Chemoprevention is a strategy to prevent the development of cancer through the administration of drugs. This approach is now being widely studied in a variety of clinical settings . The poor prognosis of lung cancer patients has led investigators to study chemoprevention as a means of decreasing lung cancer incidence. Tobacco cessation and public health efforts to prevent the next generation from becoming addicted to tobacco must be the mainstays of lung cancer prevention. Even with successful cessation, however, an increased risk of developing lung cancer persists in former smokers for more than a decade. Chemoprevention may be a means of decreasing the incidence of lung cancer among both current and former smokers.
The central idea guiding lung cancer chemoprevention efforts is the concept of the diffuse injury of respiratory epithelium resulting from chronic carcinogen exposure. Slaughter and his colleagues initially described this process as field cancerization . The widespread injury to the respiratory epithelium led investigators to consider a systemic treatment. The concept of chemoprevention is also supported by the understanding of epithelial carcinogenesis as a multistep process.
Among the potential agents for lung cancer chemoprevention, the use of retinoids, compounds with vitamin A-like activity, is supported by the results of both epidemiologic studies and animal experiments [3-6]. For clinical studies, the development of oral leukoplakia as a model of tobacco-associated carcinogenesis has also been useful in developing possible regimens for lung cancer chemoprevention . The lesions of oral leukoplakia are premalignant and, over a period of years, may develop into invasive cancers . Trials have demonstrated that retinoids, including isotretinoin(Drug information on isotretinoin) (Accutane, 13-cis-retinoic acid), could reverse these lesions .
The concept of field cancerization is supported not only by the frequent occurrence of premalignant lesions, such as leukoplakia, in carcinogen-exposed individuals, but also by the development of multiple primary tumors. For patients who survive an initial head and neck or lung cancer, development of a second primary tumor within the respiratory epithelium represents a great threat to their health [10-12]. For patients who have been treated for a head and neck cancer, the lifetime risk of developing a second primary tumor exceeds 30%. Building on the results of oral leukoplakia studies, retinoid chemoprevention trials have recently focused on the prevention of second primary tumors in patients who have been treated for a head and neck or lung cancer.
In a landmark study, Hong et al  described a reduction of second primary tumors in head and neck cancer patients treated with isotretinoin. Following surgery and/or radiation therapy for squamous cell cancer of the head and neck, patients were randomly assigned to either isotretinoin, 50 to 100 mg/m²/day, or placebo given for 1 year as an adjuvant treatment. In the initial report, with a median follow-up of 32 months, second primary tumors had developed in 12 (24%) of the placebo-treated patients, compared with only two (4%) of the isotretinoin-treated patients (P = .005). The retinoid treatment had no impact on the recurrence of the initial cancer. Isotretinoin apparently suppressed foci of damaged epithelium from progressing into invasive cancer, but was ineffective as therapy once the cancer had developed. The second primary tumors that occurred during the study were predominantly in the carcinogen-exposed field of the head and neck, lungs, and esophagus.
Although the results of this small trial (103 patients) are impressive, the high-dose retinoid treatment was associated with considerable toxicity. Side effects included dry skin, cheilitis, elevated triglycerides, and conjunctivitis. One third of the isotretinoin-treated patients were unable to complete the year of therapy as planned.
Recently the data have been reanalyzed, with the median follow-up extended to 54.5 months (Table 1) . With longer follow-up, the retinoid-treated patients have continued to have fewer second primary tumors; seven (14%) in the isotretinoin group, compared with 16 (31%) in the placebo group (P = .042). When only those second primary tumors that developed in the carcinogen-exposed field were considered, the results were more impressive, with second primary tumors developing in only three (7%) of the isotretinoin patients, compared with 13 (33%) in the placebo-treated group (P = .008). These results suggest that the beneficial chemopreventive effect of isotretinoin persisted after the year of treatment. The suppression of second primary tumors occurred despite the fact that the patients, as a group, took less of the retinoid than had initially been intended when the study was designed.
Because the high-dose isotretinoin used in this adjuvant study was poorly tolerated, there was considerable interest in determining if the chemopreventive effect could be maintained using lower, less toxic doses. The findings of a recent randomized oral leukoplakia chemoprevention trial are encouraging . Patients were initially given a 3-month induction course of high-dose isotretinoin, followed by a 9-month maintenance treatment with a low dose of isotretinoin, 0.5 mg/kg/day, or beta-carotene, 30 mg/day. The low-dose isotretinoin was effective in maintaining the benefits achieved during the induction phase; only two (8%) of the patients in the isotretinoin group had progression of the leukoplakia during the maintenance treatment, compared with 16 (55%) of the beta-carotene group (P < .001). The low-dose isotretinoin was also well tolerated; none of the isotretinoin patients discontinued therapy during the maintenance treatment due to toxicity (Table 2).
A group from France has recently reported the findings of a study that evaluated the efficacy of the synthetic retinoid etretinate (marketed as Tegison in the United States for the treatment of psoriasis) to prevent second primary tumors following squamous cell cancer of the oral cavity or oropharynx . Patients were randomly assigned to treatment with etretinate, 50 mg/day for 1 month, followed by 25 mg/day for 24 months, or placebo. Among the 316 patients studied, there was no reduction in second primary tumors associated with the retinoid treatment. The two treatment groups were equivalent both for the occurrence of second primary tumors and relapse of the initial cancer.
This prospective trial did, however, confirm the high rate of second primary tumors following head and neck cancer. With a median follow-up of 41 months, 24% of the patients had developed a second primary tumor. Consistent with the concept of field carcinogenesis, 79% of the second primary tumors occurred within the head and neck, lungs, or esophagus. Interpretation of this report was hampered by the lack of detail with regards to tobacco intake, alcohol(Drug information on alcohol) history, compliance, and toxicity .
In a trial performed among 307 patients following resection of a stage I non-small-cell lung cancer, Pastorino et al  observed a beneficial effect associated with retinyl palmitate treatment. Following surgery, patients were randomly assigned to treatment with retinyl palmitate, 300,000 IU per day for 1 year of observation. Compliance was estimated to be greater than 80%, and the retinyl palmitate was well tolerated. Only three patients dropped out of the treatment arm due to toxicity from the retinoid.
Retinyl palmitate treatment was associated with a reduction in second primary tumors. Eighteen patients in the retinyl palmitate group developed second primary tumors, compared with 29 patients in the control group. Reduction of tobacco-associated second primary tumors was more pronounced. With a median follow-up of 46 months, 13 retinyl palmitate-treated patients developed tobacco-associated second primary tumors, compared with 25 patients in the control group. The time to the development of a tobacco-associated second primary tumor also favored the retinyl palmitate-treated patients (P = .045). This study provided the rationale for the ongoing European study known as Euroscan (described below).
Another approach to retinoid chemoprevention of lung cancer has been to test the ability of these agents to reverse histologic or cytologic changes that may precede the development of invasive disease. Arnold et al , for example, evaluated the ability of etretinate, 25 mg/day, to reverse sputum atypia in specimens collected from chronic smokers. Changes in sputum atypia were assessed at the completion of a 6-month treatment period. No difference between the etretinate and placebo groups in the degree of sputum atypia was noted.
In order to evaluate the effect of the chemopreventive agent directly in the bronchial epithelium, a group of French investigators performed a study using serial bronchoscopy, with endobronchial biopsies taken from chronic smokers [20-22]. In this uncontrolled phase II study, patients found to have squamous metaplasia of the bronchial epithelium in specimens obtained during their initial bronchoscopy were treated with etretinate, 25 mg/day for 6 months. After the completion of the treatment course, the investigators noted a decline in the extent of squamous metaplasia.
As the result of this study, a randomized trial was performed in the United States . Asymptomatic chronic smokers were recruited to participate. Volunteers underwent bronchoscopy with endobronchial biopsies taken from six specific anatomic sites. Participants who were found to have a squamous metaplasia index greater than 15% and/or dysplasia were then randomized to 6 months of treatment with either isotretinoin, 1 mg/kg per day, or placebo. Of the 152 participants initially registered in the study, 93 had a metaplasia index greater than 15%.
The results of the 6-month treatment course were recently published for 69 study participants . For the group as a whole, the extent of metaplasia declined over time. There was no significant difference, however, between the isotretinoin and placebo groups in the reduction of metaplasia. The most important predictor of a decline in the extent of metaplasia was smoking cessation during the treatment period. While the results of this study were consistent with the findings in the French study with etretinate, the presence of a control group in the isotretinoin study completely altered the interpretation of the results.
Neither sputum atypia nor squamous metaplasia of the bronchial epithelium is likely to be useful as an intermediate marker of lung carcinogenesis, and both are too variable and nonspecific to guide chemoprevention trials. Sputum specimens, endobronchial biopsies, or bronchial washings may, however, be used to study other, potentially more powerful markers of lung carcinogenesis. There is now considerable interest in studying markers of proliferation, genetic injury, or oncogene expression, which may be critical in tobacco-associated carcinogenesis [24-28]. Understanding of the retinoic acid receptors may lead to the improved efficacy and decreased toxicity of retinoid chemoprevention [29,30]. No intermediate marker of lung carcinogenesis has yet been validated that could serve as a surrogate endpoint for lung cancer incidence in chemoprevention trials . Such a marker would be extremely useful in aiding the clinical development of lung cancer chemoprevention.