Vol.
-
No.
-
Month 2015
Direct immunofluorescence testing results in cases of premalignant and malignant oral lesions L.J. Montague, DMD,a I. Bhattacharyya, DDS, MSD,b M.N. Islam, DDS, BDS,b D.M. Cohen, DMD, MS, MBA,b and S.G. Fitzpatrick, DDSb Objectives. Oral premalignant and malignant lesions may mimic oral lichen planus (OLP) clinically and microscopically. OLP often shows basement membrane fibrinogen positivity on direct immunofluorescence testing (DIF). This study examined fibrinogen positivity in oral premalignant lesions and squamous cell carcinoma. Study Design. The University of Florida Oral Pathology Biopsy Service records were searched for the years 2003 to 2013 for oral premalignant lesions and squamous cell carcinoma with concurrent DIF testing. Demographic, clinical, and DIF or histologic information was collected and analyzed. Results. Sixty-eight fibrinogen positive lesions were identified within a total of 164 cases. Low-grade dysplasia and premalignant verrucous lesions made up the majority of the fibrinogen positive lesions (combined n ¼ 43; 63.2%), and the most common locations in positive cases were the buccal mucosa, tongue, and gingiva. A lichenoid distribution of the inflammatory infiltrate significantly predicted fibrinogen positivity (P < .0005). Conclusions. Fibrinogen positivity may be seen in premalignant and malignant oral lesions increasing the risk of misdiagnosis. (Oral Surg Oral Med Oral Pathol Oral Radiol 2015;-:1-9)
Within the oral cavity, significant clinical and histologic overlap may occur among oral lichenoid lesions (OLL), oral lichenoid mucositis (OLM) and oral lichen planus (OLP), which will be referred collectively throughout this article as OLP for simplicity, and premalignant lesions or squamous cell carcinoma (SCCA). Both premalignant and malignant oral lesions and OLP may occur clinically as white, red and white, or ulcerated oral lesions, and the phenomenon of premalignant and malignant oral lesions exhibiting classically lichenoid features has been well established histologically.1,2 This overlap may lead to a delay in the diagnosis and treatment of true dysplasia or SCCA of the oral cavity. Direct immunofluorescence (DIF) testing has been established as an adjunctive diagnostic tool in differentiating OLP from similar appearing lesions.3-6 Deposits of immunoreactants in OLP with “shaggy” basement membrane zone (BMZ) fibrinogen reactivity has been described as the best DIF indicator of the disease.5 However, fibrinogen positivity at the BMZ is not specific to OLP and has been reported in other inflammatory disorders.7 Because of the potential for clinicopathologic overlap between OLP and oral dysplasia or SCCA, DIF is Presented as an oral essay at the 2014 American Academy of Oral and Maxillofacial Pathology Annual Meeting in St. Augustine, Florida, USA. a Department of Oral and Maxillofacial Surgery and Pathology, University of Mississippi Medical Center, Jackson, MS, USA. b Department of Oral and Maxillofacial Diagnostic Sciences, University of Florida College of Dentistry, Gainesville, Florida, USA. Received for publication Dec 17, 2014; returned for revision Feb 19, 2015; accepted for publication Feb 20, 2015. Ó 2015 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2015.02.478
occasionally requested by the clinician for the purpose of differentiating OLP from other immune-mediated diseases, oral dysplasia, and SCCA, which may exhibit overlapping clinical and histologic features. Minimal information is found in the literature with regard to the DIF staining patterns of oral dysplasia or SCCA; thus, fibrinogen positivity on biopsy may contribute to diagnostic confusion in certain cases, especially those in which the level of premalignant features are mild or reactive. The objectives of this study were to characterize the DIF results, in particular positive fibrinogen reactivity, in cases of oral premalignant lesions and SCCA in biopsy specimens, which were incidentally submitted for DIF testing because of clinical lichenoid appearance, and to characterize the clinical and histologic patterns. The aim of the study was to prove that fibrinogen positivity may be commonly found in premalignant and malignant oral lesions.
MATERIALS AND METHODS Following institutional review board approval, the University of Florida Oral Pathology Biopsy service records were searched for the years 2003 to 2013 for
Statement of Clinical Relevance Direct immunofluorescence findings may be positive in premalignant and malignant lesions, contributing to the confusion with oral lichen planus especially where overlapping clinical and histologic similarities are seen. Clinicopathologic correlation is essential in discriminating between these entities. 1
ORAL AND MAXILLOFACIAL PATHOLOGY 2 Montague et al.
OOOO Month 2015
Fig. 1. Inflammatory pattern. A, minimal inflammation, grade 0 (hematoxylin&eosin, original magnification 10); B, diffuse lichenoid, grade 1 (hematoxylin&eosin, original magnification 10); C, focal lichenoid, grade 2 (hematoxylin&eosin, original magnification 10); D, nonlichenoid, grade 3 (hematoxylin&eosin, original magnification 10).
cases with specific diagnostic codes that also had DIF studies performed. These diagnostic codes included verrucopapillary hyperkeratosis (VPHK), dysplasia, atypical squamous epithelial proliferation (AEP), SCCA, and verrucous carcinoma (VC). Pathology reports were reviewed for all cases containing these codes, as well as age, gender, diagnosis, DIF findings, and biopsy location, were recorded. Archival hematoxylin and eosin (H&E) prepared slides were re-examined by two study participants (LM and SF) to confirm the sign-out diagnosis. All cases were originally examined and signed out by three of the study participants (IB, DC, and MI). Four of the study participants (IB, DC, MI, and SF) were board-certified oral and maxillofacial pathologists at the time of the study and the remaining study participant (LM) was at the time a senior oral and maxillofacial pathology resident. Cases in which the searched diagnosis and DIF were from different anatomic locations were excluded from the study. The cases were divided into six diagnostic categories: (1) VPHK (the term utilized at our institution for lesions compatible with the hyperkeratosis to early verrucous hyperplasia stages of proliferative verrucous leukoplakia, showing verrucoid thickening of the epithelium and orthokeratin or parakeratin layers),8 (2) low-grade dysplasia (LGD; atypia or dysplasia to mild dysplasia), (3) high-grade dysplasia (HGD; moderate to severe dysplasia), (4) AEP (defined as lesions exhibiting architectural features of malignancy but not meeting criteria for frank carcinoma), (5) SCCA, and (6) VC. Cases which were diagnosed as dysplasia in addition to
VPHK were categorized on the basis of their degree of dysplasia, rather than being included in the VPHK category. During the slide review by two study participants (SF and LM), the case was reassigned to the most appropriate category if the original diagnosis rendered was borderline between two of the designated diagnostic categories of the study. If no general agreement were to be found between the original sign-out diagnoses and those of the diagnostic reviewers, the case was to be dropped, although no cases fell into this category within this study. Multiple cases from the same patient were only included if the biopsies were from different sites within the oral cavity. If more than one biopsy was found for the same patient at the same site, the initial biopsy was included in the study, and subsequent biopsies were excluded in order to avoid counting the same lesion multiple times. Since the original slides of frozen DIF sections were not available for review (except for five cases), the DIF findings were recorded on the basis of the pathology report sign-out. The original DIF was performed at the in-house laboratory of the University of Florida College of Dentistry oral pathology biopsy service, utilizing frozen tissues cut into 4-micron sections and air dried, acetone fixed, washed with buffer, treated with protein block (Dako North America, Inc., Carpinteria, CA), incubated with antisera conjugated immunoglobulin G (IgG), IgM, IgA, C3, and fibrinogen (Thermo Scientific, Lab Vision Corporation, Fremont, CA), and then rinsed with distilled water and coverslipping.
OOOO Volume -, Number
ORIGINAL ARTICLE Montague et al. 3
-
Fig. 2. Fibrinogen reactivity on direct immunofluorescence. A/D, Verrucopapillary hyperkeratosis (A, hematoxylin&eosin, original magnification 10; D, shaggy fibrinogen along BMZ, original magnification 20). B/E, High-grade (moderate) dysplasia (B, hematoxylin&eosin, original magnification 10; E, linear fibrinogen at BMZ, original magnification 20). C/F, Atypical epithelial proliferation (C, hematoxylin&eosin, original magnification 10; F, linear fibrinogen at BMZ, original magnification 20). BMZ, basement membrane zone.
Additionally, all H&E slides for the cases included in the study were examined for the pattern of inflammation. A scale consisting of four different grades was created, and each case was assigned one of the following grades after consensus review by two reviewers (LM and SF): grade 0, minimally inflamed (noninflamed or inflammatory infiltrate present in less than 20% of tissue examined); grade 1, diffuse lichenoid pattern (bandlike inflammatory infiltrate present in greater than 50% of tissue examined); grade 2, focal lichenoid pattern (bandlike inflammatory infiltrate present in 20% -50% of tissue examined); and grade 3, nonlichenoid pattern, including ulcerated specimens, perivascular inflammation and patchy or patternless inflammation. Examples of the different patterns of inflammation can be seen in Figure 1. The data were analyzed statistically using SPSS version 22 (SPSS IBM, New York, NY). Differences between the fibrinogen-positive and fibrinogen-negative groups for age were analyzed by using independent samples t test. The chi-square test for independence was used to analyze the differences between the groups for all other variables and significance was set at the P < .05 level. The 18 cases of SCCA were excluded from statistical analysis of inflammatory pattern, since the presence of a lichenoid inflammatory pattern could not be
Table I. Demographic characteristics for all cases
Female Male Mean age (years)
Total cases n ¼ 164 (%)
Fibrinogen positive n ¼ 68 (%)
Fibrinogen negative n ¼ 96 (%)
105 (64) 59 (36) 64.9
45 (66.2)* 23 (33.8)* 65.8y
60 (62.5)* 36 (37.5)* 64.2y
*P ¼ .63 (chi-square test), no significant differences between positive and negative groups. y P ¼ .16 (t test), no significant differences between positive and negative groups.
determined in areas of invasion when the connective tissue was extensively replaced by tumor.
RESULTS DIF staining and demographic characteristics A total of 164 cases that met the inclusion criteria for the study were identified. Of these, 68 (41.5%) cases demonstrated positive staining for fibrinogen, and 96 cases (58.5%) were negative. Five fibrinogen-positive cases also demonstrated C3 positivity; 3 HGD and 1 LGD cases showed weak, granular staining along the BMZ (and 1 LGD case was C3 positive around the blood vessels). Three total fibrinogen negative cases also showed positive C3 staining; 1 HGD case stained
ORAL AND MAXILLOFACIAL PATHOLOGY 4 Montague et al.
OOOO Month 2015
Fig. 3. Fibrinogen reactivity on direct immunofluorescence. A/B, Verrucous carcinoma (A, hematoxylin&eosin, original magnification 10; B, shaggy fibrinogen at BMZ along tips of rete ridges, original magnification 20). C/D, Squamous cell carcinoma (C, hematoxylin&eosin, original magnification 10; D, focal linear fibrinogen at BMZ, original magnification 20). BMZ, basement membrane zone.
for C3 along the BMZ, 1 LGD case was C3 positive along the BMZ and around the blood vessels, and 1 SCCA case showed C3 staining only around the blood vessels. All 164 cases were negative for IgG, IgA, or IgM. Additionally, 5 of the fibrinogen-positive cases and 6 of the fibrinogen-negative cases exhibited superficial candidal colonization. The histopathologic features and fibrinogen reactivity of available cases are shown in Figures 2 and 3. The demographic features of the cases may be seen in Table I. The total cases (n ¼ 164) were composed of 105 females (64.0%) and 59 (36.0%) males. The patient age ranged from 23 to 94 years, with a mean of 64.9 years, and in one case the age was unknown. The fibrinogen-positive group (n ¼ 68) was made up of 45 (66.2%) females and 23 (33.8%) males, whereas the fibrinogen-negative group (n ¼ 96) consisted of 60 females (62.5%) and 36 (37.5%) males. The mean ages for the positive and negative groups were 65.8 years and 64.2 years, respectively. There were no statistically significant differences between the fibrinogen-positive and fibrinogen-negative groups for either gender (P ¼ .63) or age (P ¼ .16). Distribution by diagnosis The histologic diagnoses for the cases, as defined previously, are shown in Figure 4. The total diagnoses
(n ¼ 164) included 56 cases of VPHK (34.1%), 53 cases of low-grade dysplasia (32.3%), 23 cases of high-grade dysplasia (14%), 18 cases of SCCA (11%), 8 cases of AEP (4.9%), and 6 cases of VC (3.7%). The most common diagnosis among the fibrinogen-positive group was VPHK (n ¼ 22; 32.4%), followed by LGD (n ¼ 21; 30.9%). A total of 8 fibrinogen-positive cases of SCCA and 3 VC were identified. The majority of negative cases were diagnosed as VPHK (n ¼ 34; 35.4%) and low-grade dysplasia (n ¼ 32, 33.3%). Thirteen fibrinogen-negative carcinomas, including 10 SCCA and 3 VC, were identified. The difference between the fibrinogen-positive and fibrinogen-negative groups for diagnosis was not significant (P ¼ .97). The clinical and histopathologic features of cases from each diagnostic category are shown in Figure 5. Distribution by site Figure 6 demonstrates the distribution of the cases by anatomic location. The buccal mucosa or vestibule (n ¼ 56; 34.1%) and tongue (n ¼ 55; 33.5%) were the most common sites overall, followed by the gingiva or alveolar ridge (n ¼ 36; 22%), palate (n ¼ 9; 5.5%), floor of mouth (n ¼ 5; 3.0), “oral cavity” (n ¼ 2; 1.2%), and lip vermilion (n ¼ 1; 0.6%). The majority of fibrinogen-positive lesions occurred on the buccal mucosa or vestibule (n ¼ 22; 32.4%), tongue (n ¼ 21;
OOOO Volume -, Number
-
ORIGINAL ARTICLE Montague et al. 5
Fig. 4. Distribution by diagnosis.
30.9%), and gingiva (n ¼ 19; 27.9%). For the negative cases, the most common sites were the buccal mucosa or vestibule (n ¼ 34; 35.4%) and tongue (n ¼ 34; 35.4%). There was no statistically significant difference (P ¼ .31) between the positive and negative groups based on the site within the oral cavity. Distribution by inflammatory pattern The distribution of positive and negative cases based on the pattern of inflammation is provided in Figure 7. The most common inflammatory pattern among all cases (n ¼ 146; 18 SCCA excluded) was diffuse lichenoid (n ¼ 74; 50.7 %), followed by minimally inflamed (n ¼ 36; 24.7%), nonlichenoid (n ¼ 20; 13.7%), and focal lichenoid (n ¼ 16; 11%) patterns. For the fibrinogen-positive cases (n ¼ 60), a diffuse lichenoid inflammatory pattern was seen in 44 cases (73.3%), with an additional 7 cases (11.7%) demonstrating a focal lichenoid pattern, and with only 9 cases showing minimal (n ¼ 4) or nonspecific (n ¼ 5) inflammation. The most common pattern of inflammation for the negative cases (n ¼ 86) was minimally inflamed pattern (n ¼ 32; 37.2%) followed by diffuse lichenoid pattern (n ¼ 30; 34.9%). The differences in the pattern of inflammation for the fibrinogen-positive and fibrinogen-negative groups were statistically significant (P < .0005).
DISCUSSION Direct immunofluorescence testing measures immunoreactant activity through the addition of fluoresceinated antibodies directly to a frozen biopsy specimen.9 One immunoreactant commonly tested is fibrinogen, a plasma protein important to the wound healing process.10 In cutaneous lichen planus, DIF has been shown to be
highly variable, with positivity in 37% to 97% of cases, and manifests as BMZ fibrinogen positivity or cytoid body reactivity to any immunoreactant.9 Within the oral cavity, previous studies have shown the sensitivity of DIF for OLP between 48% and 66%.5,9,11-14 BMZ fibrinogen positivity may be seen in other disorders, such as systemic lupus erythematosus and erythema multiforme, as well.7 In other mucosal sites, such as the conjunctiva, fibrinogen positivity has been found to be normal to the location and has led to erroneous diagnoses of lichen planus in some cases.15 In a large study of 500 oral mucosal lesions by Helander and Rogers,5 DIF results were shown to be more useful for other oral vesiculobullous diseases, such as mucous membrane pemphigoid and pemphigus vulgaris, than for lichen planus, and the sensitivity of clinical examination and H&E staining has been shown to be higher than DIF testing for OLP. In addition, perilesional DIF positivity was found to be similar to that for locations distant to lesional tissue in OLP cases, showing that the immune response may be widespread despite the lesion being localized.11 Direct immunofluorescence fibrinogen positivity in premalignant and malignant lesions in the oral cavity has been largely unreported in the literature. Schiodt et al. first reported 30 cases of clinical leukoplakia tested with DIF and found 2 cases of fibrinogenpositive lesions.16 Later reports, including 2 cases of SCCA and 6 dysplastic lesions of the oral cavity reported by Rinaggio et al. and one case each of SCCA and atypia reported by Suresh et al., were all fibrinogen negative on DIF.17,18 Oral lichen planus and oral SCCA have both been associated with a predominantly T-cell lymphocytic inflammatory response.19,20 In OLP, this T-cell response is
ORAL AND MAXILLOFACIAL PATHOLOGY 6 Montague et al.
OOOO Month 2015
Fig. 5. Clinical and histopathologic features of fibrinogen positive cases. A/D, Verrucopapillary hyperkeratosis (hematoxylin&eosin, original magnification 10). B/E, Low-grade (mild) dysplasia (hematoxylin&eosin, original magnification 10). C/F, High-grade (moderate) dysplasia (hematoxylin&eosin, original magnification 10). G/J, Atypical epithelial proliferation (hematoxylin&eosin, original magnification 10). H/K, Verrucous carcinoma (hematoxylin&eosin, original magnification 10). I/L, Squamous cell carcinoma (hematoxylin&eosin, original magnification 10).
thought to lead to the destruction of the basement membrane.21,22 In nonulcerated lesions of OLP, there is believed to be a counterbalancing increase in cellular proliferation that prevents the occurrence of ulceration.21 Similarly, in oral dysplasia and SCCA, either ulceration or proliferation of tissue may be a feature.23 These similarities between response mechanisms may contribute to the clinical overlap and potential for misdiagnosis. The presence of fibrinogen positivity in 41% of the premalignant, borderline, and malignant lesions in this study demonstrates that fibrinogen positivity itself cannot be considered evidence for the diagnosis of OLP
without consideration for the clinical and histologic context of the lesion. No statistical significance for positivity was found in our study for age or gender. In another recent study involving one of the authors (SF), the presence of histologic lichenoid features in oral premalignant and malignant lesions was found to be similarly nonsignificant for age or gender.2 One pattern that emerged in our study was within the distribution of fibrinogen positivity by oral subsite. Although no statistically significant differences were found among the sites, the gingiva was the only site in which the positive cases outnumbered the negative
OOOO Volume -, Number
-
ORIGINAL ARTICLE Montague et al. 7
Fig. 6. Distribution by site.
Fig. 7. Distribution by inflammatory pattern.
cases. Interestingly, this phenomenon has also been reported by Helander and Rogers,5 who found that DIF sensitivity in OLP varies greatly by site, with the gingiva showing the best sensitivity to DIF. Sano et al.11 studied 79 cases of OLP and also found variability in fibrinogen positivity based on site. This study found no statistically significant difference in positivity when comparing histologic diagnoses, but this may have been caused by the sample numbers. Since invasive carcinoma has less clinical similarity to OLP compared with low-grade dysplasia, fewer biopsy specimens were submitted
with a request for DIF in this study, with borderline or malignant lesions representing only 15% (n ¼ 24, including atypical squamous epithelial proliferations) of the total biopsies included. Therefore, small numbers of malignant lesions may have prevented the emergence of a pattern. Low-grade dysplasia and VPHK are especially problematic in terms of potential for misdiagnosis as lichenoid lesions. Verrucous hyperplasia, which is a premalignant entity typified by thickened white clinical lesions sometimes appearing multifocally (proliferative verrucous leukoplakia), and VPHK typically show minimal
ORAL AND MAXILLOFACIAL PATHOLOGY 8 Montague et al.
cellular signs of dysplasia and have often been shown to have lichenoid features both clinically and histologically.8,24,25 In the present study, the only statistically significant predictor of DIF fibrinogen positivity was the presence of a histologic lichenoid (superficial bandlike lymphocytic) inflammatory pattern. The concept of “lichenoid dysplasia” was introduced in 1985, referring to the phenomenon of dysplastic lesions demonstrating lichenoid features, and since then, a lichenoid response to dysplasia has become an established phenomenon.1 The presence of fibrinogen positivity in cases of oral dysplasia and SCCA supports the lack of specificity of a lichenoid inflammatory pattern in these lesions. Although the bandlike inflammatory infiltrate seen in OLP may be perhaps the most easily recognizable feature of the disease, it may also represent a very nonspecific inflammatory response.26 Other histologic features of OLP include the presence of degenerative keratinocytes (civatte bodies), liquefactive degeneration of the basal cell layer, interface stomatitis, and sawtooth rete ridge formation.24 Several factors may increase the potential for misdiagnosis in lesions that have lichenoid histologic features. If specimens are submitted exclusively for DIF or if the DIF specimen and the formalin biopsy are submitted to separate laboratories, the only H&E-stained specimen available for review may be on frozen section. Tissue distortion and artifact present on frozen sections create diagnostic difficulties, which could lead to potential misdiagnosis, especially in small specimens and in cases of low-grade dysplasia.27 In addition, small incisional biopsies may not be representative of the overall sample, and separation of the epithelium from underlying connective tissue may lead to inconclusive or false-positive DIF findings.5 In the oral cavity, incisional biopsies have been found to underdiagnose conditions in nearly 30% of cases compared with resections.28 Finally, diagnosis by pathologists unfamiliar with the subtle nature of many oral premalignant conditions may pose a problem, considering that oral and maxillofacial pathology may represent an area of variable experience for many general pathologists. In a study of consultations to oral and maxillofacial pathologists by general pathologists, it was found that the most common need for second opinions was for oral dysplasia or squamous cell carcinoma.29
CONCLUSIONS The results of this study show the frequent occurrence of fibrinogen positivity on DIF in premalignant and malignant lesions of the oral cavity. Pathologists should be aware of this phenomenon and avoid the potential for misdiagnosis of these lesions, especially in groups with milder dysplastic change. Fibrinogen positivity on
OOOO Month 2015
DIF should not be considered pathognomonic for OLP in premalignant or borderline malignant cases, and clinical and histopathologic context, rather than DIF, should be given higher consideration in these cases. The possibility of an immune response presenting with a lichenoid pattern of inflammation to a dysplastic lesion should not be overlooked, and second opinion should be sought for confounding lesions. The authors wish to thank Douglas Damico for technical assistance on the description of the direct immunofluorescence methodology utilized.
REFERENCES 1. Krutchkoff DJ, Eisenberg E. Lichenoid dysplasia: a distinct histopathologic entity. Oral Surg Oral Med Oral Pathol. 1985;60: 308-315. 2. Fitzpatrick SG, Honda KS, Sattar A, Hirsch SA. Histologic lichenoid features in oral dysplasia and squamous cell carcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol. 2014;117: 511-520. 3. Firth NA, Rich AM, Radden BG, Reade PC. Assessment of the value of immunofluorescence microscopy in the diagnosis of oral mucosal lichen planus. J Oral Pathol Med. 1990;19:295-297. 4. Kulthanan K, Jiamton S, Varothai S, Pinkaew S, Sutthipinittharm P. Direct immunofluorescence study in patients with lichen planus. Int J Dermatol. 2007;46:1237-1241. 5. Helander SD, Rogers RS 3rd. The sensitivity and specificity of direct immunofluorescence testing in disorders of mucous membranes. J Am Acad Dermatol. 1994;30:65-75. 6. Toto PD, Nadimi HT. An immunohistochemical study of oral lichen planus. Oral Surg Oral Med Oral Pathol. 1987;63:60-67. 7. Black MM. What is going on in lichen planus? Clin Exp Dermatol. 1977;2:303-310. 8. Gillenwater AM, Vigneswaran N, Fatani H, Saintigny P, ElNaggar AK. Proliferative verrucous leukoplakia: recognition and differentiation from conventional leukoplakia and mimics. Head Neck. 2014;36:1662-1668. 9. Chhabra S, Minz RW, Saikia B. Immunofluorescence in dermatology. Indian J Dermatol Venereol Leprol. 2012;78:677-691. 10. Halper J, Kjaer M. Basic components of connective tissues and extracellular matrix: elastin, fibrillin, fibulins, fibrinogen, fibronectin, laminin, tenascins and thrombospondins. Adv Exp Med Biol. 2014;802:31-47. 11. Sano SM, Quarracino MC, Aguas SC, et al. Sensitivity of direct immunofluorescence in oral diseases. Study of 125 cases. Med Oral Patol Cir Bucal. 2008;13:E287-E291. 12. Rogers RS 3rd, Jordon RE. Immunopathology of oral mucosal inflammatory diseases. Clin Exp Dermatol. 1977;2:97-107. 13. Rogers RS 3rd, Van Hale HM. Immunopathologic diagnosis of oral mucosal inflammatory diseases. Australas J Dermatol. 1986;27:51-57. 14. Van Hale HM, Rogers RS 3rd. Immunopathology of oral mucosal inflammatory diseases. Dermatol Clin. 1987;5:739-750. 15. Mehta M, Siddique SS, Gonzalez-Gonzalez LA, Foster CS. Immunohistochemical differences between normal and chronically inflamed conjunctiva: diagnostic features. Am J Dermatopathol. 2011;33:786-789. 16. Schiodt M, Holmstrup P, Dabelsteen E, Ullman S. Deposits of immunoglobulins, complement, and fibrinogen in oral lupus erythematosus, lichen planus, and leukoplakia. Oral Surg Oral Med Oral Pathol. 1981;51:603-608.
OOOO Volume -, Number
ORIGINAL ARTICLE Montague et al. 9
-
17. Rinaggio J, Crossland DM, Zeid MY. A determination of the range of oral conditions submitted for microscopic and direct immunofluorescence analysis. J Periodontol. 2007;78:19041910. 18. Suresh L, Neiders ME. Definitive and differential diagnosis of desquamative gingivitis through direct immunofluorescence studies. J Periodontol. 2012;83:1270-1278. 19. Mattsson T, Sundqvist KG, Heimdahl A, Dahllof G, Ljungman P, Ringden O. A comparative immunologic analysis of the oral mucosa in chronic graft-versus-host disease and oral lichen planus. Arch Oral Biol. 1992;37:539-547. 20. Neuchrist C, Grasl M, Scheiner O, Lassmann H, Ehrenberger K, Kraft D. Squamous cell carcinoma: infiltrating monocyte/macrophage subpopulations express functional mature phenotype. Br J Cancer. 1990;62:748-753. 21. Georgakopoulou EA, Achtari MD, Achtaris M, Foukas PG, Kotsinas A. Oral lichen planus as a preneoplastic inflammatory model. J Biomed Biotechnol. 2012;2012:759626. 22. Sugerman PB, Savage NW, Walsh LJ, et al. The pathogenesis of oral lichen planus. Crit Rev Oral Biol Med. 2002;13:350-365. 23. Sciubba JJ. Oral cancer. The importance of early diagnosis and treatment. Am J Clin Dermatol. 2001;2:239-251. 24. Muller S. Oral manifestations of dermatologic disease: a focus on lichenoid lesions. Head Neck Pathol. 2011;5:36-40.
25. Silverman S Jr, Gorsky M. Proliferative verrucous leukoplakia: a follow-up study of 54 cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1997;84:154-157. 26. Eisenberg E, Krutchkoff DJ. Lichenoid lesions of the oral mucosa: diagnostic criteria and their importance in the alleged relationship to oral cancer. Oral Surg Oral Med Oral Pathol. 1992;73:699-704. 27. Wenig BM. Intraoperative consultation (IOC) in mucosal lesions of the upper aerodigestive tract. Head Neck Pathol. 2008;2:131-144. 28. Lee JJ, Hung HC, Cheng SJ, et al. Factors associated with underdiagnosis from incisional biopsy of oral leukoplakic lesions. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;104: 217-225. 29. Jones K, Jordan RC. Patterns of second-opinion diagnosis in oral and maxillofacial pathology. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2010;109:865-869. Reprint requests: SG Fitzpatrick, DDS Department of Oral and Maxillofacial Diagnostic Sciences University of Florida College of Dentistry 1395 Center Drive, Gainesville Florida, USA 32610 sfi[email protected]fl.edu
-
No.
-
Month 2015
Direct immunofluorescence testing results in cases of premalignant and malignant oral lesions L.J. Montague, DMD,a I. Bhattacharyya, DDS, MSD,b M.N. Islam, DDS, BDS,b D.M. Cohen, DMD, MS, MBA,b and S.G. Fitzpatrick, DDSb Objectives. Oral premalignant and malignant lesions may mimic oral lichen planus (OLP) clinically and microscopically. OLP often shows basement membrane fibrinogen positivity on direct immunofluorescence testing (DIF). This study examined fibrinogen positivity in oral premalignant lesions and squamous cell carcinoma. Study Design. The University of Florida Oral Pathology Biopsy Service records were searched for the years 2003 to 2013 for oral premalignant lesions and squamous cell carcinoma with concurrent DIF testing. Demographic, clinical, and DIF or histologic information was collected and analyzed. Results. Sixty-eight fibrinogen positive lesions were identified within a total of 164 cases. Low-grade dysplasia and premalignant verrucous lesions made up the majority of the fibrinogen positive lesions (combined n ¼ 43; 63.2%), and the most common locations in positive cases were the buccal mucosa, tongue, and gingiva. A lichenoid distribution of the inflammatory infiltrate significantly predicted fibrinogen positivity (P < .0005). Conclusions. Fibrinogen positivity may be seen in premalignant and malignant oral lesions increasing the risk of misdiagnosis. (Oral Surg Oral Med Oral Pathol Oral Radiol 2015;-:1-9)
Within the oral cavity, significant clinical and histologic overlap may occur among oral lichenoid lesions (OLL), oral lichenoid mucositis (OLM) and oral lichen planus (OLP), which will be referred collectively throughout this article as OLP for simplicity, and premalignant lesions or squamous cell carcinoma (SCCA). Both premalignant and malignant oral lesions and OLP may occur clinically as white, red and white, or ulcerated oral lesions, and the phenomenon of premalignant and malignant oral lesions exhibiting classically lichenoid features has been well established histologically.1,2 This overlap may lead to a delay in the diagnosis and treatment of true dysplasia or SCCA of the oral cavity. Direct immunofluorescence (DIF) testing has been established as an adjunctive diagnostic tool in differentiating OLP from similar appearing lesions.3-6 Deposits of immunoreactants in OLP with “shaggy” basement membrane zone (BMZ) fibrinogen reactivity has been described as the best DIF indicator of the disease.5 However, fibrinogen positivity at the BMZ is not specific to OLP and has been reported in other inflammatory disorders.7 Because of the potential for clinicopathologic overlap between OLP and oral dysplasia or SCCA, DIF is Presented as an oral essay at the 2014 American Academy of Oral and Maxillofacial Pathology Annual Meeting in St. Augustine, Florida, USA. a Department of Oral and Maxillofacial Surgery and Pathology, University of Mississippi Medical Center, Jackson, MS, USA. b Department of Oral and Maxillofacial Diagnostic Sciences, University of Florida College of Dentistry, Gainesville, Florida, USA. Received for publication Dec 17, 2014; returned for revision Feb 19, 2015; accepted for publication Feb 20, 2015. Ó 2015 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2015.02.478
occasionally requested by the clinician for the purpose of differentiating OLP from other immune-mediated diseases, oral dysplasia, and SCCA, which may exhibit overlapping clinical and histologic features. Minimal information is found in the literature with regard to the DIF staining patterns of oral dysplasia or SCCA; thus, fibrinogen positivity on biopsy may contribute to diagnostic confusion in certain cases, especially those in which the level of premalignant features are mild or reactive. The objectives of this study were to characterize the DIF results, in particular positive fibrinogen reactivity, in cases of oral premalignant lesions and SCCA in biopsy specimens, which were incidentally submitted for DIF testing because of clinical lichenoid appearance, and to characterize the clinical and histologic patterns. The aim of the study was to prove that fibrinogen positivity may be commonly found in premalignant and malignant oral lesions.
MATERIALS AND METHODS Following institutional review board approval, the University of Florida Oral Pathology Biopsy service records were searched for the years 2003 to 2013 for
Statement of Clinical Relevance Direct immunofluorescence findings may be positive in premalignant and malignant lesions, contributing to the confusion with oral lichen planus especially where overlapping clinical and histologic similarities are seen. Clinicopathologic correlation is essential in discriminating between these entities. 1
ORAL AND MAXILLOFACIAL PATHOLOGY 2 Montague et al.
OOOO Month 2015
Fig. 1. Inflammatory pattern. A, minimal inflammation, grade 0 (hematoxylin&eosin, original magnification 10); B, diffuse lichenoid, grade 1 (hematoxylin&eosin, original magnification 10); C, focal lichenoid, grade 2 (hematoxylin&eosin, original magnification 10); D, nonlichenoid, grade 3 (hematoxylin&eosin, original magnification 10).
cases with specific diagnostic codes that also had DIF studies performed. These diagnostic codes included verrucopapillary hyperkeratosis (VPHK), dysplasia, atypical squamous epithelial proliferation (AEP), SCCA, and verrucous carcinoma (VC). Pathology reports were reviewed for all cases containing these codes, as well as age, gender, diagnosis, DIF findings, and biopsy location, were recorded. Archival hematoxylin and eosin (H&E) prepared slides were re-examined by two study participants (LM and SF) to confirm the sign-out diagnosis. All cases were originally examined and signed out by three of the study participants (IB, DC, and MI). Four of the study participants (IB, DC, MI, and SF) were board-certified oral and maxillofacial pathologists at the time of the study and the remaining study participant (LM) was at the time a senior oral and maxillofacial pathology resident. Cases in which the searched diagnosis and DIF were from different anatomic locations were excluded from the study. The cases were divided into six diagnostic categories: (1) VPHK (the term utilized at our institution for lesions compatible with the hyperkeratosis to early verrucous hyperplasia stages of proliferative verrucous leukoplakia, showing verrucoid thickening of the epithelium and orthokeratin or parakeratin layers),8 (2) low-grade dysplasia (LGD; atypia or dysplasia to mild dysplasia), (3) high-grade dysplasia (HGD; moderate to severe dysplasia), (4) AEP (defined as lesions exhibiting architectural features of malignancy but not meeting criteria for frank carcinoma), (5) SCCA, and (6) VC. Cases which were diagnosed as dysplasia in addition to
VPHK were categorized on the basis of their degree of dysplasia, rather than being included in the VPHK category. During the slide review by two study participants (SF and LM), the case was reassigned to the most appropriate category if the original diagnosis rendered was borderline between two of the designated diagnostic categories of the study. If no general agreement were to be found between the original sign-out diagnoses and those of the diagnostic reviewers, the case was to be dropped, although no cases fell into this category within this study. Multiple cases from the same patient were only included if the biopsies were from different sites within the oral cavity. If more than one biopsy was found for the same patient at the same site, the initial biopsy was included in the study, and subsequent biopsies were excluded in order to avoid counting the same lesion multiple times. Since the original slides of frozen DIF sections were not available for review (except for five cases), the DIF findings were recorded on the basis of the pathology report sign-out. The original DIF was performed at the in-house laboratory of the University of Florida College of Dentistry oral pathology biopsy service, utilizing frozen tissues cut into 4-micron sections and air dried, acetone fixed, washed with buffer, treated with protein block (Dako North America, Inc., Carpinteria, CA), incubated with antisera conjugated immunoglobulin G (IgG), IgM, IgA, C3, and fibrinogen (Thermo Scientific, Lab Vision Corporation, Fremont, CA), and then rinsed with distilled water and coverslipping.
OOOO Volume -, Number
ORIGINAL ARTICLE Montague et al. 3
-
Fig. 2. Fibrinogen reactivity on direct immunofluorescence. A/D, Verrucopapillary hyperkeratosis (A, hematoxylin&eosin, original magnification 10; D, shaggy fibrinogen along BMZ, original magnification 20). B/E, High-grade (moderate) dysplasia (B, hematoxylin&eosin, original magnification 10; E, linear fibrinogen at BMZ, original magnification 20). C/F, Atypical epithelial proliferation (C, hematoxylin&eosin, original magnification 10; F, linear fibrinogen at BMZ, original magnification 20). BMZ, basement membrane zone.
Additionally, all H&E slides for the cases included in the study were examined for the pattern of inflammation. A scale consisting of four different grades was created, and each case was assigned one of the following grades after consensus review by two reviewers (LM and SF): grade 0, minimally inflamed (noninflamed or inflammatory infiltrate present in less than 20% of tissue examined); grade 1, diffuse lichenoid pattern (bandlike inflammatory infiltrate present in greater than 50% of tissue examined); grade 2, focal lichenoid pattern (bandlike inflammatory infiltrate present in 20% -50% of tissue examined); and grade 3, nonlichenoid pattern, including ulcerated specimens, perivascular inflammation and patchy or patternless inflammation. Examples of the different patterns of inflammation can be seen in Figure 1. The data were analyzed statistically using SPSS version 22 (SPSS IBM, New York, NY). Differences between the fibrinogen-positive and fibrinogen-negative groups for age were analyzed by using independent samples t test. The chi-square test for independence was used to analyze the differences between the groups for all other variables and significance was set at the P < .05 level. The 18 cases of SCCA were excluded from statistical analysis of inflammatory pattern, since the presence of a lichenoid inflammatory pattern could not be
Table I. Demographic characteristics for all cases
Female Male Mean age (years)
Total cases n ¼ 164 (%)
Fibrinogen positive n ¼ 68 (%)
Fibrinogen negative n ¼ 96 (%)
105 (64) 59 (36) 64.9
45 (66.2)* 23 (33.8)* 65.8y
60 (62.5)* 36 (37.5)* 64.2y
*P ¼ .63 (chi-square test), no significant differences between positive and negative groups. y P ¼ .16 (t test), no significant differences between positive and negative groups.
determined in areas of invasion when the connective tissue was extensively replaced by tumor.
RESULTS DIF staining and demographic characteristics A total of 164 cases that met the inclusion criteria for the study were identified. Of these, 68 (41.5%) cases demonstrated positive staining for fibrinogen, and 96 cases (58.5%) were negative. Five fibrinogen-positive cases also demonstrated C3 positivity; 3 HGD and 1 LGD cases showed weak, granular staining along the BMZ (and 1 LGD case was C3 positive around the blood vessels). Three total fibrinogen negative cases also showed positive C3 staining; 1 HGD case stained
ORAL AND MAXILLOFACIAL PATHOLOGY 4 Montague et al.
OOOO Month 2015
Fig. 3. Fibrinogen reactivity on direct immunofluorescence. A/B, Verrucous carcinoma (A, hematoxylin&eosin, original magnification 10; B, shaggy fibrinogen at BMZ along tips of rete ridges, original magnification 20). C/D, Squamous cell carcinoma (C, hematoxylin&eosin, original magnification 10; D, focal linear fibrinogen at BMZ, original magnification 20). BMZ, basement membrane zone.
for C3 along the BMZ, 1 LGD case was C3 positive along the BMZ and around the blood vessels, and 1 SCCA case showed C3 staining only around the blood vessels. All 164 cases were negative for IgG, IgA, or IgM. Additionally, 5 of the fibrinogen-positive cases and 6 of the fibrinogen-negative cases exhibited superficial candidal colonization. The histopathologic features and fibrinogen reactivity of available cases are shown in Figures 2 and 3. The demographic features of the cases may be seen in Table I. The total cases (n ¼ 164) were composed of 105 females (64.0%) and 59 (36.0%) males. The patient age ranged from 23 to 94 years, with a mean of 64.9 years, and in one case the age was unknown. The fibrinogen-positive group (n ¼ 68) was made up of 45 (66.2%) females and 23 (33.8%) males, whereas the fibrinogen-negative group (n ¼ 96) consisted of 60 females (62.5%) and 36 (37.5%) males. The mean ages for the positive and negative groups were 65.8 years and 64.2 years, respectively. There were no statistically significant differences between the fibrinogen-positive and fibrinogen-negative groups for either gender (P ¼ .63) or age (P ¼ .16). Distribution by diagnosis The histologic diagnoses for the cases, as defined previously, are shown in Figure 4. The total diagnoses
(n ¼ 164) included 56 cases of VPHK (34.1%), 53 cases of low-grade dysplasia (32.3%), 23 cases of high-grade dysplasia (14%), 18 cases of SCCA (11%), 8 cases of AEP (4.9%), and 6 cases of VC (3.7%). The most common diagnosis among the fibrinogen-positive group was VPHK (n ¼ 22; 32.4%), followed by LGD (n ¼ 21; 30.9%). A total of 8 fibrinogen-positive cases of SCCA and 3 VC were identified. The majority of negative cases were diagnosed as VPHK (n ¼ 34; 35.4%) and low-grade dysplasia (n ¼ 32, 33.3%). Thirteen fibrinogen-negative carcinomas, including 10 SCCA and 3 VC, were identified. The difference between the fibrinogen-positive and fibrinogen-negative groups for diagnosis was not significant (P ¼ .97). The clinical and histopathologic features of cases from each diagnostic category are shown in Figure 5. Distribution by site Figure 6 demonstrates the distribution of the cases by anatomic location. The buccal mucosa or vestibule (n ¼ 56; 34.1%) and tongue (n ¼ 55; 33.5%) were the most common sites overall, followed by the gingiva or alveolar ridge (n ¼ 36; 22%), palate (n ¼ 9; 5.5%), floor of mouth (n ¼ 5; 3.0), “oral cavity” (n ¼ 2; 1.2%), and lip vermilion (n ¼ 1; 0.6%). The majority of fibrinogen-positive lesions occurred on the buccal mucosa or vestibule (n ¼ 22; 32.4%), tongue (n ¼ 21;
OOOO Volume -, Number
-
ORIGINAL ARTICLE Montague et al. 5
Fig. 4. Distribution by diagnosis.
30.9%), and gingiva (n ¼ 19; 27.9%). For the negative cases, the most common sites were the buccal mucosa or vestibule (n ¼ 34; 35.4%) and tongue (n ¼ 34; 35.4%). There was no statistically significant difference (P ¼ .31) between the positive and negative groups based on the site within the oral cavity. Distribution by inflammatory pattern The distribution of positive and negative cases based on the pattern of inflammation is provided in Figure 7. The most common inflammatory pattern among all cases (n ¼ 146; 18 SCCA excluded) was diffuse lichenoid (n ¼ 74; 50.7 %), followed by minimally inflamed (n ¼ 36; 24.7%), nonlichenoid (n ¼ 20; 13.7%), and focal lichenoid (n ¼ 16; 11%) patterns. For the fibrinogen-positive cases (n ¼ 60), a diffuse lichenoid inflammatory pattern was seen in 44 cases (73.3%), with an additional 7 cases (11.7%) demonstrating a focal lichenoid pattern, and with only 9 cases showing minimal (n ¼ 4) or nonspecific (n ¼ 5) inflammation. The most common pattern of inflammation for the negative cases (n ¼ 86) was minimally inflamed pattern (n ¼ 32; 37.2%) followed by diffuse lichenoid pattern (n ¼ 30; 34.9%). The differences in the pattern of inflammation for the fibrinogen-positive and fibrinogen-negative groups were statistically significant (P < .0005).
DISCUSSION Direct immunofluorescence testing measures immunoreactant activity through the addition of fluoresceinated antibodies directly to a frozen biopsy specimen.9 One immunoreactant commonly tested is fibrinogen, a plasma protein important to the wound healing process.10 In cutaneous lichen planus, DIF has been shown to be
highly variable, with positivity in 37% to 97% of cases, and manifests as BMZ fibrinogen positivity or cytoid body reactivity to any immunoreactant.9 Within the oral cavity, previous studies have shown the sensitivity of DIF for OLP between 48% and 66%.5,9,11-14 BMZ fibrinogen positivity may be seen in other disorders, such as systemic lupus erythematosus and erythema multiforme, as well.7 In other mucosal sites, such as the conjunctiva, fibrinogen positivity has been found to be normal to the location and has led to erroneous diagnoses of lichen planus in some cases.15 In a large study of 500 oral mucosal lesions by Helander and Rogers,5 DIF results were shown to be more useful for other oral vesiculobullous diseases, such as mucous membrane pemphigoid and pemphigus vulgaris, than for lichen planus, and the sensitivity of clinical examination and H&E staining has been shown to be higher than DIF testing for OLP. In addition, perilesional DIF positivity was found to be similar to that for locations distant to lesional tissue in OLP cases, showing that the immune response may be widespread despite the lesion being localized.11 Direct immunofluorescence fibrinogen positivity in premalignant and malignant lesions in the oral cavity has been largely unreported in the literature. Schiodt et al. first reported 30 cases of clinical leukoplakia tested with DIF and found 2 cases of fibrinogenpositive lesions.16 Later reports, including 2 cases of SCCA and 6 dysplastic lesions of the oral cavity reported by Rinaggio et al. and one case each of SCCA and atypia reported by Suresh et al., were all fibrinogen negative on DIF.17,18 Oral lichen planus and oral SCCA have both been associated with a predominantly T-cell lymphocytic inflammatory response.19,20 In OLP, this T-cell response is
ORAL AND MAXILLOFACIAL PATHOLOGY 6 Montague et al.
OOOO Month 2015
Fig. 5. Clinical and histopathologic features of fibrinogen positive cases. A/D, Verrucopapillary hyperkeratosis (hematoxylin&eosin, original magnification 10). B/E, Low-grade (mild) dysplasia (hematoxylin&eosin, original magnification 10). C/F, High-grade (moderate) dysplasia (hematoxylin&eosin, original magnification 10). G/J, Atypical epithelial proliferation (hematoxylin&eosin, original magnification 10). H/K, Verrucous carcinoma (hematoxylin&eosin, original magnification 10). I/L, Squamous cell carcinoma (hematoxylin&eosin, original magnification 10).
thought to lead to the destruction of the basement membrane.21,22 In nonulcerated lesions of OLP, there is believed to be a counterbalancing increase in cellular proliferation that prevents the occurrence of ulceration.21 Similarly, in oral dysplasia and SCCA, either ulceration or proliferation of tissue may be a feature.23 These similarities between response mechanisms may contribute to the clinical overlap and potential for misdiagnosis. The presence of fibrinogen positivity in 41% of the premalignant, borderline, and malignant lesions in this study demonstrates that fibrinogen positivity itself cannot be considered evidence for the diagnosis of OLP
without consideration for the clinical and histologic context of the lesion. No statistical significance for positivity was found in our study for age or gender. In another recent study involving one of the authors (SF), the presence of histologic lichenoid features in oral premalignant and malignant lesions was found to be similarly nonsignificant for age or gender.2 One pattern that emerged in our study was within the distribution of fibrinogen positivity by oral subsite. Although no statistically significant differences were found among the sites, the gingiva was the only site in which the positive cases outnumbered the negative
OOOO Volume -, Number
-
ORIGINAL ARTICLE Montague et al. 7
Fig. 6. Distribution by site.
Fig. 7. Distribution by inflammatory pattern.
cases. Interestingly, this phenomenon has also been reported by Helander and Rogers,5 who found that DIF sensitivity in OLP varies greatly by site, with the gingiva showing the best sensitivity to DIF. Sano et al.11 studied 79 cases of OLP and also found variability in fibrinogen positivity based on site. This study found no statistically significant difference in positivity when comparing histologic diagnoses, but this may have been caused by the sample numbers. Since invasive carcinoma has less clinical similarity to OLP compared with low-grade dysplasia, fewer biopsy specimens were submitted
with a request for DIF in this study, with borderline or malignant lesions representing only 15% (n ¼ 24, including atypical squamous epithelial proliferations) of the total biopsies included. Therefore, small numbers of malignant lesions may have prevented the emergence of a pattern. Low-grade dysplasia and VPHK are especially problematic in terms of potential for misdiagnosis as lichenoid lesions. Verrucous hyperplasia, which is a premalignant entity typified by thickened white clinical lesions sometimes appearing multifocally (proliferative verrucous leukoplakia), and VPHK typically show minimal
ORAL AND MAXILLOFACIAL PATHOLOGY 8 Montague et al.
cellular signs of dysplasia and have often been shown to have lichenoid features both clinically and histologically.8,24,25 In the present study, the only statistically significant predictor of DIF fibrinogen positivity was the presence of a histologic lichenoid (superficial bandlike lymphocytic) inflammatory pattern. The concept of “lichenoid dysplasia” was introduced in 1985, referring to the phenomenon of dysplastic lesions demonstrating lichenoid features, and since then, a lichenoid response to dysplasia has become an established phenomenon.1 The presence of fibrinogen positivity in cases of oral dysplasia and SCCA supports the lack of specificity of a lichenoid inflammatory pattern in these lesions. Although the bandlike inflammatory infiltrate seen in OLP may be perhaps the most easily recognizable feature of the disease, it may also represent a very nonspecific inflammatory response.26 Other histologic features of OLP include the presence of degenerative keratinocytes (civatte bodies), liquefactive degeneration of the basal cell layer, interface stomatitis, and sawtooth rete ridge formation.24 Several factors may increase the potential for misdiagnosis in lesions that have lichenoid histologic features. If specimens are submitted exclusively for DIF or if the DIF specimen and the formalin biopsy are submitted to separate laboratories, the only H&E-stained specimen available for review may be on frozen section. Tissue distortion and artifact present on frozen sections create diagnostic difficulties, which could lead to potential misdiagnosis, especially in small specimens and in cases of low-grade dysplasia.27 In addition, small incisional biopsies may not be representative of the overall sample, and separation of the epithelium from underlying connective tissue may lead to inconclusive or false-positive DIF findings.5 In the oral cavity, incisional biopsies have been found to underdiagnose conditions in nearly 30% of cases compared with resections.28 Finally, diagnosis by pathologists unfamiliar with the subtle nature of many oral premalignant conditions may pose a problem, considering that oral and maxillofacial pathology may represent an area of variable experience for many general pathologists. In a study of consultations to oral and maxillofacial pathologists by general pathologists, it was found that the most common need for second opinions was for oral dysplasia or squamous cell carcinoma.29
CONCLUSIONS The results of this study show the frequent occurrence of fibrinogen positivity on DIF in premalignant and malignant lesions of the oral cavity. Pathologists should be aware of this phenomenon and avoid the potential for misdiagnosis of these lesions, especially in groups with milder dysplastic change. Fibrinogen positivity on
OOOO Month 2015
DIF should not be considered pathognomonic for OLP in premalignant or borderline malignant cases, and clinical and histopathologic context, rather than DIF, should be given higher consideration in these cases. The possibility of an immune response presenting with a lichenoid pattern of inflammation to a dysplastic lesion should not be overlooked, and second opinion should be sought for confounding lesions. The authors wish to thank Douglas Damico for technical assistance on the description of the direct immunofluorescence methodology utilized.
REFERENCES 1. Krutchkoff DJ, Eisenberg E. Lichenoid dysplasia: a distinct histopathologic entity. Oral Surg Oral Med Oral Pathol. 1985;60: 308-315. 2. Fitzpatrick SG, Honda KS, Sattar A, Hirsch SA. Histologic lichenoid features in oral dysplasia and squamous cell carcinoma. Oral Surg Oral Med Oral Pathol Oral Radiol. 2014;117: 511-520. 3. Firth NA, Rich AM, Radden BG, Reade PC. Assessment of the value of immunofluorescence microscopy in the diagnosis of oral mucosal lichen planus. J Oral Pathol Med. 1990;19:295-297. 4. Kulthanan K, Jiamton S, Varothai S, Pinkaew S, Sutthipinittharm P. Direct immunofluorescence study in patients with lichen planus. Int J Dermatol. 2007;46:1237-1241. 5. Helander SD, Rogers RS 3rd. The sensitivity and specificity of direct immunofluorescence testing in disorders of mucous membranes. J Am Acad Dermatol. 1994;30:65-75. 6. Toto PD, Nadimi HT. An immunohistochemical study of oral lichen planus. Oral Surg Oral Med Oral Pathol. 1987;63:60-67. 7. Black MM. What is going on in lichen planus? Clin Exp Dermatol. 1977;2:303-310. 8. Gillenwater AM, Vigneswaran N, Fatani H, Saintigny P, ElNaggar AK. Proliferative verrucous leukoplakia: recognition and differentiation from conventional leukoplakia and mimics. Head Neck. 2014;36:1662-1668. 9. Chhabra S, Minz RW, Saikia B. Immunofluorescence in dermatology. Indian J Dermatol Venereol Leprol. 2012;78:677-691. 10. Halper J, Kjaer M. Basic components of connective tissues and extracellular matrix: elastin, fibrillin, fibulins, fibrinogen, fibronectin, laminin, tenascins and thrombospondins. Adv Exp Med Biol. 2014;802:31-47. 11. Sano SM, Quarracino MC, Aguas SC, et al. Sensitivity of direct immunofluorescence in oral diseases. Study of 125 cases. Med Oral Patol Cir Bucal. 2008;13:E287-E291. 12. Rogers RS 3rd, Jordon RE. Immunopathology of oral mucosal inflammatory diseases. Clin Exp Dermatol. 1977;2:97-107. 13. Rogers RS 3rd, Van Hale HM. Immunopathologic diagnosis of oral mucosal inflammatory diseases. Australas J Dermatol. 1986;27:51-57. 14. Van Hale HM, Rogers RS 3rd. Immunopathology of oral mucosal inflammatory diseases. Dermatol Clin. 1987;5:739-750. 15. Mehta M, Siddique SS, Gonzalez-Gonzalez LA, Foster CS. Immunohistochemical differences between normal and chronically inflamed conjunctiva: diagnostic features. Am J Dermatopathol. 2011;33:786-789. 16. Schiodt M, Holmstrup P, Dabelsteen E, Ullman S. Deposits of immunoglobulins, complement, and fibrinogen in oral lupus erythematosus, lichen planus, and leukoplakia. Oral Surg Oral Med Oral Pathol. 1981;51:603-608.
OOOO Volume -, Number
ORIGINAL ARTICLE Montague et al. 9
-
17. Rinaggio J, Crossland DM, Zeid MY. A determination of the range of oral conditions submitted for microscopic and direct immunofluorescence analysis. J Periodontol. 2007;78:19041910. 18. Suresh L, Neiders ME. Definitive and differential diagnosis of desquamative gingivitis through direct immunofluorescence studies. J Periodontol. 2012;83:1270-1278. 19. Mattsson T, Sundqvist KG, Heimdahl A, Dahllof G, Ljungman P, Ringden O. A comparative immunologic analysis of the oral mucosa in chronic graft-versus-host disease and oral lichen planus. Arch Oral Biol. 1992;37:539-547. 20. Neuchrist C, Grasl M, Scheiner O, Lassmann H, Ehrenberger K, Kraft D. Squamous cell carcinoma: infiltrating monocyte/macrophage subpopulations express functional mature phenotype. Br J Cancer. 1990;62:748-753. 21. Georgakopoulou EA, Achtari MD, Achtaris M, Foukas PG, Kotsinas A. Oral lichen planus as a preneoplastic inflammatory model. J Biomed Biotechnol. 2012;2012:759626. 22. Sugerman PB, Savage NW, Walsh LJ, et al. The pathogenesis of oral lichen planus. Crit Rev Oral Biol Med. 2002;13:350-365. 23. Sciubba JJ. Oral cancer. The importance of early diagnosis and treatment. Am J Clin Dermatol. 2001;2:239-251. 24. Muller S. Oral manifestations of dermatologic disease: a focus on lichenoid lesions. Head Neck Pathol. 2011;5:36-40.
25. Silverman S Jr, Gorsky M. Proliferative verrucous leukoplakia: a follow-up study of 54 cases. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 1997;84:154-157. 26. Eisenberg E, Krutchkoff DJ. Lichenoid lesions of the oral mucosa: diagnostic criteria and their importance in the alleged relationship to oral cancer. Oral Surg Oral Med Oral Pathol. 1992;73:699-704. 27. Wenig BM. Intraoperative consultation (IOC) in mucosal lesions of the upper aerodigestive tract. Head Neck Pathol. 2008;2:131-144. 28. Lee JJ, Hung HC, Cheng SJ, et al. Factors associated with underdiagnosis from incisional biopsy of oral leukoplakic lesions. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2007;104: 217-225. 29. Jones K, Jordan RC. Patterns of second-opinion diagnosis in oral and maxillofacial pathology. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2010;109:865-869. Reprint requests: SG Fitzpatrick, DDS Department of Oral and Maxillofacial Diagnostic Sciences University of Florida College of Dentistry 1395 Center Drive, Gainesville Florida, USA 32610 sfi[email protected]fl.edu
