Abstract
There has been a continuous debate regarding the possible malignant potential of oral lichen planus (OLP). Based on the results from follow-up studies, OLP is regarded by several authors as a pre-malignant condition, and patients with OLP have been recommended to have their lesions monitored two to four times annually. This recommendation needs reconsideration, because a recall system of all patients with OLP requires substantial economic resources. In a reality where such resources are limited, a recall system must be weighed against other benefits and the fact that the malignant potential of OLP is most likely very low. The present review focuses on the diagnostic criteria for OLP, the pre-malignant potential of OLP, and the extent to which the available information can be used to reduce morbidity and mortality of oral cancer related to OLP.
Introduction
Several prospective (Holmstrup et al., 1988; Silverman et al., 1991) and retrospective studies (Warin, 1960; Altman and Perry, 1961; Janner et al., 1967; Abramova, 1968; Andreasen, 1968; Cawson, 1968; Shklar, 1972; Fulling, 1973; Kovesi and Bánóczy, 1973; Holmstrup and Pindborg, 1979; Vas’kovskaia and Abramova, 1981; Silverman et al., 1985; Murti et al., 1986; Salem, 1989; Sigurgeirsson and Lindelof, 1991; Voute et al., 1992; Barnard et al., 1993; Brown et al., 1993; Moncarz et al., 1993; Duffey et al., 1996; Markopoulos et al., 1997; Silverman and Bahl, 1997; Lo Muzio et al., 1998; Rajentheran et al., 1999) as well as case reports (Faraci et al., 1975; Marder and Deesen, 1982; Pogrel and Weldon, 1983; Fowler et al., 1987; Katz et al., 1990; Massa et al., 1990; Harland et al., 1992; Porter et al., 1997; Camisa et al., 1998; Chainani-Wu et al., 2001) have reported oral lichen planus (OLP) as having an increased potential for malignant development. The reported transformation rates vary from 0 to 9%. Several studies have been conducted recently, and there is a need for an updated review to re-assess available information to arrive at proper guidelines for the management of OLP.
The issue of a malignant potential of OLP has been a matter of serious controversy (Krutchkoff et al., 1978; Krutchkoff and Eisenberg, 1985, 1986; Holmstrup, 1992; Eisenberg, 2000). Krutchkoff and Eisenberg have stated that some of the reported OLP cases developing oral cancer were in fact not OLP, but rather dysplastic lesions with lichenoid features. The implication of this premise is that patients with lichenoid dysplasia represent a risk group which can be identified by the appropriate use of available diagnostic methods and, as such, can be distinguished from patients with OLP with no dysplasia-related increased risk of development of oral cancer.
Since OLP is considered a pre-malignant condition by several authors, a recall system for OLP patients has been recommended to facilitate the early diagnosis of oral cancer with the aim of reducing morbidity and mortality from oral cancer arising in OLP patients. One suggestion is that OLP patients should have regular follow-up examinations from two to four times annually (Scully et al., 1998). Since OLP is one of the most prevalent oral mucosal lesions, a recall program demands substantial economic resources, and hence, a reconsideration of the background of the need for a recall is warranted.
An established recall system must be based on two important basic assumptions: It is possible to identify patients with an increased risk of cancer development; and it is possible to reduce cancer morbidity and/or mortality as a result of the recall program.
These assumptions will be addressed below.
(i) Identification of Patients with Increased Risk of Cancer Development
Several studies have attempted to establish the risk of cancer development in OLP patients. Based on their design, the available studies may be divided into case reports, retrospective, and prospective follow-up studies. Case reports are of limited value and are not further considered in the present review. Retrospective studies often entail incomplete data, since the patient materials are selected after termination of the observation period, and the information available is not defined prior to the examinations and follow-ups of the patients. In an effort to base a discussion on studies with the most uniform and accurate information, the present review is restricted to the 2 prospective and 9 retrospective studies in which the diagnostic criteria have been defined (Table).
The diagnostic dilemma of OLP
The ongoing debate includes a question on what is “true OLP” (Eisenberg, 2000; Silverman, 2000). We are not sure that any type of disease entity can be “true”, because a disease entity is not a living organism acting “truly” or “falsely”. Disease processes are always modified by factors dependent on the single individual affected. If, on the other hand, diagnostic criteria are regarded as important tools for handling groups of patients, it is important to focus on whether patients included in studies of the pre-malignant nature of OLP fulfilled explicitly defined diagnostic criteria. They did so in the 11 studies mentioned in the Table. These studies involved OLP patients diagnosed on the basis of both clinical and histopathological criteria.
It is widely approved that the most characteristic clinical feature of OLP is the presence of reticular white striations and/or white papules, essentially distinguishing the disease from other diseases of the oral mucosa. In addition to these features, plaque-type, atrophic, ulcerative, and bullous lesions may be present at the time of diagnosis, or they may occur during the course of the disease (Thorn et al., 1988). None of these latter types of features is limited to OLP, and consequently they are not well-suited to distinguish OLP from other diseases. The presence of reticular white striations and/or papules was an inclusion criterion in all 11 studies in the Table.
The histopathologic features of OLP include several epithelial changes, the amount and extension of which vary. Epithelial hyperkeratosis, atrophy or hyperplasia, acantosis, saw-toothed rete ridges, liquefaction degeneration of basal cells, and single-cell keratinization are all prominent characteristics. A homogeneous cell-free zone is frequently present in the basement membrane zone. The subepithelial connective tissue shows a band-like inflammatory infiltrate dominated by lymphocytes and macrophages (Andreasen, 1968).
Biopsy is obviously an important tool in the diagnostic process for OLP (Barnard et al., 1993), but a diagnosis of OLP based solely on histopathology in some cases leads to an erroneous result. Misinterpretation by the pathologist may also be a source of error. Eisenberg and Krutchkoff (1992) have stated that some reports have shown microphotographs suggestive of diagnoses other than OLP. Whether or not this comment is justified is a matter of debate, but it emphasizes the well-known problem of subjectivity resulting in intra- and interindividual variation among pathologists (van der Meij et al., 1999a). Some diagnoses, such as, for example, lichenoid contact reactions (LCR) caused by components released from amalgam fillings, are impossible to distinguish from OLP histopathologically (Bolewska and Reibel, 1989). Since these lesions are not known to be associated with malignant transformation, their inclusion may result in a decrease in the rate of malignant potential of OLP. However, such lesions can be distinguished on the basis of clinical criteria (Bolewska et al., 1990; Bratel et al., 1996).
In the reported studies, were patients with dysplastic lesions in fact included in the risk analysis, thereby overstating the risk of malignant transformation of OLP? For obvious reasons, this question cannot be fully answered, but most of the studies have been conducted by investigators with acknowledged expertise in oral medicine and oral pathology. One of the prospective studies has specifically stated that if biopsies showed epithelial dysplastic features, the patients were excluded from the follow-up study (Holmstrup et al., 1988).
It is most important that measures be developed for the identification of the subgroup of OLP patients susceptible to the future development of oral cancer. The specific clinical features do not obviously shed light on the potential for cancer development, since the different types of OLP lesions developing oral cancer, as included in the Table, showed the following distribution: reticular 33%, plaque 29%, atrophic 13%, and ulcerative/erosive 25%.
In the past, scientists have investigated several histochemical markers to determine their relevance as prognostic tools in the evaluation of OLP, epithelial dysplasia, and oral cancer. These include altered expression of alpha9 integrin (Häkkinen et al., 1999) and laminin-5 staining (Kainulainen et al., 1997). It has also been proposed that p53-positive patients with otherwise benign lesions should be followed more closely (Crosthwaite et al., 1996), and similar results have been reported by Girod et al.(1994, 1995) implying that expression of mutated p53 is an indicator of potential malignant development in benign lesions of the oral mucosa (Girod et al., 1993). Involucrin immunoreactivity has also been proposed as a method to distinguish between lichen planus and lichenoid dysplasia (Eisenberg et al., 1987). Using microsatellite analysis to elucidate the pre-malignant potential, Zhang et al.(1997, 2000) reported that lichenoid lesions with mild dysplasia showed a higher loss of heterozygosity (54%) as compared with OLP (6%). Although the findings may reinforce the current opinion of lichenoid dysplasia as having a higher tendency of malignant development than OLP, this marker has not yet been evaluated as a prognostic marker for the development of oral cancer, and the findings do not exclude a low but significant risk for OLP patients to develop malignancy.
A recent study (Sudbo et al., 2001) may be an important advance in the molecular assessment of the risk of oral cancer development in patients with leukoplakia. A blinded investigation of DNA content in excisional-biopsy specimens of dysplastic oral leukoplakias from 150 patients who were followed for a mean of 8.6 yrs showed that the DNA content was a powerful predictor of the risk of malignant transformation of the lesion. Surprisingly, the degree of dysplasia did not correlate with DNA content or the risk of cancer. Studies are warranted to evaluate this method as a predictor for malignant transformation of OLP lesions.
The use of test methods such as vital staining with toluidine blue and brush biopsy (cytology) may enhance the possibility for the identification of signs of dysplasia or carcinoma in existing lesions (Epstein and Scully, 1997; Handlers, 2001). Therefore, in certain situations, they may serve as alternatives to biopsy in recall systems, although the value of toluidine blue staining was questioned in the recent study by Mignogna et al.(2001). It remains to be shown if these methods can be used as prognostic tools to identify OLP patients susceptible for future squamous cell carcinoma (SCC) development.
Thus, several histological markers have been examined or proposed as prognostic factors, but, to our knowledge, there is currently no specific marker that has been widely agreed upon for determination of a possible future malignant transformation.
It has been advocated that OLP patients with a risk of future malignant development can be identified with current diagnostic methods as long as the methods are used properly (Eisenberg, 2000). The comment stems from pathologists’ neglect in identifying epithelial dysplasia and clinicians’ reflexively rendering a presumptive diagnosis of OLP on the basis of relatively weak and superficial clinical criteria. Improved diagnostic skills must, however, entail sound scientific studies based on well-defined criteria which have been evaluated for their prognostic value in identifying OLP patients who might be at risk for the development of oral cancer.
In the process of evaluating diagnostic criteria, it is essential that cases with a higher potential for malignant transformation than OLP be excluded. Although existing prospective follow-up studies vary somewhat in their diagnostic criteria, their results are not very different, and they do provide a body of information for the evaluation of the pre-malignant potential of OLP.
In summary, conclusive investigations on the pre-malignant potential of OLP have been carried out as prospective or retrospective studies of patient samples characterized on the basis of explicit clinical and histopathological criteria, i.e., reticular and/or papular lesions, and demonstrating histopathologic features including hyperkeratosis, liquefaction degeneration of basal cells, and a subepithelial band-shaped inflammatory reaction dominated by lymphocytes and macrophages. Lesions with dysplastic features should be excluded on the basis of their histopathology, and this has been explicitly mentioned in one study (Holmstrup et al., 1988).
The reported malignant potential of OLP
Several studies have documented a relationship between OLP and oral cancer (Table). In addition to the diagnostic criteria, several aspects are important for the interpretation of the results obtained.
Most studies have shown that patients with OLP develop oral cancer at an increased rate as compared with the general population. There are only two studies with a prospective design and recall on a regular basis. In one of these (Holmstrup et al., 1988), all cases (n = 611) were histopathologically investigated prior to the manifestation of oral cancer. During follow-up, nine patients (1.5%: eight females, 1.9%; and one male, 0.5%) developed oral cancer. At the time of initial diagnosis, the patients had lesions with no clinical evidence of atypical features, but seven of the patients who developed oral cancer had plaque or atrophic areas as part of the initial lesions. Interestingly, the plaque form seemed to be overrepresented among patients developing oral cancer, a finding consistent with findings of other studies (Silverman et al., 1991; Mignogna et al., 2001). As mentioned above, however, the studies presented in the Table do not provide data to conclude that any clinical type of OLP is particularly prone to the future development of cancer.
Despite differences in experimental designs, it is striking that the majority of studies have reported a rate of malignant transformation of OLP between 0.5 and 2% (Table) over a five-year period. Based on age- and gender-matched general population data from the Cancer Registry of Denmark, it was estimated that patients with OLP had a 50 times higher risk of developing oral cancer than did the general population (Holmstrup et al., 1988). For women, the risk was 70 times higher; for men, 14 times higher. Others (Barnard et al., 1993) have also regarded OLP as a disorder with a potential for malignant development, with the likelihood of cancer development in lichen-planus-affected oral mucosa being 10 to 20 times greater than in the general population (MacDonald, 1975). In a recent paper, a thorough evidence-based approach was used to calculate the risk for OLP patients to develop oral cancer (Drangsholt et al., 2001). The increased risk was estimated to be roughly 10 times higher than that seen in the general population.
All studies mentioned in the Table are based on cohorts of referred patients, and, most likely, patients with symptomatic OLP are overrepresented in such cohorts. In this context, it is interesting to note that Murti et al.(1986), in a study of the general population, found an incidence of oral cancer among OLP patients that was not very different from that reported in studies of referrals.
The prevalence of OLP in the Swedish population has been estimated to be 1.89% (Axéll, 1976). In this study, which is often referred to in the on-going discussion, the diagnosis of OLP was mostly based solely on clinical appearance. Since the follow-up studies of OLP mentioned in the Table include only patients with a diagnosis of OLP based on both clinical and histopathological findings, these studies are incompatible with the Swedish study. Also, in the Swedish study, there was no distinction between OLP and LCR, because LCR was not recognized as a disease entity at that time. The implication of this is that the prevalence of OLP may include LCR. If so, the prevalence of OLP in the general population may in fact be lower than reported.
It has been concluded that if 1-2% of all OLP underwent a malignant transformation, then OLP would be the major source of oral cancer in many parts of the world (van der Meij et al., 1999b). Van der Meij et al. found this extremely unlikely, and others have expressed similar positions (Allen, 1998). It is reasonable to believe that the prevalence of OLP varies in different parts of the world (Scully et al., 1998). Consequently, it is hazardous to predict the true number of oral cancer cases associated with OLP, when the prevalence figures of OLP in most populations are merely presumed or unknown. Studies in which the transformation rate has been established may, for several reasons, not be related to these prevalence figures, because most studies are carried out in cohorts of referred OLP patients.
In a study of patients with lichen planus of the skin, an increased rate of oral but not skin cancer was discovered (Sigurgeirsson and Lindelof, 1991). In this connection, it is important to note that studies have shown that OLP in patients with cutaneous lichen planus (CLP) is not different clinically or histopathologically from OLP in patients with OLP alone (Andreasen, 1968).
To conclude, the increased incidence of oral cancer in patients with OLP merits this mucosal disease to be regarded as a pre-malignant condition. This is of primary importance when it comes to developing treatment strategies for patients suffering from OLP.
(ii) Decreased Cancer Morbidity and/or Mortality as the Result of a Recall Program
For the minimization of morbidity and mortality as a consequence of malignant transformation of OLP, an early identification of oral cancer and its precursors has been considered as important (Silverman, 1992, 1993, 1994; Fossion et al., 1994; Scully and Ward-Booth, 1995; Kabani and Dhingra, 1997; Scala et al., 1997; Shugars and Patton, 1997). In that regard, recall systems for OLP have been proposed to include professional examination 2-4 times annually (Moncarz et al., 1993; Duffey et al., 1996) or once a year (Voute et al., 1992; Barnard et al., 1993; Lo Muzio et al., 1998; Scully et al., 1998).
In a recent Italian study (Mignogna et al., 2001), a meticulous examination of OLP patients at least three times a year was advocated to improve the prognosis of patients developing oral cancer. The study is based on 502 patients with OLP where 24 cases of early oral cancer were identified during the follow-up period. In a previous study by the same group (Lo Muzio et al., 1998), one examination annually of 263 patients with OLP was conducted. With this regime, 14 cases with a more advanced form of oral cancer compared with those reported in the 2001 study were found. Surprisingly, however, the number of recurrences and the mortalities did not differ significantly when the two studies were compared. Consequently, these studies indicate that frequent follow-ups do not automatically lead to an improved prognosis for OLP patients with oral cancer. A problem similar to that mentioned above has been addressed for lung cancer, which remains a prevalent and deadly disease where only surgical resection of early-stage disease improves the prognosis for the patients (Ellis and Gleeson, 2001). The critical question is whether or not lung cancer screening produces a measurable reduction in the mortality of patients being screened. Large randomized studies have been conducted to answer this question. The Mayo Lung Project (Fontana et al., 1984, 1986) was a randomized, controlled clinical trial of lung cancer screening conducted in 9211 male smokers between 1971 and 1983. Those in the intervention arm were offered chest x-ray and sputum cytology every 4 months for 6 years; those in the usual-care arm were advised, at trial entry, to receive the same tests annually. Unfortunately, no lung cancer mortality benefit was evident at the end of the study. Needless to say, follow-ups of OLP patients have little in common with screening of lung cancer except for the fact that early diagnosis of oral as well as lung cancer undoubtedly improves the prognosis of patients developing these forms of cancer. Obviously, the measures taken have not been proven sufficient to reduce the morbidity or mortality rate of patients with lung cancer or of oral cancer related to OLP. Thus, there is definitely a demand for larger randomized studies.
What can be done in the interim? First, it can be stated categorically that OLP entails a small but significant risk for oral cancer development. However, this fact should not be used as a justification for several examinations annually, since large resources would be spent on activities with no established benefit. Under these circumstances, it seems reasonable to exploit existing resources where patients encounter health care professionals. As an example, in Scandinavia, more than 90% (Hugoson et al., 1995) of the adult population visit a dentist on a regular recall basis, which provides an obvious opportunity for regular screening of the oral mucosa. Several authors have also advocated that other dental care workers should play a role in detecting oral cancers (Leong et al., 1995; Alvi, 1996; Scala et al., 1997; Kerr, 2000), and it is therefore important that every dental care provider be properly educated for this task (Horowitz et al., 1996, 2000; Downer et al., 1998) to prevent or reduce the delay of a cancer diagnosis (Khoo et al., 1998).
In countries with few dentists or other oral health care providers, it is most likely difficult for any recall routines for OLP to be implemented (Warnakulasuriya and Nanayakkara, 1991). Available economic resources should be used to develop models to educate health care professionals and to establish nation-wide promotional campaigns to raise public awareness of precancer and oral cancer.
In summary, there are sufficient criteria to define what we presently regard as OLP with reasonable certainty. Patients fulfilling the commonly used criteria of OLP appear to have an increased risk of developing oral cancer. The observed incidence of oral cancer development is low, and there are no data at this point demonstrating that intensive follow-up of OLP patients results in reduced morbidity and mortality of oral cancer related to OLP. Consequently, on a practical and economic basis, a continuous recall of all OLP patients in specialist clinics cannot by justified. It is important, however, that every dentist and dental auxiliary be educated to detect early signs of oral cancer to ensure that these lesions are identified when the patients are seen for other purposes, such as routine examination and treatment. The patient should also be instructed to report clinical changes in the condition.
