows the Consort diagram.

Figure 1. Consort diagram. POSAS, Patient and Observer Scaring Assessment Scale.

Table 1 shows the demographic characteristics.

Table 1. Demographic characteristics.

	Prolene	Vicryl
Age—mean (range)	76 (65–88)	78 (63–87)
Gender	Male 16	Male 13
	Female 5	Female 6
Ethnicity	Caucasian 18	Caucasian 18
	Black 0	Black 0
	Asian 0	Asian 0
	Hispanic 0	Hispanic 0
	Other 3	Other 1
Antiplatelet or anticoagulants	No 17	No 17
	Yes 4	Yes 2

Table 2 shows the key histological results. There was an approximately 50% larger mean diameter of inflammation around the Vicryl than Prolene sutures, and the median histological grade sum (a summative measure of histological inflammation) was 25% higher around the Vicryl than the Prolene, but this did not reach statistical significance; the postulated effect size was 1.0 and the observed effect size was 0.51 (95% CI −0.22 to 1.21). Unfortunately, no histological analysis was possible in 11 out of 43 specimens received as no suture or suture cavity was present, which considerably reduced the power of the study. There was a median of one giant cell (a marker of foreign body-induced inflammation) around the Prolene and four around the Vicryl; this reached statistical significance (p=0.04). Figure 2 shows an example of the inflammatory reaction diameter around the two sutures.

Table 2. Primary histological outcomes.

	Prolene (n=17)	Vicryl (n=15)	P value*
Mean (SD) inflammatory reaction diameter in mm	0.39 (0.24)	0.56 (0.41)	0.19
Median (IQR) histological grade sum*	32 (22–37)	40 (27–46)	0.23
Median (IQR) number of giant cells*	1 (0–3)	4 (3-8)	0.04

^*Wilcoxon Rank Sum test.

Figure 2. Slides showing the typical histological reaction around sutures. Top: Prolene suture; bottom: Vicryl suture.

One patient in the Vicryl group developed a large suture-related clinical granuloma on the conjunctival surface of the eyelid 13 weeks following the definitive surgery (figure 3). Histology showed a nodule of immature fibrous connective tissue rich in capillaries with superficial acute inflammation. There were no associated symptoms or signs of lid infection. Surgical excision was required. There were no other complications in either group.

Figure 3. Granuloma in patient who received Vicryl sutures. Top: 2 weeks postdefinitive surgery; middle: Suture granuloma 13 weeks postdefinitive surgery; bottom: 6 months postdefinitive surgery (granuloma had been excised).

Table 3 shows the clinical scarring on the POSAS. At 2 weeks, the median sum score was higher, indicating a slightly worse scar, for the Vicryl compared with the Prolene when graded by both the patient and the masked observers, but this was not statistically significant, and by 6 months, the average score was 0 for all groups. Two masked oculoplastic surgeons with no other involvement in the study assessed the scars from clinical photographs, and there was no significant difference in grades between the two sutures at either time point.

Table 3. Clinical outcome scores on the POSAS.

POSAS	Prolene MSS (IQR) (n)	Vicryl MSS (IQR) (n)
Patient scale at 2 weeks	3 (3–12) (21)	4 (3–7) (19)
Patient scale at 6 months	0 (0–2) (21)	0 (0–2) (19)
Observer scale at 2 weeks	4 (3–5) (20)	5 (4–7) (16)
Observer scale at 6 months	0 (0–1) (20)	0 (0–2) (19)

MSS, Median of the sum scores (i.e. the median of the total POSAS scores from patients in that group); n, number of participants; POSAS, Patient and Observer Scaring Assessment Scale.

Discussion

Sutures are ubiquitous in surgical procedures and surgeons should be aware of the risk of inflammation during wound healing, which may result in dehiscence, clinical granulomas, infection and scarring. The impact of scarring in oculoplastic eyelid surgery is particularly important due to the risk of compromise or trauma to the ocular surface. Facial and eyelid scarring is highly visible and can have a significant psychosocial impact.¹

Oculoplastic surgeons use a variety of suture materials including polyester, nylon and silk, but polyglactin 910 (Vicryl) and polypropylene (Prolene) are in widespread use so were selected for this comparative study.

The ideal absorbable suture would be one that is biologically inert, causes no foreign body reaction and only loses its tensile strength as the scar tissue strengthens. Unfortunately, these properties are in direct conflict, since by definition an inert suture would not absorb. There is little information in the literature as to how Vicryl sutures absorb other than ‘by hydrolysis’, which merely implies that the material breaks down with one of the reactants being water. Intuitively, the chemical breakdown of a compound may be expected to cause inflammation, and this has been borne out in various clinical studies discussed below.

Whether absorbable or not, some materials are inherently more proinflammatory. Javed et al carried out a review of studies looking at tissue reactions to suture materials in oral surgery, the majority in animal models.⁸ Various suture materials were assessed, most commonly silk, nylon, gut and polyester; only one looked at polyglactin and none assessed polypropylene. Results were mixed, but overall, they found that silk elicits the greatest tissue response with delayed wound healing whereas nylon caused minimal tissue response.

There are a number of case reports and small series describing foreign body granulomas around Vicryl sutures causing delayed wound healing and/or dehiscence.^{2 6 12} Others have found higher rates of presumed infection²,⁶⁸ or reported clinical granulomas mimicking infection.⁷ These lesions have been described clinically as suture granulomas, pyogenic granulomas or conjunctival granulomas and may represent a foreign-body granulomatous inflammation, an imbalance in angiogenesis regulation¹³ or other error in the proliferative stage of wound healing¹⁴ precipitated by the suture and wound healing. There was a single case of a large nodular conjunctival granuloma seen at 13 weeks post-definitive surgery in the Vicryl arm of this study (figure 3). Histopathology showed inflamed highly vascular immature fibrous connective tissue consistent with a proliferative fibrovascular response to the suture.

An important question for surgeons is whether using Prolene or Vicryl sutures affects the risks of inflammation, suture granuloma formation and significant scarring. This study found that clinical scarring in eyelids is generally excellent with both suture types. The raw data suggest more inflammation and histological scarring around Vicryl sutures as they dissolve. This did not reach statistical significance for most outcomes, possibly due to the large proportion of specimens which could not be analysed histologically. Despite this, there was a statistically significantly greater number of foreign body giant cells around the Vicryl than the Prolene. Histologically, giant cells are involved in the response to a foreign body and their role includes engulfing and extricating foreign material, thus, it makes sense that more giant cells aggregate around a dissolving Vicryl suture than a non-dissolvable Prolene suture.

For logistical and ethical reasons, most previous comparative studies looking at scarring have been in animal models or considered only clinical rather than histological scarring. A study in a rabbit model found an increased tissue reactivity and higher number of giant cells in Vicryl compared with Prolene.¹⁵ Vicryl has been shown to cause a significantly greater residual wound inflammatory response compared with Prolene postcarpel tunnel decompression surgery¹⁶ and cleft lip surgery.¹⁷ Bozan and Dizdar studied outcomes of 6–0 Prolene and 6–0 Vicryl Rapide in 70 patients undergoing open rhinoplasty surgery at two centres, comparing scar assessment scales and clinical photographs.¹⁸ It is not clear on what basis patients were allocated to each group or whether it was according to surgeon preference. They found less scarring with Prolene than Vicryl, but it was not statistically significant. Parell and Becker compared the scars following facial flap repair in 41 patients using 5–0 Prolene for half the wound closure and 5–0 Vicryl for the other half.¹⁹ They found no clinical difference in scarring between the suture types at 6 months.

In a recent retrospective study involving 1079 eyes of 891 patients undergoing the lateral tarsal strip (LTS) procedure, Moledina et al found foreign body granulomas were more commonly seen with Vicryl than Prolene (2.2% vs 0.7%, p=0.03).⁴ In another retrospective study, Yoon et al reviewed 758 LTS procedures using a variety of suture materials and found the highest rate of clinical ‘granulomas’ at 6% in polyester (Mersilene) and 3% in polyester+PTFE (Polydek), which are both non-absorbable braided sutures.³ Interestingly, there were no granulomas seen with Vicryl. There was one clinical granuloma seen with Prolene in 114 procedures (0.9%), which is similar to the incidence found by Moledina (0.7%), and no granulomas with all other monofilament sutures that were studied.

The greater tissue reaction seen with Vicryl may be due to an upregulation of proinflammatory relative to anti-inflammatory cytokines reflected in the IL-1β/IL-1RA ratio. The risk of granuloma formation may be less with Prolene due to a weaker inflammatory response.²⁰

In terms of clinical outcomes, most of our patients fared well, with good or excellent scars assessed subjectively and objectively, regardless of suture type. On the POSAS, the mean score was higher, indicating more noticeable scarring, for Vicryl than Prolene when assessed by both patient and an independent observer at 2 weeks, but this did not reach statistical significance and by 6 months, the scores averaged zero (excellent scarring) for both groups. It would be interesting to measure outcomes at multiple points postoperatively and observe the clinical scarring as the inflammation resolves in the months following surgery. However, multiple extra clinic visits are likely to make recruitment harder and increase loss to follow-up. With the rise of smartphone applications that can capture and analyse high-quality photographs, there may be future opportunities for patients to submit multiple data points, although any such tool would require rigorous statistical validation testing as well as in-built mechanisms for standardisation of image capture, subject illumination and image analysis.

The power of this study was limited by the inability to locate the suture histologically in 11 out of 43 specimens received and some loss to follow-up. In addition, the demographic homogeneity means that the results may not be applicable in younger or more pigmented skin, since ethnicity and age are known to be risk factors for hypertrophic or keloid scarring.²¹ For example, 5%–10% of Africans may be affected by keloids, compared with 0.1%–1% of Asians and <1% Europeans and North Americans.²² Pregnancy and adolescence are also risk factors for worse scarring, with most hypertrophic scars occurring in people aged 11–30 years.^{21 23} Further research comparing suture reactions in these higher risk groups would be useful.

All sutures in both study arms were removed at the 2-week visit for standardisation and to allow the observers to be blinded to the suture type when carrying out POSAS assessment. Our normal practice is to remove Vicryl sutures postoperatively to minimise any suture-induced inflammation, but a second clinic visit can be avoided by leaving the sutures to absorb. An interesting avenue for research would be to study the wound healing and inflammation with sutures left in situ.

This study was consistent with other published data showing that braided sutures, particularly of non-resorbable materials, appear to cause more histological inflammation at 1 month postoperatively than monofilament sutures, although further research is required. It would be logical to avoid future use of sutures such as Vicryl in patients in whom granulomas have occurred or in patients with a personal or family history of hypertrophic or keloid scarring; monofilament non-absorbable sutures may reduce this risk. Overall, the clinical scarring was very good with both Prolene and Vicryl, and patient satisfaction was high.

Data availability statement

Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information.