Autoimmune enteropathy and nephropathy with circulating anti-epithelial cell antibodies

Autoimmune enteropathy and nephropathy with circulating anti-epithelial cell antibodies Richard B, Colletti, MD, Ann P. Guillot, MD, Seymour Rosen, MD, Atul K. Bhan, MD, C. D a m o n Hobson, Jr., A. Bernard Collins, G a r y J. Russell, MD, a n d Harland S. Winter, MD From the Combined Program in Gastroenterology and Nutrition, Children's Hospital, and the Departments of Pathology, Beth Israel Hospital and Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts; and the Department of Pediatrics, University of Vermont College of Medicine, Burlington We describe a child with circulating anti-epithelial cell antibodies, a u t o i m m u n e e n t e r o p a t h y with intestinal villous atrophy, and membranous glomerulonephritis. The patient had persistent diarrhea at 6 months of age, and a small bowel biopsy showed a c t i v e enteritis, villous atrophy, and crypt hyperplasia. When the patient was, 10 months of age, nephrotic syndrome d e v e l o p e d b e c a u s e of membranous glomerulonephritis. Results of tests for circulating immune complexes were negative. Indirect immunofluorescence studies r e v e a l e d a circulating a n t i b o d y d i r e c t e d against renal epithelial cells. Circulating antibodies directed against normal small intestine epithelial cells were also d e t e c t e d by the i m m u n o p e r o x l d a s e technique. Western blot and immunoprecipitation identified a 55-kd antigen, in both small bowel and kidney, that r e a c t e d with an ant i b o d y in the patient's serum. High-dose prednlsone therapy i n d u c e d a clinical remission, resolution of the small bowel injury, and diminished serum anti-epithelial cell antibodies; after dose reduction, clinical relapse occurred with villous atrophy and r e a p p e a r a n c e of anti-epithelial cell antibodles. When the patient was 45 months of age, persistent diarrhea recurred despite intravenous administration of corticosteroids, cyclosporine, and total parenteral nutrition. Autoantibodies to a 55-kd epithelial cell protein are temporally related to the d e v e l o p m e n t of e n t e r o p a t h y and nephropathy. Study of slmilar patients is n e e d e d to determine the role of such antibodies in this disorder. (J PEDIATR 1991;118:858-64) Features of autoimmune disease, such as circulating tissue autoantibodies or a concurrent extraintestinal disorder thought to be autoimmune, have been seen in some infants with intractable diarrhea. H ~ Two patients have been described in whom villous atrophy was associated with membranous glomerulonephritis and a circulating antibody to small intestinal mucosa. 11, 12 We describe a child with

intestinal villous atrophy and immune-mediated membranous glomerulonephritis who had a circulating anti-epithelial cell antibody directed against an epitope present in both small bowel and kidney. AECA

Anti-epithelial cell antibody

I

CASE REPORT Submitted for publication Oct. 8, 1990; accepted Dec. 12, 1990. Reprint requests: Harland S. Winter, MD, Hunnewell Ground, Gastrointestinal Division, Children's Hospital Medical Center, 300 Longwood Ave., Boston MA 02115. 9/20/27288

858

The patient was healthy until 2 months of age, when diarrhea developed and he failed to thrive while being fed cow milk-based and soy-based infant formula (Fig. 1). His condition improved when he was fed a hydrolyzed-protein formula at 3 months of age. When he was 6 months of age, the results of 30 dermal scratch tests,

Volume 118 Number 6

Autoimmune enteropathy

co, sl

Formula

H Isl

H

I

Diarrhea [Z]

Bloody Stools

AA

I

H

+4

+-

S.B. Bx.

I'n~l

++

VA VA

NV -~-

. . . .

VA ~.

+-++--+++

"t- . . . .

VA ,~

Chlorambucil

VA VA VA

C----1

Prednlsone 2.0 ( m g / k g / d a y ) "'~

-

~c~

ililillllli,mi,mJ,,,,,,,,,,,,,,,,,,,,,,,,,,,,0................ IqgJlllll,,,,,,,,,,,,u,,,Jr,,,,,,[r,,,,,,m,,,,r,,,,J,,,,,iUiJili

Anti4+ Epithelial Cell 2+ Antibodies 0

El-.

n=_

_ . .

LO 1.0 0 4+ 2+ 0

=_

20 Urinary Protein/ Craatinine Ratio

859

20

1s 10

15 10

5 0 4 3

Serum Albumin 2 1 0 MN

Renal Bx. i

,

i

i

i

i

6

,

,

i

i

i'll

MN t

12

I

i

I

t

i

[

18

I

i

i

a

i

i

24

i

i

i

~

~

30

i

L

,

i

i

36

i

i

i

i

i

J

42

i

a

i

i

i

i

48

Age (months)

Fig. t. Correlation of disease activity and treatment with circulating AECAs. High-dose prednisone treatment resulted in resolution of bloody stools, healing of villous atrophy (VA) to normal villi (NV) on small bowel biopsy (S.B.Bx), resolution of nephrotic syndrome, and reduction in circulating AECAs. Lower-dose prednisone therapy resulted in relapse. At age 45 months, patient had relapse that was resistant to high-dose intravenous therapy with corticosteroids, cyclosporine (Cy), and total parenteral nutrition (TPN). Solid bar denotes daily dose of prednisone; vertical lines denote alternate-day dose of prednisone. C, Cow milk; S, soy formula; H, hydrolyzed protein formula; AA, amino acid formula; +, presence of bloody stools; - , absence of bloody stools; MN, membranous nephropathy. including tests of cow milk and soy antigen, were negative. Two hours after he was fed 180 ml of a soy formula, vomiting, bloody diarrhea, shock, and leukocytosis developed. The urinalysis showed normal values. The patient recovered after 48 hours of supportive care, including a transfusion of packed erythrocytes. On a diet of a hydrolyzed-protein formula, rice cereal, pureed fruits and vegetables, but no dairy products, soy, or gluten, the patient had frequent loose stools, poor linear growth, and eczema. When he was 10 months of age, nephrotic syndrome developed, with ascites, peripheral edema, urinary protein level >3 gin/L, serum albumin level of 19 gm/L, and cholesterol level of 7.9 retool/ L (305 mg/dl). The blood leukocyte count was 17.7 • 109/L and the erythrocyte sedimentation rate 40 ram/hr. The urine specific gravity was 1.020, pH 6, with no leukocytes or erythrocytes but >50 hyaline cases per high-power field. The ratio of urinary protein to creatinine was 22.3 (normal <0.2). A guaiac test detected occult blood in the stool. A small bowel biopsy demonstrated villous atrophy, crypt hyperplasia, and diffuse inflammation (Fig. 2, A). Transmission electron microscopy revealed mild degenerative changes of the surface epithelium and mild focal microvillous atrophy but no evidence of immune deposits or abnormalities of the basement membrane. The renal biopsy showed glomeruli with diffuse thickening of the basement membrane but no proliferation of cells or mesangial matrix. Transmission electron microscopy revealed total effacement of the epithelial cell foot processes. Glomerular basement membrane was irregularly thickened with numerous subepithelial and intramembranous electron-dense deposits.

There was no improvement in either intestinal or renal disease after the child was fed a modified amino acid formula. Treatment with prednisone, 2 mg/kg of body weight, was begun at 12 months of age. During the subsequent 2 months the ascites, edema, and proteinuria resolved and growth resumed. A small bowel biopsy showed regenerated villi, slight crypt hyperplasia, and focal mild inflammation (Fig. 2, B). With a dose of prednisone, 0.8 mg/kg every other day, the patient remained in clinical remission of 1 year, with mild eczema and elevated serum alanine aminotransferase activity. At 27 months of age, 2 months after the dose of prednisone was reduced to 0.3 mg/kg every other day, the child had moderate proteinuria, mild hypoalbuminemia, and guaiac-positive stools. A small bowel biopsy demonstrated relapse, with crypt hyperplasia, focal partial villous atrophy, and inflammation (Fig. 2, C). Light and electron microscopic studies of the kidney showed little change. An increase in prednisone dose to 2 mg/kg per day resulted in clinical improvement. When the child was 35 months of age, 2 months after the dose of prednisone was reduced to 1.1 mg/kg per day, proteinuria, hypoalbumlnemia, and peripheral edema recurred. Treatment with chlorambucil, 0.2 mg/kg per day for 3 months (cumulative dose 270 rag), resulted in remission of the nephrotic syndrome. A relapse of bowel disease at 41 months, confirmed by small bowel biopsy, again responded to an increased dose of prednisone (1.8 mg/kg per day). However, when the patient was 45 months of age, after an acute febrile gastrointestinal illness, diarrhea with villous atrophy persisted despite intravenous treatment with corticosteroids and cyclosporine, together with total parent-

860

Colletti et al.

The Journal of Pediatrics June 1991

Fig. 2. Histopathologic features of small bowel biopsy specimen. A, Severe villous atrophy and crypt hyperplasia are present before treatment at age 12 months. B, Nearly normal villi and mild inflammation are present after high-dose prednisone treatment at age 14 months. C, Biopsy specimen during relapse at age 42 months. Severe villous atrophy and inflammation, including crypt abscess (arrow), is present. (Hematoxylin-and-eosin stain; •

Volume 118 Number 6

Autoimmune enteropathy

861

eral nutrition. He remained dependent on total parenteral nutrition. Serum electrolyte, urea nitrogen, ereatinine, bilirubin, and alkaline phosphatase values were normal. Blood culture, urine viral culture, and results of serologic tests for hepatitis viruses A, B, and C were negative. Stool for ova, parasites, bacterial pathogens, and Clostridium difficile toxin, sweat sodium and chloride, sigmoidoscopic results, and colonicand gastric antral biopsyspecimenswere negative. Contrast radiography of the small bowel showed a realabsorption pattern. Xylose absorption was zero at 1 hour. Histologic features of the liver were normal except for rare degenerated hepatocytes. IMMUNOLOGIC

STUDIES

(Table)

Serum immunoglobulin levels were normal (IgG, IgM) or elevated (IgA, IgE). Results of radioallergosorbent tests were positive for cow milk but negative for soy, wheat, rice, pork, chicken, and peanuts. Results of dermal scratch tests of cow milk and egg white were positive. The T lymphocyte numbers, subsets, and stimulation (phytohemagglutinin, pokeweed mitogen, and tetanus) were normal. Results of tests for serum immune complexes were negative. Immunohistochemical staining of the patient's small bowel biopsy specimens demonstrated CD3-, CD4-, and CD8-positive cells in the epithelium and the lamina propria. 13 Antibodies to mitochondria, smooth muscle, microsomes, nuclear antibody, parietal cells, and glomerular and renal tubular basement membrane were not detected. Human lymphocyte antigen typing showed HLA A9, A24/A11, B7, B35, DR2, DR5, DRW 52, DQW 1, and DQW 3. Direct immunofluorescence studies of the kidney demonstrated granular fluorescence along the glomerular capillary walls, and results were positive for IgG and polyvalent immunoglobulin conjugates but were negative for IgM, IgA, IgE, C3, and fibrinogen. Direct immunoftuorescence studies of the small intestine (by B. MacPherson, MD) showed no evidence that immune complexes contained IgG, IgA, IgE, complement, or fibrin. In addition, there was no evidence for anti-IgA binding to epithelial cells. However, indirect immunofluoreseence studies (by S. Rosen, MD) of the patient's serum (diluted 1:40) and normal human kidney frozen tissue sections showed positive staining of glomerular and proximal tubular epithelial cells, prompting subsequent studies to determine the nature of the antibody. Indirect immunohistology, To determine whether a circulating AECA was present, we fixed frozen tissue sections of normal human small intestine and kidney at room temperature in acetone for 10 minutes and incubated them for 1 hour with a 1:10 or 1:100 dilution of either the patient's or control serum, followed by monoelonal antibodies to human IgG and IgA (Dako Corp., Santa Barbara, Calif.). The sections were again washed, and specific binding was detected by means of biotinylated horse anti-mouse IgG

Fi9. 3. Antibodies to small intestinal epithelial cells. IgA AECAs (arrow) were detected by immunoperoxidasestaining of normal intestine after incubation with patient's serum (diluted 1:100).

(Vector Laboratories Inc., Burlingame, Calif.) and stained with immunoperoxidase with the avidin-biotin complex (Vector) and the chromogen 3-amino-9-ethylcarbazole (Aldrich Chemical Co., Inc., Milwaukee, Wis.). 13 Western transfer. To characterize the antigen of a circulating AECA, we performed polyacrylamide gel electrophoresis on mechanically solubilized normal intestinal mucosa or kidney tissue rapidly frozen in liquid nitrogen. The proteins were transferred to a Western immunoblot, incubated with serum from the patient and control subjects (celiac disease and normal), and stained with immunoperoxidase by means of the avidin-biotin complex. 14, 15 The initial serum sample was positive at 1:100 dilution. All subsequent sera were run at 1:10 dilution and compared with the initial sample at 1:100; those sera that were positive at 1:10 were retested at 1:100 dilution. The electrophoretic pattern was also compared with that of proteins of known molecular weight. The amount of AECA was graded on a scale from 0 to 4+ based on the degree of staining. RESULTS When the patient's serum or control sera were applied to normal intestinal mucosa, only the patient's serum stained brush border of small bowel epithelial cells (Fig. 3). The antibody in the serum did not appear to be restricted to a specific isotype, because staining was observed for IgG and IgA. The most intense staining was shown by IgA. When the patient's serum was applied to normal kidney, immunoper-

862

Colletti et al.

The Journal o f Pediatrics June 1991

1

Fig. 4. Identification of antigen with Western immunoblot of solubilized normal intestinal mucosa with patient's serum (diluted 1:100). Mucosal antigen (arrow) with molecular weight of approximately 55 kd reacted with antibody in patient's serum. Lane I (before treatment, 12 months of age) and Lane 2 (during relapse, 28 months of age) have 4+ staining. Lanes 3 and 4 (typical of remission) have 1+ staining. Control serum (lane 5), from patient with active celiac disease, did not stain.

Table. Immunologic studies Serum immunoglobulins, complement RAST and dermal scratch tests Peripheral blood T lymphocytes Circulating immune complexes Intestinal mucosal T lymphocytes Direct* immunofluorescence, kidney Direct* immunofluorescence, intestine Indirectt immunoftuorescence, kidney Indirectt immunohistology, intestine Western transfer + immunoblot:~

IgG 9.83 gm/L, IgA 2.27 gm/L, IgM 1.44 gm/L, IgE 1202 #g/L (501 U/ml); C3 and CHs0 normal Positive for cow milk, egg white Normal phenotype CD3, CD4, and CD8, and stimulation (PHA, PWM, tetanus) Negative CD3, CD4, and CD8 subpoputations similar to normal Positive for IgG complexes in glomerulus Negative Positive for circulating IgG anti-kidney antibody Positive for circulating IgG and IgA AECA Positive for circulating antibody to 55 kd protein in intestine and kidney

RAST, Radioallergosorbenttest; PHA, phytohemagglutinin;PWM, pokeweed mitogen. *Patient's tissue + labeled anti-human antibody. tPatient's serum + normal tissue + labeled anti-human antibody. :]:Patient's serum + solubilized normal intestine or kidney.

oxidase staining of the brush border of proximal tubules was observed for IgG. When serum was incubated with solubilized normal intestinal mucosa transferred to a Western immunoblot, the A E C A was detected in multiple serum samples from the patient but in none of the control samples (Fig. 4). The Western immunoblot demonstrated that the circulating antibody reacted with a mucosal protein with a molecular weight of approximately 55 kd. When the patient's serum was incubated with solubilized normal kidney tissue transferred to a Western immunoblot, the antibody also cross-

reacted with a renal protein of approximately 55 kd, suggesting a common epitope. In 19 of the patient's serum specimens, collected during a 38-month period, the amount of A E C A was graded on a scale from 0 to 4+ and correlated with disease activity (Fig. 1). Treatment with higher doses of prednisone correlated with clinical and histopathologic remission and a decrease in antibody titer; lower doses of prednisone were associated with relapse and an increase in antibody titer. However, when the patient was 48 months of age, persistent diarrhea recurred despite all treatment and the titer of A E C A did not

Volume 118 Number 6

rise. Exacerbations of nephrosis were not predictably concurrent with relapses of small bowel disease. One serum sample, with a high titer of circulating antibody, showed no cytotoxicity (tested by J. Roche, MD, PhD, University of Virginia, Charlottesville)) 6 DISCUSSION Although circulating AECAs were described in an infant with persistent diarrhea and villous atrophy in 1978, 2 our report correlates measurements of antibody with disease activity during a 4-year period. In the first 3 years, bowel disease activity was associated with high levels of circulating AECAs, and steroid-induced clinical and pathologic remission was associated with a reduction in antibody titer. In the fourth year, however, relapse and refractory diarrhea occurred without a rise in circulating antibody. In previously described infants with persistent diarrhea, villous atrophy, and circulating AECA, declining or disappearing levels of antibody were associated with either resolution,6 improvement,2 better prognosis,s or persistent disease7; the persistence of antibody was associated with death.6, 9 The findings in our patient indicate that identifieation of a circulating AECA in the early phase of the disease may be an indication of steroid responsiveness. The circulating antibody reacts with a 55 kd protein present in both normal small bowel mucosa and kidney. Immunofloureseence, immunohistologic, and biochemical studies suggest that the epitope is associated with a common protein on the epithelial cells of small intestine and kidney. Others have reported similar staining of the intestinal mucosa, 3 enterocyte cytoplasm,2 villous tips s and brush border 6, 9 but the antigen has not been characterized. In untreated patients with celiac disease, circulating antibodies have been found to gliadin, reticulin, endomysium, and a 90 kd protein in intestinal mucosa.17 Anti-brush border antibodies have also been detected in patients with Crohn disease is and in their relatives.16 Specific antibodies to epithelial-cell-associated components were detected in 70% of patients and 55% of their relatives with inflammatory bowel disease. Western immunoblot studies have demonstrated antibody against 160 and 137 kd molecules. An antibody has also been eluted from colonic mucosa of patients with ulcerative colitis that recognizes a 40 kd protein present on the plasma membrane of colon epithelial cells. 19 The pathogenic significance of circulating AECAs is uncertain and the role of the antibody in our patient is unknown. In Crohn disease and ulcerative colitis, antibodies have been shown to be cytotoxic. 16 In our patient, one serum sample, with a high titer of circulating antibody, showed no cytotoxicity. The simultaneous occurrence of autoimmune enteropa-

Autoimmune enteropathy

863

thy and nephropathy is rare. In one infant with intractable fatal diarrhea with villous atrophy, tubulointerstitial renal disease and membranous glomerulonephritis were associated with a circulating antibody to small intestine but not to kidney.ll Another 3-year-old child had diarrhea with villous atrophy l year after the onset of the nephrotic syndrome; there were two circulating antibodies, one to intestinal epithelial cells and one to renal tubular basement membrane. 12 Another infant with intractable diarrhea and villous atrophy, who died of renal failure, had circulating antibodies to gut epithelial cells and renal brush border, but at autopsy there was no nephropathy or interstitial nephritis.9 In two patients with Crohn disease and membranous glomerulonephritis, circulating antibodies were not described but one had a mild elevation of circulating immune complex levels. 2~ 2! In our patient, both organs to which there were autoantibodies had demonstrable injury, the antibodies recognized intestinal and renal epitopes of the same size, and, early in the course of the disease, there was a correlation with antibody titer and disease activity. Phenotypic characterization of mucosal lymphoeytes demonstrated a normal distribution of helper and eytotoxic-suppressor T lymphocytes. The possibility remains that the antibody could be an epiphenomenon, but it is probable that there is an underlying abnormality in immunoregulation,with autoimmunity against an epithelial cell-associated protein. Study of other patients with similar disorders will enhance our understanding of the pathogenesis by defining whether an antibody exists against one common epitope. We thank M. Kim, MD, and L. Guay-Woodford, MD, for care of the patient, contributions to the renal studies, and review of the manuscript; D. Antonioli, MD, and T. Trainer, MD, for review of the intestinal abnormalities; B. MacPherson, MD, for transmission electron microscopy of the kidney and selected immunofluorescence studies; J. Roche, MD, PhD, for study of cytotoxicity; and N. Roff, CGC, for care of the patient. REFERENCES

1. Larcher VF, Sheperd R, Francis DEM, Harris JT. Protracted diarrhoea of infancy: analysis Of 82 cases with particular references to diagnosis and management. Arch Dis Child 1977;52:597-605. 2. McCarthy DM, Katz SI, Gazze L, Waldmann TA, Nelson DL, Strober W. Selective IgA deficiencyassociated with total villous atrophy of the small intestine and an organ-specific anti-epithelial cell antibody. J Immunol 1978;120:932-8. 3. Fisher SE, Boyle JT, Holtzapple P. Chronic protracted diarrhea and jejunal atrophy in an infant. Dig Dis Sci 1981;26: 181-6. 4. Powell BR, Buist NR, Stenzel P. An X-linked syndrome of diarrhea, polyendocrinopathy,and fatal infection in infancy. J PEDIATR 1982;100:731-7. 5. Unsworth J, Hutchins P, Mitchell J, et al. Flat small intestinal

864

6.

7. 8.

9.

10.

11.

12.

13.

Colletti et al.

The Journal o f Pediatrics June 1991

mucosa and autoantibodies against the gut epithelium. J Pediatr Gastroenterol Nutr 1982;1:503-13. Savage MO, Mirakian R, Wozniak ER, et al. Specific autoantibodies to gut epithelium in two infants with severe protracted diarrhoea. J Pediatr Gastroenterol Nutr 1985; 4:187-95. Unsworth D J, Walker-Smith JA. Autoimmunity in diarrheal disease. J Pediatr Gastroenterol Nutr 1985;4:375-80. Mirakian R, Richardson A, Milla P J, et al. Protracted diarrhoea of infancy: evidence in support of an autoimmune variant. Br Med J 1986;293:1132-6. Mitton SG, Mirakian R, Larcher VF, Dillon M J, WalkerSmith JA. Enteropathy and renal involvement in an infant with evidence of widespread autoimmune disturbance. J Pediatr Gastroenterol Nutr 1989;8:397-400. Seidman EG, Lacaille F, Russo P, Galeano N, Murphy G, Roy CC. Successful treatment of autoimmune enteropathy with cyclosporine. J PEDIATR 1990;117:929-32. Ellis D, Fisher SE, Smith Wl, Jaffe R. Familial occurrence of renal and intestinal disease associated with tissue autoantibodies. Am J Dis Child 1982;136:323-6. Martini A, Scotta MS, Notarangelo LD, Maggiore G, Guarnaccia S, De Giacomo C. Membranous glomerulopathy and chronic small-intestinal enteropathy associated with autoantibodies directed against renal tubular basement membrane and the cytoplasm of intestinal epithelial cells. Acta Paediatr Scand 1983;72:931-4. Russell G J, Bhan AK, Winter HS. The distribution ofT and

14.

15.

16.

17. 18. 19.

20.

21.

B lymphocyte populations and MHC class II expression in human fetal and postnatal intestine. Pediatr Res 1990;27:23944. Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227:680-5. Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Proc Natl Acad Sci USA 1979;76:4350-4. Fiocchi C, Roche JK, Michener WM. High prevalence of antibodies to intestinal epithelial antigens in patients with inflammatory bowel disease and their relatives. Ann Intern Med 1989;110:786-94. Auricchio S, Greco L, Troncone R. Gluten-sensitive enteropathy in childhood. Pediatr Clin North Am 1988;35:157-87. Skogh T, Heuman R, Tagesson C. Anti-brush border antibodies in Crohn's disease. J Clin Lab Immunol 1982;9:147-50. Elson CO. The immunology of inflammatory bowel disease. In: Shorter RG, Kirsner JB, eds. Inflammatory bowel disease. Philadelphia: Lea & Febiger, 1989:97-164. O'Loughlin EV, Robson L, Scott B, Alexander F, Gall DG. Membranous glomerulonephritis in a patient with Crohn's disease of the small bowel. J Pediatr Gastroenterol Nutr 1985;4:135-9. Glassman M, Kaplan M, Spivak W. Immune-complex glomerulonephritis in Crohn's disease. J Pediatr Gastroenterol Nutr 1986;5:966-9.

FELLOWSHIPS Available fellowships in pediatric subspecialties and those for general academic pediatric training are listed once a year, in May, in THE JOURNAL OF PEDIATRICS. E a c h October, forms for listing such fellowships are sent to the C h a i r m a n of the D e p a r t m e n t of Pediatrics at most m a j o r hospitals in the U n i t e d States a n d Canada. Should you desire to list fellowships, a separate application must be m a d e each year for. each position. All applications must be returned to M o s b y - Y e a r Book, Inc., by F e b r u a r y 15 of the listing year to ensure publication. Additional forms will be supplied on request from the Journal Editing D e p a r t m e n t , M o s b y - Y e a r Book, Inc., 11830 Westline industrial Drive, St. Louis, M O 63146-3318/314-453-4317.