OUP user menu

Bacterial biofilm on monofilament suture and porcine xenograft after inguinal herniorrhaphy

Sandeep Kathju, Laura Nistico, Leslie-Ann Lasko, Paul Stoodley
DOI: http://dx.doi.org/10.1111/j.1574-695X.2010.00691.x 405-409 First published online: 1 August 2010


Bacterial biofilms have been implicated in multiple clinical scenarios involving infection of implanted foreign bodies, but have been little studied after hernia repair. We now report a case of revision inguinal herniorrhaphy complicated by chronic pain at the operated site without any external indication of infection. Computed tomographic imaging revealed a contrast-enhancing process in the left groin. Subsequent surgical exploration found an inflammatory focus centered on implanted porcine xenograft material and nonabsorbable monofilament sutures placed at the previous surgery. Confocal microscopic examination of these materials with Live/Dead staining demonstrated abundant viable bacteria in biofilm configuration. The removal of these materials and direct closure of the recurrent hernia defect eliminated the infection and resolved the patient's complaints. These results demonstrate that implanted monofilament suture and xenograft material can provide the substratum for a chronic biofilm infection.

  • biofilm
  • suture
  • confocal microscopy
  • inguinal hernia
  • xenograft


Bacterial biofilms are communities of microorganisms that can attach to both abiotic and biological (e.g. mucosal) surfaces in humans (Hall-Stoodley et al., 2004). Biofilms have been noted to be contributing or causative factors in a wide variety of infectious processes, especially those associated with implanted foreign bodies, including orthopedic prostheses (Stoodley et al., 2005, 2008), neurosurgical drains and shunts (Stoodley et al., 2010), vascular and peritoneal catheters (Gorman et al., 1994), etc. Biofilm bacteria differ from their planktonic counterparts in significant ways: they have a much higher (by orders of magnitude) resistance to conventional antibiotics, they are able to evade host humoral and cellular immunological mechanisms [largely through their encapsulating matrix of extracellular polymeric substance (EPS)], and they can frequently prove difficult to detect using standard clinical microbiological culture techniques (Hall-Stoodley et al., 2004). These properties render the diagnosis and treatment of infections with a biofilm etiology problematic (Hall-Stoodley & Stoodley, 2009).

Although biofilms have been observed on numerous types of prosthetic surfaces, there has thus far been comparatively little examination of the materials used in hernia repairs. Herniorrhaphy, the surgical repair of hernias, is usually accomplished using suture material to close the hernia defect directly, or through the use of some type of an interpositional surgical mesh. More recently, surgeons have begun to use so-called ‘biological meshes,’ that is, acellular matrices derived from human or animal donor tissues, as materials with which to reconstruct abdominal wall hernia defects (Hiles et al., 2009). There has been a hope in the surgical community that these materials may prove more biocompatible and ultimately less prone to complications than standard mesh materials such as polypropylene or polytetrafluoroethylene.

Inguinal herniorrhaphy is one of the most common surgical procedures in the United States, with some 500 000 cases performed annually. We now report a case where a patient with recurrent hernia, after two separate bilateral inguinal herniorrhaphy attempts, was reconstructed a third time with a porcine xenograft. The patient subsequently first developed a chronic draining wound in the right groin, which required surgical debridement and closure, and then 15 months later, developed chronic pain in the left groin. Subsequent evaluation and exploration of the left groin site demonstrated a live bacterial biofilm resident on the implanted xenograft and suture material. To our knowledge, this is the first demonstration of a bacterial biofilm on an implanted xenograft and on monofilament suture in the abdominal wall, and the first documentation of a biofilm as a complication of inguinal herniorrhaphy.

Case study

A 47-year-old man presented with a complicated history of repeated bilateral inguinal hernia surgeries. Inguinal hernias on both sides had initially been repaired some 23 years back prior using an external approach, but without the use of surgical mesh. One year later, the patient underwent a second surgery bilaterally as both hernias had recurred and were painful. The second repair was performed laparoscopically and polypropylene mesh implants were placed. Twenty-one years later, the patient once again underwent bilateral surgery for bilateral recurrent hernia. At this third procedure, performed via an external approach, the old mesh was reported to have been removed and the hernia defect was reconstructed with the placement of a porcine matrix xenograft (Surgisis).

Five months later, the patient presented to us with a chronic open draining wound in the right groin. The drainage was turbid, but not frankly purulent; the wound had been present for several months. He was not experiencing any fevers, chills, or other signs of systemic infection. He remained able to ambulate and function, but had some chronic pain and discomfort at the wound site itself. The left groin at this time was externally unremarkable, although the patient did complain of occasional discomfort at that site as well.

The patient was taken to surgery for exploration and debridement of the right inguinal wound. A 3-cm draining sinus aperture was excised; multiple polypropylene sutures were removed. A mass of material with the consistency of a wet tissue paper was debrided from about the abdominal wall fascia. Although it had been reported that the old polypropylene mesh had been removed, a small piece of retained mesh was discovered and explanted. After copious irrigation, the fascia was repaired directly with absorbable suture, and the skin was closed over a suction drain. Standard microbiological culture from the turbid wound fluid returned positive for Serratia marcescens, α-hemolytic streptococci, and corynebacterium. Interestingly, culture of the debrided deep tissue, likely Surgisis remnant, showed no growth at 5 days. The patient was treated postoperatively with a short course of oral ciprofloxacin, and has remained free of complaint or finding in the right groin since.

Fifteen months after his right groin exploration, the patient again presented to us with complaints of pain in his left inguinal area. This pain had become constant, and had persisted for several months. After repeated complaints from the patient, despite the absence of any generalized signs such as fever, and without any external signs of infection or recurrent hernia (see Fig. 1a), his primary physician had ordered an abdominal ultrasound, which demonstrated an abdominal wall fluid collection. A subsequent computed tomographic (CT) scan of his abdomen and pelvis revealed ‘a small superficial fluid collection measuring 4.4 × 1.6 cm. Some low attenuation fluid is also seen tracking into the lower anterior pelvic wall musculature’ (Fig. 1b). This striking radiologic finding was strong evidence for a chronic and localized inflammatory process, and the patient underwent left groin exploration.

Figure 1

Preoperative evaluation. (a) Photograph of patient's inguinal areas. Note the absence of any significant edema or erythema on the left side, which features only a well-healed scar. (A similar well-healed scar on the right is more recent.) (b) CT scan image of the inguinal areas. The blue arrow indicates a subcutaneous fluid collection on the patient's left side.

At surgery, the patient was noted again to have multiple retained polypropylene sutures, all of which were removed, and some of which were preserved for confocal microscopic examination. Just superficial to the abdominal wall fascia proper a small collection of turbid fluid was opened — this was sent for culture, and was observed to emerge from deeper in the fascia as noted in the CT report. On opening the fascia repair more widely, more cloudy (not purulent) fluid was released and a large mass of material was noted within the inguinal canal itself. This material (as on the right side previously) had the consistency of a wet tissue paper; it was clearly not incorporated or vascularized, and was removed piecemeal with a forceps until no trace remained. This material clearly represented the Surgisis implant that had been placed at a previous surgery. Finally, a hard mass of retained polypropylene mesh was discovered and was explanted. After irrigation of the surgical site, the fascia was repaired directly with absorbable sutures, and the skin was closed over a suction drain.

The patient's history and our previous experience in the right groin led us to strongly suspect a biofilm etiology to his disease in the left groin, and we therefore took multiple specimens to examine both culturally and with confocal microscopy (CM). Four separate specimens of the explanted xenograft were sent for culture, as well as a piece of the explanted polypropylene mesh. Multiple specimens were also preserved for CM. Only a single specimen of the xenograft returned positive for culture, yielding coagulase-negative staphylococci sensitive to cephalosporins; all other specimens showed no growth at 5 days. The patient was therefore treated postoperatively with a brief course of cephalexin, and has remained pain-free and complaint-free since (9 months at the time of writing).

Materials and methods

In addition to standard microbiological culture, we examined explanted suture and tissue specimens using CM to determine whether bacterial biofilms were present. Specimens were prepared as described previously (Kathju et al., 2009a, b). Briefly, suture and tissue recovered at surgery were placed in Hanks balanced salt solution (HBSS) and placed on wet ice, directly after removal. After rinsing in the HBSS (to remove unattached bacteria) and blotting on sterile paper, specimens were mounted on the bottom of a 35-mm Petri plate on partially solidified agar (Kathju et al., 2009a, b). Specimens were stained for viability assessment using Molecular Probes BacLight Live/Dead kit (Molecular Probes, Eugene, OR). The BacLight kit consists of two nucleic acid stains, Syto9 (green), which enters all bacteria, and propidium iodide (red), which can only enter bacteria with porous cell walls. Once inside the bacteria, the propidium iodide suppresses the Syto9 fluorescence so that live bacteria appear green, whereas dead or damaged cells appear red. In some cases, bacteria stain with both dyes and appear yellow — these have been interpreted as live, but nonculturable. The nuclei of human cells also take up these nucleic acid stains, but rapidly turn red. They are readily distinguished from bacteria on the basis of size and morphology. In addition, these stains have been used to stain extracellular bacterial DNA (eDNA), which is commonly found in the EPS and appears as a diffuse staining surrounding the bacterial cells (Böckelmann et al., 2006; Thomas et al., 2008).

Fully hydrated specimens were then imaged by CM using a Leica DM RXE upright microscope attached to a TCS SP2 AOBS confocal system (Leica Microsystems, Exton, PA) using either a × 20 air objective or a × 63 long working distance water immersion objective. Live (green) and ‘dead’ (red) bacteria were imaged using 488 and 594 nm lasers; the suture and xenograft were imaged using reflected CM (blue) and bright-field microscopy (gray).

Results and discussion

Examination of one of the pieces of explanted Surgisis xenograft by CM showed heterogeneously distributed patches of live and dead bacteria and evidence of associated eDNA attached to the xenograft material (Fig. 2a). These organisms had a primarily coccal appearance, consistent with the solitary finding by culture of staphylococci. Interestingly, only one of the four specimens yielded a positive culture result, illustrating the inherent difficulty in detecting biofilm infections and making the case for multiple specimens to be sent for clinical culture, as well as the utility of using independent culture-free methods. Biofilms are commonly patchy on surfaces and this heterogeneity might partly explain the inconsistency in culture data.

Figure 2

Confocal micrographs of a biofilm attached to a xenograft and associated tissue and suture material. (a) Biofilm adherent to a xenograft (blue) consisting of patches of live (green and examples indicated by white arrows) and dead (red and example indicated by black arrow) cocci. The slightly clouded appearance of the green nucleic acid stain around the bacterial cells (green arrows) is suggestive of eDNA, a component of EPS that is produced by biofilm bacteria and characteristically forms a structural and protective matrix for the biofilm. Scale bar=20 µm. (b) Live cocci (green arrow) and human cells (shown by red staining of the nuclei and indicated by the black arrow) attached to the xenograft (blue). Scale bar=20 µm. (c) Low-magnification view showing a bacterial biofilm attached to a suture knot. The knot was imaged by bright-field microscopy and appears as blue/gray. A significant accumulation of viable bacteria (green) was adherent in the crevices between individual suture knots. Dual staining with green and red resulted in a yellow appearance in some places. Scale bar=300 µm. (d) High magnification of a biofilm cluster taken between the knots of a suture. The biofilm cluster consisted of live (green and indicated by white arrows) and dead (red and indicated by black arrow) cocci. Diffuse red and green staining in between the cocci suggests eDNA in the EPS matrix. Scale bar=10 µm.

Examination of explanted suture material also showed evidence of attached and viable biofilm bacteria (Fig. 2c and d). Bright green staining, signifying living bacterial cells, was noted to accumulate in the interstices of the knots of the sutures (Fig. 2c). A higher magnification in these areas revealed biofilm clusters consisting of live and dead cocci surrounded by EPS containing eDNA (Fig. 2d). Although it has long been recognized that monofilament sutures may generally harbor fewer microorganisms than multifilament sutures (e.g. Osterberg & Blomstedt, 1979), these striking images show that the knotted area itself, unavoidable with any suture configuration, can provide an adequate microenvironment in which biofilm may accumulate.

In light of the above findings, the patient's clinical history is thrown into sharper relief and is consistent with the biofilm paradigm, fulfilling all of Parsek and Singh's suggested criteria for the clinical diagnosis of a biofilm infectious process (Parsek & Singh, 2003). These include: ‘(a) The infecting bacteria were adherent to some substratum or are surface associated’– clearly, in this case, bacteria were adherent to the xenograft and to the sutures, as demonstrated by CM. ‘(b) Direct examination of infected tissue shows bacteria living in cell clusters, or microcolonies, encased in an extracellular matrix’– again, our confocal results show just this. ‘(c) The infection is generally confined to a particular location. Although dissemination may occur, it is a secondary phenomenon’– the present case is a particularly good example of this. On the patient's left side, despite months of pain (now understood to be the result of an infectious process), no systemic spread occurred; nor was the infection visible externally. We suspect the patient likely had a similar biofilm-elicited process on the right side that did progress to development of a frank draining sinus, but even this remained a localized process, with no cellulitic or systemic spread over months. ‘(d) The infection is difficult or impossible to eradicate with antibiotics despite the fact that the responsible organisms are susceptible to killing in the planktonic state’– this characteristic was never tested in this patient. Because we suspected a biofilm etiology to the patient's infections, we relied on surgical exploration rather than antibiosis as the mainstay of intervention. Antibiotics were only administered adjuvantly, after the substrata hosting the biofilms were surgically removed.

This case also conforms to other typical features of biofilm infections. Despite numerous bacteria present and visible on explanted xenograft tissues, laboratory culture was positive in only one instance, consistent with the difficulty in recovering biofilm organisms using standard microbiological cultural techniques. The chronicity of the complaints, both on the left and on the right, is also usually a hallmark of biofilm processes, although acute exacerbations can occur, thought to derive from the known tendency of biofilms to shed populations of bacteria, some of which may be ‘planktonic’ and give rise to a more fulminant infectious course.

In both the right and the left inguinal regions, infectious complications manifested considerably after (not shortly after) the preceding surgical procedure. It therefore remains an open question as to whether the microorganisms responsible were present at the time of surgery, with a delayed presentation, or gained access to the operated fields subsequently, for example by hematogenous spread to the site. We note, however, that the two sides yielded significantly different microorganisms by final cultural analysis, which, together with the temporal interval between episodes, suggests that infection on each side derived from a separate source.

Suture material as a substrate for clinical biofilm-based infection has only been described infrequently, with most early reports coming from ocular infections. We have noted previously the role of suture-based biofilm in infections of the abdominal wall in patients who had undergone a gastric bypass surgery (Kathju et al., 2009a, b); in these previous cases, the involved sutures were all multifilament. The present report demonstrates that even monofilament suture can become a nidus for postsurgical biofilm infection, and that this can occur in the nonbariatric surgical population.

This report is also the first, to our knowledge, to document the growth of a biofilm on implanted xenograft material, and the first to document biofilm in the aftermath of inguinal herniorrhaphy. ‘Biological meshes,’ composed of organic matrices derived from both human and animal tissue sources, are becoming increasingly common in abdominal wall reconstruction. The material used in this patient is derived from porcine small intestine submucosa, and has been used in patients for diaphragmatic, perineal, ventral as well as inguinal herniorrhaphy. Reports on its success appear mixed: for example one study examining Surgisis in inguinal hernia repair noted decreased pain on coughing and movement postsurgery compared with polypropylene (Ansaloni et al., 2009). In contrast, another report compared Surgisis with Alloderm (a human acellular dermal graft) in ventral herniorrhaphy, and found postoperative pain and seroma to be significant problems with Surgisis, with seroma occurring in 13/41 patients, more commonly when a nonperforated formulation of Surgisis was used (Gupta et al., 2006). Our own findings reported here, although occurring after inguinal herniorrhaphy, are more consistent with the latter study, with the cloudy but nonpurulent fluid we observed at surgery qualifying as an infected seroma; this also suggests that a biofilm may form on Surgisis after ventral herniorrhaphy, although direct evidence is lacking. It may also be that biofilms are capable of forming on human allogeneic (as opposed to xenogenic) biological mesh implants after herniorrhaphy — further investigation will be required to address this question.

As with other biofilm infections of implanted devices, surgical explantation of the foreign body substrate (Surgisis and suture) resolved the clinical infection. It is notable that in this patient the only presenting complaint in the left groin was pain. Persistent postsurgical pain is a recognized complication of inguinal herniorrhaphy, and may be attributed to musculoskeletal causes, or to trauma or constrictive scarring of local nerves (Loos et al., 2009). Our observations here suggest that, in the case of patients with implanted foreign bodies from herniorrhaphy, a low-grade chronic infection of biofilm etiology should also be kept in mind as a potential source of ongoing pain.


We gratefully acknowledge the assistance of Ms Mary O'Toole in the preparation of this manuscript, and support from the Allegheny-Singer Research Institute.


  • Editor: John Costerton


View Abstract