Introduction

Optimal strategies for increasing cervical cancer screening may differ by screening history and healthcare setting. While mailing HPV self-sampling kits to overdue individuals increases adherence, its effectiveness among previously-adherent individuals in the U.S. is unknown. Within a large U.S. healthcare system, STEP compared different approaches for offering self-sampling (direct-mail or opt-in) to individuals who were previously-adherent, overdue, or with unknown screening history.

Methods

From November 2020-January 2022, we used electronic medical record data to identify and randomize 19,213 individuals aged 30-64 years at Kaiser Permanente Washington. Previously-adherent individuals were randomized to Education (usual care+educational materials about screening) (n=3,960), Direct-Mail (Education+mailed self-sampling kit) (n=3,956), or Opt-In (Education+option to request a kit) (n=1,482). Overdue individuals were randomized to Education (n=1,408) or Direct-Mail (n=1,415) and unknown history individuals to Education (n=3,486) or Opt-In (n=3,506). The primary outcome, screening completion within ≤6 months, was defined as: in-clinic screening, kit return with negative or HPV16/18+ results, or kit return with in-clinic reflex Pap if kit results showed other high-risk HPV+ or unsatisfactory. We used intention-to-treat binomial regression to identify absolute risk differences and 95% confidence intervals (CI).

Results

Among previously-adherent individuals, compared with Education (47.6%), screening completion was 14.2% (95%CI:11.3%-17.0%) higher with Direct-Mail (61.7%) and 3.4% (95%CI:1.1%-5.6%) higher with Opt-In (51.0%). Among overdue individuals, screening was 16.6% (95%CI:13.5%-19.7%) higher with Direct-Mail (35.5%) versus Education (18.8%). Among unknown screening history individuals, screening was 2.1% (95%CI:0.4%-3.8%) higher with Opt-In (18.0%) versus Education (15.9%).

Conclusions

Within a U.S. healthcare system, direct-mail self-sampling screening increased by >14%, with absolute effect size similar between previously-adherent and overdue groups. Conversely, the opt-in approach minimally increased screening. To maximize screening adherence, systems implementing HPV self-sampling should prioritize direct-mail outreach for due and overdue individuals. For individuals with unknown history, testing alternative outreach approaches and/or additional efforts to document screening history are warranted.

Introduction

Testing urine for high-risk human papillomavirus (hr-HPV) may be an attractive option for non-attenders of routine cervical screening. The accuracy of urine hr-HPV testing varies with different collection protocols. We hypothesised that Colli-Pee has better sensitivity for CIN2+ detection than standard pot-collected urine through reliable first-void collection, standardisation of volume collected, and immediate preservative-fixation. The aim of the Alternative CErvical Screening (ACES) Colposcopy study was to compare the sensitivity of matched urine and cervical hr-HPV testing for CIN2+ detection using two urine collection devices.

Methods

Colposcopy attendees in Manchester (UK) with abnormal cervical screening results were randomised (1:1) to Colli-Pee® 10mls with preservative or standard pot for urine collection. Urine was self-collected and matched cervical samples taken immediately prior to colposcopy; hr-HPV testing used Roche Cobas 8800. Colposcopic opinion and/or histology informed clinical diagnosis. A power calculation indicated that 480 participants (with 120 CIN2+/group) would have 89.8% power to establish a sensitivity of urine for CIN2+ detection >80%.

Results

465 participants were included in the analysis (Colli-Pee n=235, pot n=230). The groups were balanced in age (median 35.0 vs 36.3 years), ethnicity (79% vs 81% White) and referral screening results (44% vs 44% high grade; 43% vs 43% low grade/borderline; and 11% vs 12% persistent hr-HPV+/cytology-negative) in Colli-Pee and standard pot arms, respectively. Cervical hr-HPV was 97.58% sensitive (95%CI 94.81-99.11%) for CIN2+ detection (n=242/248). Urine hr-HPV sensitivity for CIN2+ was higher using Colli-pee (90.32%, 95%CI 83.84-94.38%, n=112/124) than when collected using the standard pot (73.39%, 95%CI 64.99-80.38%, n=91/124, p<0.001).

Conclusions

Hr-HPV tested Colli-Pee-collected urine shows similar clinical accuracy for CIN2+ detection compared to routine cervical screening. Further work in the general cervical screening population will establish its specificity and its potential to improve cervical screening uptake in current non-attenders.

Introduction

As the era is approaching when vaccinations will have resulted in that HPV infection is no longer transmitted, design of effective campaign strategies for targeting of screening to find infected women at risk will become important.

Methods

We used the Swedish national quality registry of cervical screening to identify cervical cancer risk profiles based on screening history. Profiles with an invasive cervical cancer risk of >1% were for the whole Sweden found for 4,918 women, and a self-sampling kit was sent to their registered address. There were 76,653 women in Sweden who had not taken a cervical smear for >20 years and these women (as well as a group of 12,038 women with risk profiles between 0.3-1%) were, by both online and physical letters, sent a link to order a self-sampling kit. Piloting during 2019-2021 also used reminders by SMS.

Results

In the 2020 pilot, we invited 6,398 women with high risk profile and 18,3% responded and ordered a self-sampling kit. In the 2021 pilot, we invited 23,318 women (mostly long-term non-attenders) and 6,4% responded. The full-scale campaign targeting 93,609 women all over Sweden was launched in September 2022. Direct send appears to have almost doubled the participation rate compared to the strategy with invitation to order a kit.

HPV prevalences varied greatly by risk group, with a noteworthy finding being that almost one third of HPV positivities were HPV16/18 in the high-risk groups. Positive women are referred to a regionally responsible gynecologist in the vicinity of the woman´s home.

Conclusions

Effective campaigns to reaching populations at high risk of cervical cancer with cervical cancer will be important for faster cervical cancer elimination. We find that a nationwide campaign using self-sampling and multiple contact strategies can be readily implemented in the whole country as a regular process for complementing the routing screening program.

Introduction

Cervical cancer is a major problem in Mexico. Only 40% attend regular screening. Self-sampling increases access and screening-coverage. Most Mexican states have access to HPV-testing. Literature has shown advantages of E6/E7-mRNA detection over HPV-DNA test in triage of abnormal cytology, however few have evaluated the performance in self-sampled vaginal material. This study compared the performance of a 14-type HPV-DNA test to a 7-type HPV-mRNA test in triage of abnormal cytology with respect to PPV, sensitivity, specificity, and number of colposcopies per CIN2+ detected.

Methods

418 Mexican women aged 25–65 years referred to colposcopy&biopsy after abnormal cytology (ASC-US+) at General Hospital of Mexico, Clinica Reina-Madre, and Clinica-Colposcopia underwent self-sampling (XytoTest, Mel-Mont Medical), completing a questionnaire evaluating acceptability. Samples were tested for HPV-DNA (Abbott-HPVm2000) and E6/E7 mRNA (PreTect HPV-Proofer`7 (16-18-31-33-45-52-58)). Study endpoint was histologically confirmed high-grade lesion (CIN2+).

Results

93.1% felt confident performing self-sampling. 95.0% (397/418) had ASC-US+; 70.8% (296/418) low-grade (ASC-US/LSIL) and 24.2% (101/418) high-grade lesions. 55.5% (232/418) had positive HPV-DNA and 25.4% (106/418) HPV-mRNA+. Prevalence of CIN2+ was 12.0% (50/418) including 4 cases of cervical cancer. Sensitivity (CIN2+) for HPV-DNA was 94.0% (47/50) versus 62.0% (31/50) for the mRNA-test, p<0.001. Both tests detected all cancers. The specificity was 49.7% (183/368) versus 79.6% (293/368), p<0.001, PPV 20.3% (47/232) and 29.2% (31/106) for DNA versus mRNA-test, p=0.09. The number of colposcopies per CIN2+ detected by low-grade/high-grade cytology, HPV-DNA and HPV-mRNA was 19.7 and 3.3 compared to 4.9 and 3.4 respectively.

Conclusions

Self-sampling is suitable for HPV-testing and highly accepted by the women. The low PPV by low-grade cytology (5.1%) reflects the need for effective triage, reducing unnecessary colposcopies. A 7-type HPV-mRNA test has higher specificity, PPV and lower number of colposcopies per CIN2+ compared to a 14-type HPV-DNA, effectively discriminating women warranted for immediate colposcopy/biopsy from return to follow-up.

Introduction

During the COVID-19 pandemic, the cervical cancer screening program was paused during 9 months as part of the public health restrictions. To compensate for the screening gap in the Region of Stockholm, Sweden, the Swedish Board of Health and Welfare allowed for a temporary transition from provider taken liquid-based cytology (LBC) with analysis of HPV and reflex cytology to HPV self-sampling in 2021. There is a paucity in data globally regarding participation rate using HPV self-sampling and an opt-out approach as the main screening strategy among participating women in the cervical cancer screening program.

Methods

Descriptive statistics from the cervical cancer screening program compared participation rate by HPV self-sampling strategy during March – December 2021 and provider taken LBC during 2018.

Results

During 2021, a total of 355,396 HPV self-sampling kits were sent by mail to eligible women of screening age in the Region of Stockholm. By comparison, in 2018 we invited 258,816 women for provider taken LBC with HPV analysis and reflex cytology. The participation rate was significantly higher with HPV-self-sampling strategy (51% vs. 38 %; p-value <0.0001). The largest increase in participation rate was among women 26-29 years (38% vs 54% ; p-value <0.0001) and 30-39 years (42% vs. 53%; p-value < 0.0001). The participation rate among non-responding women (defined as 4 or more screening invites), doubled with HPV self-sampling (12% vs 23%; p-value < 0.0001). The participation in follow-up LBC among HPV positive women after self-sampling was 90%. The time span from HPV-self-sample and appointment for a provider taken follow-up LBC was three months as per protocol.

Conclusions

Screening by HPV self-sampling in the Region of Stockholm has increased overall participation rate in the cervical cancer screening program, especially among younger women and non-responding women. However, follow-up testing among HPV positive women is crucial for early detection of pre-cancerous lesions.

Introduction

The PRESTIS trial (Prospective Evaluation of Self-Testing to Increase Screening) is the first pragmatic randomized controlled trial (RCT) in the U.S. to evaluate the effectiveness of mailed self-sample HPV testing in a safety net health system setting. Safety nets are a critical setting as they provide care to medically underserved individuals who are at disproportionate risk for cervical cancer. We hypothesized that patient navigation, a patient-centered healthcare intervention, would enhance self-sample HPV testing uptake.

Methods

PRESTIS is a 3-arm pragmatic RCT that compares screening uptake (self-sampling or clinic-based) and clinical follow-up among underscreened patients (target n=2,268) randomized to Arms: 1) recall to clinic-based screening (usual care); 2) mailed self-sample HPV testing; or 3) mailed self-sample HPV testing + patient navigation. Underscreened women (i.e., no Pap in 3.5 years or no Pap/HPV co-test in 5.5 years) were identified through the electronic medical record (EMR). Interventions were delivered by patient navigators. Mailed kits included low-literacy instructions, a vaginal swab collection kit, and a pre-paid return envelope. Kits were tested in the health system laboratory for high-risk (HR-) HPV, with reflex testing for HPV 16, 18/45. Results were documented in the EMR; HR-HPV+ patients were navigated to cytology or colposcopy.

Results

A total of n=1,955 patients (86% of target) have been enrolled and randomized as of 10/10/22. Accrual and outcomes ascertainment will be completed by 12/15/22 and 02/15/23, respectively. Currently, trial participants were predominantly Hispanic/Latina (66.6%); 52.2% indicated Spanish as their primary language; and 56.1% had healthcare coverage through the public financial assistance plan (Table 1). The average time since last screening was 9.2 years despite an average of 28.8 primary care encounters in the past 5 years.

Conclusions

PRESTIS trial results will provide critical data on the effectiveness of mailed self-sample HPV testing, alone and with patient navigation, in a medically underserved and racially/ethnically-diverse U.S. population.

Introduction

Métis Nation British Columbia (MNBC) provides programs and services that advocate for
Métis people. While Métis-specific research is limited, available data shows Indigenous women and
people with a cervix are at higher risk for cervical cancer diagnosis. Métis people are underrepresented
in both health statistics and research, and as such MNBC provides a unique opportunity to highlight the
specific health needs of the Métis people, and ultimately seeks to reduce barriers to cervix screening by
putting screening back in the hands of community members, while providing participants with a
connection to community.

Methods

Community engagement avenues were identified in partnership with MNBC and the Métis
Chartered Communities, and carried out by local MNBC personnel, community champions and
participating clinicians. This included roundtable dialogue with Elders, Community leaders and
representatives, health care providers, MNBC’s elected leadership, and other regional representatives to
evaluate cultural sensitivity and appropriate clinical pathways. As a result, Métis women and individuals
with a cervix in Northwest and Northeast BC were invited to register for cervix screening online
(www.CervixCheck.ca) and sent a self-collection kit in the mail if eligible (aged 25-65, more than 3 years
since last screen). Online feedback surveys were distributed after participation.

Results

There have been N= 44 registered, N=30 eligible and N=14 ineligible participants. Overall,
participants reported a very positive experience with self-collection. Other key themes identified through
surveys included the need for a sense of belonging to community and access to culturally safe health care
services.

Conclusions

Using existing MNBC networks including mail outs, and social media platforms, this project
has provided a unique approach to addressing barriers to cervix screening for Métis women and
individuals with a cervix. Included is Article 23 of the United Nations Declaration on the Rights of
Indigenous Peoples4. UNDRIP Article 23: Right to Development | Indigenous Rights Radio (culturalsurvival.org)

Introduction

Cervical neoplasia progression risks based on non-16/18 HPV types are not well characterized. We sought to identify the absolute risk of 12 non-16/18 HPV genotypes for cervical intraepithelial neoplasia (CIN) or worse and to predict 3-years CIN transition probabilities and progression patterns between multiple states according to genotypes.

Methods

HPV genotyping was performed based on 3 prospective cohorts. 14 high-risk HPV types (HPV16/18/31/33/35/39/45/51/52/56/58/59/66/68) were detected. The absolute immediate and 3-year cumulative risks for each HPV genotype were calculated by 3 approaches, including minimum (Min.), any type (Any.), and hierarchical attribution estimate (Hier.). Meanwhile, we applied a continuous-time multistate Markov model to predict this CIN status by addressing the probability of transitions between multiple states according to the genotypes.

Results

At baseline, eligible data were available for 15 269 women with a median age of 46 years (interquartile range, 36–53 years). There were 2622 HPV-positive women (17.16%), including 915 (5.99%) with multiple and 1707 (11.17%) with single infections. In non-16/18 infections, single HPV31, 33, 51 infections carried a high immediate risk for CIN2+/3+ (≥4%).

Among 5551 women who completed 3-year follow-ups, the cumulative risks of HPV45, 58, 51 were higher. Calculated by Min. and Hier., the cumulative risks for CIN3+ of HPV31/35/56/59/68 were 0%, though HPV31 carried high cumulative risks for CIN2+. The estimated probabilities for the 3-year transition from normal to CIN2 or more were the highest in HPV 16-positive patients (1.3%), followed by HPV 31(0.6%), 58(0.5%), 51(0.3%), 52(0.3%) and 33(0.3%).

Conclusions

Based on immediate and 3-year cumulative risks, strategy makers should reckon with genotyping HPV 31, 33, 51, 52, and 58 in screening and management. In contrast, HPV 35, 56, 59, and 68 could be considered for low intensive follow-up, to reduce the screening costs and other technologies for triage.

Introduction

HPV primary screening is a safe and efficient strategy for cervical cancer prevention. Patients with negative HPV results have low risk of cancer for years. HPV-positive patients need triage to decide who requires colposcopy. Dual stain (DS) and extended HPV genotyping have been evaluated for HPV triage. Here, we compared the clinical performance of extended genotyping combined with DS vs. cytology.

Methods

Among 3,757 HPV-positive patients from Kaiser Permanente Northern California undergoing co-testing in 2018, DS and Onclarity were conducted from residual specimens. Follow-up continued through 2021. We estimated absolute risk of CIN3, AIS, or cancer (CIN3+) using prevalence-incidence mixture models for combinations of Onclarity results (16; 18; 31; 45; 51; 52; 33/58; 35/39/68; 59/56/66 and four risk groups: 16; 18/45; alpha9; other), DS, and cytology.

Results

Among all HPV-positives, 49.1% were DS-positive, and 56.2% were ASC-US or greater (p<0.002). The sensitivity for detection of CIN3+ was higher for DS compared to cytology (91.8% vs. 77.8%; p<0.002). Across the nine genotype channels, dual stain positivity reflected underlying carcinogenicity, ranging from 33.1% for HPV59/56/66 to 68.4% for HPV16 while abnormal cytology was more similar across genotype groups and did not track with HPV carcinogenicity. Across all genotype groups, DS had higher sensitivity for CIN3+ and lower risk among dual stain negatives (from 0.1% for HPV59/56/66 to 2.0% for HPV16), compared to cytology (from 0.4% for HPV59/56/66 to 8.4% for HPV16). The risk difference between HPV16-NILM and HPV16-DS-negative would result in different management recommendations in the US (colposcopy vs. 1-year repeat).

Conclusions

DS had lower positivity compared to cytology but had greater sensitivity and specificity. Across all HPV genotypes, DS-negative patients were reassured of very low risk, which would not require colposcopy. Extended genotyping providing four HPV risk groups paired with DS provides more efficient triage compared to current cytology-based approaches.