Diagnostic significance of colloid body deposition in direct immunofluorescence
Department of Dermatology, Faculty of Medicine, Siriraj Hospital, Mahidol University, 2 Prannok Road, Bangkoknoi, Bangkok-107 00
|How to cite this article:
Chularojanamontri L, Tuchinda P, Triwongwaranat D, Pinkaew S, Kulthanan K. Diagnostic significance of colloid body deposition in direct immunofluorescence. Indian J Dermatol Venereol Leprol 2010;76:373-377
AbstractBackground: Colloid bodies (CB) in direct immunofluorescence (DIF) studies are usually found in interface dermatitis. Furthermore, CB can be found in various skin diseases and even in normal skin. Aim: To evaluate the diagnostic value of CB deposits in DIF studies. Methods: From 1996-2007, data from 502 patients where DIF studies showed immunoreactants at CB were enrolled. The definite diagnoses of these patients were based on clinical, histopathological and immunofluorescent findings. The results of DIF studies were analyzed. Results: Immunoreactants at CB were detected in 44.4%, 43.8%, 4.2%, 3.8%, and 2.2% of interface dermatitis, vasculitis, autoimmune vesiculobullous disease, panniculitis, and scleroderma/morphea, respectively. The most common immunoreactant deposit of all diseases was Immunoglobulin M (IgM). Brighter intensity and higher quantity of CB was detected frequently in the group with interface dermatitis. Conclusions: Immunoreactant deposits at CB alone can be found in various diseases but a strong intensity and high quantity favor the diagnosis of interface dermatitis. CB plus dermoepidermal junction (DEJ) deposits are more common in interface dermatitis than any other disease. Between lichen planus (LP) and discoid lupus erythematosus (DLE), CB alone is more common in LP; whereas, CB plus DEJ and superficial blood vessel (SBV) is more common in DLE. The most common pattern in both diseases is CB plus DEJ. The quantity and intensity of CB in LP is higher than in DLE.
Colloid bodies (CB) are eosinophilic hyaline ovoid bodies which are often found in the subepidermal papillary regions or sometimes in the epidermis. They are usually seen in lichen planus (LP) and lupus erythematosus (LE).  They can also be found in several dermatoses such as erythema multiforme (EM), bullous pemphigoid (BP) and diseases with suprabasal clefts.  Colloid bodies, also known as civatte bodies or cytoid bodies, are generally believed to be derived from two origins. The first type originates from apoptosis of keratinocytes causes by epithelium damage created by circulating disorders.  CB of this type are usually found locally both in the epidermis and papillary dermis. The other origin derives from the destruction of thickened basement membranes which are found only in the papillary dermis. 
Immunoglobulin (Ig) subclasses of CB: IgG, IgA, IgM, complement components (C3) and fibrinogen, can be demonstrated by direct immunofluorescence (DIF).  According to previous studies, IgM-CB is the most common immunoreactant deposit in LP, ,,, BP,  and EM.  It is still unclear whether or not CB would provide any additional on diagnostic value for overlapping skin diseases. The aim of this descriptive study was to demonstrate a possible correlation between the Ig subclasses of CB and their diagnostic yield.
Data were collected from patients who visited the Department of Dermatology, Siriraj Hospital, Mahidol University, Bangkok, Thailand from 1996-2007. Cutaneous biopsies in which DIF studies showed immunoreactants at CB were enrolled. The definite diagnoses of these patients were based on clinical, histopathological and immunofluorescent findings. DIF studies of skin lesions were performed according to the standard method described previously.  Briefly, skin biopsy specimens were embedded in Cryomatrix embedding medium (Shandon Lipshaw, Inc, Pittsburgh, PA, USA) and snap frozen at -70Ίc until sectioned. Frozen sections, 4-΅m thick, were cut on a cryostat, air dried, washed twice with phosphate-buffered saline (PBS), pH 7.4, for 10 min, and overlain for 30 min with fluorescein isothiocyanate-conjugated rabbit antihuman IgG, IgA, IgM, C3 and fibrinogen (Dako patt., Copenhagen, Denmark). Thereafter, section slides were incubated in a humidified chamber at room temperature, washed twice with PBS for 10 min, and mounted with a medium before being viewed under a fluorescent microscope. DIF patterns were interpreted according to the standard criteria.  The number of CB were counted in a high-power field (HP) (Χ400) and recorded as a few (≤5/ HP), many (6-10/ HP) or numerous (≥11/ HP). The intensity was graded in three levels (1+, 2+, and 3+).
Our study was approved by the Siriraj Institutional Review Board, Mahidol University.
Comparisons between subgroups of patients according to factors of prognostic value were performed using Chi-square test or Fisher exact test for categorical variables.
Five hundred and two patients (375 females and 127 males), age 6-87 years, were included in this study. The final diagnoses were classified into five groups including interface dermatitis, vasculitis, autoimmune vesiculobullous diseases, panniculitis and scleroderma/morphea as shown in [Table - 1]. Interface dermatitis comprises LE, LP, EM and dermatomyositis. In the LE group, the diagnoses included LE-specific skin lesions and LE-non-specific skin lesions.
[Table - 2] shows the distribution of immunoreactants in discoid lupus erythematosus (DLE), LP and EM. We included only DLE (61 cases), which is the prototype of interface dermatitis, to analyze the significance of CB in this group. The data of dermatomyositis were not used in the analysis because there were only two cases. CB deposit in addition to dermoepidermal junction (DEJ) deposit [Figure - 1] was the most common pattern in DLE and LP; whereas, CB deposit alone was the most common form in EM. IgM was the most common class of Ig deposited at CB in DLE, LP and EM in all patterns. The brightest intensity of IgM was detected in 37.9% of LP, followed by 26.2% of DLE and 20% of EM. The highest quantity of CB was detected in LP, in all classes of Ig except C3 [Table - 3].
[Table - 4] shows details of the immunoreactant deposit at the various locations in other groups. CB deposit alone was found in panniculitis, scleroderma/morphea, periarteritis nodosa (PAN), and leukocytoclastic vasculitis (LCV) in 68.4%, 54.5%, 26.1%, 2.5% of cases, respectively; whereas, it was not detected in Henoch-SchΦnlein purpura (HSP. Immunoreactants at CB plus DEJ was more commonly found in interface dermatitis (77/154 = 50%) than vasculitis (2/220 = 0.9%), pannicultis (5/19 = 26.3%), or scleroderma/morphea (3/11 = 27.3%) (P < 0.05).
The most common immunoreactant deposit at CB in all diseases was IgM.
The brightest intensity of IgM (3+) was found in 43/154 cases (27.9%), 10/220 cases (4.5%), and 1/21 cases (4.8%), of interface dermatitis, vasculitis and autoimmune vesiculobullous disease, respectively. Moreover, there was no intensity grading of 3+ in panniculitis and scleroderma/morphea. Regarding the quantity of CB, the interface dermatitis group had the highest amount when compared to other groups (vasculitis, autoimmune vesiculobullous diseases, panniculitis and scleroderma/morphea) (data not shown in the Table).
Our study demonstrated immunoreactant deposit at CB in a wide variety of diseases, which is in agreement with previous studies. ,,,,,, Previous studies proposed that CB was usually detected in interface dermatitis and might increase the index of suspicion for LE and LP.  Interestingly, our study showed that CB was commonly detected in interface dermatitis and vasculitis. However, the intensity and the number of CB in interface dermatitis were higher than those of vasculitis.
In LE and LP, which sometimes are an overlapping syndrome and difficult to differentiate, DIF can be used as an additional diagnostic clue. In our study, CB alone was more common in LP than DLE [Table - 2]. More numerous CB were detected in LP than DLE, which supports the claim that CB in LP have a tendency to cluster in groups of 10 or more in the papillary dermis. ,, A brighter intensity of IgM was also noted.
In cases that CB deposits were detected in combination with deposition at the other sites, the combination of CB with DEJ was the most common pattern in patients with LE and LP. Previous studies reported that fibrinogen was the most common immunoreactant deposit at DEJ in LP at a high percentage (91-100%). ,,, Igs (IgG, IgA, IgM and C3) deposited at DEJ ranged from 16-47%.  Kulthanan et al. reported that IgG was the most common immunoreactant at DEJ in DLE.  The second most common deposit at the DEJ was IgM, which tended to exhibit a strong intensity in association with LE. ,, Considering all the results, we suggest that any immunoreactant deposit at CB plus fibrinogen deposition, whether alone or combined with other immunoreactants at DEJ, favors LP. Meanwhile, the deposition of any immunoreactant at CB plus IgG or intense IgM at DEJ, favors LE. Among cicatricial alopecia caused by LP, LE, lichen planopilaris (LPP) and Pseudopelade of Brocq (PB), DIF findings in LP consisted of IgM deposit at CB and fibrinogen deposit at the DEJ; whereas, Ig deposit at the DEJ in a granular pattern was commonly found in LE.  Merheregan et al. reported that the DIF studies in LPP, follicular lichen planus of the scalp, and LP were similar.  With regard to PB, DIF often showed negative results or occasionally demonstrated IgM at the DEJ. Thus, CB plus different immunoreactants′ deposit at the DEJ is helpful in the diagnosis of LP, LE and cicatricial alopecia caused by LE, LPP and PB.
Finan et al. showed that the combination of CB plus DEJ was the most common DIF pattern in EM.  They correlated the findings of DIF and histopathologic findings. In the epidermal pattern, CB was present in all specimens, whereas in mixed patterns, approximately equal fluorescence of CB and DEJ were indicated. The EM in our study was of epidermal pattern in 91.4% (32/35). Thus, the most common pattern of our study was CB alone, which differs from the previous study by Finan et al. 
The quantity and intensity of CB in interface dermatitis was higher than in other groups. Surprisingly, immunoreactants′ deposit at CB alone was detected in more than 50% of panniculitis and scleroderma/morphea. CB plus DEJ was detected most commonly in the interface dermatitis group. Basically, the diagnoses of some diseases of the autoimmune vesiculobullous group require immunoreactants at DEJ. Therefore, we did not compare this pattern with autoimmune vesiculobullous diseases.
The most common immunoreactant deposit at CB was IgM, in our study. These findings were consistent with those reported by other investigators.  Thus, we concluded that the most common immunoreactant of CB might be IgM in all diseases.
Other diseases in our study which revealed CB deposits were eczema, Sweet′s syndrome, and pigmented purpuric dermatosis. Within our review, additional diseases reported by other studies to have CB deposits included lichen planopilaris,  desquamative gingivitis,  ashy dermatosis,  necrolytic migratory erythema,  elastosis perforans serpiginosa,  rheumatoid arthritis,  cutaneous amyloidosis,  pityriasis lichenoides et varioliformis acuta, , pernio,  ulcer,  actinic cheilitis,  acute generalized exanthematous pustulosis,  drug reaction with eosinophilia and systemic symptoms,  granulomatous rosacea,  lichen striatus,  prurigo pigmentosa,  pyoderma gangrenosum,  toxic epidermal necrolysis, , herpes zoster,  herpes simplex,  varicella,  dermatitis herpetiformis,  porphyria cutanea tarda,  sarcoidosis,  subcorneal pustular dermatosis,  keratosis follicularis Darier,  familial benign chronic pemphigus,  transient acantholytic dermatosis Grover,  epidermolytic hyperkeratosis,  polymorphic actinic eruption,  and even normal skin. 
In conclusion, we propose the diagnostic value of CB in DIF study as follows: (1) Immunoreactant deposits at CB alone can be found in various diseases but a strong intensity and high quantity favor interface dermatitis. (2) CB plus DEJ is more common in interface dermatitis than any other disease (not including autoimmune vesiculobullous disease). (3) Between LP and DLE, CB alone is more common in LP; whereas CB plus DEJ and superficial blood vessel (SBV) is more common in DLE. The most common pattern in both diseases is CB plus DEJ. The quantity and intensity of CB in LP is higher than in DLE.
The authors are grateful to Dr. Chulaluk Komoltri for helpful comments and Professor Punkae Mahaisavariya for all kinds of support.
Horiguchi Y, Danno K, Ikai K, Imamura S. Colloid body formation in bullous pemphigoid. Arch Dermatol Res 1985;277:167-73.[Google Scholar]
Sόmegi I. Colloid bodies in dermatoses other than lichen planus. Acta Derm Venereol 1982;62:125-31.[Google Scholar]
Ueki H. Hyaline bodies in subepidermal papillae Immunohistochemical studies in several dermatoses. Arch Dermatol 1969;100:610-7.[Google Scholar]
Abell E, Presbury DG, Marks R, Ramnarain D. The diagnostic significance of immunoglobulin and fibrin deposition in lichen planus. Br J Dermatol 1975;93:17-24.[Google Scholar]
Kulthanan K, Jiamton S, Varothai S, Pinkaew S, Sutthipinittharm P. Direct immunofluorescence study in patients with lichen planus. Int J Dermatol 2007;46:1237-41.[Google Scholar]
Laskaris G, Sklavounou A, Angelopoulos A. Direct immunofluorescence in oral lichen planus. Oral Surg Oral Med Oral Pathol 1982;53:483-7.[Google Scholar]
Wu YH, Lin YC. Cytoid bodies in cutaneous direct immunofluorescence examination. J Cutan Pathol 2007;34:481-6.[Google Scholar]
Finan MC, Schroeter AL. Cutaneous immunofluorescence study of erythema multiforme: correlation with light microscopic patterns and etiologic agents. J Am Acad Dermatol 1984;10:497-506.[Google Scholar]
Beutner EH, Kumar V , Krasny SA, Chorzelski TP. Defined immunofluorescence immunodermatology. In: Beutner EH, Chorzelski TP, Kumar V, editors. Immunopathology of the skin. 3 rd ed. New York: John Wiley and Sons; 1987. p. 3-40.[Google Scholar]
Kilpi AM, Rich AM, Radden BG, Reade PC. Direct immunofluorescence in the diagnosis of oral mucosal diseases. Int J Oral Maxillofac Surg 1988;17:6-10.[Google Scholar]
Kulthanan K, Roongphiboolsopit P, Chanjanakijskul S, Kullavanijaya P. Chronic discoid lupus erythematosus in Thailand: direct immunofluorescence study. Int J Dermatol 1996;35:711-4.[Google Scholar]
Camisa C, Neff JC, Olsen RG. Use of indirect immunofluorescence in the lupus erythematosus/lichen planus overlap syndrome: an additional diagnostic clue. J Am Acad Dermatol 1984;11:1050-9.[Google Scholar]
Daniels TE, Quadra-White C. Direct immunofluorescence in oral mucosal disease: a diagnostic analysis of 130 cases. Oral Surg Oral Med Oral Pathol 1981;51:38-47.[Google Scholar]
Schiψdt 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;51:603-8.[Google Scholar]
Kulthanan K, Pinkaew S, Suthipinittharm P. Diagnostic value of IgM deposition at the dermo-epidermal junction. Int J Dermatol 1998;37:201-5.[Google Scholar]
Isfer RS, Sanches Jϊnior JA, Festa Neto C, Cernea SS, de Oliveira ZN, Vilela MA, et al. Direct immunofluorescence in lupus erythematosus (LE). Sao Paulo Med J 1996;114:1141-7.[Google Scholar]
Weigand DA. Cutaneous immunofluorescence. Med Clin North Am 1989;73:1263-74.[Google Scholar]
Trachsler S, Trueb RM. Value of direct immunofluorescence for differential diagnosis of cicatricial alopecia. Dermatology 2005;211:98-102.[Google Scholar]
Mehregan DA, Van Hale HM, Muller SA. Lichen planopilaris: clinical and pathologic study of forty-five patients. J Am Acad Dermatol 1992;27:935-42.[Google Scholar]
Hosokawa M, Masu S, Seiji M. Immunofluorescence studies on civatte bodies and dyskeratotic cells with anti-keratin antibody. Tohoku J Exp Med 1981;135:219-29.[Google Scholar]
Laskaris G, Demetriou N, Angelopoulos A. Immunofluorescent studies in desquamative gingivitis. J Oral Pathol 1981;10:398-407.[Google Scholar]
Person JR, Rogers RS 3 rd . Ashy dermatosis. An apoptotic disease? Arch Dermatol 1981;117:701-4.[Google Scholar]
Perniciaro C, Rappaport KD, White JW Jr. Apoptosis with positive direct immunofluorescence findings in a patient with necrolytic migratory erythema. Cutis 1998;62:129-32.[Google Scholar]
Bergman R, Friedman-Birnbaum R, Hazaz B. A direct immunofluorescence study in elastosis perforans serpiginosa. Br J Dermatol 1985;113:573-9.[Google Scholar]
Ma AS, Soltani K, Bristol LA, Bernstein JE, Sorensen LB. Cutaneous immunofluorescence studies in adult rheumatoid arthritis in sun-exposed and non-sun-exposed areas. Int J Dermatol 1984;23:269-72.[Google Scholar]
MacDonald DM, Black MM, Ramnarain N. Immunofluorescence studies in primary localized cutaneous amyloidosis. Br J Dermatol 1977;96:635-41.[Google Scholar]
Blenkinsopp WK, Clayton RJ, Haffenden GP. Immunoglobulin and complement in normal skin. J Clin Pathol 1978;31:1143-6.[Google Scholar]