n-Carotene Treatment of Cervical Intraepithelial Neoplasia: A Phase II Study1

The use of Papanicolaou smears for cervical cancer screening has led to an increased detection of preinvasive conditions of the cervix, cervical intraepithelial neoplasia (CIN). Epidemiological studies have shown an association between low levels of dietary /3-carotene and CIN. In this Phase II study, we have explored the effect of p.o. 13carotene administration on CIN I and II. Patients with documented CIN I or II were treated with 30 mg daily of 13-carotene for 6 months. Response rates were determined at 0, 3, 6, and 12 months with cytology, colposcopy, and/ or biopsies. Levels of 13-carotene and vitamin E were determined at the same time intervals in vaginal mucosa cells and serum. Response rates were 18 of 30 (60%), 21 of 30 (70%), and 10 of 30 (33%) at 3, 6, and 12 months, respectively. Significant changes occurred in the serum 13-carotene levels over time. Median levels over 2200 mg/ ml were found at 3 and 6 months versus a baseline median level of 111 (P < 0.0001). Significant increases were also noted in the 13-carotene levels of the vaginal mucosa compared to baseline (P 0.01) and a significant correlation was noted between serum and vaginal 13carotene levels as well (P < 0.0001). This study indicates that a large percentage of patients with CIN I and II will respond clinically to p.o. 13-carotene supplementation. There is a positive relationship between serum and tissue levels of 13-carotene which suggests that serum levels can be used for monitoring purposes. Because of these encouraging results, prospective randomized studies are ongoing comparing the efficacy of 13-carotene against an untreated control arm. Received 4/1/96; revised 7/29/96; accepted 8/1/96. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I Supported in part by Hoffman-LaRoche. 2 To whom requests for reprints should be addressed, at Department Obstetrics and Gynecology, University of Califomia, Irvine Medical Center, 101 The City Drive, Building 23. Orange, CA 92668. Phone: (714) 456-6570; Fax: (714) 456-6463. Introduction Carcinoma of the cervix is one of the most common and deadly malignancies in the developing world. In some countries, it is the leading cause of death for women between the ages of 30 and 50. In the United States, the incidence and mortality rate of cervix cancer has declined over the last few decades with approximately 13,000 women developing the disease and 6,000 dying annually. This decreasing incidence is attributed to the effectiveness of screening programs. Because of a well-established premalignant phase, ease of use of Pap smears, low cost, and a high degree of acceptability, screening programs have been extremely successful in decreasing the incidence of invasive cervix cancer. However, the success of this program has led to a significant increase in the detection of CIN.3 Because of the easy accessibility for biopsy and inspection, cancer of the cervix is an ideal tumor for the study of chemopreventive agents. An association has been found between the incidence of CIN and dietary levels of vitamin A and/or /3-carotene (1-4). The risk of cervical cancer was also found to be higher among women with lower serum levels of total carotenoids (4). /3-carotene has been used with success in the management of premalignant lesions such as p.o. leukoplakia (5, 6). Previous clinical trials have demonstrated the potential use of topical /3-trans-retinoic acid in the management of CIN (7, 8). However, local toxicity was commonly experienced. In contrast, /3-carotene has been found to be relatively nontoxic to humans; at high doses, a slight yellowing of the skin may occur. Because of this lack of side effects, /3-carotene has received attention as a chemopreventive agent. In this Phase II trial, we have obtained preliminary data on the effect of /3-carotene on CIN. In addition, we investigated the relationship between the serum and tissue levels of micronutrients and their response to p.o. /3-carotene supplementation. Materials and Methods This nonrandomized Phase II study was intended to determine the activity and side effects of /3-carotene in the management of patients with CIN I and II. Eligibility criteria for study entry were the presence of an intraepithelial lesion recognizable using colposcopy and documented by biopsy as CIN I or II. Eligible patients had residual lesion or lesions remaining on the cervix following the biopsy. Patients with positive endocervical curettetage on previous or concomitant examinations were not eligible for this protocol. Also ineligible were patients who had been exposed in utero to diethylstilbestrol. All patients underwent initial cervical cytology and colposcopy with mapping of the cervical lesion on appropriate forms. All colposcopies were performed either by a on November 6, 2017. © 1996 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from 930 3-Carotene Treatment gynecological oncologist or a gynecological oncology fellow. Standard criteria for evaluation of CIN were used by both investigators. Quadrant extension of the lesion was also noted. Colposcopically directed biopsies were performed in all patients. An informed consent approved by the Institutional Review Board at University of California, Irvine was signed by all patients. Pathologists who were blinded as to the patients’ participation in the study graded the biopsies as normal or CIN 1, 2, or 3. Vaginal and serum levels of /3-carotene, retinol, and vitamm E were determined at 0, 3, 6, and 12 months. Analysis was performed as described previously (9). Serum samples were obtained on 86 of 122 (70%) of the total patient visits. The majority of missing samples (23/36) occurred among the first I 0 patients because of a technical failure. Cervical-vaginal cell samples were obtained from 16 patients in 46 of52 (88%) visits during the second half of the study. A cytobrush was used to collect cells from the vaginal wall. The cells were stored and transported in saline solution at 4#{176}C. Patients received 30 mg/day of /3-carotene for 6 months and were assessed for extent of disease at baseline and for response at 3, 6, 9, and 12 months. Pap smears were routinely performed at each follow-up visit. All patients had biopsies at baseline. At subsequent visits, biopsies were only performed if the colposcopy suggested progression of the dysplasia. If progression was confirmed, patients were then removed from the study and considered treatment failures. Patients with stable disease at 9 months were removed from study and treated conventionally. Patients who achieved a response were followed for an additional 6 months off therapy. Patients were considered responders if the Pap smear and colposcopy reversed to normal. In cases with negative Pap smears but suspicious colposcopy, biopsies were taken to determine response. All patients with abnormal Pap smears or positive biopsies were considered to be nonresponders. Statistical Methods. Comparison of response probability by disease stage or extent (quadrants involved) were made using the Fisher exact test. To avoid introducing a bias due to patient drop-off, patients who dropped off study were considered to be nonresponders at all subsequent times. The Kruskal-Wallis test was used to compare micronutrient levels with time. Correlations were assessed using the Spearman rank correlation method. The interquartile range gives the 25th and 75th percentile values for the variable being described, thus providing the limits for the central half of the data. Cell counts from fresh cervical-vaginal cells were scored as < I , I , 2, or 3 million cells. When the number of cells was reported as <1 million (n = 15), 0.7 million was used. Since the lowest observed /3-carotene level was 0.10 ng/million cells, a value of 0.05 ng/million cells was used in the 14 instances when the level was not detectable. Formal comparison of the serum micronutrient and cellular /3-carotene levels was done only in the patients for whom data both on (3 or 6 months) and off (0, 9, or I 2 months) /3-carotene supplementation was available using the Wilcoxon signed rank test. Whenever multiple levels were available for a patient, the values were averaged.