The midurethral sling procedure has emerged as the gold standard surgery for stress urinary incontinence (SUI) since its development in the 1990s.1 However, worries regarding the use of polypropylene mesh for pelvic reconstructive surgery have expanded to include concerns regarding polypropylene's carcinogenic properties.2 The purported increased cancer risk in patients with a history of biomaterial implantation is largely based on animal research studies and case reports.3–7 In 2000, the International Agency for Research on Cancer determined that there was no evidence for carcinogenesis with the human use of synthetic implants.8 Thin, macroporous polypropylene mesh used for synthetic suburethral sling and prolapse repairs has negligible carcinogenic properties compared with solid implants used in animal studies.9 Thus, the carcinogenicity of macroporous polypropylene mesh used in pelvic reconstructive surgery is considered by many investigators to be extremely low because of the physical form of the most commonly used pelvic floor mesh implants. This conclusion is, however, poorly substantiated, which has given rise to uncertainty and increasing concern that polypropylene mesh may give rise to cancer in the long term, although no human evidence supports this notion.10 The purpose of this long-term observational study was to determine whether there is any association between the implantation of synthetic polypropylene mesh slings for the treatment of SUI and risk of cancer utilizing well-characterized databases supported by the Swedish Board of Health and Welfare.
MATERIALS AND METHODS
This cohort study used independently and prospectively collected data from nationwide health care registers supervised by the Swedish Board of Health and Welfare (http://www.socialstyrelsen.se/english). Register linkage based on the individually unique national registration number, assigned to all nationals at birth or immigration, was used to ascertain unambiguous information on exposures and outcomes. Exposed women were identified using the Swedish Inpatient Register, which contains data on individual hospital discharges including date and discharge diagnoses according to the International Classification of Diseases (ICD), 7th and 10th Revisions, as well as operation codes according to the Swedish Classification of Operations and Major Procedures. The register has a less than 1% yearly loss to registration and correct coding for surgical procedures is achieved in 98% of cases.11
Incident cases of primary cancer were identified in the Swedish Cancer Register. The register, which includes histologically verified incident cancers, is more than 95% complete and is uniformly classified according to ICD, 7th Revision. Using national registration numbers, all individual records were also linked with The Cause of Death Register, the Swedish Medical Birth Register, the Swedish Education Register, and the Population Register. The study was approved by the Research Ethics Committee at Karolinska Institutet, Stockholm, Sweden, and conforms to the STrengthening the Reporting of OBservational studies in Epidemiology (STROBE) guidelines for reporting observational studies (www.strobe-statement.org).
From the Population Register we identified all women older than 18 years of age between January 1, 1973, and December 31, 2009 (n=5,701,207). Based on the Inpatient Register, our exposure was defined as having a recorded midurethral sling procedure according to the Swedish Classification of Operations and Major Procedures (code LEG10) with day of surgery used as the starting point for time of exposure. Women were excluded from analysis if they had any primary cancer before entering the cohort (n=76,416), if they had other inconsistencies of their data (n=14,635), and if they emigrated out of Sweden before entering the cohort (n=224,970). We derived age at surgery from the Inpatient Register, information on parity from the Medical Birth Register, and information on educational level from the Swedish Education Register.
Within the Cancer Register, we identified as our outcome measure any diagnosis of primary cancer among exposed and unexposed women based on the ICD, 7th Revision. Cancer cases diagnosed within 1 year from exposure were censored to avoid detection bias. All women entered the observational period as unexposed on January 1, 1997, and contributed person-time as unexposed unless they underwent a midurethral sling procedure after which they contributed person-time as exposed until first occurrence of any cancer, death, emigration, or end of the observational period (December 31, 2009). The starting date for the observational period was chosen specifically because the first midurethral polypropylene sling (tension-free vaginal tape) was commercially introduced in 1997. The operation code LEG10 was introduced January 1, 1997, to specifically classify midurethral tape operations using synthetic mesh. Xenografts, muscle transposition, and autologous grafts do not fall under the LEG10 classification.
We calculated incidence rates of cancer as the number of cases per 100,000 person-years with 95% CIs based on the Poisson distribution. Cox proportional hazards models were used to estimate the hazard ratios (HRs) for individual primary cancers among the exposed compared with unexposed women and adjusted for age, calendar time, parity (categorized as: nulliparous; one to two childbirths; and greater than three childbirths) and educational level (categorized as high=college or university studies; middle=attended high school; and low=only secondary school). Age, calendar year, and parity were modeled as time-dependent variables, where age was divided into 5-year intervals; calendar year was divided into 15-year periods and parity according to categories (nulliparous; one to two childbirths; and greater than three childbirths). For the Cox model, the proportional hazards assumption was checked graphically and by Schoenfeld's partial residuals as appropriate. Two-tailed 95% CIs and P values were given with P<.05 regarded as significant. The statistical software package SAS 9.2 was used for all analyses.
After exclusions, the final study population included 5,385,186 women of whom 20,905 women had undergone suburethral mesh tape surgery for SUI (exposed). The observational period encompassed a total of 44,012,936 person-years at risk, during which there were 238,476 cases of primary cancer among women with no prior cancer diagnosis.
Table 1 shows rates and overall risks of any primary cancer among exposed and unexposed women as well as their patient characteristics. There was a higher rate of primary cancer among exposed as compared with unexposed women (rate 790/100,000 person-years, 95% CI 745–838 vs 541/100,000 person-years, 95% CI 538–543). However, after adjustment for age, calendar time, parity, and educational level in the Cox regression analysis, the overall cancer risk was significantly lower among exposed as compared with unexposed women (HR 0.9, 95% CI 0.8–0.9).
The adjusted incidence rates and HRs for primary cancer according to organ system are presented in Table 2. Breast cancer was the most common cancer type afflicting 355 exposed women (1.7% in total, rate 251.0/100,000, 95% CI 226.2–278.5) and 71,243 unexposed women (1.3% in total, rate 162.4/100,000, 95% CI 161.2–163.6) (HR 0.9, 95% CI 0.8–1.1). The incidence rates per 100,000 person-years (95% CIs) for exposed compared with unexposed women were 20.5 (14.3–29.5) vs 21.0 (20.6–21.5) for rectal cancer, 25.5 (18.4–35.3) vs 19.8 (19.4–20.2) for ovarian cancer, 65.0 (53.0–79.8) vs 33.1 (32.6–33.7) for endometrial cancer, 5.7 (2.8–11.3) vs 11.9 (11.6–12.2) for cervical cancer, and 19.1 (13.1–27.8) vs 13.3 (13.0–13.7) for bladder and urethra cancer. Other than an inverse association with rectal cancer (HR 0.5, 95% CI 0.3–0.8), there were no significant differences in risk between exposed and unexposed women for pelvic organ cancers including ovarian (HR 0.8, 95% CI 0.5–1.2), endometrial (HR 1.1, 95% CI 0.8–1.4), cervical (HR 0.4, 95% CI 0.2–1.0), bladder, and urethra (HR 0.7, 95% CI 0.4–1.2). No significant association was observed between the exposed women and primary cancer in any organ system when compared with unexposed women (Table 2).
Table 3 shows the effects of the exposure over time as we performed a Cox regression analysis with overall cancer risk divided into three time bands (1–4 years; 5–9 years; and more than 10 years) after surgery. The risk for cancer after exposure showed little variation over time other than an inverse correlation within the first 4 years of surgery (HR 0.7, 95% CI 0.7–0.8). Thereafter, there were no significant differences in cancer risk between exposed and unexposed women regardless of time since surgery.
The results of this population-based cohort study suggest that midurethral polypropylene sling surgery for SUI is not associated with an increased cancer risk later in life. There were no significant associations between having the operation and cancer risk in either adjacent pelvic organs or any other specific organ system. Thus, our data do not support the notion that polypropylene mesh may increase the risk for cancer when used for SUI surgery.
A purported mechanism of carcinogenesis after the use of biomaterials involves inflammation as part of a host-vs-implant reaction and implant degradation.2 Chronic inflammation may be secondary to a variety of exposures including microbial infections, allergens, excessive alcohol intake, autoimmune diseases, and chemical exposure.12 The role of nondegradable biomaterials in promoting carcinogenesis after long-term exposure is considered negligible in general,8 yet scientifically unconfirmed with regard to paravaginal tissues in specific.2 Human in vivo studies on pelvic biocompatibility of polypropylene mesh are scarce but show that macroporous, monofilament, polypropylene mesh can induce mild but persistent periurethral tissue inflammation when used for the treatment of SUI and pelvic organ prolapse.13,14 Inflammation has been demonstrated to occur at the mesh–tissue interface,14 but there is no evidence to date in support of systemic inflammatory involvement or that persistent localized inflammation would be a precursor of cancer in remote organs. In addition, the results of the present study do not support the role of suburethral polypropylene as a cancerogenic agent in adjacent organs regardless of presumed mechanisms of action.15 Our findings corroborate a large body of evidence in other surgical fields as well as results from an uncontrolled hospital-based review of more than 2,000 patients with a midurethral sling.8,16
The introduction of a foreign body material can influence tissue composition beyond the first years after surgery and may be linked to mesh erosions and, theoretically, an increased cancer risk at long term.5 The observational period of the present study extended more than a decade for women having had surgery within the first years after introduction of the first polypropylene mesh sling in 1997. Reassuringly, we found no increased cancer risk at either mid- or long-term follow-up lending further support to the conclusion that there is no association between midurethral polypropylene sling surgery and cancer. The seemingly protective effect of surgery during the first 4 years after surgery is most likely attributed to surgical selection meaning that healthy patients are more likely to undergo an elective procedure and are less likely to develop cancer at short term as compared with women in poor health. There was no difference between exposed and unexposed women with regard to the risk for larynx and lung cancer suggesting that smoking did not confound the results. Similarly, obesity is unlikely to have confounded the results given that obesity-related cancers such as breast, endometrial, and colon cancer were not overrepresented among the unexposed women. Independent prospective data collection prevents recall and ascertainment bias, whereas the near complete registration of available women reduces risk for reporting bias. Detection bias was avoided by censoring the first year after exposure.
Adjusting the analysis for educational attainment may be considered an effective proxy for socioeconomic status, smoking, obesity as well as diabetes,17–19 which are established cancer risk factors. Some factors that could affect the decision to perform elective surgery are, however, unaccounted for in the available registers. Sources of unmeasured bias may include care-seeking behavior and a greater propensity for elective surgery among women having had surgery for SUI.20 Despite the nationwide and population-based study design, we recognize that there was limited statistical power for rare cancer types based on the counts among the exposed women. A further limitation is that the information on the number of midurethral sling procedures performed in outpatient care is unreliable and subsequently only data collected as inpatient procedures were used in the analysis. This source of nondifferential misclassification is, however, unrelated to the outcome; in other words, it is not biologically plausible that a midurethral sling performed as an inpatient procedure in and of itself carries a greater risk for cancer compared with an outpatient procedure or vice versa. Given that the operation code LEG10 was created specifically for the polypropylene midurethral sling, we consider the risk for misclassification negligible.
While acknowledging the limitations of our study, we conclude that midurethral sling surgery using polypropylene mesh does not increase the risk for cancer later in life. In view of the vast numbers of women having SUI surgery using polypropylene mesh, our findings have important public health implications. Based on our nationwide data and long-term follow-up, we find no reason for physicians to suggest inferior or more invasive alternative treatments to the midurethral polypropylene sling when SUI surgery is indicated to avoid risk of cancer later in life.
2. Ostergard DR, Azadi A. To mesh or not to mesh with polypropylene: does carcinogenesis in animals matter? Int Urogynecol J 2014;25:569–71.
3. Bischoff F, Bryson G. Carcinogenesis through solid state surfaces. Prog Exp Tumor Res 1964;5:85–133.
4. Brand KG, Buoen LC, Johnson KH, Brand I. Etiological factors, stages and the role of the foreign body in foreign body tumorigenesis: a review. Cancer Res 1975;35:279–86.
5. Oppenheimer BS, Oppenheimer ET, Stout AP, Willhite M, Danishefsky I. The latent period in carcinogenesis by plastics in rats and its relation to the presarcomatous stage. Cancer 1958;11:204–13.
6. Goldman HB, Dwyer PL. Polypropylene mesh slings and cancer: an incidental finding or association? Int Urogynecol J 2016;27:345.
7. Lin HZ, Wu FM, Low JJ, Venkateswaran K, Ng RK. A first reported case of clear cell carcinoma associated with delayed extrusion of midurethral tape. Int Urogynecol J 2016;27:377–80.
8. McGregor DB, Baan RA, Partensky C, Rice JM, Wilbourn JD. Evaluation of the carcinogenic risks to humans associated with surgical implants and other foreign bodies—a report of an IARC Monographs Programme Meeting. International Agency for Research on Cancer. Eur J Cancer 2000;36:307–13.
9. Moalli P, Brown B, Reitman MT, Nager CW. Polypropylene mesh: evidence for lack of carcinogenicity. Int Urogynecol J 2014;25:573–6.
10. King AB, Goldman HB. Current controversies regarding oncologic risk associated with polypropylene midurethral slings. Curr Urol Rep 2014;15:453.
11. Coussens LM, Werb Z. Inflammation and cancer. Nature 2002;420:860–7.
12. Nilsson A, Spetz C, Carsjö K, Nightingale R, Smedby B. Reliability of the hospital registry: the diagnostic data are better than their reputation [in Swedish]. Lakartidningen 1994;91:598, 603–5.
13. Falconer C, Söderberg M, Blomgren B, Ulmsten U. Influence of different sling materials on connective tissue metabolism in stress urinary incontinent women. Int Urogynecol J Pelvic Floor Dysfunct 2001;12(suppl 2):S19–23.
14. Elmer C, Blomgren B, Falconer C, Zhang A, Altman D. Histological inflammatory response to transvaginal polypropylene mesh for pelvic reconstructive surgery. J Urol 2009;181:1189–95.
15. Thames SF, White JB, Ong KL. The myth: in vivo degradation of polypropylene-based meshes. Int Urogynecol J 2017;28:285–97.
16. King AB, Zampini A, Vasavada S, Moore C, Rackley RR, Goldman HB. Is there an association between polypropylene midurethral slings and malignancy? Urology 2014;84:789–92.
17. Cohen AK, Rai M, Rehkopf DH, Abrams B. Educational attainment and obesity: a systematic review. Obes Rev 2013;14:989–1005.
18. Maty SC, Everson-Rose SA, Haan MN, Raghunathan TE, Kaplan GA. Education, income, occupation, and the 34-year incidence (1965–99) of type 2 diabetes in the Alameda County Study. Int J Epidemiol 2005;34:1274–81.
19. Smith BT, Lynch JW, Fox CS, Harper S, Abrahamowicz M, Almeida ND, et al. Life-course socioeconomic position and type 2 diabetes mellitus: the Framingham Offspring Study. Am J Epidemiol 2011;173:438–47.
20. Altman D, Granath F, Cnattingius S, Falconer C. Hysterectomy and risk of stress-urinary-incontinence surgery: nationwide cohort study. Lancet 2007;370:1494–9.