Increasing power in screening trials by testing control-arm specimens: Application to multicancer detection screening.

BACKGROUND: Cancer screening trials have required large sample-sizes and long time-horizons to demonstrate cancer mortality reductions, the primary goal of cancer screening. We examine assumptions and potential power gains from exploiting information from testing control-arm specimens, which we call the "Intended Effect" (IE) analysis that we explain in detail herein. The IE analysis is particularly suited to tests that can be conducted on stored specimens in the control-arm, such as stored blood for multicancer detection (MCD) tests. METHODS: We simulated hypothetical MCD screening trials to compare power and sample-size for the standard vs IE analysis. Under two assumptions that we detail herein, we projected the IE analysis for 3 existing screening trials (National Lung Screening Trial (NLST), Minnesota Colon Cancer Control Study (MINN-FOBT-A), and Prostate, Lung, Colorectal, Ovarian Cancer Screening Trial-colorectal component (PLCO-CRC)). RESULTS: Compared to the standard analysis for the 3 existing trials, the IE design could have reduced cancer-specific mortality p-values 5-fold (NLST), 33-fold (MINN-FOBT-A), or 14,160-fold (PLCO-CRC), or alternately, reduced sample-size (90% power) by 26% (NLST), 48% (MINN-FOBT-A), or 59% (PLCO-CRC). For potential MCD trial designs requiring 100,000 subjects per-arm to achieve 90% power for multi-cancer mortality for the standard analysis, the IE analysis achieves 90% power for only 37,500-50,000 per arm, depending on assumptions concerning control-arm test-positives. CONCLUSIONS: Testing stored specimens in the control arm of screening trials to conduct the IE analysis could substantially increase power to reduce sample-size or accelerate trials, and provide particularly strong power gains for MCD tests.

Cancer screening with multicancer detection tests: A translational science review.

Multicancer detection (MCD) tests use a single, easily obtainable biospecimen, such as blood, to screen for more than one cancer concurrently. MCD tests can potentially be used to improve early cancer detection, including cancers that currently lack effective screening methods. However, these tests have unknown and unquantified benefits and harms. MCD tests differ from conventional cancer screening tests in that the organ responsible for a positive test is unknown, and a broad diagnostic workup may be necessary to confirm the location and type of underlying cancer. Among two prospective studies involving greater than 16,000 individuals, MCD tests identified those who had some cancers without currently recommended screening tests, including pancreas, ovary, liver, uterus, small intestine, oropharyngeal, bone, thyroid, and hematologic malignancies, at early stages. Reported MCD test sensitivities range from 27% to 95% but differ by organ and are lower for early stage cancers, for which treatment toxicity would be lowest and the potential for cure might be highest. False reassurance from a negative MCD result may reduce screening adherence, risking a loss in proven public health benefits from standard-of-care screening. Prospective clinical trials are needed to address uncertainties about MCD accuracy to detect different cancers in asymptomatic individuals, whether these tests can detect cancer sufficiently early for effective treatment and mortality reduction, the degree to which these tests may contribute to cancer overdiagnosis and overtreatment, whether MCD tests work equally well across all populations, and the appropriate diagnostic evaluation and follow-up for patients with a positive test.

Study design considerations for trials to evaluate multicancer early detection assays for clinical utility.

Blood-based assays using various technologies and biomarkers are in commercial development for the purpose of detecting multiple cancer types concurrently at an early stage of disease. These multicancer early detection (MCED) assays have the potential to improve the detection of cancers, particularly those for which no current screening modality exists. However, the unknown clinical benefits and harms of using MCED assays for cancer screening necessitate the development and implementation of a randomized controlled trial (RCT) to ascertain their clinical effectiveness. This was the consensus of experts at a National Cancer Institute-hosted workshop to discuss initial design concepts for such a trial. Using these assays to screen simultaneously for multiple cancers poses novel uncertainties for patient care compared with conventional screening tests for single cancers, such as establishing the diagnostic workup to confirm the presence of cancer at any organ site; clarifying appropriate follow-up for a positive assay for which there is no definitive diagnosis; identifying potential harms such as overdiagnosis of indolent disease; determining clinically effective and efficient strategies for disseminating MCED screening in real-world practice; and understanding the ethical implications, such as potentially alleviating or exacerbating existing health disparities. These assays present new and complex challenges for designing an RCT. Issues that emerged from the meeting centered around the need for a flexibly designed, clinical utility RCT to rigorously capture the evidence required to fully understand the net benefit of this promising technology. Specific topic areas were endpoints, screening protocols, recruitment, diagnostic pathway, pilot phase, data elements, specimen collection, and ethical considerations.

Breast cancer overdiagnosis in stop-screen trials: More uncertainty than previously reported.

OBJECTIVE: According to the Independent UK Panel on Breast Cancer Screening, the most reliable estimates of overdiagnosis for breast cancer screening come from stop-screen trials Canada 1, Canada 2, and Malmo. The screen-interval overdiagnosis fraction is the fraction of cancers in a screening program that are overdiagnosed. We used the cumulative incidence method to estimate screen-interval overdiagnosis fraction. Our goal was to derive confidence intervals for estimated screen-interval overdiagnosis fraction and adjust for refusers in these trials. METHODS: We first show that the UK Panel's use of a 95% binomial confidence interval for estimated screen-interval overdiagnosis fraction was incorrect. We then derive a correct 95% binomial-Poisson confidence interval. We also use the method of latent-class instrumental variables to adjust for refusers. RESULTS: For the Canada 1 trial, the estimated screen-interval overdiagnosis fraction was 0.23 with a 95% binomial confidence interval of (0.18, 0.27) and a 95% binomial-Poisson confidence interval of (0.04, 0.41). For the Canada 2 trial, the estimated screen-interval overdiagnosis fraction was 0.16 with a 95% binomial confidence interval of (0.12, 0.19) and a 95% binomial-Poisson confidence interval of (-0.01, 0.32). For the Malmo trial, the estimated screen-interval overdiagnosis fraction was 0.19 with a 95% binomial confidence interval of (0.15, 0.22). Adjusting for refusers, the estimated screen-interval overdiagnosis fraction was 0.26 with a 95% binomial-Poisson confidence interval of (0.03, 0.50). CONCLUSION: The correct 95% binomial-Poisson confidence interval s for the estimated screen-interval overdiagnosis fraction based on the Canada 1, Canada 2, and Malmo stop-screen trials are much wider than the previously reported incorrect 95% binomial confidence intervals. The 95% binomial-Poisson confidence intervals widen as follow-up time increases, an unappreciated downside of longer follow-up in stop-screen trials.

Effect of flexible sigmoidoscopy screening on colorectal cancer incidence and mortality: long-term follow-up of the randomised US PLCO cancer screening trial.

BACKGROUND: Screening flexible sigmoidoscopy reduces incidence and mortality of colorectal cancer. Previously reported results from the Prostate, Lung, Colorectal, and Ovarian Cancer (PLCO) screening trial had a median follow-up of 12 years. Whether the benefit is sustained over the long term and remains so in both sexes and all age groups is uncertain. We report long-term results after an additional 5 years of follow-up. METHODS: Participants in the PLCO trial were recruited from the general population in the catchment areas of ten screening centres across the USA, without previous diagnosis of a prostate, lung, colorectal, or ovarian cancer or current cancer treatment. From 1993 to 2001, participants aged 55-74 years were randomly assigned to usual care or flexible sigmoidoscopy at baseline and again at 3 years or 5 years. Randomisation was done within blocks and stratified by centre, age, and sex. The primary endpoint was cause-specific mortality and secondary endpoints included incidence and tumour staging; cause of death was determined without knowledge of study arm. In this analysis, we assessed incidence and mortality rates overall, by time-period, and by combinations of sex, age at baseline (55-64 years/65-74 years), location (distal/proximal), and stage, on an intent-to-treat basis. This trial is registered with ClinicalTrials.gov, number NCT00002540. FINDINGS: After a median follow-up of 15·8 years (IQR 13·2-18·0) for incidence and 16·8 years (14·4-18·9) for mortality, the incidence of colorectal cancer was significantly lower in the intervention arm (1461 cases; 12·55 per 10 000 person-years) than with usual care (1761 cases; 15·33 per 10 000 person-years; relative risk [RR] 0·82, 95% CI 0·76-0·88). Similarly, mortality was lower in the intervention arm (417 deaths; 3·37 per 10 000 person-years) than the usual care arm (549; 4·48 per 10 000 person-years; RR 0·75, 95% CI 0·66-0·85). The reduction in mortality was limited to the distal colon, with no significant effect in the proximal colon. Reductions in incidence were significantly larger in men than women (pinteraction=0·04) and reductions in mortality were significantly larger in the older age group (65-74 years vs 55-64 years at baseline; pinteraction=0·01). INTERPRETATION: Reductions in colorectal cancer incidence and mortality from flexible sigmoidoscopy screening are sustained over the long term. Differences by sex and age should be examined in other ongoing trials of colorectal cancer screening to help clarify if different screening strategies would achieve greater risk reduction. FUNDING: Extended follow-up was funded under NIH contract HHSN261201600007I.

Overall and Multiphasic Findings of the Prostate, Lung, Colorectal and Ovarian (PLCO) Randomized Cancer Screening Trial.

BACKGROUND: Screening tests are typically evaluated for a single disease, but multiple tests for multiple diseases are performed in practice. The Prostate, Lung, Colorectal and Ovarian (PLCO) cancer screening trial assessed testing for four cancers simultaneously and can be viewed as a multiphasic cancer intervention. This paper presents overall and multiphasic findings of this trial. METHODS: The PLCO trial was a randomized multi-center trial conducted at ten screening centers in the US. Participants were 76,682 men and 78,215 women ages 55 - 74 and free of the target cancers at trial entry. Screening tests were PSA and digital rectal examination for prostate cancer, chest x-ray for lung cancer, flexible sigmoidoscopy for colorectal cancer, CA125 and transvaginal ultrasound for ovarian cancer. Outcomes and harms of screening were assessed including compliance, test results, incidence, mortality, false positives and overdiagnosis. RESULTS: Screening compliance was 82%, 72,820 (8%) of 906,064 exams were positive, the overall PPV was 4.2% and the cancer detection rate was 3.38/1000. A mortality reduction was observed only for colorectal cancer (RR 0.72, 95% CI 0.61 - 0.85) with no effect on all-cause mortality. Ninety-six percent of positive exams were falsely positive and there was a suggestion of overdiagnosis of prostate and possibly ovarian cancers. Multiphasic testing resulted in 7374 men and 2748 women experiencing multiple false positive results from multiple types of tests. CONCLUSION: Multiphasic cancer screening led to reduced mortality for one target cancer and imposed a burden on the health care system that included substantial false positives and likely overdiagnosis.

Data sharing in clinical trials: An experience with two large cancer screening trials.

Paul Pinsky of the US National Cancer Institute and colleagues describe the implementation and outcomes of web-based data sharing from the PLCO and NLST cancer screening trials.

Extended mortality results for prostate cancer screening in the PLCO trial with median follow-up of 15 years.

BACKGROUND: Two large-scale prostate cancer screening trials using prostate-specific antigen (PSA) have given conflicting results in terms of the efficacy of such screening. One of those trials, the Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial, previously reported outcomes with 13 years of follow-up. This study presents updated findings from the PLCO trial. METHODS: The PLCO trial randomized subjects from 1993 to 2001 to an intervention or control arm. Intervention-arm men received annual PSA tests for 6 years and digital rectal examinations for 4 years. This study used a linkage with the National Death Index to extend mortality follow-up to a maximum of 19 years after randomization. RESULTS: Men were randomized to the intervention arm (n = 38,340) or the control arm (n = 38,343). The median follow-up time was 14.8 years (25th/75th, 12.7/16.5 years) in the intervention arm and 14.7 years (25th/75th, 12.6/16.4 years) in the control arm. There were 255 deaths from prostate cancer in the intervention arm and 244 deaths from prostate cancer in the control arm; this meant a rate ratio (RR) of 1.04 (95% confidence interval [CI], 0.87-1.24). The RR for all-cause mortality was 0.977 (95% CI, 0.950-1.004). It was estimated that 86% of the men in the control arm and 99% of the men in the intervention arm received any PSA testing during the trial, and the estimated yearly screening-phase PSA testing rates were 46% and 84%, respectively. CONCLUSIONS: Extended follow-up of the PLCO trial over a median of 15 years continues to indicate no reduction in prostate cancer mortality for the intervention arm versus the control arm. Because of the high rate of control-arm PSA testing, this finding can be viewed as showing no benefit of organized screening versus opportunistic screening. Cancer 2017;123:592-599. © 2016 American Cancer Society.

Breast-Cancer Tumor Size, Overdiagnosis, and Mammography Screening Effectiveness.

BACKGROUND: The goal of screening mammography is to detect small malignant tumors before they grow large enough to cause symptoms. Effective screening should therefore lead to the detection of a greater number of small tumors, followed by fewer large tumors over time. METHODS: We used data from the Surveillance, Epidemiology, and End Results (SEER) program, 1975 through 2012, to calculate the tumor-size distribution and size-specific incidence of breast cancer among women 40 years of age or older. We then calculated the size-specific cancer case fatality rate for two time periods: a baseline period before the implementation of widespread screening mammography (1975 through 1979) and a period encompassing the most recent years for which 10 years of follow-up data were available (2000 through 2002). RESULTS: After the advent of screening mammography, the proportion of detected breast tumors that were small (invasive tumors measuring <2 cm or in situ carcinomas) increased from 36% to 68%; the proportion of detected tumors that were large (invasive tumors measuring ≥2 cm) decreased from 64% to 32%. However, this trend was less the result of a substantial decrease in the incidence of large tumors (with 30 fewer cases of cancer observed per 100,000 women in the period after the advent of screening than in the period before screening) and more the result of a substantial increase in the detection of small tumors (with 162 more cases of cancer observed per 100,000 women). Assuming that the underlying disease burden was stable, only 30 of the 162 additional small tumors per 100,000 women that were diagnosed were expected to progress to become large, which implied that the remaining 132 cases of cancer per 100,000 women were overdiagnosed (i.e., cases of cancer were detected on screening that never would have led to clinical symptoms). The potential of screening to lower breast cancer mortality is reflected in the declining incidence of larger tumors. However, with respect to only these large tumors, the decline in the size-specific case fatality rate suggests that improved treatment was responsible for at least two thirds of the reduction in breast cancer mortality. CONCLUSIONS: Although the rate of detection of large tumors fell after the introduction of screening mammography, the more favorable size distribution was primarily the result of the additional detection of small tumors. Women were more likely to have breast cancer that was overdiagnosed than to have earlier detection of a tumor that was destined to become large. The reduction in breast cancer mortality after the implementation of screening mammography was predominantly the result of improved systemic therapy.

Extended mortality results for ovarian cancer screening in the PLCO trial with median 15years follow-up.

BACKGROUND: The Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial originally reported no mortality benefit of ovarian cancer screening after a median of 12.4years of follow-up. The UKCTOCS screening trial failed to show a statistically significant mortality reduction in the primary analysis but reported an apparent increased mortality benefit in trial years 7-14 compared to 0-7. Here we report an updated analysis of PLCO with extended mortality follow-up. METHODS: Participants were randomized from 1993 to 2001 at ten U.S. centers to an intervention or usual care arm. Intervention arm women were screened for ovarian cancer with annual trans-vaginal ultrasound (TVU) (4years) and CA-125 (6years), with a fixed cutoff at 35U/mL for CA-125. The original follow-up period was for up to 13years (median follow-up 12.4years); in this analysis follow-up for mortality was extended by up to 6years. RESULTS: 39,105 (intervention) and 39,111 (usual care) women were randomized, of which 34,253 and 34,304, respectively, had at least one ovary at baseline. Median follow-up was 14.7years in each arm and maximum follow-up 19.2years in each arm. A total of 187 (intervention) and 176 (usual care) deaths from ovarian cancer were observed, for a risk-ratio of 1.06 (95% CI: 0.87-1.30). Risk-ratios were similar for study years 0-7 (RR=1.04), 7-14 (RR=1.06) and 14+ (RR=1.09). The risk ratio for all-cause mortality was 1.01 (95% CI: 0.97-1.05). Ovarian cancer specific survival was not significantly different across trial arms (p=0.16). CONCLUSION: Extended follow-up of PLCO indicated no mortality benefit from screening for ovarian cancer with CA-125 and TVU.

Managing Multi-Center Recruitment in the PLCO Cancer Screening Trial.

There were significant recruitment challenges specific to the PLCO Cancer Screening Trial. Large numbers of participants were to be randomized from ten catchment areas nationwide within time and budgetary constraints. The eligible population was elderly and had to meet health and behavioral thresholds. Informed consent was required to participate and be randomized to screening for three cancers at periodic clinic visits or to a usual care arm that included no clinical visits. Consenting required special efforts to fully explain the trial and its potential scientific benefit to future patients with potentially no benefits but possible harms to PLCO participants. Participation would include continued follow-up for at least 13 years after randomization. Strong collaborative investments were required by the NCI and screening centers (SCs) to assure timely recruitment and appropriate racial participation. A trial-wide pilot phase tested recruitment and protocol follow through at SCs and produced a vanguard population of 11,406 participants. NCI announced the trial nationally in advance of the pilot and followed with an even more intense collaborative role with SCs for the main phase to facilitate trial-wide efficient and timely recruitment. Special efforts to enhance recruitment in the main phase included centralized and local monitoring of progress, cross-linking SCs to share experiences in problem solving, centralized training, substantial additional funding dedicated to recruitment and retention, including specialized programs for minority recruitment, obtaining national endorsement by the American Cancer Society, launching satellite recruitment and screening centers, including minority focused satellites, and adding a new SC dedicated to minority recruitment.

PLCO: Evolution of an Epidemiologic Resource and Opportunities for Future Studies.

The Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial (PLCO), a large-scale, multi-institutional, randomized controlled trial, was launched in 1992 to evaluate the effectiveness of screening modalities for prostate, lung, colorectal, and ovarian cancer. However, PLCO was additionally designed to serve as an epidemiologic resource and the National Cancer Institute has invested substantial resources over the years to accomplish this goal. In this report, we provide a summary of changes to PLCO's follow-up after conclusion of the screening phase of the trial and highlight recent data and biospecimen collections, including ancillary studies, geocoding, administration of a new medication use questionnaire, consent for linkage to Medicare, and additional tissue collection that enhance the richness of the PLCO resource and provide further opportunities for scientific investigation into the prevention, early detection, etiology and treatment of cancer.

Study designs for determining and comparing sensitivities of disease screening tests.

OBJECTIVE: To investigate the capability of various study designs to determine the sensitivity of a disease screening test. METHODS: Quantities that can be calculated from these designs were derived and examined for their relationship to true sensitivity (the ability to detect unrecognized disease that would surface clinically in the absence of screening) and overdiagnosis. RESULTS: To examine the sensitivity of one test, the single cohort design, in which all participants receive the test, is particularly weak, providing only an upper bound on the true sensitivity, and yields no information about overdiagnosis. A randomized design, with one control arm and participants tested in the other, that includes sufficient post-screening follow-up, allows calculation of bounds on, and an approximation to, true sensitivity and also determination of overdiagnosis. Without follow-up, bounds on the true sensitivity can be calculated. To compare two tests, the single cohort paired design in which all participants receive both tests is precarious. The three arm randomized design with post screening follow-up is preferred, yielding an approximation to the true sensitivity, bounds on the true sensitivity, and the extent of overdiagnosis of each test. Without post screening follow-up, bounds on the true sensitivities can be calculated. When an unscreened control arm is not possible, the two-arm randomized design is recommended. Individual test sensitivities cannot be determined, but with sufficient post-screening follow-up, an order relationship can be established, as can the difference in overdiagnosis between the two tests.

The PLCO Cancer Screening Trial: Background, Goals, Organization, Operations, Results.

The randomized PLCO trial was designed to answer four primary questions: does screening for these cancers using often promoted tests reduce cancer-specific mortality? Nearly 155,000 men and women were allocated to screening or usual care arms in a 1:1 ratio under a centralized, secure randomization algorithm at ten competitively selected screening centers nationwide. Screened men received PSA blood tests and digital rectal examinations. Screened women received CA125 blood tests and trans-vaginal ultrasound. Both men and women in the screened arm received anterolateral view chest x-ray and 60 cm flexible sigmoidoscopy. Blood specimens were collected at each screening visit and buccal cell DNA was collected once from the usual care participants. Histology slides were collected for cancer cases. Participants completed a baseline questionnaire covering health and risk factors and a dietary questionnaire. Data collected on standardized machine-readable forms were scanned remotely at screening and laboratory sites utilizing PLCO dedicated, NCI provided and configured computer systems for quality checks, archiving, and analysis. Comprehensive quality assurance was implemented over recruitment, consenting, randomization, screening, data management, records keeping, patient-specific screening results reporting, follow-up, and data analysis. Performance and data quality were monitored on-site and remotely by data edits, site visits, and random record audits. Specially trained and certified professionals performed screening procedures and medical record abstracting. An independent committee of medical specialists reviewed and certified case-specific cause of death. Scientific leadership was provided by NCI Project Officers, PLCO principal investigators, external consultants, and an independent data and safety monitoring board.