Estimating sensitivity and sojourn time in screening for colorectal cancer: a comparison of statistical approaches

Overcoming underestimation of the association of excess weight with pancreatic cancer due to prediagnostic weight loss: Umbrella review of systematic reviews, meta-analyses, and pooled-analyses

Elevated body mass index (BMI) is linked to increased pancreatic cancer (PC) risk. Cancer-associated weight loss can occur years before the malignancy is diagnosed. This might have led to underestimation of the BMI-PC association. However, it is unknown if and to what extent this issue has been considered in previous epidemiological studies. We searched two databases through February 19, 2024 for systematic reviews, meta-analyses, and pooled analyses examining the BMI-PC association. We extracted information on study design with a special focus on the article's examination of prediagnostic weight loss as a potential source of bias, as well as how included cohort studies addressed this concern. Thirteen review articles, meta-analyses, and pooled analyses were identified. Only five (four pooled analyses, one systematic review) considered prediagnostic weight loss in their analyses. Twenty-four of 32 identified cohort studies reported having excluded initial years of follow-up. However, only 13 studies reported results after such exclusions, and effect estimates generally increased with longer periods of exclusion. We conclude that the association of overweight and obesity with PC risk is likely larger than suggested by published epidemiological evidence. Future studies should pay careful attention to avoid or minimize potential bias resulting from prediagnostic weight loss.

The underestimated preventive effects of flexible sigmoidoscopy screening: re-analysis and meta-analysis of randomized trials

Flexible sigmoidoscopy (FS), which is less invasive, resource intensive and costly than colonoscopy, is among the recommended screening options for colorectal cancer (CRC). Four large randomized trials consistently reported statistically significant, albeit modest effects of screening by FS on CRC incidence. However, their effect estimates included cancers that were already prevalent at recruitment and could not have been prevented by screening. We performed a re-analysis and meta-analysis of two of the trials (including the largest one) to estimate reduction of truly incident cases by a single FS offered between 55 and 64 years of age among the "at risk study population" without prevalent CRC at recruitment. In meta-analyses of data reported after more than 15 years of follow-up, relative risk (95% CI) in intention-to-screen and per-protocol analyses were 0.71 (0.66-0.76) and 0.59 (0.55-0.65) for any CRC, and 0.52 (0.47-0.57) and 0.34 (0.30-0.39) for distal CRC, respectively. These results indicate much stronger effects than those suggested by the original reports and imply that a single screening FS can prevent approximately two out of three distal incident CRC cases within 15 + years of follow-up.

Incidences of colorectal adenomas and cancers under colonoscopy surveillance suggest an accelerated "Big Bang" pathway to CRC in three of the four Lynch syndromes

BACKGROUND

Colorectal cancers (CRCs) in the Lynch syndromes have been assumed to emerge through an accelerated adenoma-carcinoma pathway. In this model adenomas with deficient mismatch repair have an increased probability of acquiring additional cancer driver mutation(s) resulting in more rapid progression to malignancy. If this model was accurate, the success of colonoscopy in preventing CRC would be a function of the intervals between colonoscopies and mean sojourn time of detectable adenomas. Contrary to expectations, colonoscopy did not decrease incidence of CRC in the Lynch syndromes and shorter colonoscopy intervals have not been effective in reducing CRC incidence. The prospective Lynch Syndrome Database (PLSD) was designed to examine these issues in carriers of pathogenic variants of the mis-match repair (path_MMR) genes.

MATERIALS AND METHODS

We examined the CRC and colorectal adenoma incidences in 3,574 path_MLH1, path_MSH2, path_MSH6 and path_PMS2 carriers subjected to regular colonoscopy with polypectomy, and considered the results based on sojourn times and stochastic probability paradigms.

RESULTS

Most of the path_MMR carriers in each genetic group had no adenomas. There was no association between incidences of CRC and the presence of adenomas. There was no CRC observed in path_PMS2 carriers.

CONCLUSIONS

Colonoscopy prevented CRC in path_PMS2 carriers but not in the others. Our findings are consistent with colonoscopy surveillance blocking the adenoma-carcinoma pathway by removing identified adenomas which might otherwise become CRCs. However, in the other carriers most CRCs likely arised from dMMR cells in the crypts that have an increased mutation rate with increased stochastic chaotic probabilities for mutations. Therefore, this mechanism, that may be associated with no or only a short sojourn time of MSI tumours as adenomas, could explain the findings in our previous and current reports.

Reduction in colorectal cancer incidence by screening endoscopy

Colorectal cancer (CRC) incidence rates decreased by up to 50% in older age groups in the USA in the era of the widespread uptake of screening colonoscopy, despite adverse trends in CRC risk factors and increasing CRC incidence at younger ages. However, reported first results from a randomized trial, the NordICC study, suggested rather modest effects of screening colonoscopy. As outlined in this Perspective, the apparent discrepancy between real-world and trial evidence could be explained by strong attenuation of effect estimates from screening endoscopy trials by several factors, including limited screening adherence, widespread uptake of colonoscopy outside the screening offers and the inclusion of prevalent, non-preventable CRC cases in reported numbers of incident cases. Alternative interpretations of screening endoscopy trial results accounting for prevalence bias are in line with trends in CRC incidence reduction in countries offering CRC screening, and should encourage more widespread implementation and uptake of effective CRC screening.

When gold standards are not so golden: prevalence bias in randomized trials on endoscopic colorectal cancer screening

Randomized trials on the effectiveness of screening endoscopy in reducing colorectal cancer (CRC) risk have reported statistically significant, but rather modest reduction of CRC risk by the screening offer. However, risk estimates in these trials included substantial proportions of prevalent CRC cases which were early detected, but could not possibly have been prevented by screening. Thereby, a key principle of randomized prevention trials is violated that only "at risk" persons who do not yet have the disease one aims to prevent should be included in measures of preventive effects. Using recently published data from the Nordic-European Initiative on Colorectal Cancer (NordICC) trial as an example, we illustrate that approaches aimed to account for "prevalence bias" lead to effect estimates that are substantially larger than those reported in the trial and more in line with results from observational studies and real life settings. More rigorous methodological work is needed to develop effective and user-friendly tools to prevent or adjust for prevalence bias in future screening studies.

Test sensitivity in a prospective cancer screening program: A critique of a common proxy measure

The true sensitivity of a cancer screening test, defined as the frequency with which the test returns a positive result if the cancer is present, is a key indicator of diagnostic performance. Given the challenges of directly assessing test sensitivity in a prospective screening program, proxy measures for true sensitivity are frequently reported. We call one such proxy empirical sensitivity, as it is given by the observed ratio of screen-detected cancers to the sum of screen-detected and interval cancers. In the setting of the canonical three-state Markov model for progression from preclinical onset to clinical diagnosis, we formulate a mathematical relationship for how empirical sensitivity varies with the screening interval and the mean preclinical sojourn time and identify conditions under which empirical sensitivity exceeds or falls short of true sensitivity. In particular, when the inter-screening interval is short relative to the mean sojourn time, empirical sensitivity tends to exceed true sensitivity, unless true sensitivity is high. The Breast Cancer Surveillance Consortium (BCSC) has reported an estimate of 0.87 for the empirical sensitivity of digital mammography. We show that this corresponds to a true sensitivity of 0.82 under a mean sojourn time of 3.6 years estimated based on breast cancer screening trials. However, the BCSC estimate of empirical sensitivity corresponds to even lower true sensitivity under more contemporary, longer estimates of mean sojourn time. Consistently applied nomenclature that distinguishes empirical sensitivity from true sensitivity is needed to ensure that published estimates of sensitivity from prospective screening studies are properly interpreted.

Association of Overweight, Obesity, and Recent Weight Loss With Colorectal Cancer Risk

Importance

Overweight and obesity, conditions with rising prevalence in many countries, are associated with increased colorectal cancer (CRC) risk. However, many patients with CRC lose weight before diagnosis, which may lead the association to be underestimated.

Objective

To evaluate the association of body mass index (BMI) and weight change with CRC risk when considering BMI at different time frames, with the intention to account for prediagnostic weight loss.

Design, Setting, and Participants

This population-based case-control study was conducted in southwestern Germany between 2003 and 2021. Cases with a first diagnosis of CRC and controls (frequency matched by age, sex, and county) with comprehensive risk factor information and self-reported weight at different time points were included. Data were analyzed between October 2022 and March 2023.

Exposure

BMI and weight change at different time frames before the time of diagnosis (cases) or recruitment (controls).

Main Outcomes and Measures

Association of BMI and weight change at various points in time before and up to diagnosis with CRC, assessed by multivariable logistic regression with comprehensive confounder adjustment.

Results

A total of 11 887 participants (6434 CRC cases, 5453 controls; median [IQR] age, 69 [61-77] years; 7173 male [60.3%]) were included. At the time of diagnosis, 3998 cases (62.1%) and 3601 controls (66.0%) were overweight or obese, suggesting an inverse association between excess weight and CRC risk. Conversely, we found significant positive associations of overweight (adjusted odds ratio [aOR], 1.27; 95% CI, 1.03-1.56), obesity (aOR, 2.09; 95% CI, 1.61-2.70), and a 5-unit increase in BMI (aOR, 1.35; 95% CI, 1.21-1.50) with CRC risk when using BMI measured 8 to 10 years before diagnosis. High BMI as a risk factor for CRC was increased as earlier periods before diagnosis were examined, with the association being particularly pronounced using BMI at least 8 years before diagnosis. An opposite trend was found for the association of weight loss (at or exceeding 2 kg) with CRC, with the greatest effect sizes occurring for weight loss within 2 years before diagnosis (aOR, 7.52; 95% CI, 5.61-10.09), and gradually decreased for earlier intervals.

Conclusions and Relevance

In this population-based case-control study, accounting for substantial prediagnostic weight loss further highlighted the association of overweight and obesity with CRC risk.

Is the association of overweight and obesity with colorectal cancer underestimated? An umbrella review of systematic reviews and meta-analyses

Although high body-mass index (BMI) is associated with increased risk of developing colorectal cancer (CRC), many CRC patients lose weight before diagnosis. BMI is often reported close to diagnosis, which may have led to underestimation or even reversal of direction of the BMI-CRC association. We aimed to assess if and to what extent potential bias from prediagnostic weight loss has been considered in available epidemiological evidence. We searched PubMed and Web of Science until May 2022 for systematic reviews and meta-analyses investigating the BMI-CRC association. Information on design aspects and results was extracted, including if and how the reviews handled prediagnostic weight loss as a potential source of bias. Additionally, we analyzed how individual cohort studies included in the latest systematic review handled the issue. Overall, 18 reviews were identified. None of them thoroughly considered or discussed prediagnostic weight loss as a potential source of bias. The majority (15/21) of cohorts included in the latest review did not exclude any initial years of follow-up from their main analysis. Although the majority of studies reported having conducted sensitivity analyses in which initial years of follow-up were excluded, results were reported very heterogeneously and mostly for additional exclusions of 1-2 years only. Where explicitly reported, effect estimates mostly increased with increasing length of exclusion. The impact of overweight and obesity on CRC risk may be larger than suggested by the existing epidemiological evidence.

Tumour Growth Models of Breast Cancer for Evaluating Early Detection-A Summary and a Simulation Study

With the advent of nationwide mammography screening programmes, a number of natural history models of breast cancers have been developed and used to assess the effects of screening. The first half of this article provides an overview of a class of these models and describes how they can be used to study latent processes of tumour progression from observational data. The second half of the article describes a simulation study which applies a continuous growth model to illustrate how effects of extending the maximum age of the current Swedish screening programme from 74 to 80 can be evaluated. Compared to no screening, the current and extended programmes reduced breast cancer mortality by 18.5% and 21.7%, respectively. The proportion of screen-detected invasive cancers which were overdiagnosed was estimated to be 1.9% in the current programme and 2.9% in the extended programme. With the help of these breast cancer natural history models, we can better understand the latent processes, and better study the effects of breast cancer screening.

Estimating the Length of the Preclinical Detectable Phase for Open-Angle Glaucoma

Importance

A 50% reduction of glaucoma-related blindness has previously been demonstrated in a population that was screened for open-angle glaucoma. Ongoing screening trials of high-risk populations and forthcoming low-cost screening methods suggest that such screening may become more common in the future. One would then need to estimate a key component of the natural history of chronic disease, the mean preclinical detectable phase (PCDP). Knowledge of the PCDP is essential for the planning and early evaluation of screening programs and has been estimated for several types of cancer that are screened for.

Objective

To estimate the mean PCDP for open-angle glaucoma.

Design, Setting, and Participants

A large population-based screening for open-angle glaucoma was conducted from October 1992 to January 1997 in Malmö, Sweden, including 32 918 participants aged 57 to 77 years. A retrospective medical record review was conducted to assess the prevalence of newly detected cases at the screening, incidence of new cases after the screening, and the expected clinical incidence, ie, the number of new glaucoma cases expected to be detected without a screening. The latter was derived from incident cases in the screened age cohorts before the screening started and from older cohorts not invited to the screening. A total of 2029 patients were included in the current study. Data were analyzed from March 2020 to October 2021.

Main Outcomes and Measures

The length of the mean PCDP was calculated by 2 different methods: first, by dividing the prevalence of screen-detected glaucoma with the clinical incidence, assuming that the screening sensitivity was 100% and second, by using a Markov chain Monte Carlo (MCMC) model simulation that simultaneously derived both the length of the mean PCDP and the sensitivity of the screening.

Results

Of 2029 included patients, 1352 (66.6%) were female. Of 1420 screened patients, the mean age at screening was 67.4 years (95% CI, 67.2-67.7). The mean length of the PCDP of the whole study population was 10.7 years (95% CI, 8.7-13.0) by the prevalence/incidence method and 10.1 years (95% credible interval, 8.9-11.2) by the MCMC method.

Conclusions and Relevance

The mean PCDP was similar for both methods of analysis, approximately 10 years. A mean PCDP of 10 years found in the current study allows for screening with reasonably long intervals, eg, 5 years.

Impact of changes to the interscreening interval and faecal immunochemical test threshold in the national bowel cancer screening programme in England: results from the FIT pilot study

INTRODUCTION

The NHS Bowel Cancer Screening Programme (BCSP) faces endoscopy capacity challenges from the COVID-19 pandemic and plans to lower the screening starting age. This may necessitate modifying the interscreening interval or threshold.

METHODS

We analysed data from the English Faecal Immunochemical Testing (FIT) pilot, comprising 27,238 individuals aged 59-75, screened for colorectal cancer (CRC) using FIT. We estimated screening sensitivity to CRC, adenomas, advanced adenomas (AA) and mean sojourn time of each pathology by faecal haemoglobin (f-Hb) thresholds, then predicted the detection of these abnormalities by interscreening interval and f-Hb threshold.

RESULTS

Current 2-yearly screening with a f-Hb threshold of 120 μg/g was estimated to generate 16,092 colonoscopies, prevent 186 CRCs, detect 1142 CRCs, 7086 adenomas and 4259 AAs per 100,000 screened over 15 years. A higher threshold at 180 μg/g would reduce required colonoscopies to 11,500, prevent 131 CRCs, detect 1077 CRCs, 4961 adenomas and 3184 AAs. A longer interscreening interval of 3 years would reduce required colonoscopies to 10,283, prevent 126 and detect 909 CRCs, 4796 adenomas and 2986 AAs.

CONCLUSION

Increasing the f-Hb threshold was estimated to be more efficient than increasing the interscreening interval regarding overall colonoscopies per screen-benefited cancer. Increasing the interval was more efficient regarding colonoscopies per cancer prevented.

Multistate models for the natural history of cancer progression

BACKGROUND

Multistate models can be effectively used to characterise the natural history of cancer. Inference from such models has previously been useful for setting screening policies.

METHODS

We introduce the basic elements of multistate models and the challenges of applying these models to cancer data. Through simulation studies, we examine (1) the impact of assuming time-homogeneous Markov transition intensities when the intensities depend on the time since entry to the current state (i.e., the process is time-inhomogenous semi-Markov) and (2) the effect on precancer risk estimation when observation times depend on an unmodelled intermediate disease state.

RESULTS

In the settings we examined, we found that misspecifying a time-inhomogenous semi-Markov process as a time-homogeneous Markov process resulted in biased estimates of the mean sojourn times. When screen-detection of the intermediate disease leads to more frequent future screening assessments, there was minimal bias induced compared to when screen-detection of the intermediate disease leads to less frequent screening.

CONCLUSIONS

Multistate models are useful for estimating parameters governing the process dynamics in cancer such as transition rates, sojourn time distributions, and absolute and relative risks. As with most statistical models, to avoid incorrect inference, care should be given to use the appropriate specifications and assumptions.

Faecal immunochemical test after negative colonoscopy may reduce the risk of incident colorectal cancer in a population-based screening programme

OBJECTIVE

Subjects with a positive faecal immunochemical test (FIT) have a much higher likelihood of advanced neoplasms than the general population. Whether FIT-positive subjects with negative colonoscopy should receive subsequent FIT screening remain unclear.

DESIGN

Subjects with a negative colonoscopy after positive FIT in the first screening in the Taiwanese Colorectal Cancer (CRC) Screening Program 2004-2009 were followed until the end of 2014. CRC incidence was compared between those who did and did not receive subsequent FIT screening. Cox regression analysis was conducted, adjusting for major confounders to investigate whether subsequent FIT was associated with lower risk of incident CRC.

RESULTS

The study cohort was comprised of 9179 subjects who had negative diagnostic colonoscopy after positive FIT in 2004-2009, of whom 6195 received subsequent FIT during the study period. The CRC incidence (per 1000 person years) was 1.34 in those who received subsequent FIT and 2.69 in those who did not, with corresponding adjusted HR (aHR) of 0.47 (95% CI 0.31 to 0.71). Lower adenoma detection rate of diagnostic colonoscopy was associated with higher risk of incident CRC but became non-significant in multivariable analysis after adjustment for subsequent FIT. Higher baseline faecal haemoglobin concentration (FHbC, μg haemoglobin/g faeces) was associated with increased risk of incident CRC (reference: FHbC=20-39; aHR=1.93 (1.04-3.56), 0.95 (0.45-2.00), 2.26 (1.16-4.43) and 2.44 (1.44-4.12) for FHbC=40-59, 60-99, 100-149 and ≥150, respectively).

CONCLUSION

Subsequent FIT should be scheduled after negative colonoscopy to detect missed neoplasms and reduce the risk of incident CRC in a national FIT screening programme.

Classifying interval cancers as false negatives or newly occurring in fecal immunochemical testing

OBJECTIVE

To classify interval colorectal cancers as false negatives or newly occurring cases in a biennial Fecal immunochemical test (FIT) screening program and by various interscreening intervals.

SETTING

Data from the Taiwanese biennial colorectal cancer screening program involving FIT from 2004 to 2014 were used to estimate the incidence rate of asymptomatic colorectal cancer and the rate of its subsequent progression to clinical mode.

METHODS

The sensitivity of detecting asymptomatic colorectal cancers excluding newly developed colorectal cancers was compared to the conventional estimate of sensitivity, the complementary FIT interval cancer rate as a percentage of the expected incidence rate ((1-I/E)%). The relative contribution of newly developed or false-negative cases to FIT interval colorectal cancers was estimated by age and interscreening intervals.

RESULTS

The Taiwanese biennial fecal immunochemical test screening program had a conventional sensitivity estimate of 70.2%. After newly developed colorectal cancers were separated from FIT interval cancers, the ability to detect asymptomatic colorectal cancers increased to 75.5%. FIT interval colorectal cancers from the biennial program mainly resulted from newly developed colorectal cancers (68.8%). The corresponding figures decreased to 61.1% for the annual program but increased to 74.7% for the triennial program. The preponderance of newly developed colorectal cancers among FIT interval cancers was more prominent in screenees aged 50-59  than in those aged 60-69.

CONCLUSIONS

Newly developed colorectal cancers showed a predominance among the FIT interval colorectal cancers in particular in the younger population screened. It is desirable to identify high-risk individuals to offer them a short interscreening interval or advanced detection methods to reduce their odds of developing interval cancer.

Surveillance for Hepatocellular Carcinoma Also Improves Survival of Incidentally Detected Intrahepatic Cholangiocarcinoma Arisen in Liver Cirrhosis

BACKGROUND

Due to its poor survival, intrahepatic cholangiocarcinoma (ICC) is held to be a much more aggressive cancer than hepatocellular carcinoma (HCC). In most published series, patients were diagnosed when symptomatic. However, ICC is now increasingly being discovered during the surveillance for HCC in cirrhosis. Whether this earlier detection of ICC is associated with an equally dismal prognosis or not is unknown.

METHODS

This is amulticenter retrospective study of consecutive ICC patients. Patients were stratified into subgroups according to the absence/presence of cirrhosis. A propensity score matching was performed to reduce the potential biases. Cirrhotic patients were further stratified according to their surveillance status. The lead-time bias and its potential effects were also estimated.

RESULTS

We gathered 184 patients. Eighty-five patients (46.2%) were cirrhotic. Liver cirrhosis was not related to a worse overall survival (33.0 vs. 32.0 months, p = 0.800) even after the propensity score analysis (43.0 in vs. 44.0 months in 54 pairs of patients, p = 0.878). Among the cirrhotic population, 47 (55.3%) patients had received a diagnosis of ICC during a surveillance programme. The 2 subgroups differed in maximum tumour dimensions (30 vs. 48 mm in surveyed and non-surveyed patients, respectively). Surveyed patients were more likely to receive surgical treatments (59.8 vs. 28.9%, p = 0.003). Overall survival was higher in surveyed patients (51.0 vs. 21.0 months, p < 0.001). These benefits were confirmed after correcting for the lead-time bias.

CONCLUSIONS

Cirrhotic patients have different clinical presentation and outcomes of ICC according to their surveillance status. In our series, ICC in cirrhosis was not associated with worse OS. Cirrhosis itself should not discourage either surgical or non-surgical treatments.

Incidence, Characteristics, and Outcomes of Interval Breast Cancers Compared With Screening-Detected Breast Cancers

IMPORTANCE

Breast cancer comprises a highly heterogeneous group of diseases. Many breast cancers, particularly the more lethal ones, may not satisfy the assumptions about biology and natural history of breast cancer necessary for screening mammography to be effective.

OBJECTIVES

To compare tumor characteristics of breast cancers diagnosed within 2 years of a normal screening mammogram (interval breast cancer [IBC]) with those of screen-detected breast cancers (SBC) and to compare breast cancer-specific mortality of IBC with SBC.

DESIGN, SETTING, AND PARTICIPANTS

In this registry-based cohort study, we collected data about relevant tumor- and patient-related variables on women diagnosed with breast cancer between January 2004 and June 2010 who participated in the population-based screening program in Manitoba, Canada, and those diagnosed with breast cancer outside the screening program in the province. We performed multinomial logistic regression analysis to assess tumor and patient characteristics associated with a diagnosis of IBC compared with SBC. Competing risk analysis was performed to examine risk of death by cancer detection method.

EXPOSURES

Breast cancer diagnosis.

MAIN OUTCOMES AND MEASURES

Differences in tumor characteristics and breast cancer-specific mortality.

RESULTS

A total of 69 025 women aged 50 to 64 years had 212 screening mammograms during the study period. There were 1687 breast cancer diagnoses (705 SBC, 206 IBC, 275 were noncompliant, and 501 were detected outside the screening program), and 225 deaths (170 breast cancer-specific deaths). Interval cancers were more likely than SBC to be of high grade and estrogen receptor negative (odds ratio [OR], 6.33; 95% CI, 3.73-10.75; P < .001; and OR, 2.88; 95% CI, 2.01-4.13; P < .001, respectively). After a median follow-up of 7 years, breast cancer-specific mortality was significantly higher for IBC compared with SBC cancers (hazard ratio [HR] 3.55; 95% CI, 2.01-6.28; P < .001), for a sojorn time of 2 years. Non-breast cancer mortality was similar between IBC and SBC (HR, 1.33; 95% CI, 0.43-4.15).

CONCLUSIONS AND RELEVANCE

In this cohort study, interval cancers were highly prevalent in women participating in population screening, represented a worse biology, and had a hazard for breast cancer death more than 3-fold that for SBC. Strategies beyond current mammographic screening practices are needed to reduce incidence, improve detection, and reduce deaths from these potentially lethal breast cancers.

Estimating the natural progression of non-invasive ductal carcinoma in situ breast cancer lesions using screening data

OBJECTIVES

In addition to invasive breast cancer, mammography screening often detects preinvasive ductal carcinoma in situ (DCIS) lesions. The natural progression of DCIS is largely unknown, leading to uncertainty regarding treatment. The natural history of invasive breast cancer has been studied using screening data. DCIS modeling is more complicated because lesions might progress to clinical DCIS, preclinical invasive cancer, or may also regress to a state undetectable by screening. We have here developed a Markov model for DCIS progression, building on the established invasive breast cancer model.

METHODS

We present formulas for the probability of DCIS detection by time since last screening under a Markov model of DCIS progression. Progression rates were estimated by maximum likelihood estimation using BreastScreen Norway data from 1995-2002 for 336,533 women (including 399 DCIS cases) aged 50-69. As DCIS incidence varies by age, county, and mammography modality (digital vs. analog film), a Poisson regression approach was used to align the input data.

RESULTS

Estimated mean sojourn time in preclinical, screening-detectable DCIS phase was 3.1 years (95% confidence interval: 1.3, 7.6) with a screening sensitivity of 60% (95% confidence interval: 32%, 93%). No DCIS was estimated to be non-progressive.

CONCLUSION

Most preclinical DCIS lesions progress or regress with a moderate sojourn time in the screening-detectable phase. While DCIS mean sojourn time could be deduced from DCIS data, any estimate of preclinical DCIS progressing to invasive breast cancer must include data on invasive cancers to avoid strong, probably unrealistic, assumptions.

Overestimated Sensitivity of Fecal Immunochemical Tests in Screening Cohorts With Registry-Based Follow-up

OBJECTIVES

Several recent studies have reported very high estimates of sensitivity and specificity of fecal immunochemical tests (FITs) at seemingly high levels of precision using registry-based follow-up of participants in very large FIT-based screening programs. We aimed to assess the validity of estimates of diagnostic performance parameters derived by this indirect approach.

METHODS

We modeled expected values of sensitivity and specificity of colorectal cancer detection in studies using the indirect approach and their deviation from true values under a broad range of plausible assumptions, and we compared these expected values with recently reported estimates of FIT sensitivity and specificity from such studies.

RESULTS

Using a sensitivity of 75% and specificity of 93.6% (from studies using a direct approach, i.e., colonoscopy follow-up of all participants), the indirect approach would be expected to yield sensitivities between 84.5% and 91.1% and specificities between 93.4% and 93.6% under a range of realistic assumptions regarding colonoscopic follow-up rates of positive FITs and clinical manifestation rates of preclinical colorectal cancer.

DISCUSSION

Very high sensitivities of FITs recently reported with seemingly very high levels of precision by several large-scale registry-based studies, which are in line with expected results based on our model calculations, are likely to be strongly overestimated and need to be interpreted with due caution.

Statistical models of tumour onset and growth for modern breast cancer screening cohorts

Historically, multi-state Markov models have been used to study breast cancer incidence and mammography screening effectiveness. In recent years, more biologically motivated continuous tumour growth models have emerged as alternatives. However, a number of challenges remain for these models to make use of the wealth of information available in large mammography cohort data. In particular, methodology is needed to address random left truncation and individual, asynchronous screening. We present a comprehensive continuous random effects model for the natural history of breast cancer. It models the unobservable processes of tumour onset, tumour growth, screening sensitivity, and symptomatic detection. We show how the unknown model parameter values can be jointly estimated using a prospective cohort with diagnostic data on age and tumour size at diagnosis, and individual screening histories. We also present a microsimulation study calibrated to population breast cancer incidence data, and to data on mode of detection and tumour size. We highlight the importance of adjusting for random left truncation, derive tumour doubling time distributions for screen-detected and interval cancers, and present results concerning the relationship between tumour presence time and age at diagnosis.

Test sensitivity of mammography and mean sojourn time over 40 years of breast cancer screening in Nijmegen (The Netherlands)

Objectives

We investigated whether changes in mammographic technique and screening policy have improved mammographic sensitivity, and elongated the mean sojourn time, since the introduction of biennial breast cancer screening in Nijmegen, the Netherlands, in 1975.

Methods

Maximum likelihood estimation, non-linear regression, and Markov Chain Monte Carlo simulation were used to estimate test sensitivity, mean sojourn time, and underlying breast cancer incidence in four time periods, covering 40 years of breast cancer screening in Nijmegen (1975–2012).

Results

Maximum likelihood estimation generated an estimated test sensitivity of approximately 90% and a mean sojourn time around three years, while the estimates based on non-linear regression and Markov Chain Monte Carlo simulation were 80% and four years, respectively. All three methods estimated a rise in the underlying breast cancer incidence over time, with approximately one case more per 1000 women per year in the final period compared with the first period.

Conclusions

The three methods showed a slightly higher mammographic sensitivity and a longer mean sojourn time in the last period, after the introduction of digital mammography. Estimates were more realistic for the more sophisticated methods, non-linear regression and Markov Chain Monte Carlo simulation, while the simple closed form approximation of maximum likelihood estimation led to rather high estimates for sensitivity in the early periods.

Estimating age group-dependent sensitivity and mean sojourn time in colorectal cancer screening

OBJECTIVE

In evaluating the efficacy of cancer screening programmes, sojourn time (duration of the preclinical detectable phase) and sensitivity of the screening test are the two key parameters. Studies suggest that in breast cancer screening, both parameters may vary depending on age at the time of screening, but few studies have examined other cancers. We expanded an existing probability model for periodic screening by performing simultaneous estimation of age group-dependent and sensitivity at preclinical onset time, and tested the expanded model using data from the Korean National Colorectal Cancer Screening Programme.

METHODS

Simulation studies were conducted to assess the performance of the proposed probability model. The method was then applied to the analysis of 376,542 participants aged 50 or over who underwent fecal occult blood testing (FOBT) as part of the National Colorectal Cancer Screening Programme between 2004 and 2007. Age group-dependent mean sojourn time and screening sensitivity of FOBT for colorectal cancer were derived using maximum likelihood estimation.

RESULTS

The method performed well in terms of bias, standard deviation, and coverage probability. National Colorectal Cancer Screening Programme data results indicated that the sensitivity of FOBT to detect colorectal cancer increases with age, while mean sojourn time decreases with age (approximately 4.3 years for participants aged 50-54, 3.9 years at age 55-59, 3.4 years at age 60-64, and 3.6 years at age 65-69, with corresponding sensitivity estimates around 41%, 47%, 45%, and 51%, respectively).

CONCLUSION

Simulation studies showed that the proposed stochastic model considering both mean sojourn time and sensitivity yields highly accurate results.

Critical Analysis of Markov Models Used for the Economic Evaluation of Colorectal Cancer Screening: A Systematic Review

BACKGROUND

The economic evaluation of colorectal cancer screening is challenging because of the need to model the underlying unobservable natural history of the disease.

OBJECTIVES

To describe the available Markov models and to critically analyze their main structural assumptions.

METHODS

A systematic search was performed in eight relevant databases (MEDLINE, Embase, Econlit, National Health Service Economic Evaluation Database, Health Economic Evaluations Database, Health Technology Assessment database, Cost-Effective Analysis Registry, and European Network of Health Economics Evaluation Databases), identifying 34 models that met the inclusion criteria. A comparative analysis of model structure and parameterization was conducted using two checklists and guidelines for cost-effectiveness screening models.

RESULTS

Two modeling techniques were identified. One strategy used a Markov model to reproduce the natural history of the disease and an overlaying model that reproduced the screening process, whereas the other used a single model to represent a screening program. Most of the studies included only adenoma-carcinoma sequences, a few included de novo cancer, and none included the serrated pathway. Parameterization of adenoma dwell time, sojourn time, and surveillance differed between studies, and there was a lack of validation and statistical calibration against local epidemiological data. Most of the studies analyzed failed to perform an adequate literature review and synthesis of diagnostic accuracy properties of the screening tests modeled.

CONCLUSIONS

Several strategies to model colorectal cancer screening have been developed, but many challenges remain to adequately represent the natural history of the disease and the screening process. Structural uncertainty analysis could be a useful strategy for understanding the impact of the assumptions of different models on cost-effectiveness results.

International variation in the prevalence of preclinical colorectal cancer: Implications for predictive values of noninvasive screening tests and potential target populations for screening

Screening for colorectal cancer (CRC) is implemented in an increasing number of countries. We aimed to assess international variation in the prevalence of preclinical CRC and the resulting variation in positive and negative predictive values (PPVs, NPVs) of existing and potential CRC screening tests in various countries. Using age‐ and sex‐specific CRC incidence data and transition rates from preclinical to clinical CRC we estimated overall and age‐ and sex‐specific prevalence of preclinical CRC in the target population aged 50–74 years in different parts of the world. These prevalence estimates were used to derive PPVs and NPVs for existing and potential noninvasive screening tests with varying levels of sensitivity and specificity. Within all regions and countries, prevalence strongly increases with age and is higher in men than in women. In addition, major variation was seen between regions and countries, with overall prevalence varying between 1 and 0.1%. As a result, PPVs are expected to strongly vary between ∼10% for men in high incidence countries, such as Australia and Germany, and 1% for women in low incidence countries, whereas NPVs are expected to be consistently well above 99%. Variation in CRC prevalence profoundly affects expected PPVs of screening tests, and PPVs should be carefully considered when decisions on screening tests and strategies are made for specific populations and health care systems. Here, we provide estimates of preclinical CRC and expected PPVs and NPVs of noninvasive screening tests, which may enhance the empirical basis for planning of population‐based CRC screening strategies.

What's new?

Colorectal cancer (CRC) screening is implemented in an increasing number of countries, usually in a two‐step approach consisting in a noninvasive test followed by colonoscopy in case of a positive result. Prevalence of preclinical colorectal cancer strongly affects screening efficiency, but such data is scarce. Here, the authors provide detailed age‐ and sex‐specific preclinical CRC prevalence estimates for various countries and geographical regions and show their implications on expected positive and negative predictive values of existing and potential noninvasive screening tests. Knowledge of these predictive values should enhance the empirical basis for decisions on CRC screening tests and target populations.

A Review of Estimation of Key Parameters and Lead Time in Cancer Screening

Early detection combined with effective treatments are the only ways to fight against cancer, and cancer screening is the primary technique for early detection. Although mass cancer screening has been carried out for decades, there are many unsolved problems, and the statistical theory of cancer screening is still under developed. Screening sensitivity, time duration in the preclinical state, and time duration in the disease free state are the three key parameters, which are critical in cancer screening, since all other estimates are functions of the three key parameters. Lead time is the diagnosis time advanced by screening, and it serves as a measurement of effectiveness of screening programs. In this article, we provide a review for major probability models and statistical methodologies that have been developed on the estimation of the three key parameters and the lead timedistributions. These methods can be applied to screening of other chronic diseases after slight modifications.

Predicting the effectiveness of the Finnish population-based colorectal cancer screening programme

Objective

Because colorectal cancer (CRC) has a long natural history, estimating the effectiveness of CRC screening programmes requires long-term follow-up. As an alternative, we here demonstrate the use of a temporal multi-state natural history model to predict the effectiveness of CRC screening.

Methods

In the Finnish population-based biennial CRC screening programme using faecal occult blood tests (FOBT), which was conducted in a randomised health services study, we estimated the pre-clinical incidence, the mean sojourn time (MST), and the sensitivity of FOBT using a Markov model to analyse data from 2004 to 2007. These estimates were applied to predict, through simulation, the effects of five rounds of screening on the relative rate of reducing advanced CRC with 6 years of follow-up, and on the reduction in mortality with 10 years of follow-up, in a cohort of 500,000 subjects aged 60 to 69.

Results

For localised and non-localised CRC, respectively, the MST was 2.06 and 1.36 years and the sensitivity estimates were 65.12% and 73.70%. The predicted relative risk of non-localised CRC and death from CRC in the screened compared with the control population was 0.86 (95% CI: 0.79–0.98) and 0.91 (95% CI: 0.85–1.02), respectively.

Conclusion

Based on the preliminary results of the Finnish CRC screening programme, our model predicted a 9% reduction in CRC mortality and a 14% reduction in advanced CRC.

Comparison between alcohol- and hepatitis C virus-related hepatocellular carcinoma: clinical presentation, treatment and outcome

BACKGROUND

Hepatitis C virus (HCV) and alcohol abuse are the main risk factors for hepatocellular carcinoma (HCC) in Western countries.

AIM

To investigate the role of alcoholic aetiology on clinical presentation, treatment and outcome of HCC as well as on each Barcelona Clinic Liver Cancer (BCLC) stage, as compared to HCV-related HCCs.

METHODS

A total of 1642 HCV and 573 alcoholic patients from the Italian Liver Cancer (ITA.LI.CA) database, diagnosed with HCC between January 2000 and December 2012 were compared for age, gender, type of diagnosis, tumour burden, portal vein thrombosis (PVT), oesophageal varices, liver function tests, alpha-fetoprotein, BCLC, treatment and survival. Aetiology was tested as predictor of survival in multivariate Cox regression models and according to HCC stages.

RESULTS

Cirrhosis was present in 96% of cases in both groups. Alcoholic patients were younger, more likely male, with HCC diagnosed outside surveillance, in intermediate/terminal BCLC stage and had worse liver function. After adjustment for the lead-time, median (95% CI) overall survival (OS) was 27.4 months (21.5-33.2) in alcoholic and 33.6 months (30.7-36.5) in HCV patients (P = 0.021). The prognostic role of aetiology disappeared when survival was assessed in each BCLC stage and in the Cox regression multivariate models.

CONCLUSIONS

Alcoholic aetiology affects survival of HCC patients through its negative effects on secondary prevention and cancer presentation but not through a greater cancer aggressiveness or worse treatment result. In fact, survival adjusted for confounding factors was similar in alcoholic and HCV patients.

[Interval cancers and episode sensitivity in population-based screening programmes for colorectal cancer: a systematic review]

OBJECTIVE

To describe interval cancers (IC) and the sensitivity of colorectal cancer (CRC) screening programmes.

METHODS

A systematic review of the literature was conducted through a MEDLINE (PubMed) search. The search strategy combined the terms 'interval cancer', 'false negative', 'mass screening', 'screening' 'early detection of cancer', 'colorectal cancer' and 'bowel cancer'. Inclusion criteria consisted of population-based screening programmes, original articles written in English or Spanish and publication dates between 1999/01/01 and 2015/02/28. A narrative synthesis of the included articles was performed detailing the characteristics of the screening programmes, the IC rate, and the information sources used in each study.

RESULTS

Thirteen articles were included. The episode sensitivity of CRC screening programmes ranged from 42.2% to 65.3% in programmes using the guaiac test and between 59.1% and 87.0% with the immunochemical test. We found a higher proportion of women who were diagnosed with IC and these lesions were mainly located in the proximal colon.

CONCLUSION

There is wide variability in the IC rate in CRC programmes. To ensure comparability between programmes, there is a need for consensus on the working definition of IC and the methods used for their identification and quantification.

Interval cancers in a population-based screening program for colorectal cancer in catalonia, Spain

Objective. To analyze interval cancers among participants in a screening program for colorectal cancer (CRC) during four screening rounds. Methods. The study population consisted of participants of a fecal occult blood test-based screening program from February 2000 to September 2010, with a 30-month follow-up (n = 30,480). We used hospital administration data to identify CRC. An interval cancer was defined as an invasive cancer diagnosed within 30 months of a negative screening result and before the next recommended examination. Gender, age, stage, and site distribution of interval cancers were compared with those in the screen-detected group. Results. Within the study period, 97 tumors were screen-detected and 74 tumors were diagnosed after a negative screening. In addition, 17 CRC (18.3%) were found after an inconclusive result and 2 cases were diagnosed within the surveillance interval (2.1%). There was an increase of interval cancers over the four rounds (from 32.4% to 46.0%). When compared with screen-detected cancers, interval cancers were found predominantly in the rectum (OR: 3.66; 95% CI: 1.51-8.88) and at more advanced stages (P = 0.025). Conclusion. There are large numbers of cancer that are not detected through fecal occult blood test-based screening. The low sensitivity should be emphasized to ensure that individuals with symptoms are not falsely reassured.

Symptom lead times in lung and colorectal cancers: what are the benefits of symptom-based approaches to early diagnosis?

BACKGROUND

Individuals with undiagnosed lung and colorectal cancers present with non-specific symptoms in primary care more often than matched controls. Increased access to diagnostic services for patients with symptoms generates more early-stage diagnoses, but the mechanisms for this are only partially understood.

METHODS

We re-analysed a UK-based case-control study to estimate the Symptom Lead Time (SLT) distribution for a range of potential symptom criteria for investigation. Symptom Lead Time is the time between symptoms caused by cancer and eventual diagnosis, and is analogous to Lead Time in a screening programme. We also estimated the proportion of symptoms in lung and colorectal cancer cases that are actually caused by the cancer.

RESULTS

Mean Symptom Lead Times were between 4.1 and 6.0 months, with medians between 2.0 and 3.2 months. Symptom Lead Time did not depend on stage at diagnosis, nor which criteria for investigation are adopted. Depending on the criteria, an estimated 27-48% of symptoms in individuals with as yet undiagnosed lung cancer, and 12-32% with undiagnosed colorectal cancer are not caused by the cancer.

CONCLUSIONS

In most cancer cases detected by a symptom-based programme, the symptoms are caused by cancer. These cases have a short lead time and benefit relatively little. However, in a significant minority of cases cancer detection is serendipitous. This group experiences the benefits of a standard screening programme, a substantial mean lead time and a higher probability of early-stage diagnosis.

Symptom lead time distribution in lung cancer: natural history and prospects for early diagnosis

BACKGROUND

Before their diagnosis, patients with cancer present in primary care more frequently than do matched controls. This has raised hopes that earlier investigation in primary care could lead to earlier stage at diagnosis.

METHODS

We re-analysed primary care symptom data collected from 247 lung cancer cases and 1235 matched controls in Devon, UK. We identified the most sensitive and specific definition of symptoms, and estimated its incidence in cases and controls prior to diagnosis. We estimated the symptom lead time (SLT) distribution (the time between symptoms attributable to cancer and diagnosis), taking account of the investigations already carried out in primary care. The impact of route of diagnosis on stage at diagnosis was also examined.

RESULTS

Symptom incidence in cases was higher than in controls 2 years before diagnosis, accelerating markedly in the last 6 months. The median SLT was under 3 months, with mean 5.3 months [95% credible interval (CrI) 4.5-6.1] and did not differ by stage at diagnosis. An earlier stage at diagnosis was observed in patients identified through chest X-ray originated in primary care.

CONCLUSIONS

Most symptoms preceded clinical diagnosis by only a few months. Symptom-based investigation would lengthen lead times and result in earlier stage at diagnosis in a small proportion of cases, but would be far less effective than standard screening targeted at smokers.

Estimation of lead-time bias and its impact on the outcome of surveillance for the early diagnosis of hepatocellular carcinoma

BACKGROUND & AIMS

Lead-time is the time by which diagnosis is anticipated by screening/surveillance with respect to the symptomatic detection of a disease. Any screening program, including surveillance for hepatocellular carcinoma (HCC), is subject to lead-time bias. Data regarding lead-time for HCC are lacking. Aims of the present study were to calculate lead-time and to assess its impact on the benefit obtainable from the surveillance of cirrhotic patients.

METHODS

One-thousand three-hundred and eighty Child-Pugh class A/B patients from the ITA.LI.CA database, in whom HCC was detected during semiannual surveillance (n = 850), annual surveillance (n = 234) or when patients came when symptomatic (n = 296), were selected. Lead-time was estimated by means of appropriate formulas and Monte Carlo simulation, including 1000 patients for each arm.

RESULTS

The 5-year overall survival after HCC diagnosis was 32.7% in semiannually surveilled patients, 25.2% in annually surveilled patients, and 12.2% in symptomatic patients (p<0.001). In a 10-year follow-up perspective, the median lead-time calculated for all surveilled patients was 6.5 months (7.2 for semiannual and 4.1 for annual surveillance). Lead-time bias accounted for most of the surveillance benefit until the third year of follow-up after HCC diagnosis. However, even after lead-time adjustment, semiannual surveillance maintained a survival benefit over symptomatic diagnosis (number of patients needed to screen = 13), as did annual surveillance (18 patients).

CONCLUSIONS

Lead-time bias is the main determinant of the short-term benefit provided by surveillance for HCC, but this benefit becomes factual in a long-term perspective, confirming the clinical utility of an anticipated diagnosis of HCC.

Parameter estimates for invasive breast cancer progression in the Canadian National Breast Screening Study

Background:

The aim of screening is to detect a cancer in the preclinical state. However, a false-positive or a false-negative test result is a real possibility.

Methods:

We describe invasive breast cancer progression in the Canadian National Breast Screening Study and construct progression models with and without covariates. The effect of risk factors on transition intensities and false-negative probability is investigated. We estimate the transition rates, the sojourn time and sensitivity of diagnostic tests for women aged 40–49 and 50–59.

Results:

Although younger women have a slower transition rate from healthy state to preclinical, their screen-detected tumour becomes evident sooner. Women aged 50–59 have a higher mortality rate compared with younger women. The mean sojourn times for women aged 40–49 and 50–59 are 2.5 years (95% CI: 1.7, 3.8) and 3.0 years (95% CI: 2.1, 4.3), respectively. Sensitivity of diagnostic procedures for older women is estimated to be 0.75 (95% CI: 0.55, 0.88), while women aged 40–49 have a lower sensitivity (0.61, 95% CI: 0.42, 0.77). Age is the only factor that affects the false-negative probability. For women aged 40–49, ‘age at entry', ‘history of breast disease' and ‘families with breast cancer' are found to be significant for some of the transition rates. For the age-group 50–59, ‘age at entry', ‘history of breast disease', ‘menstruation length' and ‘number of live births' are found to affect the transition rates.

Conclusion:

Modelling and estimating the parameters of cancer progression are essential steps towards evaluating the effectiveness of screening policies. The parameters include the transition rates, the preclinical sojourn time, the sensitivity, and the effect of different risk factors on cancer progression.

Modelling the overdiagnosis of breast cancer due to mammography screening in women aged 40 to 49 in the United Kingdom

INTRODUCTION

Overdiagnosis of breast cancer due to mammography screening, defined as the diagnosis of screen-detected cancers that would not have presented clinically in a women's lifetime in the absence of screening, has emerged as a highly contentious issue, as harm caused may question the benefit of mammographic screening. Most studies included women over 50 years old and little information is available for younger women.

METHODS

We estimated the overdiagnosis of breast cancer due to screening in women aged 40 to 49 years using data from a randomised trial of annual mammographic screening starting at age 40 conducted in the UK. A six-state Markov model was constructed to estimate the sensitivity of mammography for invasive and in situ breast cancer and the screen-detectable mean sojourn time for non-progressive in situ, progressive in situ, and invasive breast cancer. Then, a 10-state simulation model of cancer progression, screening, and death, was developed to estimate overdiagnosis attributable to screening.

RESULTS

The sensitivity of mammography for invasive and in situ breast cancers was 90% (95% CI, 72 to 99) and 82% (43 to 99), respectively. The screen-detectable mean sojourn time of preclinical non-progressive and progressive in situ cancers was 1.3 (0.4 to 3.4) and 0.11 (0.05 to 0.19) years, respectively, and 0.8 years (0.6 to 1.2) for preclinical invasive breast cancer. The proportion of screen-detected in situ cancers that were non-progressive was 55% (25 to 77) for the first and 40% (22 to 60) for subsequent screens. In our main analysis, overdiagnosis was estimated as 0.7% of screen-detected cancers. A sensitivity analysis, covering a wide range of alternative scenarios, yielded a range of 0.5% to 2.9%.

CONCLUSION

Although a high proportion of screen-detected in situ cancers were non-progressive, a majority of these would have presented clinically in the absence of screening. The extent of overdiagnosis due to screening in women aged 40 to 49 was small. Results also suggest annual screening is most suitable for women aged 40 to 49 in the United Kingdom due to short cancer sojourn times.

Estimated mean sojourn time associated with hemoccult SENSA for detection of proximal and distal colorectal cancer

BACKGROUND

Sojourn time is the length of the preclinical screen-detectable phase, a period when a test can detect asymptomatic disease. Mean sojourn time (MST) is an important factor in determining appropriate screening intervals. Available estimates of MST for colorectal cancer (CRC) are imprecise and are associated with the older Hemoccult II test. This article presents MST estimates associated with the newer Hemoccult SENSA test and describes differences in MST by the location of cancer in the colorectum and age at the time of screening.

METHODS

MST was estimated from a cohort of 42,079 patients who underwent Hemoccult SENSA between January 1, 1997 and December 31, 2010. The precision of MST estimates was improved by incorporating information from a meta-analysis of the sensitivity of Hemoccult SENSA into the analytic model.

RESULTS

Estimated MST for cancers in the proximal and distal colorectum, with 95% credible intervals (CrI) in years, were: 3.86 (1.55-6.91) and 3.35 (2.11-4.93) among 45- to 54-year olds; 3.78 (2.18-5.77) and 2.24 (1.48-3.17) among 55- to 64-year olds; and 2.70 (1.41-4.31) and 2.10 (1.34-3.04) among 65- to 74-year olds.

CONCLUSIONS

MST associated with Hemoccult SENSA was longer for CRC in the proximal versus distal colon. We found no evidence that MST increases with age and some evidence that it may decrease.

IMPACT

These results add new information about the natural history of CRC and information about the performance of Hemoccult SENSA.

Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths

BACKGROUND

In the National Polyp Study (NPS), colorectal cancer was prevented by colonoscopic removal of adenomatous polyps. We evaluated the long-term effect of colonoscopic polypectomy in a study on mortality from colorectal cancer.

METHODS

We included in this analysis all patients prospectively referred for initial colonoscopy (between 1980 and 1990) at NPS clinical centers who had polyps (adenomas and nonadenomas). The National Death Index was used to identify deaths and to determine the cause of death; follow-up time was as long as 23 years. Mortality from colorectal cancer among patients with adenomas removed was compared with the expected incidence-based mortality from colorectal cancer in the general population, as estimated from the Surveillance Epidemiology and End Results (SEER) Program, and with the observed mortality from colorectal cancer among patients with nonadenomatous polyps (internal control group).

RESULTS

Among 2602 patients who had adenomas removed during participation in the study, after a median of 15.8 years, 1246 patients had died from any cause and 12 had died from colorectal cancer. Given an estimated 25.4 expected deaths from colorectal cancer in the general population, the standardized incidence-based mortality ratio was 0.47 (95% confidence interval [CI], 0.26 to 0.80) with colonoscopic polypectomy, suggesting a 53% reduction in mortality. Mortality from colorectal cancer was similar among patients with adenomas and those with nonadenomatous polyps during the first 10 years after polypectomy (relative risk, 1.2; 95% CI, 0.1 to 10.6).

CONCLUSIONS

These findings support the hypothesis that colonoscopic removal of adenomatous polyps prevents death from colorectal cancer. (Funded by the National Cancer Institute and others.).

Sojourn time of preclinical colorectal cancer by sex and age: estimates from the German national screening colonoscopy database

The sojourn time of preclinical colorectal cancer is a critical parameter in modeling effectiveness and cost-effectiveness of colorectal cancer screening. For ethical reasons, it cannot be observed directly, and available estimates are based mostly on relatively small historic data sets that do not include differentiation by age and sex. The authors derived sex- and age-specific estimates (age groups: 55-59, 60-64, 65-69, 70-74, 75-79, and ≥80 years) of mean sojourn time, combining data from the German national screening colonoscopy registry (based on 1.88 million records) and data from population-based cancer registries (population base: 37.9 million people) for the years 2003-2006. Estimates of mean sojourn time were similar for both sexes and all age groups and ranged from 4.5 years (95% confidence interval: 4.1, 4.8) to 5.8 years (95% confidence interval: 5.3, 6.3) for the subgroups assessed. Sensitivity analyses indicated that mean sojourn time might be approximately 1.5 years longer if colorectal cancer prevalence in nonparticipants of screening colonoscopy is 20% lower than prevalence in participants or 1 year shorter if it exceeds the prevalence in participants by 20%. This study provides, for the first time, precise estimates of sojourn time by age and sex, and it suggests that sojourn times are remarkably consistent across age groups and in both sexes.

Evaluating the long-term effect of FOBT in colorectal cancer screening

BACKGROUND

Cancer screening has been effective in detecting tumors early before symptoms appear. However, the effectiveness of the regular fecal occult blood test (FOBT) in colorectal cancer in the long term has not been quantified.

METHODS

We applied the statistical method developed by Wu and Rosner using data from the Minnesota Colon Cancer Control Study (MCCCS). All initially asymptomatic participants were classified into four mutually exclusive groups: true-early-detection, no-early-detection, over-diagnosis, and symptom-free life; human lifetime was treated as a random variable and is subject to competing risks. All participants in the screening program will eventually fall into one of the four outcomes above. Predictive inferences on the percentages of the four outcomes for both genders were made using the Minnesota study data.

RESULTS

Depending on gender, screening frequency and age at the initial screening, for all participants the probability of "symptom-free-life" varies between 95.3% and 96.6%; the probability of "true-early-detection" is 1.9-3.8%; the probability of no-early-detection is 0.3-2.0%; the probability of over-diagnosis is 0.16-0.3%. Among those with colorectal cancer detected by regular FOBT, the probability of over-diagnosis is lower than expected and is between 6% and 9%, with 95% CI (2.5%, 21.3%) for females and (1.9%, 44.7%) for males. The probability of true-early-detection increases as screening interval decreases. The probability of no-early-detection decreases as screening interval decreases.

CONCLUSION

The probability of over-diagnosis among the screen-detected cases is not as high as previously thought. We hope this outcome can provide valuable information on the effectiveness of the FOBT in colorectal cancer detection in the long term.

Evaluating risk factor assumptions: a simulation-based approach

BACKGROUND

Microsimulation models are an important tool for estimating the comparative effectiveness of interventions through prediction of individual-level disease outcomes for a hypothetical population. To estimate the effectiveness of interventions targeted toward high risk groups, the mechanism by which risk factors influence the natural history of disease must be specified. We propose a method for evaluating these risk factor assumptions as part of model-building.

METHODS

We used simulation studies to examine the impact of risk factor assumptions on the relative rate (RR) of colorectal cancer (CRC) incidence and mortality for a cohort with a risk factor compared to a cohort without the risk factor using an extension of the CRC-SPIN model for colorectal cancer. We also compared the impact of changing age at initiation of screening colonoscopy for different risk mechanisms.

RESULTS

Across CRC-specific risk factor mechanisms, the RR of CRC incidence and mortality decreased (towards one) with increasing age. The rate of change in RRs across age groups depended on both the risk factor mechanism and the strength of the risk factor effect. Increased non-CRC mortality attenuated the effect of CRC-specific risk factors on the RR of CRC when both were present. For each risk factor mechanism, earlier initiation of screening resulted in more life years gained, though the magnitude of life years gained varied across risk mechanisms.

CONCLUSIONS

Simulation studies can provide insight into both the effect of risk factor assumptions on model predictions and the type of data needed to calibrate risk factor models.

Incorporating frailty in a multi-state model: application to disease natural history modelling of adenoma-carcinoma in the large bowel

Homogeneous multi-state models of disease progression have been widely used for designing and evaluating cancer screening programs. However, in screening for premalignant conditions of the cervix or large bowel, it is unlikely that all premalignant lesions have the same underlying propensity for progression. Incorporating frailty into multi-state models raises practical difficulties as it precludes the derivation of finite transition probabilities by matrix solution of the Kolmogorov equations. We address this problem by formulating a heterogeneous process as a series of homogeneous processes linked by transitions which are subject to heterogeneity (frailty). Continuous frailty and discrete mover—stayer models were developed. We applied these to the example of progression of adenoma to colorectal cancer in a three-state model and to a five-state model including consideration of adenoma size. Results were compared with those of purely homogeneous models in a previous study in terms of cumulative risk of malignant transformation from adenoma to invasive colorectal cancer.

Estimating screening test sensitivity and tumour progression using tumour size and time since previous screening

As mammography screening aims to improve the prognosis through earlier detection/treatment, tumour progression and screening test sensitivity (STS) represent key parameters in the evaluation of screening programs. We will here study some methods for estimation of tumour progression and STS, and show how previously used methods can be combined and developed to utilise more of the data available in modern screening programs. Weedon-Fekjaer et al. recently suggested a study design using interview data about time since previous screening to estimate tumour progression and STS in a stepwise Markov model. While useful, the approach does not utilise tumour size measurements, nor link tumour progression to tumour size. Hence, we will here propose formulas for estimating tumour progression and STS using a continuous tumour growth model. To estimate tumour progression and STS, tumour growth curves are followed from one screening to the next, and probabilities for all combinations of tumour sizes at repeated screening examinations calculated. Based on the probabilities for screening detection on subsequent screening examinations, maximum likelihood estimates are calculated. Applied to Norwegian data, the new approach gives similar results to previously published results based on interval data, confirming the earlier estimated large variation in tumour growth rates.

Multi-state Markov models in cancer screening evaluation: a brief review and case study

This work presents a brief overview of Markov models in cancer screening evaluation and focuses on two specific models. A three-state model was first proposed to estimate jointly the sensitivity of the screening procedure and the average duration in the preclinical phase, i.e. the period when the cancer is asymptomatic but detectable by screening. A five-state model, incorporating lymph node involvement as a prognostic factor, was later proposed combined with a survival analysis to predict the mortality reduction associated with screening. The strengths and limitations of these two models are illustrated using data from French breast cancer service screening programmes. The three-state model is a useful frame but parameter estimates should be interpreted with caution. They are highly correlated and depend heavily on the parametric assumptions of the model. Our results pointed out a serious limitation to the five-state model, due to implicit assumptions which are not always verified. Although it may still be useful, there is a need for more flexible models. Over-diagnosis is an important issue for both models and induces bias in parameter estimates. It can be addressed by adding a non-progressive state, but this may provide an uncertain estimation of over-diagnosis. When the primary goal is to avoid bias, rather than to estimate over-diagnosis, it may be more appropriate to correct for over-diagnosis assuming different levels in a sensitivity analysis. This would be particularly relevant in a perspective of mortality reduction estimation.

A projection of benefits due to fecal occult blood test for colorectal cancer

OBJECTIVES

A prospective study to estimate benefits due to fecal occult blood tests for colorectal cancer are carried out for both males and females, under different screening frequencies.

METHODS

We apply the statistical method developed by Wu et al. (2007) [1] using the Minnesota colorectal cancer study group data, to make Bayesian inference for the lead time, the time of diagnosis advanced by screening for both male and female participants in a periodic screening program. The lead time is distributed as a mixture of a point mass at zero and a piecewise continuous distribution. The two parts of the mixture correspond to two aspects of the screening: the probability of no benefit, or the percentage of interval cases; and the probability distribution of the early diagnosis time. We present estimates of these two measures for males and females by simulation studies using the Minnesota study group data. We also provide the mean, mode, variance, and density curve of the program's lead time by gender. This may provide policy makers important information on the effectiveness of the FOBT screening in colorectal cancer early detection.

RESULTS

The mean lead time increases as the screening time interval decreases for both males and females. The standard error of the lead time also increases as the screening time interval decreases for both genders. Females seem get more benefit than males, in that females usually have a longer lead time than males if both take the test at the same time interval and the lead time mode for females is greater than that of males in the same screening time interval.

CONCLUSION

According to the predictive estimation of the lead time distribution, to guarantee a 90% chance of early detection, it seems necessary for the males to take the fecal occult blood test every 9 months, while the females can take it annually.

Estimating mean sojourn time and screening sensitivity using questionnaire data on time since previous screening

OBJECTIVES

Mean sojourn time (MST) and screening test sensitivity (STS), is usually estimated by Markov models using incidence data from the first screening round and the interval between screening examinations. However, several screening programmes do not have full registration of cancers submerging after screening, and increased use of opportunistic screening over time can raise questions regarding the quality of interval cancer registration. Methods/settings Based on the earlier used Markov model, formulas for expected number of cases given time since former screening activity was developed. Using questionnaire data for 336,533 women in the Norwegian Breast Cancer Screening Programme (NBCSP), mean square regression estimates of MST and STS were calculated.

RESULTS

In contrast to the previously used method, the new approach gave satisfactory model fit. MST was estimated to 5.6 years for women aged 50-59 years, and 6.9 years for women aged 60-69 years, and STS was estimated to 55% and 60%, respectively. Attempts to add separate parameters for breast cancer incidence without screening, or previous STS, resulted in wide confidence intervals if estimated separately, and non-identifiably if combined.

CONCLUSION

Previously published results of long MST and low screen test sensitivity were confirmed with the new approach. Questionnaire data on time since previous screening can be used to estimate MST and STS, but the approach is sensitive to relaxing the assumptions regarding the expected breast cancer incidence without screening and constant STS over time.

A deterministic model for estimating the reduction in colorectal cancer incidence due to endoscopic surveillance

There is evidence that the removal of adenomas, by endoscopy, from the large bowel can prevent the occurrence of colorectal cancer (CRC). However, the reduction in cancer incidence due to endoscopic surveillance is difficult to estimate. Studies of cohorts of adenoma patients typically rely on comparisons with groups of historical controls. We present a model for disease progression which enables estimation of this quantity without direct comparison to a reference group. Models are applied to data from the National Polyp Study. Rates of adenoma recurrence and progression to carcinoma are estimated based on study data and relevant literature. This allows calculation of the number of cancers expected in the absence of surveillance and, thus, the number of cancers prevented. Results are compared with the original analysis. Models estimate that surveillance reduced CRC incidence by at least 97% in this cohort. The majority of the effect was due to the initial removal of adenomas rather than the follow-up surveillance. These results are similar to those produced in the original analysis when using the most appropriate reference groups. They indicate that polypectomy and follow-up surveillance can lead to large reductions in cancer incidence which may have been under-estimated in previous studies.

Estimating key parameters in FOBT screening for colorectal cancer

OBJECTIVES

The association between screening sensitivity, transition probability, and individual's age in FOBT for colorectal cancer are explored, for both males and females.

METHODS

We apply the statistical method developed by Wu et al. [1] using the Minnesota colorectal cancer study group data, to make Bayesian inference for the age-dependent screening test sensitivity, the age-dependent transition probability from disease-free to preclinical state, and the sojourn time distribution, for both male and female participants in a periodic screening program. This gives us more information on the effectiveness of the fecal occult blood test in colorectal cancer detection.

RESULTS

The sensitivity appears to increase with age for both genders. However, the posterior mean sensitivity is not monotonic with age for males; it has a peak around age 74. The standard errors of the sensitivity are not monotone either; there is a minimum at age 69 for males and at age 78 for females. The age-dependent transition probability is not a monotone function of age; it has a single maximum at age 72 for males and a single maximum at age 75 for females. The age dependency seems more dramatic for females than for males. The posterior mean sojourn time is 4.08 years for males and 2.41 years for females, with a posterior median of 1.66 years for males and 1.88 years for females. The 95% highest posterior density (HPD) interval is (0.97, 20.28) for males and (1.15, 5.96) for females, which are very large ranges, especially for males. The reason might be that there were fewer men than women in the annual screening program.

CONCLUSION

Reliable estimates of age-dependent sensitivity and transition probability are of great value to policy-makers regarding the initial age for colorectal cancer screening exams. We found that the mean sojourn time for males is much longer than that for females, which may imply that FOBT screening for colorectal cancer may be more effective for males than for females.

Breast cancer tumor growth estimated through mammography screening data

Introduction

Knowledge of tumor growth is important in the planning and evaluation of screening programs, clinical trials, and epidemiological studies. Studies of tumor growth rates in humans are usually based on small and selected samples. In the present study based on the Norwegian Breast Cancer Screening Program, tumor growth was estimated from a large population using a new estimating procedure/model.

Methods

A likelihood-based estimating procedure was used, where both tumor growth and the screen test sensitivity were modeled as continuously increasing functions of tumor size. The method was applied to cancer incidence and tumor measurement data from 395,188 women aged 50 to 69 years.

Results

Tumor growth varied considerably between subjects, with 5% of tumors taking less than 1.2 months to grow from 10 mm to 20 mm in diameter, and another 5% taking more than 6.3 years. The mean time a tumor needed to grow from 10 mm to 20 mm in diameter was estimated as 1.7 years, increasing with age. The screen test sensitivity was estimated to increase sharply with tumor size, rising from 26% at 5 mm to 91% at 10 mm. Compared with previously used Markov models for tumor progression, the applied model gave considerably higher model fit (85% increased predictive power) and provided estimates directly linked to tumor size.

Conclusion

Screening data with tumor measurements can provide population-based estimates of tumor growth and screen test sensitivity directly linked to tumor size. There is a large variation in breast cancer tumor growth, with faster growth among younger women.