Recipients of liver transplantation (LT) may be predisposed to various oral and dental diseases because of permanent immunosuppression. Drug-induced gingival overgrowth (DIGO) is strongly associated with cyclosporine A (CSA), and this risk is further increased if an organ transplant patient is using nifedipine or other calcium channel blockers (CACHs) and has poor oral hygiene and untreated periodontitis.[1, 2] Other oral manifestations in LT recipients include fissured tongue, candidiasis, an increased incidence of viral infections by herpes simplex types 1 and 2 or cytomegalovirus, graft-versus-host disease, and oral cancer.[3-8] In addition to immunosuppressants, LT recipients often take several other medications that may contribute to hyposalivation and xerostomia, which can lead to an increased risk for oral infections and also cause subjective dry mouth–related symptoms.[3, 9]
The dental treatment of LT recipients is aimed at preventing infections.[6, 10] After transplantation, a careful examination of the lips and oral mucosal tissues for possible neoplastic lesions is important. Studies on oral side effects in LT recipients are lacking. We, therefore, studied oral mucosal lesions (OMLs) in groups of LT recipients with different etiologies and immunosuppressive medications, and we compared these groups with a control population. We also investigated risk factors associated with OMLs to better understand the special requirements for the dental management of LT recipients.
PATIENTS AND METHODS
LT in Finland is centralized at Helsinki University Central Hospital. This observational, cross-sectional study involved 84 adult LT recipients (>18 years old; 51 men and 33 women; median age = 55.1 years, range = 24.6-70.9 years). The same cohort of patients was included in our previous study. The main indication for LT was chronic liver disease: primary sclerosing cholangitis (27%), alcohol cirrhosis (14%), primary biliary cirrhosis (7%), cryptogenic cirrhosis (7%), other known cirrhosis (11%; including autoimmune hepatitis and viral hepatitis), other liver diseases (5%; including Budd-Chiari syndrome), and liver tumors (5%). Patients with acute and subacute liver disease (24%) were generally healthy before they experienced sudden liver insufficiency. Two patients underwent combination transplantation (liver and kidney), and 3 patients underwent retransplantation.
An oral examination is a prerequisite for transplantation in our country, and all of the chronic LT recipients and some of the acute LT recipients had undergone pretransplant clinical oral examinations with necessary dental treatment before they were accepted onto an LT waiting list. The patients were recruited for a new clinical oral examination after transplantation in connection with a protocol medical follow-up examination at the transplant center. Both pretransplant and posttransplant oral health data were available for a subgroup of 71 patients, but oral health before transplantation was investigated somewhat differently.
The posttransplant oral examinations were performed between September 2008 and October 2011. The inclusion criteria required LT to have been originally performed between 2000 and 2006 with a minimum follow-up time of 2 years (n = 252). In that study population, 223 LT recipients were alive in 2008 when the recruitment started. All consecutive patients would have been included in this study except for logistic limitations such as difficulties in scheduling a dental examination for the day of the follow-up visit (especially for an LT recipient living at a distance). Furthermore, the fact that the protocol follow-up visit was every 2 to 3 years for those LT recipients who had undergone LT more than 3 years previously (n = 164) made it impossible to include all potential LT recipients. More than half of the eligible LT recipients were included, however. The patients were categorized as either acute LT recipients (n = 20) or chronic LT recipients (n = 64) in accordance with the etiology of their liver disease. The groups also reflected the different types of immunosuppression therapy.
LT recipient data for systemic diseases (according to the World Health Organization International Classification of Diseases) and prescribed daily medications (according to the World Health Organization Anatomical Therapeutic Chemical Classification System) came from patient records at the time of the protocol follow-up at the transplant center. The type of underlying liver disease in LT recipients was also recorded from medical files.
Information regarding dysphagia, smoking, and alcohol use came from either a questionnaire or an interview for LT recipients and controls. Current smoking and current alcohol use were assessed for any individuals responding that they were currently smoking or using any alcohol. Data on smoking and alcohol use before transplantation among LT recipients were not available.
Control data (n = 252) originated from the Health 2000 survey in Finland: from 2000 to 2001, 6335 individuals participated in a clinical oral health examination, and 6986 participated in a home interview. This large, nationwide survey provided information on the most important public health problems in the country and their causes and treatment as well as the population's functional and working capacity; Heistaro reported its methodology in detail. Three control subjects were matched to each LT recipient with respect to age, sex, and area of residence.
Liver disease was an exclusion criterion for the control population, but individuals with other systemic diseases such as cardiovascular disease and diabetes were not excluded. Data for systemic diseases (diagnosed by an M.D.) and the total number of prescribed daily medications were assessed with home interviews. The assessment for diabetes was complemented by blood sugar levels and also by data from the drug reimbursement registry.
Clinical Oral Examinations in the Study and Control Groups
The clinical oral examinations of the patients were performed by 2 hospital dental specialists (H.Ru. and L.G.). The oral examination also included/consisted information about dentures and how well they fit. Each oral examination included a careful clinical visual inspection of the mucosal and gingival tissues of the oral cavity, including the lips, hard and soft palates, floor of the mouth, inner cheeks, tongue, labial sulcus, and gingiva, for any mucosal or gingival pathology. The presence of different OMLs was reported according to a structured protocol.
Clinical oral examinations were performed similarly for the control group. The quality assurance for the clinical measurements in the Health 2000 survey included parallel measurements by the field dentists and the reference dentist. The percentage of unanimous diagnoses regarding oral mucous membrane findings was 87, and the κ value was 0.47.
OMLs included DIGO; mucosal hyperplasia (caused by dentures); denture stomatitis (clinical entity: redness of the oral mucosa under a full denture); oral lichen planus (OLP)–like lesions; lichenoid reactions; leukoplakias (which refer to predominantly white lesions of the oral mucosa that cannot be scraped off or characterized as any other definable lesion); erythroplakias; ulcers; and other OMLs such as angular cheilitis, echymoses, petechiae, fistulas, tumorlike structures (eg, fibromas, hemangiomas, and mucoceles), and other inflammatory OMLs. No amalgam tattoos, frictional keratosis, or materia alba was recorded. For LT recipients, tongue lesions were reported separately (atrophic tongue, fissured tongue, hairy tongue, and geographic tongue).
Salivary and Mycological Sampling in the Study Group
Measurements of saliva flow and oral fungal cultivation were performed only for LT recipients. Xerostomia refers to a subjective sensation of dry mouth, whereas hyposalivation is objectively measured via salivary flow (SF).
Unstimulated and paraffin-stimulated whole saliva was collected from LT recipients for 5 minutes before the oral examination. For unstimulated saliva, each patient was instructed to passively drool saliva into a collection tube while he or she sat quietly in a dental chair with his or her head slightly bent. For stimulated saliva, a piece of paraffin wax was first chewed for 30 seconds, and then the saliva in the mouth was swallowed; after that, the patient drooled saliva into a collection tube while he or she simultaneously chewed on paraffin approximately once per second. This method of saliva collection was documented by Navazesh and Kumar. Hyposalivation was diagnosed when unstimulated and stimulated SF rates were significantly reduced. Reduced unstimulated and stimulated SF rates here refer to ≤0.1 and ≤0.7 mL/minute, respectively.
The presence of Candida was evaluated via direct mucosal swabs from the oral mucosa of LT recipients, and these swabs were taken routinely from all LT recipients according to the study protocol. Samples from suspected candidiasis lesions in the oral cavity (and from denture surfaces when this was deemed necessary) were transferred with sterile cotton-tipped swabs to the Transpocult transport medium. These samples were then immediately transported to the hospital laboratory for analysis.
This study was approved by the ethics committee of Helsinki University Central Hospital (192/13/03/02/2008, 16.8.2008), is in agreement with the Declaration of Helsinki, and is registered in the Helsinki University Central Hospital database for clinical trials. Permission for the Health 2000 survey was given by the ethics committees of the University Hospital Region of Helsinki and Uusimaa and the National Public Health Institute. The LT recipients and also the Health 2000 participants were informed about the studies, and each signed an informed consent form.
Statistical associations were studied with the chi-square test and Fisher's exact test (2-tailed) for categorical variables and with the Mann-Whitney U test for continuous variables. A multivariate binary logistic regression analysis (backward Wald) allowed the assessment of possible risk factors for the occurrence of any OML. The following independent variables were entered into the model: case/control, age at oral examination, dysphagia, presence of removable dentures, number of teeth, current alcohol use, current smoking, total number of daily medications, and cardiovascular medications (including beta-blockers, diuretics, angiotensin-converting enzyme inhibitors, and CACHs). These variables were chosen because findings in other studies indicated that they might predispose patients to OMLs. All confidence intervals (CIs) were calculated at the 95% level, and P values ≤ 0.05 were considered statistically significant. The statistical software was PASW 18.0 (IBM Co., Armonk, NY).
The basic characteristics of the LT recipients and the controls are presented in Table 1, and those results overlap with the results of our earlier study. Twenty-five percent of all LT recipients had diabetes after LT. Most already had diabetes before LT; only 3 developed posttransplant diabetes. All OMLs were significantly more frequent in LT recipients versus controls (42.9% versus 14.7%, P < 0.001); comparisons between chronic patients, acute patients, and controls are provided in Table 2. DIGO occurred significantly more often in chronic LT patients versus controls (16.1% versus 1.0%, P < 0.001) and in acute LT patients versus controls (20.0% versus 0.0%, P = 0.003). OMLs with malignant potential (OLP-like lesions, lichenoid reactions, leukoplakias, erythroplakias, and ulcers) were more common in chronic LT recipients versus controls, but the difference was not statistically significant (12.5% versus 5.7%, P = 0.10). Reduced unstimulated saliva (≤0.1 mL/minute) was diagnosed in 31.3% of the chronic LT recipients and in 10.0% of the acute LT recipients (not significant). The findings for Candida cultivation were positive for 56.5% of the chronic LT recipients and for 55.0% of the acute LT recipients.
Table 1. Basic Characteristics of the 2 Study Groups and Controls
Chronic LT Recipients (n = 64)
Controls (n = 192)
Acute LT Recipients (n = 20)
Controls (n = 60)
The data are presented as medians and ranges.
The data are presented as means and standard deviations.
Including beta-blockers, diuretics, angiotensin-converting enzyme inhibitors, and CACHs.
In the subgroup of 71 patients, oral mucosal pathologies were more common during the pretransplant stage versus the posttransplant stage (52% versus 45%, not significant). The oral hygiene status was good for 62% of all LT recipients. DIGO was more than twice as common in patients with poor oral hygiene versus those with good oral hygiene (28% versus 12%, P = 0.08).
Table 3 compares OMLs in the different immunosuppressant groups. Most OMLs were more common in the CSA group versus the tacrolimus (TAC) group, and it is particularly significant that DIGO was found more frequently in the CSA group versus the TAC group (29.3% versus 5.4%, P = 0.007). The prevalence was even higher when a CACH was used concomitantly with CSA versus TAC (46.7% versus 8.3%, P = 0.04). A calcineurin inhibitor (CNI) with a steroid raised the prevalence of any OML to 52.6%; furthermore, Candida was found more with steroid use, but this was not significant in comparison with a CNI alone (68.4% versus 49.2%).
Table 3. Prevalence of Oral Manifestations in the Different Immunosuppressant Groups
Oral Mucosal Status
CNI Without Steroid (n = 61)
CNI and Steroid (n = 19)
CSA (n = 43)
TAC (n = 37)
CSA and CACH (n = 15)
TAC and CACH (n = 12)
mTOR Inhibitor (n = 4)
NOTE: Percentages have been calculated for individual groups.
A single patient could have had several OMLs.
P = 0.007 for CSA versus TAC (chi-square test, exact significance, 2-tailed).
P = 0.04 for CSA and CACH versus TAC and CACH (chi-square test, exact significance, 2-tailed).
Hairy tongue, atrophic tongue, fissured tongue, and geographic tongue.
OLP-like lesions, lichenoid reactions, leukoplakias, erythroplakias, and ulcers (excluding frictional keratosis and materia alba).
Angular cheilitis, ecchymoses (bruises), petechiae, fistulas, tumorlike structures (eg, fibromas, hemangiomas, and mucoceles), and other inflammatory OMLs.
The risk factors for OMLs are listed in Table 4. The risk was highest for LT recipients with an odds ratio (OR) of 9.2 (95% CI = 4.5-19.0, P < 0.001). Significant risk factors also included current alcohol use (OR = 2.4, 95% CI = 1.2-5.0, P = 0.02), current smoking (OR = 2.0, 95% CI = 1.0-3.8, P = 0.04), and the number of teeth (OR = 0.9, 95% CI = 0.9-1.0, P < 0.001).
Table 4. Risk Factors for the Occurrence of Any OMLs in LT Recipients (n = 84) and Controls (n = 252)
NOTE: The following independent variables were entered into a multivariate binary logistic regression analysis: case/control, age, dysphagia, presence of removable dentures, number of teeth, current alcohol use, current smoking, total number of daily medications, and cardiovascular medications (including beta-blockers, diuretics, angiotensin-converting enzyme inhibitors, and CACHs).
Number of teeth
Alcohol use (yes/no)
Clinical characteristics associated with 5 cases of non-albicansCandida strains are detailed in Table 5. SF was numerically but not statistically less frequent among those patients with non-albicans Candida versus those patients with Candida albicans or no Candida.
Table 5. Oral Mucosal Health for Different Candida Groups and Clinical Characteristics of LT Recipients With Non-albicansCandida
To the best of our knowledge, this is the only study focusing solely on the oral health of LT recipients, and the study group is the largest among protocol studies with the inclusion of a control group. Among LT recipients, all OMLs were generally more frequent in comparison with controls, and this difference was highly significant between the chronic LT recipients and the control population. The difference between the acute LT recipients and the controls remained nonsignificant, possibly because of the small sample size. Nevertheless, our findings agree with those of Díaz-Ortiz et al., who reported a high frequency of OMLs, except that our study population showed only a low prevalence of fissured tongue. This difference could stem from different etiologies of liver disease; viral hepatitis is not a common indication for LT in Finland. A recent study of 40 solid organ transplant recipients found OMLs to be rare, but that study by Rojas et al. included only 17 LT recipients with a shorter follow-up time; such methodological differences could explain the different outcomes of OMLs.
As suggested, DIGO (Fig. 1B) was the most common type of OML in LT recipients, and there was a significant difference between both the chronic and acute LT groups and the controls. Although commonly associated with CSA and CACHs, this side effect may also be present in patients receiving mammalian target of rapamycin (mTOR) inhibitors; however, our few patients with mTOR inhibitors were free of DIGO. Poor oral hygiene seemed to correlate with DIGO here; evidently, proper oral self-care should always be reinforced. Still, recurrent cases may require a change from CSA to TAC[18, 19] or the replacement of a CACH with another antihypertensive drug. If DIGO requires annual surgery, this is not only annoying to the patient but is also a burden to the operating dentist.
One study has performed oral yeast cultivations according to protocol. Immunosuppressive medications predispose patients to both superficial and systemic fungal infections, which may have serious consequences. Systemic antifungal prophylaxis has been beneficial in preventing recurring C. albicans infections in these patients, but the non-albicans strains are an increasing concern because of their resistance to common antifungals. Among the non-albicans strains, Candida glabrata is associated with the highest mortality rate because of its hemolytic activity and tissue-invasive capacity; it thus brings an increased risk for bloodstream infections.[23, 24] In Finland, the non-albicans strains seem to be fairly common. More than half of our LT study population was positive for oral Candida, and the combination of CNIs and steroids seemed to promote increased occurrence. Although C. albicans was the most common strain, there seemed to be a shift toward non-albicansCandida after transplantation. One patient with positive C. glabrata findings presented with a hairy, black tongue that was associated with heavy smoking and moderate use of alcohol (Fig. 1A).
Instead of systemic antifungals, a topical antifungal drug is generally used in dentistry to treat oral candidosis (Fig. 1D). In our clinic, this has proven to be an effective treatment with fewer drug interactions and adverse effects. The treatment regimen includes 1 mL of nystatin (Mycostatin) administered 4 times daily for 4 to 6 weeks or as pulse therapy in highly susceptible patients with recurring oral candidosis. Edentulism, ill-fitting dentures, alcohol use, smoking, older age, and a reduced SF rate are all known predisposing factors for oral candidiasis and OMLs,[26, 27] and those risks emerged also with the regression analysis in the present study.
Sufficient SF is crucial for maintaining healthy oral tissues. It may be assumed that reduced salivary secretion could be one explanatory factor for our high prevalence of OMLs. The chronic LT recipients in particular often presented with reduced salivary secretion, which may have been due to several xerogenic medications such as the cardiovascular medications that the LT recipients were often already using during the pretransplant stage. Moreover, dry mouth–related subjective symptoms, especially xerostomia and dysphagia, were common among the chronic LT recipients, as shown in our previous study, and the present study includes the same cohort of patients. Dry mouth is a major concern and may have serious consequences; it predisposes patients to dental and oral diseases and causes difficulties in speaking, eating, and wearing dentures. Preventive measures for dry mouth include sufficient water intake, stimulation of saliva by the use of sugarfree suction tablets or sugarfree chewing gum, saliva substitutes, fluoride supplements, and possibly changes to some medications that are less xerogenic.
LT recipients are also at risk for neoplasias in the oral cavity and its surroundings. Two LT recipients in our study had skin lesions removed that were diagnosed as basal cell carcinoma, and several patients were on continuous follow-up with a dermatologist for fibromas or molds on their facial skin. OMLs with malignant potential such as OLP-like lesions, lichenoid lesions, leukoplakias, and erythroplakias occurred more frequently, especially among chronic LT recipients, in comparison with controls, but what must be noted is that their diagnoses were made clinically by visualization. A biopsy specimen is indicated when there is an ulceration or other clinical suspicion of a potentially malignant transformation of the lesion. OLP has been associated with liver disease and has been reported among immunocompromised patients in general,[8, 31, 32] but none emerged in the present study. Instead, a few cases of lichenoid lesions occurred (Fig. 1C). These may have been induced by antihypertensives because drugs such as beta-blockers, diuretics, and angiotensin-converting enzyme inhibitors are capable of triggering a lichenoid drug reaction, and these drugs were commonly used by the LT recipients. There should be a special focus on the atrophic and erosive forms of OLP, however, which could have a 0.5% to 4% chance of becoming malignant.[34, 35] A definite diagnosis should always be made via the biopsying of a suspected lesion. Among our LT recipients, 5 cases of unspecified, homogeneous leukoplakias, mainly on the alveolar ridge where a tooth was missing, were detectable. Alcohol use, smoking, poor oral health, and oral fungal infections are known risk factors for oral cancer, and clearly, long-term exposure to certain harmful lifestyle habits plays a role in the possible development of malignancies after transplantation.[36, 37]
The strength of our investigation is the prospective nature of the study and the inclusion of an age- and sex-matched control population. The LT patients were examined by 2 experienced hospital dental specialists according to a fixed protocol. All patients were LT recipients, whereas the patients in an uncontrolled study included other organ transplant recipients as well. We compared OMLs in a control population representative of the whole adult population (≥30 years old) in Finland. In general, the studies in this field are few, and many of them include children or LT recipients with liver disease etiologies very different from those in our investigation; these basic differences make comparisons of the results difficult.
The limitations of this study, however, also need to be addressed. In the control group, SF was not measured, nor were fungal samples cultivated. Possible candidosis in the control group was assessed clinically as a pseudomembranous fungal infection by a visual inspection, but no cases were reported. The relatively small number of acute LT recipients is another weakness, although we did include all potential LT recipients in this one and only LT center in Finland. The ratio of chronic LT recipients to acute LT recipients was comparable to the distribution of chronic and acute LT recipients at large in Finland. A bias toward a high frequency of OMLs could have resulted from the fact that some of the lesions may have already existed during the pretransplant period. Overall, LT seemed to lead to an improved oral health status because fewer oral pathologies were apparent after transplantation, but it must be noted that data concerning pretransplant oral health were collected retrospectively.
In conclusion, OMLs were more prevalent in LT recipients versus population controls, and this perhaps can be explained by immunosuppressive drugs and several other daily nontransplant medications. Reduced SF among LT recipients, on the other hand, may have been prevalent already before LT. As is generally accepted, CSA offered a risk for DIGO, and CACHs further increased its prevalence. Steroid use seemed to elevate the risk for OMLs and for oral candidosis. OMLs with malignant potential were more common among chronic LT recipients in comparison with controls. In LT patients on immunosuppressive medications, however, these lesions should be followed up carefully. Because OMLs are a risk for oral malignancies in LT recipients, annual oral examinations are recommended for all LT recipients. The follow-up period should be re-evaluated if changes occur in the clinical appearance of OMLs.
The authors thank Anna-Lisa Söderholm for the contribution of her expertise in oral mucosal diseases to the design and data collection for the Health 2000 health examination survey in Finland. That study was organized by the National Institute for Health and Welfare of Finland (formerly known as the National Public Health Institute). Carol Norris is warmly thanked for her prompt, expert English language editing.