Clinical Evaluation of Oral Chronic Graft-Versus-Host Disease

Article Outline

• Abstract
• Introduction
• Methods
• Results
• Discussion
• References
• Copyright

Abstract 

Oral chronic graft-versus-host disease (cGVHD) is a significant and serious complication following allogeneic hematopoietic stem cell transplantation (HSCT). The purpose of this study was to characterize the distribution, type, and extent of lesions and their correlation with patient-reported symptoms such as pain and discomfort. The effect of time since transplantation on these measures was also assessed. Consecutive patients with oral cGVHD referred to the Center for Oral Disease at Brigham and Women's Hospital, Boston, MA, were evaluated over a 2-year period. Subjective data included the responses to 4 targeted symptom questions (yes/no) and a visual analog scale pain score (0-10). Objective data included the location and extent of reticulation, erythema, and ulcerations using a previously published scoring system as well as time since HSCT. Multiple linear regression analyses were performed using SAS. We evaluated 27 patients, for a total of 79 clinic visits (median 2, range: 1-8). The median time since HSCT was 18 months (range: 5-157 months). The buccal and labial mucosa and tongue were the sites of 93% of all ulcerations, 72% of all erythematous lesions, and 76% of all reticular lesions, and were the most frequently affected sites. The gingiva, floor of mouth, and hard and soft palate were infrequently affected. Although uncommon, ulceration of the soft palate was the objective finding most highly correlated with increased pain (P < .0001), and there was a generalized significant trend for increased pain scores with increased extent of ulceration. Overall, 95% of pain scores were ≤5 (scale from 0-10, range: 0-7), with 40% reporting a score of zero. However, 80% admitted to avoiding certain foods because of mouth pain. After controlling for the presence and extent of ulcerations, we found that time since HSCT was inversely related to the pain score (P < .04). There was a statistically significant inverse relationship between the overall presence of ulceration and time since HSCT. We found that oral cGVHD most frequently affects the buccal and labial mucosa and the tongue. The functional impact was significant, as most patients had to restrict oral intake because of discomfort. Both the signs and symptoms associated with oral cGVHD tend to decrease over time. The association between ulceration of the soft palate and patient-reported pain highlights the significance of the location of involvement and the need for targeted approaches to therapy. Our findings, in large part, support the recently introduced National Institutes of Health response criteria for oral cGVHD, which is critical for the conduct of effective and meaningful research in this field; however, prospective application in clinical and investigative settings is necessary for evaluating its utility and efficacy in practice.

Key Words: Hematopoietic cell transplantation, Graft-versus-host disease, Oral mucosal disease, Salivary gland disease

Back to Article Outline

Introduction 

Chronic graft-versus-host disease (cGVHD) is a significant and serious complication following allogeneic hematopoietic stem cell transplantation (HSCT), that affects upward of 70% of patients surviving more than 100 days [1]. Typically presenting 3 to 4 months after HSCT, cGVHD develops when immunocompetent donor T cells recognize putative host tissue antigens in target organs as “foreign,” resulting in an immune-mediated inflammatory response and subsequent tissue destruction 2, 3. The skin, gastrointestinal tract, and liver are most commonly involved; however, almost any tissue/organ may be affected [4]. Despite tremendous advances in prevention and management, cGVHD is the leading cause of nonrelapse treatment-related mortality (TRM) >2 years after allogeneic transplantation 3, 5, 6, 7.

The oral cavity is affected in >70% of individuals who develop cGVHD 6, 8. The mucosa and/or salivary glands may be affected, causing significant pain and discomfort, dysfunction, limitation of oral intake and nutritional deficits, and an increased risk of dental caries and other oral infections 2, 9, 10. Oral mucosal disease has been described clinically and histopathologically, with reticular (hyperkeratotic) and/or erythematous (atrophic) being the most common forms, and ulcerative lesions, typically the most painful, presenting less frequently 9, 11, 12, 13, 14, 15, 16, 17. Despite aggressive conventional and novel systemic immunomodulatory therapies that are often effective in managing other clinical manifestations of cGVHD, the oral cavity is frequently refractory to treatment and may be the primary source of morbidity. In some cases, the oral cavity is the only site of clinical involvement.

Although a number of studies have described its various clinical and histopathologic findings, oral cGVHD has not been characterized systematically with respect to the location, distribution, type, and extent of lesions. Very little has been reported regarding symptomatic findings, and importantly, to what extent objective findings correlate with patient reported measures of disease. With respect to the subjective and objective features of oral cGVHD, the effect of time elapsed since HSCT has not been previously reported. This knowledge is critical for the development and refinement of appropriate and effective measures of disease activity and severity.

Back to Article Outline

Methods 

Consecutive patients with oral cGVHD who were referred to the Center for Oral Disease at Brigham and Women's Hospital/Dana-Farber Cancer Institute, Boston, MA, were evaluated from 2003-2005. Patients were typically referred for either definitive diagnosis and/or for treatment of disease not adequately managed by systemic therapies alone. Standardized subjective and objective data were prospectively collected on all patients evaluated during this period. Examinations were performed by 1 of 2 experienced and calibrated oral medicine specialists (N.T. or S.W.). Subjective data included the responses (yes/no) to 4 targeted symptom questions and a visual analog scale pain score to report current pain (0-10). Objective data included the location and extent of reticulation (typically considered to be least symptomatic), erythema (generally more symptomatic), and ulcerations (generally most symptomatic; Figure 1) using a previously published clinical scoring system, and a comprehensive weighted mucosal disease severity score was calculated as previously described [18]. Ulcerative lesions were cultured for herpes simplex virus (HSV). Concurrent immunosuppressive medications were recorded.

• 
  • View Large Image
  • Download to PowerPoint

• Figure 1 

  Oral mucosal cGVHD of the right buccal mucosa with reticular, erythematous, and ulcerative features. The unaffected mucosa is normal and pink.

Because of the small sample size, nonparametric analyses were performed with P values <.05 considered statistically significant. Descriptive statistics were calculated to evaluate the location, type, extent, and frequency of lesions as well as symptomatic data. As the signs and symptoms of oral cGVHD are known to fluctuate considerably over time, each individual patient visit was considered as the unit of analysis; therefore, individuals could account for multiple visits. The association of median pain score and presence/severity of ulcerations was investigated. Because outcomes recorded on the same patient at different visits might exhibit some degree of correlation (especially for visits that are close temporally), analyses were performed based on fitting linear regression models using generalized estimating equations [19]. All statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC). This study was approved by the institutional review board of the Dana-Farber Cancer Institute/Harvard Cancer Center, Boston, MA.

Back to Article Outline

Results 

Twenty-seven patients transplanted for a variety of underlying malignancies were evaluated over the 2-year study period (Table 1). The median time since HSCT was 18 months (range: 5-157 months). All patients were being treated with topical therapy (dexamethasone and/or tacrolimus rinses), and all but one patient was taking systemic immunomodulatory medications that included prednisone, rapamycin, mycophenylate mofetil (MMF), cyclosporine, and tacrolimus (median of 2 medications per patient). Mouth pain was reported at 41% of visits, and subjects admitted to having to avoid certain foods (eg, spicy and acidic foods) because of mouth sensitivity at nearly 80% of visits (Table 2). Mouth tightness was reported in 23% of visits (Table 2). There were no reported pain scores >7 (scale from 0-10, range: 0-7; Table 1). Overall, 95% of pain scores were ≤5, with 40% reporting a score of zero (data not shown). There were no positive HSV cultures.

Table 1. Descriptive Statistics and Pre-HSCT Diagnoses

Patients	27 (male = 13, female = 14)
Patient visits	79
Median number of visits (range)	2 (1-8)
Median age (range)	44 (28-68)
Median months since transplant (range)	18 (5-157)
Median pain score (range)	1.0 (0-7)
Diagnosis	n = 27
Chronic myelogenous leukemia	10
Acute myelogenous leukemia	9
Chronic lymphocytic leukemia	3
Chronic neutrophilic leukemia	1
Hodgkin disease	1
Myelodysplastic syndrome	1
Non-Hodgkin lymphoma	1
Polycythemia vera	1

Table 2. Responses to Symptomatic Questions

The presence and extent of oral lesions based on type, extent, and location are illustrated in Table 3, Table 4. The buccal mucosa, labial mucosa, and tongue were most frequently affected (Table 3), accounting for 93% of ulcerations, 72% of erythematous lesions, and 76% of reticular lesions (Table 4). The gingiva, floor of mouth, and hard/soft palate were infrequently affected; when involved, these sites were always associated with buccal, labial, or lingual lesions. When correlated with the pain score, the composite weighted score of extent of mucosal involvement was borderline statistically significant (Spearman correlation coefficient 0.35, P = .08; data not shown).

Table 3. Presence of Reticulation, Erythema, and Ulceration by Site

Table 4. Extent of Erythema and Ulceration by Site

Extent of reticulation not reported as only recorded as present/not present.

As ulcerations are considered to be the most symptomatic oral lesions, we investigated the extent to which presence and/or extent of ulcerations was associated with pain. A significant positive relationship was found between the presence of ulcerations and the overall pain score (Table 5). Although the majority of anatomic sites reached statistical significance for this association, the actual range of pain scores (from no ulcer to ulcer) was small. Although uncommonly affected, ulceration of the soft palate (Figure 2) was the objective finding that correlated with the greatest level of pain (5.5 with ulcers, 1.0 without ulcers; P < .0001). Furthermore, a Spearman correlation demonstrated that there was a generalized statistically significant trend for increased pain scores with increased extent of ulceration (Table 6). Twenty-two percent (n = 14) of visits were limited to reticular disease; in this subgroup, 86% of pain scores were zero and none were >1; however, 64% still admitted to having to avoid foods because of sensitivity.

Table 5. Association between Presence of Ulcerations and Pain Score

• 
  • View Large Image
  • Download to PowerPoint

• Figure 2 

  Ulcerative and erythematous cGVHD of the palate.

Table 6. Spearman Correlations of Size of Ulceration with Pain Score

As cGVHD tends to develop during a relatively well-defined period of time, we evaluated the extent to which time since HSCT affected the above-reported outcomes, again using the presence and extent of ulceration as a surrogate marker. There was a generalized inverse relationship between the overall prevalence of ulcerations in the study population and time since HSCT, with two anatomic sites reaching statistical significance (eg, soft palate, P < .0001; data not shown). For example, taking into account those sites that were statistically significant (P < .05), the median number of months since HSCT associated with ulcers ranged from 11 to 16, compared to 19 to 26 without ulcers. Extent of ulceration was inversely correlated with time since HSCT (eg, left and right buccal mucosa Spearman correlation coefficients of −0.22 and −0.24, P = .05 and .04, respectively). After controlling for the presence and extent of ulcerations in the overall study population, we found that time since HSCT was inversely related to the pain score (P < .04).

Back to Article Outline

Discussion 

This study represents one of the largest series of patients with oral cGVHD and the most comprehensively detailed report with respect to both objective and subjective findings. The clinical findings of distribution and type of lesions were for the most part consistent with previous reports. Schubert et al. [9] found that both erythema (73%) and lichenoid reticulation (36%) of the buccal and labial mucosa were common in 33 patients with cGVHD. They also found that ulcerative lesions were seen infrequently but most commonly affected the buccal mucosa, palate, and tongue dorsum, and that pain was reported in 56% of patients but not exclusively those with ulcerations. Busca et al. [20] followed 147 adult patients surviving until at least day 100, and found that 71% developed cGVHD, and of these, 54% developed oral cGVHD. They found that 44% had erythema, 68% had lichenoid/reticular changes, and 26% presented with ulcerations. The buccal mucosa was involved 98% of the time followed by the tongue (30%), lips (18%), and palate (16%). Nakamura et al. [16] found that of 18 patients diagnosed with cGVHD, 61% (n = 11) had lichenoid lesions on the buccal, labial, and lingual mucosa, although extent, severity, and symptoms were not reported.

We found that the buccal, labial, and lingual mucosa accounted for greater than three-quarters of all lesions, suggesting that greater emphasis should be placed on examining these sentinel sites as part of routine surveillance of cGVHD. These data support the recently published response criteria for oral cGVHD (http://www.asbmt.org/GvHDForms), which includes these sites but excludes the gingiva and floor of the mouth, which were rarely affected [21].

Approximately 60% of ulcerative lesions were <1.0 cm in diameter, and most frequently affected the buccal mucosa, palate, and tongue dorsum, similar to previous findings [9]. The National Institutes of Health (NIH) response criteria measures ulcerations as either not present (mild), ≤20% of the evaluated mucosal surfaces (moderate), or >20% (severe). Given our findings, very few if any patients would be scored as “severe” as the ulcerations have to account for >20% of the total mucosal surface areas evaluated. Our data demonstrates that the presence of ulcerations has a profound impact on the overall symptoms of mouth pain; however, mucosal pain and sensitivity may be present in the absence of ulcerations, as even purely reticular lesions are associated with an underlying inflammatory response and may be symptomatic.

Oral cGVHD can have a tremendous impact on quality of life, especially with respect to one's ability to speak and eat. Forty-one percent admitted to having mouth pain, compared to previous reports of 5% [20] and 56% [9]. In contrast, 80% admitted to having to avoid certain foods (eg, spicy, acidic, and hard/crusty foods) because of sensitivity, indicating that mouth discomfort was, in fact, far more prevalent with a significant impact on oral intake. The NIH Consensus Development Project on Criteria for Clinical Trials in Chronic Graft-versus-Host Disease includes patient reported outcomes (0-10 scale, most severe during the previous week) for both mouth pain in the absence of stimulation as well as mouth sensitivity, which is defined as irritation from normally tolerated spices, foods, liquids, and/or flavors (eg, mint toothpaste, carbonated drinks) [21]. We did not assess which foods specifically were avoided, or the extent to which oral intake was limited by oral cGVHD symptoms; therefore, the dietary and nutritional impacts, critical elements in the maintenance of overall health, are unknown. Future investigations should incorporate the NIH response criteria to more accurately and uniformly assess pain and its impact on function, and may also include a more detailed assessment of the extent to which oral intake is restricted, and specifically which foods are avoided 22, 23.

In nearly all patients, signs and symptoms of oral cGVHD were restricted to the oral cavity and did not extend to the posterior oropharynx or esophagus. Very few patients reported pain when swallowing, a symptom most likely attributable to the presence of soft palate ulcers (Table 5). Another possible but less likely explanation for such pain could be the presence of superficial mucoceles on the soft palate (not reported). Superficial mucoceles are common in patients with oral cGVHD and are typically described more as a “nuisance” than actually being painful 2, 17, 24, 25. The question of tightness was intended to evaluate sclerotic changes. Tightness may be noted in the context of inflamed mucosa and in more severe cases palpable fibrotic bands of tissue in the posterior buccal mucosa occur; a combination of both findings may have explained the 20% “yes” response in this series.

cGVHD typically develops between 4 and 7 months following HSCT, and patients are treated for extended periods of time with often unpredictable and incomplete responses 4, 23. The median duration of immunosuppressive therapy is 19-23 months, with approximately 50% continuing to require therapy at 5 years, and 15% of those surviving without relapse continuing treatment for more than 7 years 26, 27. All but one of our patients remained on systemic immunosuppressive therapies at 19-23 months; however, both the size of ulcerations and pain score were inversely related to time since HSCT. The explanation for this finding is unclear, although it is likely related to the underlying pathophysiology of cGVHD (which is still poorly understood), namely that the level of activity tends to decrease as tolerance develops. As most subjects were being actively treated systemically and topically throughout the study period, it would be difficult to explain by efficacy of treatment alone.

Our findings, in large part support the recently introduced NIH response criteria for oral cGVHD, which represents an important step forward for the conduct of effective and meaningful research in this field. Prospective application in clinical and investigative settings will be critical for evaluating its utility and efficacy in practice. The impact and utilization of systemic and localized therapies on the measures of oral cGVHD activity over time are unknown; only with well-designed, multicenter interdisciplinary research will these questions begin to be answered.

Back to Article Outline

References 

1. Lee SJ. New approaches for preventing and treating chronic graft-versus-host disease. Blood. 2005;105:4200–4206
   • View In Article
   • MEDLINE
   • CrossRef
2. Woo SB, Lee SJ, Schubert MM. Graft-vs.-host disease. Crit Rev Oral Biol Med. 1997;8:201–216
   • View In Article
   • MEDLINE
   • CrossRef
3. Lee SJ, Vogelsang G, Flowers ME. Chronic graft-versus-host disease. Biol Blood Marrow Transplant. 2003;9:215–233
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (186 KB)
   • CrossRef
4. Bhushan V, Collins RH. Chronic graft-vs-host disease. JAMA. 2003;290:2599–2603
   • View In Article
   • CrossRef
5. Socie G, Stone JV, Wingard JR, et al. Long-term survival and late deaths after allogeneic bone marrow transplantation. Late Effects Working Committee of the International Bone Marrow Transplant Registry. N Engl J Med. 1999;341:14–21
   • View In Article
   • MEDLINE
   • CrossRef
6. Lee SJ, Klein JP, Barrett AJ, et al. Severity of chronic graft-versus-host disease: association with treatment-related mortality and relapse. Blood. 2002;100:406–414
   • View In Article
   • MEDLINE
   • CrossRef
7. Curtis RE, Metayer C, Rizzo JD, et al. Impact of chronic GVHD therapy on the development of squamous-cell cancers after hematopoietic stem-cell transplantation: an international case-control study. Blood. 2005;105:3802–3811
   • View In Article
   • MEDLINE
   • CrossRef
8. Flowers ME, Parker PM, Johnston LJ, et al. Comparison of chronic graft-versus-host disease after transplantation of peripheral blood stem cells versus bone marrow in allogeneic recipients: long-term follow-up of a randomized trial. Blood. 2002;100:415–419
   • View In Article
   • MEDLINE
   • CrossRef
9. Schubert MM, Sullivan KM, Morton TH, et al. Oral manifestations of chronic graft-v-host disease. Arch Intern Med. 1984;144:1591–1595
   • View In Article
   • MEDLINE
10. Mattsson T, Arvidson K, Heimdahl A, Ljungman P, Dahllof G, Ringden O. Alterations in taste acuity associated with allogeneic bone marrow transplantation. J Oral Pathol Med. 1992;21:33–37
    • View In Article
    • MEDLINE
    • CrossRef
11. Rodu B, Gockerman JP. Oral manifestations of the chronic graft-v-host reaction. JAMA. 1983;249:504–507
    • View In Article
    • MEDLINE
12. Barrett AP, Bilous AM. Oral patterns of acute and chronic graft-v-host disease. Arch Dermatol. 1984;120:1461–1465
    • View In Article
    • MEDLINE
13. Berkowitz RJ, Strandjord S, Jones P, et al. Stomatologic complications of bone marrow transplantation in a pediatric population. Pediatr Dent. 1987;9:105–110
    • View In Article
    • MEDLINE
14. Dahllof G, Heimdahl A, Modeer T, Twetman S, Bolme P, Ringden O. Oral mucous membrane lesions in children treated with bone marrow transplantation. Scand J Dent Res. 1989;97:268–277
    • View In Article
    • MEDLINE
15. Hiroki A, Nakamura S, Shinohara M, Oka M. Significance of oral examination in chronic graft-versus-host disease. J Oral Pathol Med. 1994;23:209–215
    • View In Article
    • MEDLINE
    • CrossRef
16. Nakamura S, Hiroki A, Shinohara M, et al. Oral involvement in chronic graft-versus-host disease after allogeneic bone marrow transplantation. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1996;82:556–563
    • View In Article
    • Abstract
    • Full-Text PDF (3665 KB)
    • CrossRef
17. Treister NS, Woo SB, O'Holleran EW, Lehmann LE, Parsons SK, Guinan EC. Oral chronic graft-versus-host disease in pediatric patients after hematopoietic stem cell transplantation. Biol Blood Marrow Transplant. 2005;11:721–731
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (392 KB)
    • CrossRef
18. Piboonniyom SO, Treister N, Pitiphat W, Woo SB. Scoring system for monitoring oral lichenoid lesions: a preliminary study. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2005;99:696–703
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (280 KB)
    • CrossRef
19. Liang K, Zeger S. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:12–22
    • View In Article
20. Busca A, Locatelli F, Vai S, Dall'Omo AM, Gargiulo A, Falda M. Clinical grading of oral chronic graft-versus-host disease in 104 consecutive adult patients. Haematologica. 2005;90:567–569
    • View In Article
21. Pavletic SZ, Martin P, Lee SJ, et al. Measuring therapeutic response in chronic graft-versus-host disease: National Institutes of Health Consensus Development Project on Criteria for Clinical Trials in Chronic Graft-versus-Host Disease: IV. Response Criteria Working Group report. Biol Blood Marrow Transplant. 2006;12:252–266
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (334 KB)
    • CrossRef
22. Pavletic SZ, Lee SJ, Socie G, Vogelsang G. Chronic graft-versus-host disease: implications of the National Institutes of Health consensus development project on criteria for clinical trials. Bone Marrow Transplant. 2006;38:645–651
    • View In Article
    • MEDLINE
    • CrossRef
23. Shlomchik WD, Lee SJ, Couriel D, Pavletic SZ. Transplantation's greatest challenges: advances in chronic graft-versus-host disease. Biol Blood Marrow Transplant. 2007;13:2–10
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (370 KB)
    • CrossRef
24. Nicolatou-Galitis O, Kitra V, Van Vliet-Constantinidou C, et al. The oral manifestations of chronic graft-versus-host disease (cGVHD) in paediatric allogeneic bone marrow transplant recipients. J Oral Pathol Med. 2001;30:148–153
    • View In Article
    • MEDLINE
25. Campana F, Sibaud V, Chauvel A, Boiron JM, Taieb A, Fricain JC. Recurrent superficial mucoceles associated with lichenoid disorders. J Oral Maxillofac Surg. 2006;64:1830–1833
    • View In Article
    • Full Text
    • Full-Text PDF (420 KB)
    • CrossRef
26. Koc S, Leisenring W, Flowers ME, et al. Therapy for chronic graft-versus-host disease: a randomized trial comparing cyclosporine plus prednisone versus prednisone alone. Blood. 2002;100:48–51
    • View In Article
    • MEDLINE
    • CrossRef
27. Stewart BL, Storer B, Storek J, et al. Duration of immunosuppressive treatment for chronic graft-versus-host disease. Blood. 2004;104:3501–3506
    • View In Article
    • MEDLINE
    • CrossRef