Risk of human papillomavirus infection in women with rheumatic disease: cervical cancer screening and prevention

Abstract

Human Papillomavirus (HPV) is the most common sexually transmitted infection in the USA, with over 14 million people acquiring HPV each year. HPV is also the cause of most anogenital cancers. About 90% of HPV infections spontaneously resolve over 3 years. However, about 10% remain as persistent infection defined as repeatedly detected in cervical samples. As HPV is controlled by local and systemic immune responses, individuals with immunosuppression are at risk for cervical cancer. It is hypothesized that immunosuppressed individuals are more likely to have HPV persistence, which is necessary for malignant transformation. Accordingly, women with rheumatic diseases such as SLE and RA are likely vulnerable to HPV infection and the progression of cervical disease. The HPV vaccine, given as a series of vaccinations, is safe and effective that can prevent HPV infection and cervical cancer. There is no contraindication to HPV vaccination for women to age 26 with rheumatic disease, as it is not live. As in the general population, timing is key for the efficacy of the HPV vaccine as the goal is to vaccinate prior to sexual debut and exposure to HPV. There are no formal recommendations for cervical cancer screening in women with rheumatic disease but recommendations for the HIV-positive population can be adopted, meaning to screen with a Pap test annually for three consecutive years and if all normal, to extend the interval to every 3 years with the option of co-testing with HPV at 30 years and older.

Rheumatology key message

• Risk of human papillomavirus infection is increased among women with autoimmune disease, likely associated with immunosuppression.

Introduction

Human papillomavirus (HPV) is the most common sexually transmitted infection in the USA, with over 14 million people acquiring HPV each year [1]. For the 1984 discovery of the HPV as the cause of cervical cancer, Harald zur Hausen was awarded the 2008 Nobel Prize in Physiology or Medicine [2]. HPV is also the cause of most anogenital cancers. Over 200 types of HPV have been identified and categorized based on DNA homology into alpha, beta, gamma and mu types [3]. Alpha types have been linked to the anogenital cancers. Of the 40 alpha types detected, 15 have been associated with the development of cervical cancer. HPVs 16 and 18 are responsible for ∼70% of cervical cancers. HPV 16 is the predominant type in anal and oropharyngeal cancers. Among the HPV-associated vulvar, vaginal and penile cancers, HPV 16 also predominates [4]. HPVs 6 and 11 are responsible for the majority of anogenital warts.

Natural history of HPV infection

The HPV has capsid proteins surrounding circular, dsDNA. The natural history of HPV shows that initially HPV infects the basal cells in the cervical mucosa. There is some evidence that infection of basal stem cells results in true precancerous lesions whereas infection of other epithelial cells will result in loss of viral replication [5]. HPV requires cell differentiation and replication for its own survival depending on the host cell machinery for its replication. HPV does not enter the bloodstream and does not trigger a significant immune response compared with other infections such as chlamydia or gonorrhoea [6]. Antibodies to HPV occur 3–9 months post-infection. Their presence represents past exposure. In general, antibodies are not indicative of strong protection. However, high antibody titres do appear to be protective from re-infection whereas lower antibody levels show no protection [7]. For this reason, blood tests for HPV infection or HPV antibodies are currently not clinically useful. Only testing of the potentially infected epithelial cells uncovers an HPV infection. About 90% of HPV infections spontaneously resolve over 3 years [8]. However, about 10% remain as persistent infection defined as repeatedly detected in cervical samples [9]. An estimated 0.8% of infections result in the transformation of an infected epithelial cell into an invasive cancer [2] (Fig. 1).

Fig. 1.

Human papillomavirus

From https://www.nobelprize.org/nobel_prizes/medicine/laureates/2008/press.html [2]; used with permission © The Nobel Committee for Physiology or Medicine. Illustrator: Annika Röhl.

Epidemiology of HPV infection

HPV is highly transmissible through sexual contact, with a median probability of transmission around 40%, and studies suggest female to male transmission is more common than male to female [10, 11]. The probability of transmission between partners reaches 100% for all HPV types after 11 episodes of sexual intercourse [10]. In the USA, the prevalence of genital HPV infection in females peaks at around 40–50% in the 20- to 24-year-old age range and decreases to remain around 30% for the subsequent years [12, 13]. While women in their 30s and 40s have similar rates of high and low risk HPV infection, more women in the 20- to 24-year-old age range are infected with high-risk HPV types.

Cervical intraepithelial neoplasia and cervical cancer

Active replicative infections result in typical cervical epithelial changes that include basal cell proliferation, abnormal mitotic figures and nuclear enlargement. Historically, premalignant squamous changes of the cervix were described as cervical dysplasia. The 2012 Lower Anogenital Squamous Terminology system, developed to describe HPV-associated squamous lesions of the anogenital tract [14, 15], uses low-grade or high-grade squamous intraepithelial lesion (SIL) for both cytological and histological findings. Low-grade SIL, formerly mild dysplasia, refers to the changes described above and is limited to the bottom one-third of the cervical epithelium. High-grade SIL, formerly moderate or severe dysplasia, refers to the presence of these dysplastic changes in greater than one-third of the cervical epithelium. The progression of HPV infection to cervical cancer is accompanied by a sequence of histological changes. The majority (>70%) of low-grade SILs regress spontaneously, whereas 30–70% of high-grade SILs regress without treatment, with higher rates of regression in young women [16, 17]. Thus, low-grade SIL is currently considered a benign HPV infection with high rates of regression, and high-grade SIL is considered precancerous [18, 19]. Due to the great advancements in screening with the Papanicolaou (Pap) test, cervical cancer is uncommon in the USA, with an estimated 11 000 cased diagnosed annually (Incidence rate, IR = 7.9/100 000 person-years) [20]. For feasibility and efficiency, high-grade SIL has been therefore used as a surrogate end point for cervical cancer in clinical trials of HPV vaccines [21, 22].

Although HPV persistence is the strongest (and necessary) risk for cervical cancer, other factors have also been implicated in increased risk. These include older age, HPV type (with the greatest risk for type 16), smoking, prolonged use of oral contraceptives, number of pregnancies, Chlamydia trachomatis, immunosuppression and family history [16, 23].

Risk of HPV infection in women with rheumatic disease

As HPV is controlled by local and systemic immune responses, individuals with immunosuppression, whether iatrogenic, genetic or infectious, are at risk for cervical cancer. It is thought that immunosuppressed individuals are more likely to have HPV persistence, which is necessary for malignant transformation [16, 24, 25]. In line with these observations, women with rheumatic diseases are likely vulnerable to HPV infection and the progression of cervical disease [26–31].

Several studies have shown an increased the risk of cervical SIL and HPV infection in patients with SLE [32–34]. Infection with more than two types of HPV as well as a greater number of infections with high-risk HPV types have been found in SLE patients compared with controls [33–35]. In a large Swedish cohort study, SLE (n = 4976) was a risk factor for cervical intraepithelial neoplasia but not invasive cervical cancer compared with the matched general population (n = 29 703) [34]. The absence of risk of cervical cancer was likely due to the fact that these women were screened regularly and treated when appropriate, abating the risk of invasive cancer. Among women with other rheumatic disease including RA, and SS, the risk of cervical SIL was reported to be elevated as well [36–38]. In a US population-based cohort study, women with SLE and RA, but not those with psoriasis or IBD, had a 1.5 times increased risk of high-grade cervical SIL or cervical cancer compared with women without systemic inflammatory disease. Similar results were seen in a Swedish cohort study in which biologic-naive women with RA (n = 34 984) had a 40–50% increased risk of both low- and high-grade SIL compared with the general population [39]. However, this study found no increased risk of invasive cervical cancer among biologic-naive RA patients.

These observed increased risks of cervical SIL or cancer in women with rheumatic disease are likely related to known risk factors for HPV infection in the general population to some degree, but the disease itself or exposure to immunosuppressive medications may confer further risk. In a survey study of 289 female patients enrolled in a prospective registry of RA [40], 97% reported having at least one Papanicolaou test during their lives. Abnormal Papanicolaou test results were reported by 29% of the women; in a comparison of women with and without an abnormal Papanicolaou test, the main risk factors were more than three sexual partners, prior use of birth control and a prior diagnosis of another sexually transmitted infection. Prior use of steroids, race, marital status, level of education and income did not impact the risk of having an abnormal Papanicolaou test result. Similar findings were seen in a North America-based study of 1015 SLE female patients with a median age of 42 years [41]. In this study, 13.3% had an abnormal Papanicolaou report, which appeared to be associated with a history of sexually transmitted infection and use of oral contraceptives. Few studies examined the risk of cervical intraepithelial neoplasia or cervical cancer related to types of immunosuppressive drugs. More common occurrence of cervical SIL was reported in women with SLE treated with immunosuppressive agents such as steroids and intravenous CYC [42, 43]. A US population-based cohort study of 20 282 women with SLE reported a 1.4 times greater risk of cervical SIL or cervical cancer associated with initiation of immunosuppressive drugs vs HCQ [44]. Women taking TNF-inhibitors for RA (n = 9629) were found to have a 1.4-fold greater risk for high-grade SIL and a doubled risk for invasive cervical cancer vs biologic-naive RA patients (n = 34 984) [45]. In a large cohort study (n = 44 534) using claims data from both US commercial and federal health plans, women with RA who initiated a treatment with a biologic DMARD had a 1.3 times increased risk, albeit not statistically significant, of cervical SIL or cervical cancer versus those on non-biologic DMARDs [46]. Due to the low incidence of high-grade SIL or cervical cancer, even these large population-based studies were not sufficiently powered.

Cervical cancer screening

Some studies have suggested that patients with chronic diseases such as RA or SLE do not receive optimal cancer screening tests and other preventive medical services [47–50]. Data from a historical cohort of 1335 adults with RA enrolled in a national fee-for-service insurance plan showed that RA patients received inadequate quality of care including health maintenance tests such as mammography and Papanicolaou smears [48]. In a more recent US population-based cohort study of RA patients, patients with RA (n = 13 314) were screened on average once every 3 years for cervical cancer and did not appear to be at risk for receiving fewer cancer screenings (including Papanicolaou tests) than non-RA patients (n = 212 324) [51]. A small Canadian survey study of 48 SLE patients suggested women with SLE were less likely to undergo appropriate cancer screening including Papanicolaou tests compared with the general population [52].

To date, there is no specific guideline for cervical cancer screening in women with rheumatic disease. Current cervical cancer screening guidelines from the American College of Obstetricians and Gynecologists, US Preventive Services Task Force, American Cancer Society and American Society for Colposcopy and Cervical Pathology [53–55] apply to ‘low risk women’ and recommend the onset of screening at age 21 with Papanicolaou testing every 3 years until the age of 30, at which point testing can be spaced out to every 5 years if Papanicolaou and HPV testing are performed together. If all results are normal, the patient remains ‘low risk’. The definition of low risk varies by organization, but generally excludes women who are immunosuppressed. However, ‘immunosuppression’ is not defined and alternative screening approaches are not provided, except for women infected with HIV. For HIV-positive women, the recommendation is to screen with the onset of sexual activity or by age 21, and to screen annually with Papanicolaou testing for 3 years, and if all normal, to extend the interval to 3 years, adding HPV co-testing at the age of 30, but continuing at 3-year intervals [56]. Recent EULAR recommendations for women’s health issues in SLE and/or aPL syndrome suggest annual peroxidase–antiperoxidase testing in heavily immunosuppressed patients or according to the general cervical cancer screening guidelines in low-risk patients [57].

Whether more frequent screening for cervical cancer or more aggressive monitoring of cervical SIL is needed in immunosuppressed women secondary to rheumatic disease is unknown. It is also unclear whether discontinuation of immunosuppressive drugs in patients with persistent HPV infection or high-grade SIL would lead to a better outcome.

HPV vaccine

There are three HPV vaccines now approved by the FDA in the USA. All are based on non-infectious virus-like particles comprising the L1 HPV protein. The first approved vaccine was a quadrivalent vaccine (4vHPV) covering HPV 6, 11, 16 and 18; the second approved was a bivalent vaccine (2vHPV) covering HPV 16 and 18; the most recent is a 9-valent (9vHPV) covering HPV 6, 11, 16, 18, 31, 33, 45, 52 and 58. Currently, only 9vHPV2 is available in the USA. The most recent American Committee on Immunization Practices recommendation is for routine HPV vaccination at age 11–12 for both girls and boys, but may be given as early as 9 years of age [58]. The HPV vaccine is approved for catch-up in women up to age 26 and to age 21 in men. Vaccination is recommended up to age 26 for men who have sex with men. If the first dose is given at or prior to age 14, only two doses are needed, with the second dose given 6–12 months after the first dose. If the patient is 15 years or older, three doses are recommended, with the second dose given at least 1–2 months after the first dose and the third dose given at least 6 months after the first dose. If the schedule is interrupted, there is not a need to restart the cycle: the two or three-dose schedule should be completed as scheduled. If the series was begun with 2vHPV2 or 4vHPV, 9vHPV can be used to complete the series. For persons who have completed a series of 2vHPV or 4vHPV, there is currently no American Committee on Immunization Practices recommendation regarding additional vaccination with 9vHPV.

Timing is key for the efficacy of the HPV vaccine with the goal being vaccination prior to sexual debut and exposure to HPV. The efficacy of the vaccine drops dramatically based on the age of vaccination due to prior HPV exposure. For example, in the PATRICIA trial of 18 644 young women aged 9–26 years [22], the effectiveness of the 4vHPV to prevent high-grade SIL varied dramatically by age: for those vaccinated prior to age 18, effectiveness was 79.1%; for those vaccinated after age 20, effectiveness was just 26.4%. The ability of the vaccine to prevent genital warts decreases with age, rendering it largely ineffective when given over age 26 [59].

HPV vaccination in patients with rheumatic disease

There is no contraindication to HPV vaccination for immunocompromised patients, as it is not a live vaccine. The 2012 ACR panel recommends HPV vaccination for people with RA before starting or currently receiving DMARDs or biologics, if indicated [60]. Unfortunately, rates of vaccination in the USA remain well below the goal of 75% of 11–12 year olds completing the series—and this includes patients with rheumatic disease. A study using US commercial claims data from 2006 to 2012 showed that only 23% of healthy people and 21% of people with rheumatic disease aged 9–26 years started an HPV vaccine series and only half of these completed the cycle [61]. More recent data from the CDC show that almost 60% of girls and 50% of boys aged 13–17 years have received at least one HPV vaccination, while almost 40% of girls and 28% of boys have received all three doses [62].

In most, HPV infection does not result in a robust antibody response. Consequently, the presence of HPV antibodies in unvaccinated individuals is not considered protective, but rather a marker of past exposure. In addition, men are less likely to seroconvert than women, suggesting their genital infections are often superficial and transient and therefore never eliciting an immune response. In contrast, the HPV vaccines all induce robust responses 100–1000-fold higher than natural infections. Protection from infection is thought to reflect a transudate of serum antibodies in the cervix preventing viral attachment and cell entry [63, 64]. HPV2 and HPV4 have been shown to maintain high levels of anti-HPV antibodies for up to 10 years [65, 66].

In children with HIV infection, antibody responses to the HPV4 vaccine were lower than for HPVs 6 and 18 age-similar historical controls comparable to healthy children [64, 67]. However, titres were still above natural infection. Decay of antibody titres suggests that 50% of perinatally HIV infected children will become seronegative after 5–10 years [68, 69]. Among women with SLE, at 7 and 12 months after vaccine, the levels of antibodies were comparable to healthy controls [70]. In this study, the only immunosuppressant that appeared to impact vaccine response was MMF, with only 33% of patients taking mycophenolate becoming seropositive 18 months after vaccination. Patients on steroids, AZA, ciclosporin, tacrolimus and HCQ did not have a significant drop in antibody titres over this period. No vaccine efficacy studies have been published to date in patients with immunosuppression or autoimmune disease.

Does the HPV vaccine induce autoimmune disease?

The main side effects from the HPV vaccine are localized swelling, erythema and pain at the injection site underscoring the immunostimulatory effect of the virus-like particles. The virus-like particle of the HPV vaccine is very immunogenic, leading to a strong B cell immune response [71]. By having a repetitive viral surface, it strongly activates B cells and prompts complement fixation. Some patients are concerned that this dramatic response to viral particles may also induce an autoimmune reaction.

While studying the potential link between HPV vaccination and autoimmune disease is challenging, several studies have reported generally low rates of autoimmune disease following HPV vaccination. The US Vaccine Adverse Event Reporting System received 12 424 voluntary reports of adverse events following immunization with 4vHPV between 2006 and 2008 [72]. Of those, there were 51 cases of autoimmune diseases, with a reporting rate of 0.2/100 000 4vHPV doses distributed [72]. In a systematic review of 13 randomized controlled trials of HPV vaccine [73], two studies (n = 966) reported new-onset autoimmune disease following injection with HPV vaccine or the control [74, 75]. However, the rate of autoimmune disease was similar between the vaccine group (3.36%) and the controls (4.85%) [73]. Other studies found much lower frequencies with two studies each finding eight cases among over 600 000 vaccinated people in the USA, six cases of autoimmune disease among 5.8 million teen girls vaccinated in France and 23 cases of autoimmune disease of 700 000 people vaccinated in Denmark and Sweden [76]. A study of the Kaiser Permanente data from Southern California suggested possibly higher risks for Hashimoto’s thyroiditis (rate ratio (RR) = 1.29) and optic neuritis (RR = 1.45), though these associations did not bear out with more rigorous medical record review [77].

Whether HPV vaccination would lead to a greater risk of developing a new autoimmune disease in women with a known diagnosis of autoimmune disease is not fully understood. It is also unclear whether the vaccination would affect the disease activity in women with a known diagnosis of autoimmune disease. Among 60 women with SLE, erythema and pain at the injection site was reported by 6%, compared with 4% of control women [70]. Interestingly, these rates were much lower than reported in the clinical trials. The rates of rash, nausea, headache, irregular menses and upper respiratory infection also did not differ [70] and there were no clinically significant changes in lupus laboratory parameters (complement, dsDNA) or disease activity in the 2, 6 and 12 months following HPV vaccination. A systematic review of five studies of young women with autoimmune diseases found no changes in disease exacerbations with vaccination [78]. In a Swedish cohort study, 4vHPV vaccination was not associated with increased incidence of new-onset autoimmune disease in female patients with preexisting autoimmune disease aged between 10 and 30 years compared with unvaccinated girls and women [79].

Conclusions

In summary, HPV is the most common sexually transmitted disease and can cause anogenital intraepithelial neoplasia and cancer including cervical, vaginal, vulvar, anal, penile and oropharyngeal cancers. The prevalence of HPV infection and SIL is higher among women with autoimmune disease, likely associated with iatrogenic immunosuppression. However, immune dysfunction associated with these diseases may also play a significant role. While the national guidelines have deemed annual cervical cancer screening unnecessary for many women, for HIV-positive women it is now recommended to screen with Papanicolaou test annually for three consecutive years and if all normal, to extend the interval to every 3 years with the option of co-testing with HPV at 30 years and older. Although to date, no clinical studies are available specifically focused on women with rheumatic disease, recommendations for the HIV-positive population can be used in women with other immunosuppressed conditions including rheumatic disease.

The HPV vaccine is a safe and effective series of vaccines that can prevent HPV infection and cervical cancer. The key to the success of the vaccine is dosing prior to HPV exposure, before sexual debut. The current rates of vaccination are very low in the USA, but a targeted push towards vaccinating our teenage female and male patients may, in the coming years, make HPV-related disease and cancer even less frequent.

Supplement: This project was approved by the Duke University Institutional Review Board.

Funding: This work was supported by National Institutes of Health [R13AR070007].

Disclosure statement: S.K. has received grants/research support from Pfizer, Lilly, Bristol-Myers Squibb, AstraZeneca and Genentech. A.-B.M. has served on an advisory board for Merck. The other author has declared no conflicts of interest.