¹University of North Carolina School of Medicine, Chapel Hill, North Carolina

Correspondence: Robert S Sandler, MD, MPH, Division of Gastroenterology and Hepatology, CB 7555, 4111 Bioinformatics Bldg, University of North Carolina, Chapel Hill, NC 25599-7555

Received 11 October 2004; Revised  0000; Accepted 12 October 2004.

Abstract

Calcium supplements have been shown to decrease the risk of colorectal adenomas. In this issue of the Journal, Shaukat et al. report the results of a systematic review and metaanalysis of randomized controlled trials of calcium supplementation. The authors statistically combined the data from the three trials that met strict eligibility criteria. The overall relative risk was 0.80 (95% CI: 0.68–0.93) and the number needed to treat was 14. The results of this metaanalysis support a preventive role for calcium supplements.

Chemoprevention is defined as the use of chemical compounds to prevent, inhibit, or reverse carcinogenesis. These compounds can be pharmacologic agents or natural chemicals. Because cancer chemopreventive agents are given to healthy individuals for long periods of time, the agents must have minimal or no toxicity. This contrasts with cancer chemotherapeutic agents that are given to cancer patients who have life-threatening disease. We are willing to tolerate considerable toxicity from cancer therapy drugs because the agents are used for a short period of time and because the stakes are high.

In many respects calcium is the perfect agent for chemoprevention. Calcium is safe, inexpensive, and has beneficial effects beyond the colon, namely prevention of osteoporosis. Calcium can only be a perfect agent if it really works. In this issue of the Journal Shaukat et al. (¹) provide additional evidence supporting the role of supplemental calcium in the recurrence of colorectal adenomas.

There is no shortage of studies that have evaluated calcium for the prevention of both colorectal cancer and colorectal adenomas. The studies have evaluated both dietary sources of calcium and supplements. The vast majority of these studies are observational in nature. Observational studies, whether they are prospective or retrospective, are subject to certain biases that make them a less reliable source of information. Moreover, in the case of calcium, the results of the studies have been inconsistent. More solid proof for the potential protective effect would come from randomized trials. Shaukat et al. (¹) therefore conducted a systematic review and a metaanalysis of randomized controlled trials of calcium to prevent recurrent adenomas. It would be preferable to use randomized controlled trials with cancer as the endpoint. Unfortunately, there are no randomized trials of calcium to prevent cancer because it takes decades for cancer to develop making a randomized trial impossible. The authors selected trials that used adenomas as the endpoint because adenomas are assumed to be the precursors for the vast majority of colorectal cancers. It is important to recognize, however, that using a surrogate endpoint biomarker such as an adenoma is not without pitfalls. We know, for example, that cancers may sometimes develop from flat lesions, rather than the raised lesions that are typically targeted in adenoma prevention studies. Perhaps, calcium is effective for raised adenomas but not flat ones thereby overestimating the beneficial effect that has been demonstrated in trials.

Shaukat et al. (¹) performed a systematic review as a necessary first step for a high-quality metaanalysis. Systematic reviews use explicit and rigorous methods to identify, appraise, and synthesize relevant studies. Using a replicable, scientific, and transparent approach minimizes bias. The intent is to identify all admissible evidence. The authors consulted seven electronic bibliographic databases in order to locate as many studies as possible. Systematic reviews often restrict themselves to English language publications that can sometimes bias the results (²). The authors did not restrict by language. They also hand searched the list of references of the articles they retrieved in their search to identify additional data. Finally, they contacted the authors of the articles they had selected to obtain information that may not have been published. They apparently did not consult the "fugitive" literature. The fugitive literature consists of conference proceedings, government reports, published dissertations, and other unreferenced documents that are difficult to retrieve. It is not likely that the results would have been altered.

The authors used very stringent criteria—they limited the metaanalysis to randomized controlled trials that compared supplemental calcium versus placebo to prevent recurrent adenomas. Of the 2,053 citations identified in the systematic review, there were only three randomized trials that met the eligibility criteria for the metaanalysis. The pooled relative risk was 0.80 (95% CI: 0.68–0.93) with a risk difference of 7%. This corresponds to a number, needed to treat (NNT), of 14 (95% CI: 8–100). Because calcium is cheap and safe this NNT makes calcium supplements a reasonable strategy.

Metaanalysis are most useful when the individual studies are too small to provide precise estimates of benefit or harm. Metaanalysis may also be useful when the published study results are heterogeneous or when there is a need for a precise estimate that would be used in a decision analysis or to set public policy. In the metaanalysis by Shaukat et al. (¹) there are only three studies that were eligible to be included in the metaanalysis. The risk ratios in the three trials were consistently in the range from 0.71 to 0.83. Only one of the studies did not reach conventional levels of statistical significance. Because one of the studies was so much larger than the others, it contributed most of the weight to the overall summary estimate.

Based on the results of the published literature and this metaanalysis there is little question that calcium supplements decrease the risk of recurrent adenomas. A number of possible mechanisms have been suggested. The most popular hypothesis suggests that the beneficial effect is a consequence of binding and precipitation of bile acids and soluble fatty acids by calcium (³). Calcium has been shown to decrease cell proliferation in rats (⁴), although proliferation was not correlated with adenoma development in a randomized trial of calcium in humans (⁵). There has been a suggestion that calcium has a direct effect on colon cells that might be responsible for the beneficial effect (⁶).

What are the implications for our patients? It would be reasonable to recommend a calcium supplement to patients with prior adenomas with the expectation of a modest decrease in future adenomas. In addition to decreasing the proportion of individuals with adenomas, calcium has been shown to decrease the number of adenomas (⁷), potentially making colonoscopy quicker and safer. To the extent that colonoscopists miss lesions (⁸), an effective chemopreventive agent will offer protection. Calcium supplements would also be expected to decrease the risk of osteoporosis. There would appear to be very little risk. There have been reports linking increased calcium intake with prostate cancer in some studies (^9,10) but not others (¹¹).

Research on calcium and colorectal neoplasia continues. The Women's Health Initiative is a large randomized trial in the United States that is evaluating calcium and vitamin D primarily to prevent osteoporosis, but secondarily to evaluate malignancies such as colorectal cancer (¹²). A total of 45,000 women are expected to participate in the calcium arms of the study. The endpoint will be lesions that are discovered during routine medical care rather than identified during regular colonoscopy exams that are a feature of randomized chemoprevention studies such as the ones summarized by Shaukat et al. (¹). The Polyp Prevention Study Group, the group responsible for the largest of the calcium trials (⁷), is conducting a randomized trial of calcium and vitamin D to prevent colorectal adenomas.

While the concept of chemoprevention is attractive to our patients it is important that they understand that agents such as calcium offer incomplete protection. For that reason chemopreventive agents must remain adjuncts to colonoscopy. It would be unfortunate if our patients were falsely reassured by taking calcium and neglected their surveillance colonoscopies.