Vol. 1, 155-159, January/February 1992 Cancer Epidemiology, Biomarkers & Prevention 155

Review

Retinoids and Carotenoids in the Prevention of Oral Cancer:A Critical Appraisal1

Harinder S. Garewal2 and Frank Meyskens, Jr.3

Section of Hematology-Oncology, Tucson VA Medical Center,

Tucson, Arizona 85723

Introduction

The term “oral cancer” refers to malignant tumors arisingin the oral and oropharyngeal mucosa, including thoseof the tongue and lips. Squamous cancers account forover 90% of oral cancer. Internationally, oral and phamyn-geal cancers constitute the sixth commonest site in bothsexes combined (1). The highest reported incidencesoccur in pants of South and Southeast Asia, Brazil, andother developing countries, where up to 25% of allcancers are found in the oral cavity (2, 3). Recently,increases in some types of oral cancer have been notedin Scandinavia, the United States, and Scotland, espe-ciabby among younger males (4-6). In the United Statesapproximately 30,000 new cases occur annually, result-ing in about 9,000 deaths (7). This variation in regionalincidence has been related to known etiobogicab agents,the most important of which are tobacco and alcohol. Indeveloping countries, tobacco and betel nut chewing,often mixed with other substances such as slaked lime,are responsible for most cases of this disease (8).

Although management of oral cavity cancers hasimproved, there is little to no evidence that treatmenthas resulted in an increased survival mate. Patients whopresent with advanced disease and local metastases havea 5-year survival of about 30%. Unfortunately, mostpatients are diagnosed bate, even with easily detectablelesions such as those of the lip or tongue. Early tumorshave a high cure rate by radiotherapy and/or surgery.More advanced cancers have traditionally been treatedby radiation, neoadjuvant chemotherapy, and/or surgery.Although producing rather dramatic responses in therange of 50-85%, neoadjuvant chemotherapy has notchanged overall patient survival. The major causes ofdeath of advanced-disease patients include local recur-rence (about 50-60%), distant metastasis usually accom-panying local recurrence (20-30%), and, especially forearly-stage patients cured of their initial lesion, develop-ment of second primary tumors (10-40%) (9, 10). Thelatter has been attributed to the concept of “field can-cenization,” presumably by exposure to carcinogens bead-

Received 4/16/91.1 Supported in part by USPHS Grant CA27502 and a Career DevelopmentAward from the American Cancer Society (H. S. G.).2 To whom requests for reprints should be addressed, at Section of

Hematology-Oncology (hiD), Tucson VA Medical Center, 3601 S. 6thAvenue, Tucson, AZ 85723.3 Present address: UCI Cancer Center, 101 The City Drive, Building 44,Route 81, Orange, CA 92668.

ing to tumor formation in the epithebium of the upper

aerodigestive tract (11).As stated in a recent editorial in Lancet, “Prevention

is likely to prove the most fruitful avenue in reducing themorbidity and mortality from oral cancer” (12). This state-ment is based on the rather well-defined etiologicalassociations with respect to tobacco use, betel quidchewing, and alcohol consumption. Furthermore, con-sidemable recent evidence suggests a robe for nutritionalagents, particularly those rebated to vitamin A (retinoidsand camotenoids), in the inhibition of oral cancinogenesis.In fact, intervention trials with these compounds in sub-jects at risk for oral cancer have produced some of themost promising data on the potential for cancerchemoprevention.

The epidemiobogicab and biological evidence of amole for these agents in the inhibition of upper aemodiges-tive tract carcinogenesis has been the subject of recentreviews (13, 14). Similarly, experiments using animalmodels also have shown a marked inhibitory effect onoral carcinogenesis, especially in the hamster cheekpouch model system (15-17). In this review we willconcentrate mainly on human intervention trials con-ducted in individuals at high risk for oral cancer. Anumber of such studies describing improvement in pu-tative “intermediate markers” of cancer development orreversal or inhibition of premabignant lesions have beenreported over the past several years. The results indicatethat vitamin A-rebated compounds may indeed have arole in the prevention of this malignancy.

Intermediate Marker Studies. A number of intermediatemarkers of potential use in defining high-risk populationsare under study. These include micronucleated exfol-iated cells, specific cytokeratin expression, proliferationindices, growth factor expression, transglutaminase, in-vobucnin, and oncogene expression. Of all these, only thefrequency of micronucleated exfoliated cells has so fanbeen reported in published intervention trials. Micronu-cleated oral mucosal cells are considered to be indicativeof genotoxic damage induced by carcinogens (18). Theirfrequency has been increased in individuals at increasedrisk of cancer, particularly in studies conducted in Indiaand Southeast Asia, where betel nut, tobacco, and limechewing are prevalent (18). Orally administered fl-camo-tene, especially when given in conjunction with vitaminA, reduced the proportion of micronucleated cells frommucosa containing premabignant lesions (18, 19). Thisreduction can be demonstrated within a few weeks ofbeginning therapy and was sustained during the courseof treatment. At present, however, there are no studiesclearly linking the reduction in frequency of micronucbeito the reversal of premalignant lesions or prevention ofcancer.

on March 6, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

156 Review: Retinoids and Carotenoids in Oral Cancer Prevention

Studies in Premalignant Lesions. The development ofcancer occurs through several steps involving initiation,promotion, and progression to malignancy. Premalignantlesions are the first clinically identifiable lesions alongthis pathway. Therefore, the search for interventionscausing reversal or suppression of premalignant lesionsconstitutes an important strategy for cancer prevention,e.g., reduction in incidence or removal of colonic polypsas a preventive approach for colon cancer.

We should emphasize at this point that the ultimategoal of this strategy is to develop interventions applicableto the prevention of cancer and not merely to eradicatepremalignant lesions. In general, the latter are not lethalor even morbid by themselves and are associated withrather low rates of transformation to cancer. Therefore,it is imperative that any agents selected on the basis oftrials in premalignant lesions that are likely to proveworthwhile for cancer prevention have minimal, or pref-crabby no toxicity, since a large number of subjects whoselesions are unlikely to progress to cancer will necessarilybe exposed to the intervention. Cleanly, the types andamount of side effects considered acceptable for anytherapy depend on the severity of the condition beingtreated. High levels of toxicity are quite acceptable intreatments for overt malignancy. Similarly, a moderatedegree of toxicity can be tolerated for some premalignantdiseases such as familial pobyposis of the colon which areassociated with a very high cancer risk. However, for themajority of the more common premalignant lesions, thecancer risk is often very low, and almost any side effectsproduced by a drug will generally be unacceptable.

The majority of oral cavity pmemabignant lesionscome under the category of leukoplakia, i.e., a whitepatch on plaque on the mucosa that cannot be rubbedoff and is not attributable to a specific disease entity. In

general, they have a rather low malignant potential (20).Oral emythroplakia and speckled leukoplakia have ahigher transformation mate but are relatively mare lesions.Similarly, the presence of severe dysplasia demands amore aggressive treatment strategy (20).

The ultimate objectives of intervention trials involv-ing leukoplakia must therefore be kept in mind whendesigning chemoprevention studies. If the objective is todevelop a treatment applicable to a small minority ofpatients with emythmoplakia and/or high-grade dysplasiathat are not amenable to standard treatments such asreduction in local irritants, surgical excision, or cryosum-gemy, then some degree of toxicity in the therapy may beacceptable. In this category would be the use of agentssuch as topical bleomycin and 5-fluomoumacil, which have

been shown to be effective (21). However, ifthe objec-tive is to develop agents for generalized, population-

based use for the primary prevention of oral cancer, then

clearly such agents are not practical.

Retinoid and Carotenoid Trials in Leukoplakia

Several trials have been conducted demonstrating theefficacy of various retinoids, both natural and synthetic,in reversing leukoplakia. In earlier studies Koch (22, 23)showed that 13-cis-netinoic acid, trans-metinoic acid, andetmetinate each could produce objective responses in60-90% of patients. The treatment was associated withconsiderable (particularly mucocutaneous) toxicity, withfrequent relapses after therapy was discontinued. Resultsof more recent trials are summarized in Table 1 . Shah etal. (24) also reported a high response rate (82%) in 1 1patients treated with 1 3-cis-retinoic acid lozenges. It is ofinterest that 16 patients were entered on this trial, butfive could not complete the treatment, with three stop-ping because of toxicity. (One of five showed carcinomain situ of the lip along with leukoplakia of the tongue andwas considered inevaluable. No change in the tongue

lesion was noted during 7 weeks of treatment in thispatient.) Therefore, the total response rate based on 16entered patients was nine of 1 1 evaluable (82%) onperhaps more correctly nine of 16 treated (56%). Of thethree patients with complete response, a follow-up bi-opsy showed normal epithelium with residual leuko-plakic changes in a single microscopic focus in onepatient, another showed improvement of leukoplakia tonormal epithelium, while the remaining case showedhypenkenatosis only. Varying histological findings in the

six patients with partial responses were also reported.In a randomized, double-blind, controlled trial of 13-

cis-metinoic acid in 44 patients, a response rate of 67%(8% complete, 59% partial response) was reported in 24patients in the treatment arm versus two partial responsesin 20 patients (10%) in the placebo arm (25). Mucocuta-neous toxicity including cheilitis, facial erythema, and

dryness and peeling of the skin occurred in 79% andconjunctivitis in 54% of patients in the treatment arm.These results compared with 20% or less reported sideeffects in the placebo arm, emphasizing the problemsand futility of “blinded” randomized trials of toxic drugsversus placebo on other nontoxic agents.

“Histological responses” were reported in 39 of the44 patients entered in whom pretreatment and posttreat-ment biopsy specimens had been obtained. Dysplasiawas “reversed” in 13 of 24 (54%) of patients in the 13-

Table 1 Recent retinoid/carot enoid trials in or aI leukoplakia

Investigator(Ref.)

AgentNo. of

patientsCR’ % PR % OR % Country

Shah et al. (24)

Hong et al. (25)Stich et al. (19)C

Stich et al. )26)CGarewal et al. (27)

13-cRA (lozenges)

1 3-cRABCBC+VitA

Vit A

BC

ii (16)”

242751

21

24

27 (19)8

1527

57

8

55 (37)59NSNSNS

63

82 56)67NSNSNS

71

United StatesUnited SlatesIndia

India

United Stales

a CR, complete remission; NS, not stated; PR, partial response; OR, overall response; 1 3-cRA, 1 3-cis-retinoic acid; BC, �J-carotene.

b Numbers in parentheses are results based on patients entered (see text).C Only complete remission rate reported.

on March 6, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

Cancer Epidemiology, Biomarkers & Prevention 157

cis-metinoic acid group, as compared with 10% (2 of 20)in the placebo group. However, close scrutiny of the

reported histological change data (Ref. 25, Fig. 3) showsthat in the placebo arm the majority of patients withpretreatment and posttreatment biopsies were in theatypical hyperplasia (i.e., free of dysplasia) category andremained so at the end of treatment. Only five patientshad dysplasia on their pretreatment biopsy, with threeremaining in the same severity category, two showingimprovement (including one severely dysplastic becom-ing nondysplastic), and one showing worsening. In addi-

tion, two patients without dysplasia on pretreatmentbiopsy reportedly had dysplasia on their posttmeatmentbiopsy. Also, a larger number of patients in the treatmentgroup (15) had dysplasia compared with the placebogroup (8). As discussed further below, one must becautious in interpreting “improvement” in histology be-cause this may merely reflect sampling error, especiallyin the partial responders, which constituted the vastmajority of responding patients in this trial.

Stich et al. (19, 26) have conducted a series of trialsin India using vitamin A alone or in combination with fi-carotene. It should be emphasized that this study popu-lation is different from that enrolled in other trials, sincethe majority of lesions here result from tobacco and betelquid chewing. Nevertheless, from a worldwide pemspec-tive, these studies are particularly relevant, since they arebeing conducted in countries where the disease is amajor health problem. In one trial treatment consisted of/3-carotene (180 mg/week) (group I), �3-carotene (180 mg/week) plus vitamin A (100,000 lU/week) (group II), andplacebo (group Ill) given twice weekly for 6 months (19).After 6 months of treatment complete remission of eu-koplakias was observed in 1 5% of patients in group I,27% in group II, and only 3% in group Ill. Similarly,

development of new leukoplakias during the 6-monthperiod was strongly inhibited in group II (8%) and inhib-ited to a lesser degree in group I (15%) in comparisonwith group Ill (21%). In a second trial by the sameinvestigators, subjects were randomly divided into twogroups, one receiving 200,000 IU of vitamin A/week for6 months and the other receiving placebo (26). A 57%complete remission rate was observed in the vitamin Agmoup with total suppression of the development of newleukoplakias. In contrast a 3% complete remission and2 1% new lesion rate in the placebo group were observed.Biopsies were performed at the beginning and the endof treatment in this second trial. Improvement in varioushistological criteria such as number of layers of spinouscells, loss of polarity of basal cells, and subepidemmallymphocytic infiltration was significant in the vitamin A-treated but not in the placebo group.

In a pilot study conducted by our group, a response

rate of approximately 70% (8% complete, 63% partialresponse) was observed in 24 patients treated with �3-carotene at a dose of 30 mg/day (27). A majority of thepatients used tobacco and/or alcohol. Biopsies of alllesions were done at entry, with dysplasia present in 1 1of the 24 subjects. Follow-up biopsies were not pen-formed in this preliminary trial.

In interpreting the above trials as well as developinga strategy for the chemoprevention of oral cancer severalissues need to be kept in proper perspective. Some ofthe more important ones are briefly discussed below.

Clinical Response Criteria. Leukoplakia is a lesion thatcan be directly visualized. Therefore, marked improve-ment and thinning of the lesion are often apparent onvisual examination but are difficult to quantitate in amanner similar to that usually used in therapeutic trialsof anticancer agents, where often the responding lesionis visualized and measured on a radiograph or otherscanning modality. Therefore, although we and others(e.g., Ref. 25) have tried to apply the standard 50%reduction in the product of the two longest diameters oflesions to define so-called partial responses, this is by nomeans easy to do with any degree of precision in adisease such as beukoplakia in which a response mayclearly be evident but difficult to measure.

One approach is to consider only complete me-sponses in the evaluation of leukoplakia trials. This ap-proach has been taken in the studies reported by Stichet al. (19, 26) with fl-carotene and vitamin A. If this isdone, then the response rates of f3-carotene in our trial(8%), 13-cis-retinoic acid in the trial by Hong et al. (8%),and �3-carotene alone observed by Stich et a!. (15%) arequite similar. Higher complete response rates of 27%with f3-camotene in combination with 100,000 lU/weekof vitamin A and 57% using 200,000 bU of vitamin A werereported by Stich et al. (19, 26). Adopting this approachof estimating complete response only does have the riskof discarding potentially valuable modalities in short-duration trials since, even though they are difficult toquantitate, many lesions do improve dramatically, andthis may signify considerable activity by the agent beingtested.

Histological Response. The interpretation of so-calledhistological responses deserves considerable scrutiny. Itis known that dysplasia or other histological abnormalitiesare not evenly distributed throughout premalignant orother lesions, including those of the oral cavity. In otherprecancerous conditions such as ulcerative colitis, Bar-rett’s esophagus, esophageal dysplasia, and cervical dys-plasia, it is well established that the distribution of dys-plastic changes is patchy, with respect not only to theirpresence or absence but also to their degree (28, 29). Inaddition, theme is considerable intra- and interobservervariability in quantitating the degree of such change inan individual biopsy (28). Therefore, comparing twobiopsies often taken weeks to months apart, with theobjective of comparing the degree of dysplasia, is anapproach potentially subject to serious error. This is notto say that follow-up biopsies are of no utility at all. Theirgreatest usefulness lies in the setting of complete clinicalresponse where it would be important to show thatepithebial maturation has also normalized at the histobog-ical bevel. Patients with partial or no clinical response totreatment that are alleged to have improved histologicallyin their degree of dysplasia may perhaps be acceptableif this improvement is demonstrated in a series of follow-up biopsies that show few to no “oscillations” of dysplasiagrades from biopsy to biopsy. Otherwise, changes froma single pretreatment biopsy to a single follow-up post-treatment biopsy will most probably reflect sampling andinterpretation error rather than a true change in the gradeof dysplasia.

Recognizing these difficulties with quantitatingchanges in hypemplasia and dysplasia, Stich et al. (26) intheir recent publications have attempted to developother histological and cytological criteria. These too need

on March 6, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

1 58 Review: Retinoids and Carotenoids in Oral Cancer Prevention

to be validated before they can be accepted as responsecriteria on their own. Another approach that has beenapplied to some organs, such as the stomach or thecolon, would involve doing multiple biopsies at eachevaluation visit and comparing the composite scores ofseveral biopsy readings to try and reduce the effect ofsampling. However, because of the nature and locationof oral cavity lesions, this is not generally feasible formost patients entered into oral leukoplakia trials.

Suitability of Agents for Chemoprevention. Preventionand treatment are two different goals in the control ofcancer. Potential chemopreventive agents must haveproperties that are not necessarily required for drugsused in advanced cancer. They must be essentially non-toxic to be acceptable for widespread use. Availability ofa chemopreventive agent as a nutrient has a definiteadvantage, since it allows consideration of supplemen-tation by dietary adjustments. Finally, the cost of theagent should not be totally prohibitive for use in under-developed countries, which face the most serious prob-bems with oral cancer.

With the above considerations in mind, it is evidentthat toxic agents such as bleomycin, 13-cis-retinoic acid,and trans-metinoic acid, even if active, are not suitable foruse except in the mane case with severe dysplasia that isnot amenable to treatment by presently available tech-niques. At the doses used, these retinoids produce der-matobogical, hepatic, and other side effects too severefor general population use. Their teratogenicity, even atlower doses, is also a serious problem. Perhaps similarconsiderations apply to higher doses of vitamin A. Al-though Stich et al. report no toxicity at the 200,000 IU/day dose that has produced the highest complete me-sponse rate (57%) reported to date, it should be empha-sized that only a small number of patients were treatedand no results of liver function tests, serum lipids, andother biochemical measurements were reported. In ad-dition, these trials were conducted in a population thatis probably deficient in vitamin A. /3-Carotene, on theother hand, does fulfill all the criteria for a suitablechemopreventive agent in that it is nontoxic, it is reba-tively inexpensive, and it is a nutrient.

Future Directions

From currently available data, both vitamin A and fi-carotene appear to have some activity in reducing thefrequency of micronucleated exfoliated oral cells and inthe reversal and suppression of leukoplakia. Although fi-carotene appears to have some activity on its own, thetrials conducted by Stich et al. suggest that the combi-nation of this agent with bow doses of vitamin A mayresult in a greater complete response mate. It should benoted, however, that a biweekly dosing schedule wasused in these trials because of practical considerationsrather than the usual daily doses used in trials conductedby other groups. Owing to the known toxicities of vitaminA, especially when used over a prolonged period of timeat doses in excess of about 25,000 lU/day, caution mustbe exercised when selecting a dose for use in chemopre-vention trials. Vitamin E is another relatively nontoxicagent that is effective in animal models (17, 30). Noclinical trials with vitamin E have been reported to date.

Target populations suitable for screening trials in-dude not only those subjects with leukoplakia but also

patients who have been cured of their primary head andneck cancer. Most patients with early cancer are curedby current treatment modalities but are at a high risk forsecond primary neoplasms. Agents active in reversingpreneoplastic lesions might indeed prove to be effectivein reversing the “field cancenization” defect thought tounderlie the increased incidence of second malignanciesof the aerodigestive tract. Nontoxic combinations are tobe preferred in this setting also, since prolonged treat-ment is anticipated and many of these patients havereceived radiation treatment resulting in mucosal injury.Therefore, they may be unable to tolerate the mucocu-taneous toxicity associated with the currently availablesynthetic retinoids. Nevertheless, a somewhat greaterdegree of toxicity may be acceptable in this group, sincecancer risk appears to be higher than in the usual patientwith leukoplakia.

A recent report of an adjuvant trial using high-dose13-cis-metinoic acid in patients with all stages (I to IV) ofcancer showed a remarkable reduction in the incidenceof second primary tumors, although it failed to show asignificant adjuvant effect, i.e., reduction in recurrence(31). Toxicity was a major problem, leading to prematurediscontinuation of therapy in one-third of the study pa-tients. Trials using lower doses of 13-cis-metinoic acid onfl-carotene (± metinol) in early-stage patients have me-cently been initiated.

Finally, the mole of maintenance on long-term therapyneeds to be studied. In most leukoplakia trials reportedso far, the lesions tend to recur once treatment is discon-tinued. In fact, in the study by Hong et al., the reportedtime course of response data suggests that some patientsbegan to relapse while still receiving 1 3-cis-metinoic acidat 1-2 mg/kg/day, i.e., within the 3-month treatmentperiod. It is not yet clear whether continued treatmentwith bower doses of vitamin A on /3-carotene, alone or incombination, will indeed be effective for this purpose inthe trials of Stich et al., after an initial response has beenobtained with higher doses of vitamin A.

Nevertheless, in view of these early exciting results,which do establish the ability of these agents to affectestablished precancerous lesions, the time is appropriateto begin planning strategies to test whether oral cancercan actually be reduced in a targeted population bysupplementation with chemoprevention agents. Theidentification of a suitable population and the training ofpersonnel to conduct such a trial over a prolonged periodof intervention will clearly be an immense undertaking.However, the ultimate objective is prevention of oralcancer and not merely treatment of leukoplakia. It ap-pears likely from current trials that a promising nontoxicagent or combination of agents will be identified fromongoing studies in patients with oral premabignant Ic-sions. But will their supplementation actually preventoral cancer? This can be definitively answered only byconducting randomized controlled phase Ill interventiontrials in suitable populations, i.e., high-risk populations inendemic areas, for prevention of “primary” cancers, andearly-stage head and neck cancer patients, cured of theirprimary tumor, for prevention of second malignancy. Inthe United States, a number of trials in the latter categoryhave recently begun on are being planned, reflecting theincreasing interest in this field of cancer research. Theseinclude interventions with /3-carotene alone (Yale), �

carotene plus low-dose vitamin A (Southwest Oncology

on March 6, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

Cancer Epidemiology, Biomarkers & Prevention 159

Group), and 13-cis-metinoic acid (Northern California On-cology Group).

References

1. Parkin, S. M., Laara, E., and Muir, C. S. Estimates of the worldwidefrequency of sixteen major cancers. Int. J. Cancer, 41: 184-197, 1988.

2. Dunham, L. I. A geographic study of a relationship between oral

cancer and plants. Cancer Res., 28: 2369-2371, 1968.

3. Baden, E. Tabac et cancers de Ia region oropharyng#{233}e et des bronches.

Donn#{233}es actuelles. Rev. Med. Toulouse, 14: 549-560, 1978.

4. Davis, S., and Severson, R. K. Increasing incidence of cancer of thetongue in the US among young adults. Lancet, 2: 910-911, 1987.

5. Hakulinen, 1., Anderson, A. A., Malker, B., et al. Trends in cancer

incidence in the Nordic countries. Ada Pathol. Scand., 288(Suppl): 94,

1986.

6. McFarlane, G. I., Boyle, P., and Scully, C. Rising mortality from cancerof the tongue in young Scottish males. Lancet, 2: 912, 1987.

7. Silverberg, E., Boring, C. C., and Squires, T. S. Cancer statistics. CA

Cancer I. Clin., 40: 9-26, 1990.

8. Muir, C. S.. and Kirk, R. Betel, tobacco and cancer of the mouth. Br.

I. Cancer, 14: 597-608, 1960.

9. McGuirt, W. F., Mathews, B., and Kaufman, J. A. Multiple simultaneous

tumors in patients with head and neck cancer. Cancer (Phila.), 50: 1195-1199, 1982.

10. Kotwall, C., Razack, M. S., Sako, K., et al. Multiple primary cancersin squamous cell cancers of the head and neck. I. Surg. Oncol., 40: 97-99, 1990.

11 . Slaughter, D. P., et a). Field cancerization in oral stratified squamous

epithelium, clinical implications of multicentric origin. Cancer (Phila.), 5:963-968, 1953.

12. Oral cancer (Editorial). Lancet, 2: 311-312, 1989.

13. Lippman, S. M., Kessler, I. F., and Meyskens, F. L. Retinoids aspreventive and therapeutic anticancer agents. Part I. Cancer Treat. Rep.,

71: 391-405, 1987.

14. Ziegler, R. G. A review of epidemiologic evidence that carotenoids

reduce the risk of cancer. J. Nutr., 1 19: 116-122, 1989.

1 5. Burge-Bottenbley, A., and Shklar, G. Retardation of experimental oral

cancer development by retinyl acetate. Nutr. Cancer, 5: 121-129, 1983.

16. Schwartz, J., Suda, D., and Light, G. /3-Carotene is associated with

the regression of hamster buccal pouch carcinoma and induction of

tumor necrosis factor in macrophages. Biochem. Biophys. Res. Commun.,136: 1130-1135, 1986.

1 7. Shklar, G., Schwartz, J., TrickIer, D., el al. Regression of experimentalcancer by oral administration of combined ce-tocopherol and f3-carotene.Nutr. Cancer, 12: 321-325, 1989.

18. Stich, H. F. Micronucleated exfoliated cells as indicators for genotoxicdamage and markers in chemoprevention trials. J. Nutr. Growth Cancer,4:9-18, 1989.

19. Stich, H. F., Rosin, M. P., Hornby, A. P., et al. Remission of oral

leukoplakias and micronuclei in tobacco/betel quid chewers treated with

fl-carotene and with fl-carotene plus vitamin A. Int. J. Cancer, 42: 195-199, 1988.

20. Silverman, S., and Shillitoe, E. J. In: S. Silverman fed.) Etiology andPredisposing Factors in Oral Cancer, Ed. 2, pp. 7-36, New York: American

Cancer Society, 1985.

21. Wong, F., Epstein, J., and Millner, A. Treatment of oral leukoplakia

with topical bleomycin. Cancer (Phila.), 64: 361-365, 1989.

22. Koch, H. F. Biochemical treatment of precancerous oral lesions: theeffectiveness of various analogues of retinoic acid. J. Maxillofac. Surg., 6:

59-63, 1978.

23. Koch, H. F. Effect of retinoids on precancerous lesions of oral

mucosa. In: C. E. Orfanos, 0. Braun-Falco, E. M. Farber, et al. (eds.),Retinoids: Advances in Basic Research and Therapy, pp. 307-312. Berlin:

Springer-Verlag, 1981.

24. Shah, J. P., Strong, E. W., DeCosse, J. J., et a!. Effect of retinoids on

oral leukoplakia. Am. J. Surg., 146: 466-470, 1983.

25. Hong, W. K., Endicott, J., Itri, L. M., et a!. 13-cis-Retinoic acid in the

treatment of oral leukoplakia. N. EngI. J. Med., 315: 1501-1505, 1986.

26. Stich, H. F., Hornby, A. P., Mathew, B., et al. Response of oral

leukoplakiastothe administration ofvitarnin A. Cancer Lett., 40: 93-101,1988.

27. Garewal, H. S., Meyskens, F. L., Killen, D., et a!. Response of oralleukoplakia to fl-carotene. J. Clin. Oncol., 8: 1715-1720, 1990.

28. Reid, B. J., Haggitt, R. C., Rubin, C. E., et al. Observer variation inthe diagnosis of dysplasia in Barrett’s esophagus. Hum. Pathol., 19: 166-178, 1988.

29. Crissman, J. D. Upper aerodigestive tract. In: E. D. Henson and J.

Albores-Saavedra (eds.), The Pathology of Incipient Neoplasia, pp. 39-

55. Philadelphia: W. B. Saunders Co., 1986.

30. Odukoya, 0., Hawach, F., and Shklar, G. Retardation of experimental

oral cancer by topical vitamin E. Nutr. Cancer, 6: 98-104, 1984.

31. Hong, W. K., Lippman, S. M., tn, L. M., et a!. Prevention of second

primary tumors with isotretinoin in squamous-cell carcinoma of the head

and neck. N. EngI. J. Med., 323: 795-801, 1990.

on March 6, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

1992;1:155-159. Cancer Epidemiol Biomarkers Prev   H S Garewal and F Meyskens, Jr  critical appraisal.Retinoids and carotenoids in the prevention of oral cancer: a

  Updated version

  http://cebp.aacrjournals.org/content/1/2/155.citation

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cebp.aacrjournals.org/content/1/2/155.citationTo request permission to re-use all or part of this article, use this link

on March 6, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from