Chemoprevention of Second Primary Lung Cancers

Abstract & Commentary

Synopsis: Animal models have shown that vitamin A can reverse squamous metaplasia in the setting of vitamin A deficiency. Furthermore, published data from a randomized trial from Milan showed that patients who took vitamin A following treatment for stage I non-small cell lung cancer realized a statistically significant decrease in second primary tumors of the upper aerodigestive tract. An NCI-sponsored, randomized, intergroup placebo-controlled trial based on these types of data sought to confirm whether, indeed, second primary tumors and recurrences could be prevented with adjuvant isotretinoin. It was concluded that no protective effect could be identified, and that isotretinoin might be harmful if taken by current smokers.

Source: Lippman SM, et al. J Natl Cancer Inst. 2001;

During the years 1992 through 1997, 1265 patients were randomized from 6 weeks to 3 years post-stage I non-small cell lung cancer resection to placebo vs. 30 mg isotretinoin (13-cis-retinoic acid) daily for 3 years to determine whether second primary cancers and lung cancer recurrences could be prevented. Patients were stratified based on tumor histology, T-stage, and smoking status. Participating groups in this NCI-sponsored Intergroup 91-0001 trial included the RTOG, ECOG, SWOG, CALGB, NCCTG, and M.D. Anderson along with its affiliates. Based on an expected annual incidence of 2-3% for second primary tumors (SPTs), the study possessed at least an 80% power to detect a 50% reduction in SPTs. SPTs were defined as those tumors with different histology, lobe of origin, or incidence beyond 5 years as compared to the index lung tumor. After excluding patients who were implicated in a computer randomization error and those who were ineligible for participation, 1166 patients formed the basis of the study.

The median age of the study participants was 61 years (31-86). There were 57% males, 92% whites, 39% current smokers, 53% former smokers, and 68% non-squamous histologies among the participants. Median follow-up was 3.5 years. The trial was closed in February 2000 after 7 years. Multivariate analysis indicated that there were no statistically significant differences between the 2 study arms with regard to the study endpoints of SPTs and recurrences, or mortality. Both nonsmoking related SPTs (GI, prostate, breast) and smoking-related SPTs (lung, head & neck, esophagus, bladder) occurred with equal frequency in both groups. Sixty-two percent of the SPTs were considered to be smoking-related, including 47 in the isotretinoin group and 44 in the placebo group. Subset analysis revealed that current smokers in the vitamin A arm exhibited a trend toward higher recurrence rates (P = .15) and a statistically significant higher mortality rate (P = .01) than their counterparts in the placebo arm. Affected participants were advised to stop taking isotretinoin despite any remaining time in their 3-year study schedule.

Lippman and colleagues concluded that the trial results were disappointing, and were a reflection of the current lack of understanding of the interaction between retinoids and bronchial epithelial cells. Longer follow-up studies are planned.


Randomized trial results published by Pastorino and associates from Italy in 1993 indicated that adjuvant retinol palmitate, at 300,000 IU daily for up to 2 years, was effective in reducing the number of smoking-related upper aerodigestive tract SPTs and might improve disease-free survival in patients who underwent resection of stage I non-small-cell lung cancers.1 Three hundred and seven patients were followed for a median of 46 months. Thirteen smoking-related SPTs occurred in the vitamin A arm vs. 25 in the control arm. The estimated 5-year disease-free survival was 64% vs. 51% in favor of the treatment arm (P = .054).

Early results of the Italian study formed the basis of another randomized trial, the European Study on Chemoprevention with Vitamin A and N-acetylcysteine (EUROSCAN) study, which was begun in 1988.2 Once again, 300,000 IU of vitamin A daily over 2 years was tested as a chemopreventive agent in patients treated for head and neck or lung cancer. At a median follow-up of 49 months, van Zandwijk and associates did not find any benefit associated with either vitamin A or N-acetylcysteine. They concluded that study subjects should be followed for a minimum of 10 years since the latent period for carcinogenesis is at least that long.

The Lippman study, despite its reportedly negative outcome, raises some interesting issues. They state at the end of the paper that they are not certain why the isotretinoin failed to exert the anticipated effect. Unfortunately, little or no attention was paid to the potential drawbacks of the study, or to how a follow-up study might be improved. An easy mark for criticism is the mishandled randomization, which Lippman et al themselves pointed out. Other possible confounding factors are the 15% (n = 174) of patients who were lost to follow-up, and the 40% of participants in the vitamin A arm who were noncompliant with the study design. These numbers seem to be rather high, particularly since the analysis was carried out on an intent-to-treat basis.

The intergroup trial study design called for a chest x-ray up-front to rule-out existing SPTs in the lungs. However, based on the recently published Early Lung Cancer Action Project (ELCAP) work by Henschke and colleagues from Cornell,3 we saw that malignancies are detected 4 times as frequently on low-dose CT as on CXR. Eighty-three percent of the stage I malignancies were not detected on CXR. One can thus appreciate that there is no way to know whether the vitamin A cohort in the intergroup trial was destined to fail based on a higher number of unrecognized lung cancers at the time of randomization. Low-dose CT screening would be advisable in any new effort.

It is fairly certain, as mentioned in the EUROSCAN paper, that longer follow-up is needed before pronouncements of failure can be offered. This is highlighted by data in Pastorino et al’s paper which described the average time to developing a second malignancy as 16-42 months for smoking-related SPTs and, specifically, about 2 years for lung cancers. Lippman et al’s study suffers from a short (3.5 year) follow-up in relation to the timing of expected second events, which can take many years.

Just as the mechanism of action of retinoids is unclear, it is also unclear who will derive the most benefit from adjuvant vitamin A. Perhaps ongoing smoking neutralizes a beneficial effect, and perhaps never-smokers who develop lung cancer possess an intrinsic defect in their mucosa that is not amenable to correction by retinoid ingestion. A study focusing on initiation of vitamin A within 6 months of early stage lung cancer treatment in former smokers might be interesting. Lippman et al’s study enrolled participants up to 3 years removed from surgery, which may have been too late to reverse incipient tumor development.

In summary, no one knows the optimal time for initiation, dosage, duration, or formulation of vitamin A that will bestow the maximum benefit on patients at risk for SPTs or recurrences. It appears that continued follow-up, as suggested by Lippman et al, and improved study design eventually will help sort out whether vitamin A is ultimately proven to offer a preventive effect against second tumors and recurrences.


1. Pastorino U, et al. J Clin Oncol. 1993;11:1216-1222.

2. van Zandwijk N, et al. J Natl Cancer Ist. 2000;92:

3. Henschke CI, et al. Cancer. 2000;89:2474-2482.