Although the overall incidence of cervical cancer is decreasing, it is estimated that over 200,000 women are living with cervical cancer in the US, and it is expected to lead to over 4,000 deaths in 2018.¹ Cervical cancer is predominantly caused by infection of the cervical mucosa with human papilloma virus (HPV), particularly by several pro-oncogenic subtypes.

Multiple HPV genotypes can infect the genital tract mucosa, but types 16 and 18 are responsible for the majority of cervical cancers. HPV infection is highly prevalent in sexually active young woman, and the majority will clear the infection within 8–24 months.² Although cleared, HPV infection can lie dormant for prolonged periods. It can recur and be detected again decades later, mandating the need for cervical screening throughout a patient’s lifetime.³ In addition, observational data indicates that the relative risk of abnormal cervical cytology is markedly increased in those with persistent HPV infection, particularly where infection is with a high-risk type of HPV.²

HPV vaccination was introduced in 2006 and it is anticipated that widespread uptake of the vaccine will considerably alter the prevalence of HPV infection and, therefore, cervical cancer in the coming decades.  However, a substantial proportion of adult women have previously been exposed to HPV and remain at risk of cervical cancer. Defining an optimal screening strategy that combines efficacy and cost-effectiveness is still important. Current US guidelines recommend co-screening with cervical cytology (pap smear) and HPV testing every five years for low-risk women between 30 and 65 years of age (if both initial tests are negative) or pap smear every three years.^4,5

Several professional societies have expressed interest in the use of primary HPV testing, i.e., HPV testing in the absence of pap smear. Prior studies have suggested this is effective in detecting high-grade lesions early but could lead to overdiagnosis of lower risk cervical changes and increase the number of patients referred for colposcopy.⁶ Furthermore, the longer term effect of this approach has not been clear.

The Study

In this context, a large multicenter trial was carried out across Canada where 19,000 women were randomized to either primary HPV testing or liquid-based cytology (LBC).⁷ The primary outcome of the study was the cumulative incidence of cervical intraepithelial neoplasia (CIN) grade 3 or higher at 48 months after randomization.

Enrolled patients were aged 25–65 years, had not had a smear in the previous year, and were at lower risk of cervical cancer—not immunosuppressed, with no history of CIN grade 2 or higher, or cervical cancer. Patients were assigned to primary HPV testing or LBC at enrollment. All patients underwent HPV and cytology testing at 48 months, the end of the study.

The follow-up plan for this study was complex, and incorporated a detailed management protocol for those in whom abnormalities were detected.⁷ Those who had negative primary HPV test at enrollment were not recalled until 48 months. Those in the LBC group with negative baseline results were recalled for LBC testing at 24 months, as per standard practice in Canada. Those with abnormalities on HPV screening at enrollment testing underwent LBC. If HPV was positive and LBC negative, they were recalled at 12 months for repeat dual testing. If either test was positive at 12 months, patients were referred for colposcopy. If both were negative at 12 months, patients reverted to standard management and were recalled at 48 months. Patients with abnormalities in the control group were managed in a similar manner.

The Results

In the first round of screening, significantly more CIN3+ cases were detected in the primary HPV screening arm (an increase in cases of 2.67 per 1,000 women screened). However, at 48 months, the opposite was true, with significantly fewer cases of CIN3+ in the primary HPV screening arm, with a risk ratio of 0.42 when compared to control management.⁷ The absolute difference in incidence rate was a reduction of 3.22/1,000 (95% CI -5.12 to -1.48). In subgroup analyses, women who were HPV-negative at baseline had a significantly reduced risk of CIN3+ at 48 months when compared to women who had negative cytology at baseline.

Colposcopy referrals were assessed as a secondary end point. At baseline, higher referral rates were seen in the HPV screening group, but the opposite was true at 48 months, where rates were lower in the HPV group than LBC group. Interestingly, cumulative colposcopy rates were similar between the two groups over the entire trial, suggesting the HPV group may have been referred at an earlier time point, as opposed to simply at higher rates. Furthermore, this effect gives some insight into the expected colposcopy referral rates when an HPV screening program is introduced—it is likely that initially both prevalent and incident cases were detected, and as time goes on, referrals reflect new cases only.

Indications

Taken together, these data suggest that primary HPV screening is effective and detects dysplastic changes in the cervical mucosa earlier and more accurately than cytology. Women with negative-HPV testing at baseline appear to be at reduced risk of CIN3+ on follow-up, suggesting the HPV screening may more accurately detect the lowest risk individuals who could benefit from increased intervals between screening events. Furthermore, this study provides reassurance that at least within a two-year follow-up window, HPV screening alone is safe and effective. Longer-term studies are required to assess its efficacy when compared to pap smear.

However, given the widespread availability of HPV vaccination, it is hoped that for the next generation of women, baseline rates of cervical dysplasia will be significantly lower. This welcome change in disease demographics will require ongoing review of the benefits and cost-effectiveness of all screening interventions.