Background: Cytology-based cervical cancer screening has been one of the most widely used and successful
public health screening interventions in North America for decades. However, there are significant limitations to
cytology alone as a screening test leading to thousands of preventable deaths. Highly sensitive molecular tools
for detection of high-risk HPV, the cause of cervical cancer, are poised to transform cervical cancer screening
programs in the United States and Canada. HPV-based screening is expected to reduce cervical cancer rates
by 30% compared to cytology, due to improved sensitivity of HPV for pre-cancerous lesions. With the improved
negative predictive value of HPV testing, screening intervals can be extended to every 4-5 years leading to
substantial health care savings. Despite these advantages, the loss of specificity with HPV due to detection of
transient infections that do not result in disease, particularly in young women, leads to twice the number of
colposcopy treatments in otherwise well women. Thus, while preventing more cervical cancer cases with less
frequent screening, HPV-based screening could lead to unintended harm for well women of reproductive age
and unsustainable health care system costs if deployed without appropriate screening and triage algorithms.
Triage with alternate molecular methods (ie. mRNA and genotyping) in HPV positive women or adopting these
methods as a primary screen could offer improved specificity without reducing sensitivity. However, if, how and
when to optimally integrate these tests into screening algorithms remains a critical knowledge gap globally.
Aims: This study addresses priority questions from North American policy makers to evaluate the
effectiveness of HPV testing with and without cytology co-testing, determine adverse effects of primary HPV
testing, and to ultimately inform optimal screening algorithms for cervical cancer screening. Specifically the
project will: 1) determine the long-term efficacy (120 months) of HPV-based primary screening after a single
and multiple screening rounds, compared to cytology and co-testing; 2) determine the efficacy of 3 different
HPV assays for triage of HPV positive specimens, and primary screening for precancerous lesions; 3) define
parameters for modeling population and systems-level outcomes of different protocols on cervical cancer rates
Methods: Participants of an established, highly engaged cohort (n=25,223) from a longitudinal randomized
controlled trial comparing primary HPV testing to cytology will be followed to 120 months. Trial participants are
currently 48 months years from baseline and have complete retrospective and prospective cervical cancer
screening, colposcopy and treatment records, with linkage to a population-based cancer registry. Baseline
negative women will be followed to 120 months though the centralized screening program, while a subset of
participants will receive prospective additional HPV testing. Clinical endpoints, sensitivity/specificity and other
parameters will be used for mathematic modelling & cost-effectiveness analysis.
