Surgical mesh and cancer risk

There is worldwide controversy about the effects of surgical mesh; particularly following adverse events associated with its use for repair of pelvic organ prolapse in women. ¹ This led to the USA Food and Drug Administration (USFDA) issuing a Public Health Notification in 2008. ² These and similar actions in other jurisdictions began a debate on risk versus benefit of surgical procedures using surgical mesh as opposed to traditional suture-based repairs. Interest has focused on urogyecological procedures where there is a large body of data showing the problems associated with mesh use, including erosion through the vaginal epithelium and vaginal scarring. ^3,4 The formation of scar tissue involves the deposition of collagen from fibroblasts which are stimulated to divide by transforming growth factor (TGF)-β₂ in response to tissue damage and the resulting inflammatory response. ⁵

The reported side effect of surgical mesh use are predominantly medium-term (i.e. 1–5 y); these follow up studies focused on quality of life, pain, return of symptoms, dysparuria, etc. and thus are not relevant to tumorigenesis. ^6,7 Indeed, it is very unlikely that tumours would manifest in a 5-year timescale. Few data are available to facilitate robust assessment of long-term (>10 y) risks, and for other surgical procedures utilising surgical mesh, including abdominal hernia repair (for which surgical mesh was first used in the 1950s). ² Long-term hernia repair follow-up studies focused on hernia recurrence, scar pain and cosmetic effects and concluded that mesh was superior to conventional suture-based surgical procedures based on these risk factors. ^8,9 A 17-year follow-up study involved palpation to check for ‘signs of erosion, extrusion, fistulation and other tissue reactions’ after tension-free vaginal tape (TVT) procedures did not report any tumours. ¹⁰ However, it is likely that the palpation method would not detect small tumours, and the authors make the point that the amount of mesh implanted when TVT is used is much less than is used for prolapse surgery. ¹⁰

It remains possible that the inflammatory and cell proliferative responses resulting from surgical mesh implantation could lead to cellular responses causing hyperplasia and perhaps neoplasia in the long-term by a non-genotoxic mechanism of carcinogenesis. ¹¹ Indeed, it has been suggested more recently that induction of cell proliferation might be a useful indicator of non-genotoxic carcinogenic risk. ¹² The link between implanted foreign bodies, cell proliferation and induction of tumours was first appreciated in, largely forgotten, studies carried out in the 1940s. Despite their historical obscurity these studies formed the basis of our developing understanding of tissue responses to persistent foreign implants, and initiated our thinking about non-genotoxic mechanisms of carcinogenesis. ¹¹

In 1942 Turner implanted a Bakelite disc subcutaneously into a rat and found a sarcoma 23 months after implantation. ¹³ Surprised by this, he carried out a follow-up study and found 3/8 (37.5%) implanted rats developed sarcomas within 18 months. ¹³ In 1948 Oppenheimer and Stout implanted plastic films around rats’ kidneys and found 8/23 (35%) animals developed sarcomas within 11 months. ¹⁴ Then in 1952, Zolinger repeated the plastic film work and found 8/21 (38%) sarcomas. ¹⁵ There is a remarkable agreement between the incidences of sarcomas in these plastic implant experiments.

Originally it was thought that a carcinogenic impurity leaching from the plastics was responsible for the sarcomas (i.e. chemical carcinogenesis), but further experiments in animals disproved this and showed that the implanted plastic itself was responsible. ¹⁶ Later, it was shown that it was not a particular plastic that resulted in tumours, but the physical nature of the plastic implant itself. ¹⁷

Studies with perforated plastic implants showed that more perforations meant fewer sarcomas. ¹⁷ This reinforced the suspicion that the physical nature rather than the chemistry of the implant was responsible for implant-related carcinogenesis. It is possible that this is a non-genotoxic mechanism of carcinogenesis related to an inflammatory response elicited by contact between the plastic implant and the animals’ tissues.

It is interesting to relate these largely forgotten findings to surgical mesh. Non-absorbable synthetic surgical meshes are usually made of polypropylene, polyester or polytetrafluoroethene ^2,18 and so are rigid plastics akin to those used in the early experiments outlined above. Their implantation might initiate local damage with a concomitant inflammatory response leading to TGF-β₂ release, cell proliferation and scar tissue formation. The cell proliferation response might present the signal for non-genotoxic carcinogenesis (Figure 1). Since the physical form of surgical mesh is akin to a multi-perforated implant this would likely reduce, but not mitigate, its potential to induce tumours. These changes are exactly the indicators that are now believed to signal non-genotoxic carcinogenesis, and have been suggested as early screening indicators for non-genotoxic carcinogens. ¹²

Figure 1.

Schematic representation of a plastic implant-induced inflammatory response leading to TGF-β₂ release which stimulates fibroblast proliferation and collagen deposition to form scar tissue. A generalised cell proliferation might lead to hyperplasia and provide the DNA replicative target for spontaneous mutations which might lead to neoplasia.

Studies on the pathology of tissues surrounding surgical mesh implants for hernia repair surgery found that approximately 23 months after implantation they had ‘a persisting inflammatory proliferative foreign body reaction with increased cell turnover’. ¹⁸ This is just the response that Clayson described in 1962 in rats implanted with plastic films – he describes ‘the normal foreign-body reaction of the tissue’ that led to ‘rearrangement of the fibroblasts to parallel the surface of the film’ with later fibroblast activity and formation of granulation tissue. The persistent inflammatory response could be the underlying basis of non-genotoxic carcinogenesis. Indeed, Klosterhalfen recognises the risk that the consequences of the mesh-induced chronic wound could lead to malignant transformation. ¹⁸ Mally and Chipman went further by suggesting that cell proliferation might be a reliable screen for non-genotoxic carcinogens. ¹² If surgical meshes were subjected to such a test they would likely fail.

More recently, it has been suggested that the balance between apoptosis and cell proliferation is a key determinant of non-genotoxic carcinogenesis, and that ‘signature’ omic effects might be used to indicate non-genotoxic carcinogens in a screening context. In this respect gene expression related to peroxisome proliferation is interesting because peroxisome proliferation is associated with non-genotoxic carcinogenesis. Peroxisome proliferator-activated receptor alpha (PPARα) has been suggested as a target gene because PPARα is involved with cell cycle control leading to cell proliferation. ¹⁹ Interestingly, whole genome expression analysis of peri-implant tissue in maxillofacial surgery has shown upregulation of PPARβ. ²⁰ While maxillofacial implants are not plastic, these results show a gene response to an implanted foreign body which could indicate the potential for non-genotoxic carcinogenesis.

There is controversy about the human relevance of sarcomas in rats in toxicology studies. However, in the context of plastic implants, sarcomas are the neoplastic endpoints of an inflammatory, cell proliferative tissue reaction to the implant. It is interesting that in one study in humans implanted with surgical mesh the same sequence of events described for plastic implanted rats was seen. ¹⁸ Mesh-based surgical implants in humans have not been followed up for long enough to determine whether local neoplastic changes result from mesh-induced chronic tissue reactions. Ellington and Richter reviewed the complications and risks associated with the use of surgical mesh in pelvic organ prolapse surgery in 2013. They concluded that physicians should ‘continue to review the literature and report their own outcomes’ relating to complications. ⁴ In addition, there is a need for long-term (>10 years) follow-up of mesh-implanted patients to determine whether mesh-initiated proliferative responses might leads to tumours.