Working Procedures

B. SCIENTIFIC REVIEW AND EVALUATION

2. Studies of cancer prevention in humans

This section includes all pertinent experimental and observational epidemiological studies of cancer prevention in humans, with cancer as an outcome (see Part A, Section 4). Studies of biomarkers as indicators of the intervention are included in Section 4 when they are relevant to an evaluation of the cancer-preventive effect in humans.

(a) Types of study considered

This section focuses on studies that assess the prevention of cancer as an outcome in humans. Relevant evidence is normally provided by experimental studies (for example, randomized clinical trials and community intervention trials), and analytical observational studies, primarily cohort studies and case-control studies. For certain interventions applied at the population level, well-designed ecological studies (studies measuring both outcome and exposure on the aggregate, or population, level) or interrupted time-series studies may also be informative. Cross-sectional studies, descriptive epidemiological studies, case-series and case reports are usually not reviewed. The uncertainties that surround the interpretation of such studies make them inadequate, except in exceptional circumstances, to form the basis for inferring a preventive relationship. However, when considered together with experimental and analytical observational studies, these types of study can sometimes contribute to the decision of the Working Group as to whether or not a causal relationship exists.

Intervention studies are experimental in design - that is, the use of, or exposure to, the intervention is assigned by the investigator. Experimental studies can provide the strongest and most direct evidence of a protective or preventive effect; however, the use of such studies is limited for practical and ethical reasons and the subjects are often drawn from select groups that may not represent the population at large.

In exceptional cases, epidemiological studies on advanced preneoplastic lesions and other end-points thought to be relevant to cancer are also reviewed in this section. The results of such studies may strengthen inferences drawn from other studies.

(b) Quality of studies considered

In considering whether a particular study should contribute to the evaluation of an intervention, the Working Group considers the following aspects:

  • The relevance of the study;
  • The appropriateness of the design and analysis to the question being asked;
  • The adequacy and completeness of the presentation of the data; and
  • The degree to which chance, bias, and confounding may have affected the results; for drugs or other marketed products, this bias assessment should include review of the funding source.

Aspects that are particularly important in evaluating randomized controlled trials are: the selection of participants, the nature and adequacy of the randomization procedure, evidence that randomization achieved an adequate balance between the groups, exclusion criteria used before and after randomization, compliance with the intervention in the intervention group, and “contamination” of the control group with the intervention. Other considerations are the means by which the end-point was determined and validated (either by screening or by other means of detection of the disease), the length and completeness of follow-up of the groups, and the adequacy of the analysis.

It is necessary to take into account the possible roles of bias, confounding, and chance in the interpretation of cohort and case-control studies. Bias is the effect of factors in study design or execution that leads erroneously to a stronger or weaker association than in fact exists between an intervention and outcome. Confounding is a form of bias that occurs when the relationship with the outcome is made to appear stronger or weaker than it is in reality, due to an association between the apparent causal factor and another factor that is associated with either an increase or decrease in the incidence of the disease. The role of chance is related to biological variability and the influence of sample size on the precision of estimates of effect.

In evaluating the extent to which these factors have been minimized in an individual study, consideration is given to a number of aspects of design and analysis as described in the report of the study. Most of these considerations apply equally to all types of study. Lack of clarity regarding any of these aspects in the reporting of a study can decrease its credibility and the weight given to it in the final evaluation.

Firstly, the study population, target organ, and exposure should have been well defined by the authors. Cancer occurrence in the study population should have been identified in a way that was independent of the intervention of interest, and exposure to the intervention should have been assessed in a way that was not related to disease status.

Secondly, the authors should have taken into account - in the study design and analysis - other variables that could influence the risk of disease and may have been related to the exposure of interest. Potential confounding by such variables should have been dealt with either in the design of the study (e.g. by matching), or in the analysis (by statistical adjustment). Internal comparisons of frequency of disease among individuals with different levels of exposure are desirable in cohort studies, since they minimize the potential for confounding related to the difference in risk factors between an external reference group and the study population.

Thirdly, the authors should have reported the basic data on which the conclusions are founded, even if sophisticated statistical analyses were employed. At the very least, they should have given the numbers of exposed and unexposed cases and controls in a case-control study, and the numbers of cases observed and expected in a cohort study. Further tabulations by duration of exposure and other temporal factors are also important. In a cohort study, data on all cancer sites and all causes of death should have been given, to reveal the possibility of reporting bias. In a case-control study, the effects of investigated factors other than the exposure of interest should have been reported.

Finally, the statistical methods used to obtain estimates of relative risk, absolute rates of cancer, confidence intervals, and significance tests, and to adjust for confounding should have been clearly stated by the authors. These methods have been reviewed for case-control studies (Breslow & Day, 1980) and for cohort studies (Breslow & Day, 1987).

(c) Quantitative aspects

The Working Group gives special attention to quantitative assessment of the preventive effect of the intervention under study, by assessing data from studies investigating different doses or levels of exposure. The Working Group also addresses issues of timing and duration of use or exposure. Such quantitative assessment is important to clarify the circumstances under which a preventive effect can be achieved, as well as the dose or level of exposure at which a toxic effect has been shown.

(d) Criteria for preventive effects

After summarizing and assessing the individual studies, the Working Group makes a judgement concerning the strength of the evidence that the intervention in question prevents cancer in humans. In making its judgement, the Working Group considers several criteria for each relevant cancer site.

Evidence is frequently available from different types of study and is evaluated as a whole. Findings that are replicated in several studies of the same design or in studies using different approaches are more likely to provide evidence of a true protective effect than are isolated observations from single studies.

Evidence of protection derived from intervention studies of good quality is particularly informative. Evidence of a substantial and significant reduction in risk, including a dose-response relationship, is more likely to indicate a true effect. Nevertheless, a small effect, or an effect without a dose-response relationship, does not imply lack of real benefit and may be important for public health if the cancer is common.

The Working Group evaluates possible explanations for inconsistencies across studies, including differences in use of, or exposure to, the intervention, differences in the underlying risk of cancer, and metabolism and genetic differences in the population, as well as differences in study methodology. The results of studies judged to be of high quality are given more weight. Note is taken of both the applicability of preventive action to several cancers and of possible differences in activity, including the possibility of different findings between cancer sites.

Posted 5 July 2016