UMN-Cervical (Minnesota)

Basic Model Attributes
Cancer site	Cervical
Host institution	University of Minnesota
Purpose	The group of UMN models (a standalone Markov model, a microsimulation model and a transmission model) were developed to reflect HPV-induced cervical carcinogenesis stratified by HPV 16, 18 and other high-risk HPV types.
Contact	Shalini Kulasingam (kulas016@umn.edu)

Model overview. The University of Minnesota (UMN) models are a group of models that reflect HPV-induced cervical carcinogenesis stratified by human papillomavirus (HPV)-16 and/or 18 and other high-risk types. (1-3)

HPV transmission. Transmission of HPV infections in males and females is modeled with a dynamic individual-based model, with individual partnerships characterized by sex, age, and sexual activity. Females and males form heterosexual partnerships as they age, and transmission of type-specific HPV can occur as a function of sexual behavior patterns in the population, prevalence of HPV in the population, and female-to-male or male-to-female transmission probabilities of HPV per susceptible-infected partnership. Following clearance of HPV, individuals develop natural immunity, reducing future risk of that same-type of infection. Women with high-risk infection can develop precancerous lesions (i.e., cervical intraepithelial neoplasia (CIN)1, CIN2 or CIN 3), which may regress naturally, and those with CIN 2 or 3 may develop invasive cancer. Death can occur from age- and sex-specific background mortality or excess mortality in women with invasive cervical cancer.

Cervical carcinogenesis. Both the individual-level and Markov cohort models include health states that reflect cervical carcinogenesis associated with HPV-16 and/or 18 and other high-risk types). In these models, women transition between health states, which reflect the cohort’s underlying true health and include HPV infection status, grade of CIN (CIN 1, CIN 2 and CIN 3), and stage of invasive cancer (I through IV). In the cohort model, women enter the model before sexual debut and transition between health states according to probabilities that depend on age, HPV type, type-specific natural immunity, CIN status, and treatment history. Natural immunity is modeled as a reduction in future type-specific infection. Death can occur each year from non-cervical cancer causes from all health states, or from cervical cancer after its onset. Hysterectomy is modeled as a competing risk.

Vaccination.The dynamic model is used to project the effects of HPV vaccination in reducing HPV-16, HPV-18 and other high-risk type infections over time, capturing both direct and indirect benefits. The dynamic model also accounts for the impact of these effects on CIN and cancer.

Screening, diagnosis and treatment of CIN. The dynamic and cohort models can accommodate detailed features of screening strategies, including algorithms that are based on a single test or multiple tests (either in parallel or serial). The models reflect screening, follow-up and treatment recommendations based on American Cancer Society (ACS), US Preventive Services Task Force (USPSTF) and American Society for Colposcopy and Cervical Pathology (ASCCP) guidelines, but assumptions can be modified flexibly. The models both incorporate a detailed post-treatment surveillance component. (3)

Cancer treatment and survival. The models include cancer states by stage (I through IV) and conditional probabilities of survival based on stage of detection. The models also include a separate state for survivors and cancer-related deaths based on data from the Surveillance, Epidemiology, and End Results (SEER) Program.

Calibration and validation. The Markov model was calibrated by varying HPV incidence, CIN progression and regression rates, and probability of symptoms by cancer stage. The parameter set that achieved best visual fit to historic data (in the absence of screening) was selected for analysis. The face validity of the models is assessed by comparing model-projected estimates of age-specific HPV prevalence and age-specific cervical cancer incidence, as well as the lifetime risk of cervical cancer, to empirically observed values from SEER.

References

Kulasingam S, Connelly L, Conway E, Hocking JS, Myers E, Regan DG, et al. A cost-effectiveness analysis of adding a human papillomavirus vaccine to the Australian National Cervical Cancer Screening Program. Sex Health. 2007;4(3):165-75. [Abstract]
Melnikow J, Kulasingam S, Slee C, Helms LJ, Kuppermann M, Birch S, et al. Surveillance after treatment for cervical intraepithelial neoplasia: outcomes, costs, and cost-effectiveness. Obstet Gynecol. 2010;116(5):1158-70. [Abstract]
Kulasingam SL, Havrilesky LJ, Ghebre R, Myers ER. Screening for cervical cancer: a modeling study for the US Preventive Services Task Force. J Low Genit Tract Dis. 2013;17(2):193-202. [Abstract]

Tip: Hover your cursor over the dashed attribute links below for more information. View the details of this model in a grid with other cervical models.

Detailed Package Attributes
Attribute Category	Attribute
Approach
Primary Purpose	Screening evaluation, Epidemiological analysis, Policy evaluation, Population trends,
Features
Intervention	Prevention, Screening, Treatment, Vaccination,
Natural History	Precancer, Dysplasia (CIN1, CIN2 and CIN3 that can either progress or regress.), Virus (HPV 16, -18, 5 HR types (-31, -33, -45, -52, -58) and 9 other HR types (-35, -39, -51, -56, -59, -66, -68, -82) naturally immunity by type.),
Construction
Approach	Micro Simulation, Agent-based, Dynamic Transmission,
Methods	Longitudinal, Likelihood optimization, Stochastic process, State Transition,
Unit of Analysis	Tumor,
Data Source	KPNC,
Census	US,
Cancer Registry	SEER, NMHPVPR,
Linked
Clinical Trial
Survey	NHANES, NHDS,
Meta Analysis
Assumptions
Benefit Factors
Screening	Stage Shift, Temporal Trends, Prevention,
Treatment	Precancer,
Vaccination	Prevention,
Inputs	Incidence, Disease, Grade Distribution (CIN1, CIN2, CIN3), Stage Distribution (Stage 1, 2, 3, 4), Other Conditions,
Screening	Attendance (The model has the potential to model this, but we haven't incorporated it at this stage.), Test Performance, Effect, Risk Adaptive Factors, Incidental Finding Surveillance,
Diagnosis
Precancer	Test Performance,
Treatment
Precancer	Attendance, Efficacy,
Survival	Relative,
Mortality	Other cause, Disease-specific,
Risk Factor	Age, Personal History (Natural immunity and sexual behavior), Sexual behavior, Demography (Hysterectomy due to reasons other than cervical cancer), Natural History, Cost,
Vaccination	Quadrivalent, Bivalent, Nonavalent, Efficacy (Variable),
Outputs	Cost, Incidence,
Disease	Stage Distribution, Grade Distribution (CIN 1, CIN 2, CIN 3), HPV Infection,
Prevalence	Other Conditions (HPV infection (by genotype)),
Treatment
Precancer
Screening
Risk Factor	Smoking, Natural History,
Outcomes	Survival, Life years, QALY, Cause-specific Mortality, All-cause Mortality, Temporal trends, Vaccination effect,
Screening	False Positives, True Positives, False Negatives, True Negatives, Biopsies performed, Unnecessary biopsies, History,
Treatment	History, Overtreatment,
Implementation
Development
Tested Platforms	Windows, Mac OS X,
Language	Java,

2020

Simms KT, Yuill S, Killen J, Smith MA, Kulasingam S, de Kok IMCM, van Ballegooijen M, Burger EA, Regan C, Kim JJ, et al., Historical and projected hysterectomy rates in the USA: Implications for future observed cervical cancer rates and evaluating prevention interventions., Gynecol Oncol, Sept. 1, 2020 [Abstract]
de Kok IMCM, Burger EA, Naber SK, Canfell K, Killen J, Simms K, Kulasingam S, Groene E, Sy S, Kim JJ, et al., The Impact of Different Screening Model Structures on Cervical Cancer Incidence and Mortality Predictions: The Maximum Clinical Incidence Reduction (MCLIR) Methodology., Med Decis Making, May 1, 2020 [Abstract]

2019

Burger EA, de Kok IMCM, Groene E, Killen J, Canfell K, Kulasingam S, Kuntz KM, Matthijsse S, Regan C, Simms K, Smith M, et al., Estimating the Natural History of Cervical Carcinogenesis Using Simulation Models: A CISNET Comparative Analysis, J Natl Cancer Inst, Dec. 10, 2019 [Abstract]

Additional publications by this modeling group

You may also be interested in these publications by this modeling group, which were not supported by the CISNET grant.

2014

Roura E, Castellsagué X, Pawlita M, Travier N, Waterboer T, Margall N, Bosch FX, de Sanjosé S, Dillner J, Gram IT, Tjønneland A, et al., Smoking as a major risk factor for cervical cancer and pre-cancer: results from the EPIC cohort, Int J Cancer, July 15, 2014 [Abstract]

2013

Kulasingam SL, Havrilesky LJ, Ghebre R, Myers ER, Screening for cervical cancer: a modeling study for the US Preventive Services Task Force, J Low Genit Tract Dis, April 1, 2013 [Abstract]

2010

Melnikow J, Kulasingam S, Slee C, Helms LJ, Kuppermann M, Birch S, McGahan CE, Coldman A, Chan BK, Sawaya GF, Surveillance after treatment for cervical intraepithelial neoplasia: outcomes, costs, and cost-effectiveness, Obstet Gynecol, Nov. 1, 2010 [Abstract]

2007

Kulasingam S, Connelly L, Conway E, Hocking JS, Myers E, Regan DG, Roder D, Ross J, Wain G, A cost-effectiveness analysis of adding a human papillomavirus vaccine to the Australian National Cervical Cancer Screening Program, Sex Health, Sept. 1, 2007 [Abstract]