Comparative Histopathology of Pemphigus Foliaceus and Superficial Folliculitis in the Dog

Transcription

1 ~ Vet Pathol 31:19-27 (1994) Comparative Histopathology of Pemphigus Foliaceus and Superficial Folliculitis in the Dog K. A. KUHL, F. S. SHOFER, AND M. H. GOLDSCHMDT Section of Dermatology, Department of Clinical Studies, University of Pennsylvania, School of Veterinary Medicine, Philadelphia, PA (KAK); Emergency Services, Department of Medicine, University of Pennsylvania School of Medicine, Hospital of the University of Pennsylvania, Philadelphia, PA (FSS); and Laboratory of Pathology, Department of Pathobiology, University of Pennsylvania, School of Veterinary Medicine, Philadelphia, PA (MHG) Abstract. A comparative histopathologic study of cases of pemphigus foliaceus and 47 cases of superficial folliculitis in the dog was undertaken to identify those histopathologic features important in differentiating these diseases. All cases were diagnosed by the Surgical Pathology Service of the Laboratory of Pathology of the University of Pennsylvania, School of Veterinary Medicine, between January 1986 and March These cases were followed clinically to determine the accuracy of the initial histopathologic diagnosis. Data were analyzed by Fisher's exact test for discrete data and the Mann-Whitney U-test for continuous data. A probability of <0.05 was considered statistically significant. The following breeds were affected with pemphigus foliaceus: Akita, Brittany Spaniel, Chow Chow, Collie, Dachshund, Doberman Pinscher, English Cocker Spaniel, English Setter, Golden Retriever, Siberian Husky, Laborador Retriever, Miniature Schnauzer, Old English Sheepdog, Scottish Terrier, Chinese Shar Pei, Spitz, Shetland Sheep Dog, Weimeraner, and West Highland White Temer. An increased risk of developing pemphigus foliaceus was noted in the Akita (OR = 37.8), English Springer Spaniel (OR = 20.7), Chow Chow (OR = 12.3), Chinese Shar Pei (OR = 7.9), and Collie (OR = 3.9). Pemphigus foliaceus had a higher average density of acantholytic cells (226 k 22.9) than superficial folliculitis (1 1.8 * 4.6), and acantholytic cells were 183 times more likely to be present in pemphigus foliaceus. The presence of rafts of acantholytic cells occurred in pemphigus foliaceus (23/) more often than superficial folliculitis (1/47). Recornification and reformation of pustules were, respectively, 13.4 and 3.6 times more likely to occur in pemphigus foliaceus. Finally, a longer pustule that bridges multiple hair follicles (2.3 k 0.20 mm and 1.31 k 0.10 mm, respectively) was seen in pemphigus foliaceus. Key words: Acantholysis; pemphigus foliaceus; superficial folliculitis. Pemphigus foliaceus (PF) is an autoimmune disease in which antibodies bind to desmoglein,6,l8 a transmembrane desmosomal protein that mediates intercellular adhesi~n.~j This antigen has been characterized in humans, but not in animals. The antibody binding induces plasminogen activator release by the keratinocyte. This has been documented in pemphigus vulgaris in animals.26 The plasminogen activator subsequently activates plasminogen to plasmin. The plasmin then degrades cell surface proteins, resulting in cellular dyshesion and acantholysi~.~-*.~~.~~ Cellular dyshesion and the resultant acantholysis may also occur in superficial folliculitis (SF) as neutrophils migrate through the epidermis or follicular outer root sheath and release hydrolytic enzymes that disrupt the inter- cellular adhesion^.^,^' Therefore, SF, which is a common dermatologic disorder, can be difficult to differentiate from PF clinically and histopathological- 1y.9,12,15,23.25 Because PF and SF are treated differently, it is important that the histopathologic evaluation be accu- 19 rate. n the past, the criteria used to diagnose PF have been subjective, such as marked intragranular or subcorneal acantholysis with production of a cleft and vesicle or pustule formation; 1,5,6,9,10,12,13,15,21-23,25 stratum granulosum cells adherent to the overlying stratum corneum ("cling-on~");'.~.'~,~~.~~,~~ lichenoid cellular infiltrate;'5,2',22 eosinophilic exocytosis within the epidermis or follicular outer root sheath or b ~th;~~'~' involvement of the follicular outer root sheath in the acantholytic process;',9.13'15'21-23 and acantholytic dyskeratotic granular epidermal cells at the surface of the ero~ions.~.~~j~,~~ Therefore, the study reported here is to establish objective criteria that can be applied reliably to differentiate these two specific diseases. Materials and Methods Cases of PF and SF were ascertained from specimens submitted to the Surgical Pathology Service of the Laboratory of Pathology of the University of Pennsylvania, School of Veterinary Medicine, between January 1986 and March The specific diagnoses were based on the following histo-

2 20 Kuhl, Shofer, and Goldschrnidt Vet Pathol 3 :, 994 pathologic parameters: marked intragranular or subcorneal acantholysis with production of a cleft and vesicle or pustule formation, stratum granulosum cells adherent to the overlying stratum corneum ( cling-ons ), lichenoid cellular infiltrate, eosinophilic exocytosis within the epidermis or follicular outer root sheath or both, involvement ofthe follicular outer root sheath in the acantholytic process, and acantholytic dyskeratotic granular epidermal cells at the surface of the erosions. Only those cases that were submitted by the Section of Dermatology at the Veterinary Hospital of the University of Pennsylvania (VHUP) or by Dr. Barbara Kummel of the Animal Dermatology Clinic of Rockville, Maryland, were considered for inclusion in the study. Additionally, cases of PF and SF had to be followed clinically for a minimum of 2 months. n the cases of PF, the histopathologic diagnosis was verified by lack of response to antibiotics and response to treatment with immunosuppressive drugs, lo.3.s.2.zz Only of the 70 cases of PF reviewed met this criteria. n cases of SF, the original histopathologic diagnosis was verified by appropriate response to antibiotics.5 Forty-seven cases of SF were identified that met these criteria. The following histopathologic parameters were compared: density of acantholytic cells; rafts of acantholytic cells (three or more acantholytic cells attached to one another within the pustule) and cells from the stratum granulosum adherent to the roof of the pustule per square mm of superficial pustule; the density of erythrocytic, granulocytic and lymphocytic exocytosis per square mm of epidermis beneath the pustule; the density of acantholytic cells and rafts of acantholytic cells per square mm of pustule within hair follicles (lumina or external root sheaths); and the number of follicles bridged per mm of pustule length and the actual length (mm) of the pustules. The density measurements were made by first determining the area of the pustule (either superficial or follicular) in square mm. Then each of the cell types was counted within the pustule. This gave a cell per square mm of pustule or density measurement. The numbers of cells per square mm of epidermis directly beneath the pustule were determined similarly. The frequency measurements were determined by first measuring the length of the base of the pustule in mm. Then the number of hair follicles bridged by this pustule was counted, thus determining the frequency of hair follicles bridged per mm of pustule length. These parameters were determined on a video image analysis system at the Laboratory of Dermatopathology at the Hospital of the University of Pennsylvania (HUP). By evaluating numerous sections on the video image analysis system, we have attempted to remove any bias that would be produced by cutting through different cross sections of the pustule. Also evaluated were the percentage of intact neutrophils, intact eosinophils, and degenerated granulocytes within pustules; presence or absence of recornification and reformation of the pustules; and presence or absence of parakeratosis and orthokeratosis and presence or absence of follicular involvement (follicular lumen or external root sheath). Follicular pustules were defined as pustules present in either the follicular lumen or external root sheath or both, with no extension of the pustule onto the perifollicular epidermis. Superficial pustules were defined as subcorneal or intragranular pustules involving the epi- Fig. 1. Skin; high magnification of a subcorneal pustule containing single or sheets of acantholytic cells. HE. Bar = pm. dermis with or without extension to involve the follicular lumen or external root sheath. Recornification was defined as at least one layer of cornified epithelial cells re-forming along 75% of the base of the pustule. These parameters were measured on the diagnostic histopathologic sections from each disease although multiple biopsy samples may have been submitted from each PF and SF case. The continuous parameters from each section were then averaged for each case to determine a mean value for each parameter per case. Categorical parameters were considered to be present in a case if present in at least one of the sections from the case. To determine age, breed, or sex predilections, all PF histopathology submissions (n = 70) were compared to all biopsy specimens submitted to the Surgical Pathology Service during the same time period (n = 64,462). Odds ratios with 95% confidence limits were calculated for all breeds at risk ofdeveloping PF. Statistical significance, defined as P < 0.05, was determined by the Pearson s chi-square test. Only breeds with significantly increased risk of developing PF (OR > 1.O) were reported. Those breeds with a reduction in PF risk (OR < 1.O) were not reported. n addition, odds ratios were calculated only for breeds in which there were at least two cases of PF. Data were analyzed by standard methods for case-comparison studies. Odds ratios with 95% confidence limits were calculated for discrete data (e.g., presence of orthokera-

3 Vet Pathol 3l:l, 1994 Pemphigus Foliaceus and Superficial Folliculitis 21 Table 1. Comparison of the density of acantholytic cells, rafts of acantholytic cells, cellular exocytosis, and frequency of follicles bridged per mm of pustule length in pemphigus foliaceus and superficial folliculitis. Pemphigus Foliaceus Histopathologic Meant Parameters n* (range) n* Acantholytic 226 f cells/mm2q (15-794) Rafts/mmz f (0-3 7) Acantholytic k cells/mm2 11 (0-1,068) Rafts/mmz ? (0-10.5) Granulocytic 562 f exocytosis/mmz (0-1,785) Red blood cell 4.5 k diapedesis/mm2 ( ) Lymphocytic 76.4 f exocytosis/mm2 ( ) Follicles/mm of k pustule (0-3.1) Pustule length 2.3 f (mm) ( ) * Number of cases in which parameter could be measured. t Mean * SE. 4 Probability > Parameters of superficial pustules. 11 Parameters of follicular pustules. Superficial Folliculitis Meant (range) 11.8 f 4.6 (0-164) 0.0 f 0.0 (0) 8.0 f 4.0 (0-44) (0-4) 263 f 49.8 (0-1,200) 1.2 f 1.2 (0-57.5) 69.6 f 21.3 (0-7 02) 0.86 f 0.12 (0-3.6) 1.18 f 0.15 (0.14.6) Probability$ tosis). Statistical significance (P < 0.05) was determined by Pearson s chi-square test. Mann-Whitney U test was used for continuous data (e.g., Yo intact eosinophils). Multiple logistic regression as described by Breslow and Day2 and implemented by BMDP3 was used to identify histopathologic parameters that were independently associated with PF. All parameters that were significant at P < 0.1 on the initial univariate analysis were tested for inclusion in the logistic models. Analyses other than logistic regression were performed using SAS software.9 Results All cases of PF had subcorneal to intragranular pustules consisting of moderate to large numbers of acantholytic cells and both intact and degenerating granulocytes (Fig. 1). PF cases had a greater average density of acantholytic cells (226 * 22.9/mm2) in the superficial pustule than did SF cases (1 1.8 t 4.7/mm2) (Table 1). Acantholytic cells were 183 times more likely to be present in PF than in cases of SF with superficial pustules (P < 0.01). Rafts of acantholytic cells were more dense in both the superficial pustules and follicular pustules in PF. No rafts were found in the superficial pustules in SF, but one of 47 cases of SF had rafts within the external root sheath of the hair follicle. Acantholytic cells are 45 times more likely to be present in PF follicular pustules than in SF follicular pustules (P < 0.01). The cells from the stratum gran- ulosum adherent to the roof of the pustule (Fig. 2) were present in only six of the PF cases and none of the cases of SF. The density of erythrocytic exocytosis was not statistically significantly different between the PF and SF cases (Table 2), but the density of granulocytic exocytosis (Fig. 3) was significantly greater in cases of PF (562 k 67/mm2) as opposed to cases of SF (263 f 49.8/mm2). The percentage of intact eosinophils within the follicular pustules was not statistically different between the two specific diseases. ntact neutrophils tended to predominate in cases of PF, whereas degenerated granulocytes tended to predominate in cases of SF in both the superficial and follicular pustules. The follicles bridged per mm of pustule and the actual length of the pustule (Figs. 4, 6) were greater in PF ( O.l/mm; mm, respectively) than in SF (0.86 k 0.12/mm; 1.2 k 0.15 mm, respectively). The follicles bridged per mm of pustule length could not be determined in alopecic areas or on areas such as the planum nasale or the footpads. The risk of developing PF (relative to SF) was 3.6 times higher in dogs with recornification than in dogs without recornification (Table 3) (P < 0.01; C = ). The risk of developing PF (relative to SF) was 13.4 times higher in dogs with pustule reformation (Fig. 5), defined as one or more pustules forming be-

4 22 Kuhl, Shofer. and Goldschmidt Vet Pathol 31:l Fig. 2. Skin; multiple subcorneal pustules bridging a hair follicle. Adhesion of keratinocytes to the roof of the pustule is evident (arrow). HE. Bar = 200 pm. neath the original pustule, than in dogs without pustule PF (relative to SF) was 8.75 times higher when the reformation (P < 0.01; C = ). Parakeratosis external root sheath was involved and no associated was more likely to occur in cases of SF than in cases superficial pustule was present (P < 0.01; C = 1.9- of PF (P < 0.01; C = ). The risk of developing 4 1.O). Table 2. Percentage of cell types within the pustules of pemphigus foliaceus and superficial folliculitis. Pemphigus Foliaceus Superficial Folliculitis Histopathologic Meant Meant Parameters n* (range) n* (range) Probability$ Percent intact 7.66 i t eosinophilsg (041) (0-95) Percent intact ? neutrophils4 (0-1 00) (0-90) Percent degenerated 45.9? ? granulocytes$ (0-100) (0-100) Percent intact i f eosinophils ( (0-70) (0-80) Percent intact ? k neutrophilsll (0-00) (0-80) Percent degenerated i ? granulocytes / (0-00) (0-100) * Number of cases in which parameter could be measured. t Mean i SE. 4 Probability 3. Z 1. Parameters of superficial pustules. 11 Parameters of follicular pustules.

5 Vet Pathol 3i:l Pemphigus Foliaceus and Superficial Folliculitis 23 Fig. 4. Skin; elongated subcorneal; pustule with two overlying degenerated pustules in the stratum corneum. HE. Bar = pm. Siberian Husky, Labrador Retriever, Miniature Schnauzer, Old English Sheepdog, Scottish Terrier, Chinese Shar Pei, Spitz, Shetland Sheep Dog, Weimeraner, and West Highland White Terrier. Although there were no age or sex predilections, the Akita (n = 5), English Springer Spaniel (n = 2), Chow Chow (n = 2), Chinese Shar Pei (n = 2), and Collie (n = 3) had an increased risk of developing PF (Table 4). Fig. 3. Multiple intracorneal pustules with a layered appearance. The epidermis is hyperplastic with intercellular edema and granulocytic exocytosis. HE. Bar = 100 pm. Using logistic regression, no interrelationships between multiple factors were more predictive than any single factor for identifying PF. When all submissions definitively diagnosed as PF (n = 70) were compared to all submissions (n = 64,462) to the Surgical Pathology Service, the following breeds were affected with PF: Akita, Brittany Spaniel, Chow Chow, Collie, Dachshund, Doberman Pinscher, English Cocker Spaniel, English Setter, Golden Retriever, Discussion n agreement with previous studies, we have identified the Akita and Chow9,15,22 to have an increased risk of developing PF. Additionally, we have determined that the English Springer Spaniel, Chinese Shar Pei, and Collie have an increased risk of developing PF. Whereas the Akita and Chow may be predisposed breeds, the other breeds stated here and in the literature may reflect regional variations in breed preferences or lineages. Some of the factors that are most significant in identifying PF on histopathology are in agreement with the previous l i t e r a t ~ r e and ~ ~ were ~ ~ expect- ~ ~ ~ ~ ~ ~ ~ ed due to the pathogenesis of PF. One of these factors, Table 3. Presence or absence of recornification and reformation of pustules, orthokeratosis, parakeratosis, and external root sheath involvement in pemphigus foliaceus (n = ) and superficial folliculitis (n = 47). Pemphigus Superficial Foliaceus Folliculitis 95 '/o Odds Confidence Histopathologic Parameter Present Absent Present Absent Ratio ntervals Probability Recornification of pustule base Reformation of pustules Orthokeratosis NC* NC Parakeratosis External root sheath involvement O * NC = not calculable.

6 24 Kuhl, Shofer, and Goldschmidt Vet Pathol 31:l Fig. 5. Skin; higher magnification of Fig. 4 to demonstrate overlying degenerating pustule with several acantholytic cells while the lower subcorneal pustules contain acantholytic cells and granulocytes. HE. Bar = 100 wm. the presence of acantholytic cells, makes a case much more likely to be PF. However, acantholytic cells may be present in either disease. Large numbers of acantholytic cells are more often associated with PF, and rafts of acantholytic cells are also helpful in differentiating the two diseases, because they occurred in only 2.3% of SF cases and 44% of PF cases. The greater density of acantholytic cells and rafts of acantholytic cells is due to the difference in the pathogenesis of each of the specific diseases, as discussed previously. More importantly, this study was designed to differentiate between these two diseases in those cases lacking rafts or having few acantholytic cells. n these cases other parameters must be considered. Two of the more important factors are the size of the pustule and the number of follicles bridged by the pustule. PF pustules are consistently of greater length than the pustules found in SF. Additionally, the PF pustules were found to bridge more hair follicles per mm of pustule length. Thus, the histopathology supports the clinical findings that SF is an inflammation of the infundibulum of the hair follicle,5 whereas PF lacks a follicular orientation. Table 4. Breed predilections for developing pemphigus foliaceus. Breeds 95% Pemphigus Histopathology Odds Confidence (n = 70) (n = 64,462) Ratio ntervals Probability Akita English Cocker Spaniel Chow Chow Chinese Shar Pei Collie

7 Vet Pathol 3l:l, 1994 Pemphigus Foliaceus and Superficial Folliculitis 25 Fig. 6. Skin; multifocal folliculitis involving several follicles, including the infundibulum (arrows). HE. Bar = 400 pm. Therefore, in PF, the same length of pustule might bridge two hair follicles, whereas in SF the pustule will center around one hair follicle and bridge only one hair follicle. Two other important factors that are helpful in differentiating PF from SF cases are pustule recornification and reformation. Pustule recornification occurs more often in PF, but it occurs with sufficient frequency in SF that it cannot be used as a single diagnostic criterion. Pustule reformation is also more likely to be seen in PF than in SF. The latter factor is one that we have considered most useful in differentiating these two specific diseases. PF has clinically been observed to have a cyclic clinical cou~se,~ which agrees with the histopathologic concept of pustule reformation and recornification. As the base of the pustule recornifies, the disease shows signs of clinical improvement, whereas pustule reformation correlates with worse clinical disease. Although granulocytic exocytosis was greater in cases of PF, this is a difficult parameter to apply from a practical standpoint because the difference was only found by using the photoimaging system. This is not an application that is found in most diagnostic laboratories. The increased density of granulocytic exocy- tosis is an interesting feature and may be the basis for future research into the specific cytokines produced by the epidermal keratinocytes in each of these disease processes. The cell types within the pustule and their morphology should also be evaluated. The percentage of intact eosinophils was greater in the superficial pustules of PF, but this would not be a reliable histopathologic parameter to differentiate these two diseases due to the small absolute difference in values. However, the percentage of intact neutrophils in the PF pustules was significantly greater than in SF pustules and should prove a useful parameter to differentiate SF and PF. This fact should only be used in conjunction with other diagnostic features because the average percentage of intact neutrophils and degenerated granulocytes was nearly equal in PF. The difference in intact neutrophils versus degenerated granulocytes may be due to the processes by which the pustules are forming. n SF, many hydrolytic enzymes are released from the neutrophils as they attack the follicular wall, causing dis- solution of the intercellular adhesi~ns.~j This results in the degenerated granulocytes and occasional acantholytic cells associated with SF. n PF, antibodies bind to plakoglobin and desmoglein and induce plasmino-

8 26 Kuhl, Shofer, and Goldschmidt Vet Pathol 31:l gen activator synthesis in the keratinocyte. This is then secreted and subsequently activates plasminogen to plasmin. The plasmin then degrades the cell surface proteins, resulting in cellular dyshesion and acantholysis.7,8,14,24 n this process, neutrophils are secondarily percolating through the previously damaged epidermis or external root sheath and are not actively secreting lysosomal enzymes. These factors may account for the larger percentage of intact neutrophils in PF. Parakeratosis occurred more often in SF. The reason for this is unknown, but many cases of SF are pruritic either from the disease itself or from an underlying allergic etiology, and focal parakeratosis may be seen histopathologically as a result of self-trauma. Therefore, this histopathologic feature should not be used in establishing a specific diagnosis. External root sheath involvement with extensive acantholysis has always been considered to be more prevalent in cases of PF.9,' Our findings support this. However, acantholysis involving the external root sheath was also seen in one case of SF. This may have been a severely pathogenic stage in the development of perforating folliculitis, because in other areas of this section perforating folliculitis was apparent. n this specific case, the dog did respond to antibiotics and immunosuppressive therapy was not necessary. n conclusion, the most reliable factors in differentiating PF from SF are the density of acantholytic cells within the pustule and the presence of rafts of acantholytic cells. When the density of acantholytic cells is low and no rafts are present, other factors that may be used to differentiate PF and SF are longer pustule length, greater number of hair follicles per mm of pustule length, the presence of pustule recornification and reformation, and the greater percentage of intact neutrophils versus degenerated granulocytes within the pustules in PF. Acknowledgements The authors thank Drs. George Murphy and Mike Kamina for helping us to learn and use the video image analysis system in the Laboratory of Dermatopathology at HUP, and Drs. Barbara Kummel, Kevin Shanley, Robert Schick, James Jeffers, and David Duclos for the management of cases. References 1 Ackerman LJ: Canine and feline pemphigus and pemphigoid Part. Pemphigus. Compend Contin Educ Pract Vet 7:89-97, Breslow NE, Day NE: Statistical Methods in Cancer Research, Volume 1, The Analysis of Case-control Studies. nternational Agency for Research in Cancer, Lyon, France, Engelman L: Stepwise logistic regression. n; BMDP Statistical Software, ed. Dixon WJ, pp University of California Press, Los Angeles, CA, Goodwin L, Hill JE, Raynor K, Raszi L, Manage M, Cowin P: Desmoglein shows extensive homology to the cadherin family ofcell adhesion molecules. Biochem Biophys Res Commun 173: , Halliwell REW: Skin diseases associated with autoimmunity Part. The bullous autoimmune skin diseases. Compend Contin Educ Pract Vet 2: , Halliwell REW, Goldschmidt MH: Pemphigus foliaceus in the canine: a case report and discussion. J Am Anim Hosp Assoc 13: , Hashimoto K, Shafran KM, Webber PS, Lazarus GS, Singer KH: Anti-cell surface pemphigus autoantibody stimulates plasminogen activator activity of human epidermal cells. J Exp Med 157: , Hashimoto K, Wun TC, Baird J, Lazarus GS, Jensen PJ: Characterization of keratinocyte plasminogen activator inhibitors and demonstration of the prevention of pemphigus gg-induced acantholysis by a purified plasminogen inhibitor. J nvest Dermatol 92: , hrke PJ, Stannard AA, Ardans AA, Griffin CE: Pemphigus foliaceus in dogs: a review of 37 cases. J Am Vet Med Assoc 186:59-66, hrke PJ, Stannard AA, Ardans AA, Griffin CE, Kallet AJ: Pemphigus foliaceus of the footpads in three dogs. J Am Vet Med Assoc 186:67-69, Koch PJ, Walsh MJ, Schmelz M, Goldschmidt MD, Zimbelmann R, Franke WW: dentification of desmoglein, a constitiutive desmosomal glycoprotein, as a member of the cadherin family of cell adhesion molecules. Eur J Cell Biol 53: 1-1 2, Lever WF: Histopathology ofthe Skin, 6th ed., pp JB Lippincott Co, Philadelphia, PA, Manning TO, Scott DW, Kruth SA, Sozanski M, Lewis RM: Three cases of canine pemphigus foliaceus, and observation on chrysotherapy. J Am Anim Hosp Assoc 16: , Morioka S, Lazarus GS, Jensen PF: nvolvement of urokinase-type plasminogen activator in acantholysis induced by pemphigus gg. J nvest Dermatol 89: , Muller GH, Kirk RW, Scott DW: Small Animal Dermatology, 4th ed., pp , 2-257, WB Saunders Co., Philadelphia, PA, Razsi L, Raynor K, Aharon Y, Manabe M, Cowin P: Molecular cloning of desmoglein. J nvest Dermatoi 94: 570, Rubin E, Farber JL: Pathology, pp JB Lippincott Co., Philadelphia, PA, Rubinstein N, Stanley JR: Pemphigus foliaceus antibodies and a monoclonal antibody to desmoglein 1 demonstrate stratified squamous epithelia-specific epitopes of desmosomes. Am J Dermatopathol 9: , SAS nstitute nc.: SAWSTAT Users Guide, Release 6.03 ed., 1,028 p. SAS nstitute nc, Cary, NC, Schlesselman JJ: Case Control Studies. Design, Conduct, Analysis. Oxford University Press, New York, NY, Scott DW, Lewis RM: Pemphigus and pemphigoid in dog and man: comparative aspects. J Am Acad Dermatol , 1981

9 Vet Pathol 31:l Pemphigus Foliaceus and Superficial Folliculitis Scott DW, Smith CA: mmune-mediated dermatoses in domestic animals: ten years after- Part. Compend Contin Educ Pract Vet 9: , Scott DW, Wolfe MJ, Smith CA, Lewis RM: The comparative pathology of non-viral bullous skin diseases in domestic animals. Vet Pathol 17: , Singer KH, Hashimoto K, Jensen PJ, Morioka S, Lazarus GS: Pathogenesis ofautoimmunity in pemphigus. Annu Rev mmunol3:87-108, Werner LL, Brown KA, Halliwell REW: Diagnosis of autoimmune skin disease in the dog: correlation between histopathologic, direct immunofluorescent and clinical findings. Vet mmunol mmunopathol , 1983/ Wilkinson JE, Smith CA, Suter MM, Falchek W, Lewis RM: Role of plasminogen activator in pemphigus vulgaris. Am J Pathol 134: , 1989 Request reprints from Dr. Michael H. Goldschmidt, Laboratory of Pathology, Department of Pathobiology, University of Pennsylvania, School of Veterinary Medicine, 3800 Spruce Street, Philadelphia, PA (USA).