Screening of Cancer
This information is produced and provided by the National Cancer Institute
In 1995, an estimated 1,252,000 people in the United States will be diagnosed with cancer and 547,000 will die of this disease. How many of these deaths could have been avoided through screening? Estimates vary from 3-35%, depending on the criteria. Screening for breast and cervical cancer alone is projected to be capable of a 3% reduction in total cancer deaths.
In developing the cancer screening statements, the PDQ Screening Editorial Board used the following definitions:
The purpose of this section is to present the explicit evidence-based approach to the development of the state-of-the-art screening statements and the associated assignment of levels of evidence.
The statements are based on various levels of published scientific evidence and collective clinical experience. The highest level of evidence is taken as mortality reduction in controlled, randomized clinical trials. The results of clinical studies, case-control studies, and clinical information were considered in formulating the statements. In addition, the incidence of cancer, stage distribution, treatment, survival rates, and mortality were considered. The statements are subject to modification as new evidence becomes available.
The Scientific Basis
Two requirements must be met for screening to be useful:
These requirements are necessary, but not sufficient to prove the efficacy of screening, which requires demonstration of a decrease in cause-specific mortality. For example, these two criteria are met in the case of screening for childhood neuroblastoma by assessment of urinary catecholamine metabolites. On the basis of these criteria, a mass screening program was conducted in Saitama Prefecture, Japan, from 1981 to 1992 for six-month-old infants. Over that 12 year period, the annual incidence of neuroblastoma in children under age 1 increased from about 28 to 260 per million. No significant reduction in incidence occurred for children over age 1. There was no reduction in mortality for the disease. This experience gave strong evidence of the phenomenon of overdiagnosis—diagnosis of a subset of disease detectable by mass screening but which would not be clinically diagnosed later.
Observation is the most widely available examination for the detection of cancer. It is useful in identifying suspicious lesions in the skin, lip, mouth, larynx, external genitalia, and cervix.
The second most available detection procedure is palpation. It is particularly valuable in detecting lumps, nodules, or tumors in the breast, mouth, salivary glands, thyroid, subcutaneous tissues, anus, rectum, prostate, testes, ovaries, and uterus and enlarged lymph nodes in the neck, axilla, or groin.
Internal cancers require an extension of observation through endoscopes, x-rays, magnetic resonance imaging, and ultrasound. Laboratory tests, such as the Pap smear, and occult blood testing of the feces have also proven helpful for some of these cancers. However, concerns regarding effectiveness and yield play a particularly important role in decisions to screen for cancers not easily amenable to earlier detection through physical examination in asymptomatic individuals. The performance of these tests is usually measured in terms of sensitivity, specificity, and positive and negative predictive values.
There are some cancers where screening does not appear useful: those where no early detection tests exist, as in cancer of the pancreas; and in cancers with no apparent localized stage, as in leukemia.
The type, periodicity, and commencement of screening in high-risk populations for most cancers reflect the judgment of expert practitioners rather than evidence from scientifically-conducted tests. Some individuals are known to be at high risk for cancer, such as those with a strong family history of cancer (in two or more first-degree relatives). Physician judgment is needed in such circumstances to determine the most appropriate application of available screening methods. Prudence dictates increased vigilance in the higher-risk populations. This primarily means that the high-risk person is identified, is counseled appropriately, and regularly undergoes the screening procedures of known benefit.
The relation of stage to survival and mortality is the basis of clinical cancer management. It is the major factor in prognosis, in the determination of treatment, and in the evaluation of end results. In the 1940's, a generalized staging classification of localized, regional, and distant (LRD) was developed to show long-term trends, and it is still useful. In the more detailed TNM system, which has been periodically modified, the (T)umor size, the status of the lymph (N)odes, and the status of distant (M)etastases are also categorized. These elements are then grouped into Stages 0-IV according to their association with survival. Classification of cancer by prognostic factors has developed through the extensive study of thousands of cases followed for many years. As malignant tumors increase in size, they have a greater propensity to metastasize to regional lymph nodes and to distant sites (Table 1). Stage has such a profound effect that all randomized treatment trials require the comparison of similar stages in evaluating differences in outcome. In the Health Insurance Plan (HIP) breast cancer randomized screening trial, the reductions in mortality have been shown to be partly due to shifts in stage.[4,5] In other cancer sites, shifts in stage may also herald improved survival and decreased mortality, though stage shift alone does not establish benefit.
Table 1: Median relative survival time in years by stage (local, regional, or distant) of disease and size of tumor - SEER program data 1973-1986
The Natural Experiment
The Surveillance, Epidemiology, and End Results (SEER) Program of the NCI gathers data from 11 geographic areas, covering approximately 10% of the US population. These data, because of their population coverage in these areas and long duration (1973-present), are a unique and important resource in considering the potential for early cancer detection. Patients with localized cancers have a better survival than those with regional or distant spread. In considering over 1,000,000 cancers (excluding in situ carcinomas, and squamous and basal cell cancers of the skin) by stage, the 5-year relative survival rate is 78% for localized cancer, 45% for regional cancer, and 12% for distant metastatic cancer.
In geographic population-based tumor registries that include all cases, studies of incidence, stage, treatment and survival may be subject to selection, lead-time, length, and healthy volunteer biases during periods of active screening; studies using deaths due to the cancer may not be affected as much by these biases. When differences in overall survival exist between two geographic areas, races, sexes, or religious or socioeconomic groups, it may be due to stage differences reflecting detection practices.
There is a possibility through screening to find small cancers that may never have surfaced in life.[8-10] This is particularly true in prostate cancer where autopsy series have shown a high percentage of occult carcinomas in elderly men. The discovery of these cancers could increase the number of cases, give the appearance of stage shift, and increase survival without necessarily reducing mortality. Therefore, the measures of an improved outcome are in order from the strongest to the weakest: a decrease in cause-specific mortality, reduction in incidence of advanced stage cancers, an increase in survival, and a shift in stage.
There are varying levels of evidence that support a given statement. The strongest evidence would be that obtained from a well-designed and well-conducted randomized controlled trial. It is, however, not always practical to conduct such a trial to address every question surrounding the field of screening. As in all other aspects of medicine, practice must be based upon information that falls short of a randomized trial. For each summary of evidence statement, the associated levels of evidence are listed. In order of strength of evidence, the five levels are as follows:
Experimental trials are designed to correct for or eliminate selection, lead-time, length, healthy volunteer, and other biases when prospectively testing a detection procedure to determine its effect on outcome. The highest level of evidence and greatest benefit is mortality reduction in a randomized controlled trial. For most sites, such evidence is not, and may never be, available. Theoretically it is possible, but the sample size that is needed, the expense, and the duration for such trials in other sites, such as melanoma or gastric cancers, make this approach impractical at present. Therefore, evidence obtained by other methods is often used.
Case-control and cohort studies provide indirect evidence for the effectiveness of screening. Such studies do not prove a mortality reduction effect, but they can suggest a mortality reduction. Such evidence is particularly compelling for the effectiveness of screening for cervical cancer. However, the potential for bias to invalidate inferences from case-control and cohort studies must be recognized.[13-16]
Descriptive uncontrolled studies based on the experience of individual physicians, hospitals, and non-population-based registries may yield some information for screening. The performance of various detection tests, such as sensitivity, specificity, and positive predictive values, are generally first reported in descriptive studies. The first evidence that screening may be successful is an increase in the number of early cancers with shifts in stage and increased survival rates; later, a reduction in deaths may occur. These descriptive studies do not establish efficacy because of the absence of an appropriate control group.