The effect of retinoids on premalignant oral lesions

Focus on topical therapy


  • Meir Gorsky D.M.D.,

    1. The Maruice and Gabriela Goldschleger School of Dental Medicine, Tel Aviv University, Tel Aviv, Israel
    Search for more papers by this author
  • Joel B. Epstein D.M.D., M.S.D.

    Corresponding author
    1. British Columbia Cancer Agency, Vancouver, British Columbia, Canada
    2. Department of Oral Medicine, University of Washington, Seattle Washington
    • British Columbia Cancer Agency, 600 W 10th Avenue, Vancouver, British Columbia, Canada V5Z 4E6
    Search for more papers by this author
    • Fax: 604-872-4596



Retinoids have been studied as chemopreventive treatment for patients with oropharyngeal carcinoma. Vitamin A modulates growth and differentiation of cells, and its deficiency enhances susceptibility to carcinogenesis. The chemopreventive mechanism of action of vitamin A is discussed, and a review of clinical results and side effects of the systemic use of vitamin A is included. The objective of the current report was to review the possible role of topical vitamin A and vitamin A derivatives in the management of patients with oral lesions with a risk of transformation to carcinoma.


A Medline search was conducted and references identified within the identified papers were also reviewed.


Only four studies using topical vitamin A for patients with oral leukoplakia have been reported. A complete response was achieved in 10–27% of patients, and a partial response was achieved in 54–90% of patients; however, recurrence of leukoplakia was reported after withdrawing the medication in approximately 50% of patients. The side effects of the topical use were minimal.


Although the direct application of higher concentrations of retinoic acid results in suppression of oral leukoplakias only, its use in the treatment of patients with recurrent and persistent lesions may be justified for controlling lesions that otherwise may progress. Further controlled clinical studies are needed. Cancer 2002;95:1258–64. © 2002 American Cancer Society.

DOI 10.1002/cncr.10874

The term leukoplakia is a clinical descriptive term for a white lesion of the oral mucosa that does not represent any specific histologic condition.1, 2 Any lesion in which the etiology is known should be described with the etiology as a modifier, for example, tobacco-associated leukoplakia and frictional or traumatic leukoplakia. Leukoplakia may represent histologic findings of a wide variety of conditions from epithelial hyperkeratosis, hyperplasia, and dysplasia to carcinoma. If a specific histologic diagnosis is made, then these findings should be included in the diagnosis, for example, dysplastic leukoplakia and hyperplastic candidiasis. Although only up to 10% of leukoplakias may represent dysplastic lesions, up to 17.5% of the lesions may transform into carcinoma, depending on the severity of dysplasia.3 Because a transformation risk as high as 43% was reported for dysplastic leukoplakias,4 patients with such lesions must be managed and followed closely, so that, if malignant degeneration of the lesion occurs, early diagnosis and treatment of the malignancy is possible.

Recent studies have documented molecular profiles of oral dysplasia and the risk of progression of dysplastic lesions to malignant disease.5–19 These molecular markers include loss of heterozygosity often at 3p, 9p, 17p, 8q, 11q, and 13q and other makers, such as p53. These findings that predict the progression of oral lesions to dysplasia and carcinoma may allow the identification of lesions for treatment, may serve as intermediate markers of interventions, and may lead to new treatment approaches.

Current treatment approaches for patients with oral dysplastic lesions include excision that may be indicated for localized and accessible lesions. However, surgical management may be limited for patients with widespread lesions when they involve sensitive anatomic structures, such as over the ducts of major salivary glands, or when lesions recur after repeated excision. Excision may cause postoperative morbidity. In addition, recurrence after local excision is common and has been reported in 10–35% of patients.3, 4, 20–23 Chemoprevention may provide an additional approach to the management of patients with leukoplakia who have premalignant oropharyngeal lesions and cannot be managed surgically or in patients who develop recurrent lesions after undergoing excision. It has been reported that topical bleomycin provides an additional approach to the management of patients with oral dysplastic lesions.24 Initial studies of topical retinoids also have been reported in case series.25–27 The objective of this article was to review the current understanding of the role of vitamin A and its derivatives in the management of patients with oral lesions who have an increased risk of malignant transformation, with a focus on topical approaches.


A Medline search was conducted using the following keywords: Vitamin A, retinoic acid, retinoids, oral leukoplakia and dysplasia. References identified within papers identified on Medline were also reviewed.


Vitamin A

Vitamin A is required in the normal pathway of epithelial cell differentiation. Growth and differentiation of normal and malignant cells in vitro have been modulated by retinoids by their effects on gene expression.28 The retinoids induce apoptosis, leading to normal maturation of dividing cells, and suppress carcinogenesis;29, 30 therefore they have been assessed in the control and suppression of carcinogenesis. A decrease in growth rate after exposure of epithelial cells to retinoids also has been documented.31

Free-radical reactions can cause changes in enzymatic functions and DNA mutations, increasing the risk of developing malignant cell lines. Reducing free radicals using antioxidants, such as vitamin A, may prevent such cellular changes.32 An association between vitamin A deficiency and the enhanced susceptibility to carcinogenesis was reported with an increased risk for developing different epithelial carcinomas of the lung, colon, pharynx, larynx, and esophagus.33

Systemic Vitamin A, Oral Carcinoma, and Oral Leukoplakia

Epidemiological studies have demonstrated that nutritional factors may represent a strategic approach to decreasing the development of squamous cell carcinoma,34 with the retinoids showing potential as promising molecules.35, 36 Retinoids are derivative compounds of natural vitamin A (retinol). More than 1500 such compounds have been synthesized, and several of these suppress cell mitosis. Retinoids have the ability to maintain adequate balance between growth, differentiation, and cell loss. This homeostatic balance is disturbed in malignant disease and has to be restored to obtain equilibrium between proliferation and the death of cells. Although it is believed that retinoids have the potential to control an imbalance, the exact mechanism is not known precisely.

13-cis retinoic acid may prevent the development of second primary squamous cell carcinoma.37 The expression of retinoic acid receptor β (RAR-β) mRNA, which is lost in premalignancy, can be restored by treatment with 13-cis retinoic acid,38 suggesting an encouraging response to the treatment.

It is well established that p53 alteration is associated with genetic instability and tumorigenesis. The level of p53 protein is also is associated directly with histologic grade and response to retinoids.39 Shin et al.40 reported a lower response rate to isotretinoin in patients with lesions that had high baseline p53 protein levels compared with lesions in patients with low p53 protein levels. The investigators assumed that the high level of p53 proteins in their patients may have been associated with mutant p53 proteins that usually do not undergo apoptosis and have long half-lives. It is speculated that the success of chemoprevention is associated with specific biomarkers, including the p53 alterations, nuclear antigens, and epidermal growth factor receptor (EGFr).41

The mechanism of action of the retinoids probably is linked to different nuclear RARs.42 The availability of RARs and retinoids is involved in the regulation of cell growth.43 After binding to the receptors, transcription factors (heterodimers or homodimers) are formed that bind to specific DNA sequences, leading to an up-regulation or down-regulation that ultimately affects gene transcription. It is believed that retinoids act by regulating gene expression through RAR nuclear receptors.44 A recent study by Beenken et al.45 showed that pretreatment expression levels of transforming growth factor α (TGF-α) and EGFr were increased significantly in dysplastic leukoplakias compared with normal appearing mucosa. The expression ratios of TGF-α in dysplastic leukoplakias deceased significantly after patients received systemic treatment using the 13-cis retinoic acid. EGFr serves as a biomarker for the prognostic value of retinoid treatment for leukoplakia. The EGFr expression ratio in patients with dysplastic oral leukoplakia, even though it was decreased, was not altered significantly by systemic retinoid treatment in one study.45 However, other studies showed a reduction of EGRr expression after exposure of cell lines to retinoic acid.46 The mechanisms underlying the effect of retinoids on TGF-α and EGFr expression are not understood fully, although, it has been suggested46 that it occurs primarily through decreased gene transcription and by normalizing the rates of TGF-α and EGFr mRNA in the cells.

The homeostasis of normal tissue represents a balance between cell proliferation and cell apoptosis. It has been demonstrated that retinoids alter gene expression and induce apoptosis, as mentioned above. Scher et al.30 investigated the effect of (4-hydroxyphenyl)retinamide) (4-HPR) on the growth of head and neck squamous cell carcinoma cell lines in vitro and found a significant antiproliferative effect. The effect of 4-HPR occurred in a dose and time dependent fashion. DNA fragmentation into multiple base pairs of oligonucleosomes, which is one of the features of apoptosis, was detected in some of the cell lines treated with 4-HPR.

Because 4-HPR is well tolerated and may be less toxic than other retinoic acids,50 and because it had greater growth inhibition and induction of apoptosis compared with 13-cis-retinoic acid,47, 48 it was speculated30 that 4-HPR controls proliferation by an additional mechanism of action. Epithelial malignancies were suppressed in vitro by the synthetic retinoid fenretinide (4-HPR).47, 48 Other studies reported that this retinoid was an effective chemotherapeutic agent for nonoral tumors49 and that it had a preventive and a well-tolerated effect on oral leukoplakia.50 It has been shown that 4-HPR/retinamide induces apoptosis and enhances the generation of reactive oxygen species.51–53

The expression of apoptotic bodies was examined in patients with oral leukoplakia who were treated topically with 13-cis-retinoic acid.54 A significant increase in apoptotic bodies was observed after the topical treatment, indicating that 13-cis retinoic acid has potential in the topical management of patients with oral mucosal leukoplakia. Further studies will be essential for the identification of specific retinoid-sensitive markers and nuclear receptors for a more successful outcome of chemoprevention therapy with retinoids in patients with premalignant oral conditions.

A decrease in the size of oral leukoplakia was reported in 67% of patients who were provided systemic 13-cis retinoic acid (1–2 mg/kg per day) for 3 months.55 The leukoplakia was reversed in > 80% of patients after 6 months without the medication. Side effects, mainly skin and mucosal dryness as well as hypertriglyceridemia, were reported in > 50% of patients, requiring isotretinoin dose reduction; however, the side effects reversed after the patients discontinued therapy. It was demonstrated that lower doses of 13-cis retinoic acid (0.5 mg/kg per day) administered over an extended period were less toxic compared with high doses56 but were associated with a lower recurrence rate after cessation of treatment. Another study57 reported clinical improvement of oral leukoplakia in 62% of patients. Reversal of dysplastic laryngeal lesions was reported using 13-cis-retinoic acid, α-tocopherol, and interferon-α in approximately 50% of patients at 6 months, compared with patients who had oral lesions, in whom the protocol lead to a response rate of < 10% at 6 months, and there were no responses at 1 year.58, 59 Recently, Leuing et al.60 used high doses of retinyl palmitate for the treatment of six patients with oral dysplastic leukoplakia and reported complete histologic resolution of the dysplastic lesions.

The effect of Spirulina fusiformis, which is a dietary source of vitamin A, was evaluated in India for chemoprevention.61 Although a significant regression of oral leukoplakias (45% of patients) was found in tobacco chewers, compared with only 7% of individuals in a placebo control group, approximately 50% of lesions recurred after patients discontinued the supplements. Another study performed in India62 reported the complete regression of oral precancerous lesions with systemic vitamin A in 52% of patients; in that study, two-thirds of the lesions recurred after the patients stopped supplementation. Stich et al.63 compared the outcome of 200,000 IU of vitamin A daily for 6 months on oral leukoplakia of 65 Indian tobacco/betel nut chewers versus placebo treatment. Fifty-seven percent of the patients who received vitamin A demonstrated a complete response and a complete preventive effect for the possible development of new lesions compared with a response rate of only 3% in the control group. Over 21% of individuals in the placebo control group developed new lesions. Silverman et al.64 reported partial or complete resolution of leukoplakia in 44% of their patients who were treated using systemic vitamin A. Discontinuation of the systemic therapy resulted in recurrence of the leukoplakia in 75% of the patients who experienced complete resolution. Only 30% of 10 patients with oral leukoplakia who received 50 mg per day of isotretinoin experienced some response;65 however, drug toxicity forced a dose reduction in 60% of their patients.

In conclusion, a complete resolution was achieved in 44–57% of patients who were treated with systemic retinoids. The recurrence rate of those patients ranged between 66% and 80%. A decrease in the size of the lesions was reported in 45–67% of patients.

Considerable dose dependant toxicity has been seen with use of systemic vitamin A and its derivatives (Table 1). In prophylactic doses, the toxicities include mainly skin rash, dryness and bleeding of the nasal mucosa, conjunctivitis, oral mucositis, cheilitis, hypertriglycerinemia, and teratogenic effects.55, 66 It was reported that the systemic use of vitamin A and its derivatives in patients with leukoplakia,64, 67 as well as in patients with other oral conditions,68 was associated with a high prevalence of side effects. Natural vitamin A at clinically tolerated doses has only limited activity against human neoplastic conditions.69 Therefore, clinical work has focused on the systemic derivatives with a higher therapeutic index.

Table 1. Toxicity Profile of Systemic Vitamin A and Retinoids
  1. HDL: high-density lipoprotein.

 Dry skin and dermatitis
 Thinning of hair
 Nail dystrophy
 Facial erythema and skin rash
 Peeling of palms and soles
 Chest pain and tachycardia
 Skeltal hyperostosis
Mucosal and internal
 Dry lips (chielitis)
 Burning lips/tongue
 Dry eyes
 Inflammatory bowel diseases
 Bleeding of gums
 Conjunctivitis and photophobia
 Abnormal menses
 Headache and malaise
 Decreased night vision
 Liver toxicity
 Decrease in serum HDL level
 Elevated sedimentation rate
 Elevated fasting sugar
 Weight loss
 Depression or psychosis
 Elevation of serum cholesterol
 Anemia and leukopenia
 Proteinuria and hematuria

In some patients, systemic treatment is much less tolerable than the disease itself and the lesion is suppressed, with recurrence if treatment is discontinued. It was suggested44 that high-dose induction using retinoic acid followed by low-dose therapy as maintenance may reduce toxicity, control the lesion, and provide be a viable approach to management.

Topical Vitamin A

The topical use of vitamin A compounds for oral white lesions is well known. Although vitamin A was used for controlling oral lesions of lichen planus,70, 71 to the best of our knowledge, only preliminary studies using topical vitamin A for oral leukoplakia have been reported (Table 2). Topical therapy has the potential advantages of high local dose with low systemic exposure, limiting the risk of systemic side effects. However, topical application requires patient willingness and ability to comply with repeated local application. In addition, there may be sites where topical application is difficult, and the medication may be diluted by saliva. Local tissue irritation may occur in the form of tissue sensitivity and burning with current topical applications due to the acidic nature of the topical products or the base in which the medication is placed.

Table 2. Impact of Topical Vitamin A on Oral Leukoplakia
StudyLocationPatientsMean age (yrs)Complete response (%)Partial response (%)aClinical recurrence (%)Follow-up (mos)
  • N/A: not available.

  • a

    At least 50% reduction in clinical size.

Shah et al.72USA11N/A27549011
Epstein and Gorsky25Canada26622754403.5 yrs
Piattelli et al.54Italy10611090N/A4
Tete et al.73Italy14N/A21.578.5N/A4

A 6-month dose ranging trial of isotretinoin lozenges was performed that included 16 patents with oral leukoplakia.72 When the effect of retinoids (in the form of 1–5 mg 13-cis retinoic acid oral lozenges) on oral leukoplakia was evaluated clinically,72 thinning of the leukoplakia and a reduction in the white surface was noted as the initial visible response. Three of 11 patients who completed the trial had a complete, visible clinical response (27%), and 6 of 11 patients (54%) had a partial response, which was defined as a decrease ≥ 50% in the size of the lesion. In 2 of 3 patients (67%) who achieved a complete response, leukoplakia returned within 5 weeks. In 1 patient (33%), leukoplakia has not recurred during 11 months of follow-up, although his final histopathologic examination showed hyperkeratosis. Toxicity, consisting of skin and mucosal dryness, epistaxis, and mucositis, was mild. Although the trial by Shah et al.72 is considered a topical approach to the management of patients with leukoplakia, this form of drug delivery represents a combination of topical and systemic dosing. Indeed, the systemic side effects may serve as a confirmation of such a systemic effect.

Although 80% of the patients who were treated by Shah et al., along with other patients with leukoplakia who were treated systemically, had documented remission that was credited directly to the use of retinoids, spontaneous remission does occur. Pindborg et al.4 reported that 37.4% of patients with leukoplakia will reverse spontaneously either totally or partially without any medical clinical intervention.

Epstein and Gorsky25 used 0.05% tretinoin gel that was applied topically 4 times per day for the management of nonmalignant oral white lesions in 26 patients (17 men and 9 women) with a mean age of 62 years. The vitamin A acid gel was applied locally for a mean of 3.5 years in patients who experienced clinical improvement. Although a complete clinical remission was reported in 27% of patients, a partial response was noted in 54% of patients, and clinical recurrence was experienced in about 50% of patients after the topical treatment was discontinued. Side effects of only a localized soreness were reported by only 19% of patients.

In a study by Piattelli et al.,54 topical 0.1% 13-cis retinoic acid (isotretinoin) gel was applied 3 times per day in 10 patients (6 males and 4 females; mean age, 61 years) with oral leukoplakia. The medication was used for 4 months, and a complete response was noted in 1 patient (10%), although 90% of patients experienced a reduction in size of at least 50%, representing a partial response. No side effects were reported. Tete et al.73 evaluated the effect of 0.1% isotretinoin gel applied topically 3 times per day on 14 patients with oral leukoplakia, and complete remission was noted in 3 patients (21.5%). A marked improvement (a reduction in size ≥ 50%) was achieved in 11 patients (78.5%) in 4 months of treatment.

The combined results of the two similar studies from Italy54, 73 indicate a complete response in 16.5% of patients who were treated with topical 0.1% isotretinoin and complete resolution in > 25% of patients who applied 0.05% tretinoin in the study reported by Epstein and Gorsky.25 Further studies will be necessary to confirm the complete and partial response rates with topical retinoid applications.

It is possible that the local application of retinoic acid results in a higher concentration of the substance directly in the target tissue, while avoiding systemic side effects. It is important to recognize that, even with clinical effect, cellular (phenotypic) and molecular effects may or may not occur, and the use of molecular markers may provide additional information for assessing the intermediate effects of intervention. Studies of systemic and topical delivery of vitamin A and retinoids show that, in patients who respond, lesions frequently will recur when the application is discontinued, suggesting that suppression of the condition may be achieved but that cure is not certain, even with clinical improvement or, indeed, clinical resolution. However, in patients with recurring and persisting lesions for whom other treatments have failed or are not provided, prevention of disease progression with a therapy of limited toxicity, such as topical retinoids, may control some lesions that otherwise may progress.


Although vitamin A and derivative retinoids have been successful in temporarily decreasing the clinical presence of leukoplakia, the toxicity of the medications with systemic administration remains a significant problem and has an important impact on common and long-term usage. The long-term effectiveness of chemoprevention has not been established to date, and the choice of agents requires further study. Advances in single or combined chemopreventive medications for the topical or systemic control of leukoplakia may decrease the potential of malignant transformation. Similar outcomes appear to have been achieved with systemic and topical retinoids, with resolution in 27–57%, of patients, a partial response in 45–90% of patients, and recurrences in approximately 50% of patients if the retinoid is discontinued. Prospective controlled studies are needed to establish the value of chemoprevention in patients with oral leukoplakia and to assess the potential role of topically applied retinoids. The preliminary studies to date provide support for further study, and double-blind, randomized, controlled clinical trials with topical retinoids are needed. Further studies should assess the mucosal lesions according to accepted criteria,2, 74, 75 including the size and nature of lesions, and they should include analysis of the histology and possibly of molecular components.