- Email: [email protected]
Liver disease associated with alpha1-antitrypsin deficiency in childhood
January 1976
The J o u r n a l o f P E D I A T R I C S
19
Liver disease associated with alphal-antitrypsin deficiency in childhood Liver disease in children with alphal-antitrypsin deficiency and protease inhibitor type ZZ does not necessarily carry a bad prognosis. Fourteen of our 18 patients presented with the neonatal hepatitis syndrome and four had hepatomegaly without jaundice. Although four patients have died of cirrhosis and its complications, and three have severe liver disease, most of the 11 others, of whom four are over 13 years of age, have relatively little clinical, biochemical, or histologic evidence of liver disease. Persistent elevation of SGOT during the third year of life and renal or pulmonary problems were associated with a poor prognosis. Liver biopsy early in the course o f the disease was not helpful Frognostieally but was useful in assessment of the severity of Rver disease and demonstration of alA T storage, alA T deficiency was found in 29~ of our patients who presented with the neonatal hepatitis syndrome. One of seven apparently healthy Pi type ZZ sibs o f our patients had significant liver disease which had not been suspected previously.
Stanley P. Moroz, M.D., F.R.C.P. (C), Ernest Cutz, M.D., F.R.C.P. (C), Diane Wilson Cox, Ph.D., and Andrew Sass-Kortsak, M.D., F.R.C.P. (C),* Toronto, Ont., C a n a d a
ThE ASSOCIATIONS between alphal-antitrypsin deficiency and familial pulmonary emphysema in adults ~ and between alAT deficiency and neonatal hepatitis or cirrhosis in children 2 are well recognized. A recent report from our center described two alAT-deficient sibs who had both pulmonary and hepatic disease? Since then hepatic involvement has been reported in a~AT-deficient adults with or without pulmonary disease. 4-7 Most aiAT-deficient children with hepatic involvement have presented with the neonatal hepatitis syndrome. Some, however, had no history of early jaundice and presented with cirrhosis. Sharp ~ and others 9-12 noted a uniformly bad prognosis in these children. Although the jaundice cleared, the liver disease progressed to cirrhosis From the Departments of Pediatrics and Pathology, the University of Toronto, and the Research Institute, The Hospital for Sick Children. Supported by a National Health Grant (No. 605-7606)from the Department of Health and Welfare of Canada. *Reprint address: A. S.-K., The Hospitalfor Sick Children, 555 UniversityAve., Toronto, Ont., M5G 1X8, Canada.
and its complications. Because of the poor outcome of these patients, the factors that may affect prognosis have not been considered in previous reports. Two alAT-deficient sibs who had had the neonatal hepatitis syndrome were recently described as clinically Abbreviations used alAT: alpha,-antitrypsin Pi: protease inhibitor PAS-D: PAS after diastase PTAH: phosphotungstic acid hematoxylin FITC: fluorescein isothiocynanate well, 13 suggesting that the course of the liver disease associated with cqAT deficiency is not necessarily progressively downhill. These two children, however, had been followed only until age 18 months and 3 years, respectively, and their fiver function tests had remained abnormal. We had in our care 18 children who had liver disease in early life associated with a~AT deficiency and protease inhibitor type ZZ. Four of the surviving 14 patients have been followed by us for a period of 13 years or longer.
Vol. 88, No. 1, pp. 19-25
20
Moroz et al.
The Journal of Pediatrics January 1976
PATIENT t t t
1 2 3 5689I0-
111215161718-
,/ '
'
i
I
"
I
2
'
I
'
4
I 6
'
I
8
'
I 10
A G E (months)
Fig. 1. Duration of jaundice in 14 alAT-deficient children who had had the neonatal hepatitis syndrome. Patients who died (t) and those who have severe liver disease (*) are compared to those who are now relatively well. Seven seemingly healthy Pi type ZZ sibs of these patients were also examined. Hepatic histology of 21 of the 25 Pi type ZZ children were studied. Our experience with these 25 children is the subject of this report. MATERIALS AND METHODS Pi type was determined by starch gel electrophoresis at pH 4.9 a4 followed by crossed antigen-antibody electrophoresis? 5 Serum alAT concentration was determined in most Pi type ZZ individuals by electroimmunoassay?" Results were expressed as a percentage of normal with 100% representing the concentration of alAT in a pool of sera obtained from 1,000 normal donors by Dr. M. Fagerhol in Norway. Serum a~AT concentration in five patients early in the study was measured by radial immunodiffusion in antibody-impregnated agarose plates (Partigen, Behringwerke). Using this latter technique, the mean • SD for 30 healthy children was 2.80 +_ 0:18 rag/ ml? Since 1969, Pi typing has been performed regularly on the sera of all patients with liver disease who have been under the care of our gastroenterology service either as inor outpatients. In addition to this prospective study, we have screened all admissions to our hospital with liver disease during the period 1959 to 1969 by examining tracings of cellulose acetate paper electrophoresis of serum proteins for absence of the al-globulin peak and by reviewing fiver biopsies and liver tissue obtained at autopsy from patients who died with liver disease for the presence of cytoplasmic inclusions characteristic of a~AT deficiency? The suspicion of alAT deficiency raised by
these screening techniques was confirmed by determining the Pi type of the patient, whenever possible, and that of the parents and sibs. Liver function was assessed by routine biochemical techniques. Specifically, SGOT, alkaline phosphatase, serum proteins and their electrophoretic fractions, total and direct-reacting serum bilirubin and bromsulphalein retention were measured periodically in all patients. The patients with liver disease included in this study were carefully investigated for other possible causes of their liver disease. Those who presented with directreacting hyperbilirubinemia in early infancy were investigated as outlined in a recent publication. 17 Older children with liver disease were tested for the possibility of Wilson disease, cystic fibrosis, microcystic disease of the liver and kidney (renal tubular ectasia), toxic factors, etc. Investigation of both infants and older children included tests for hepatitis B~ antigen by crossed immunoelectrophoresis and hepatitis B~ antibody by passive he• Hepatic tissue was obtained by percutaneous needle biopsy, at laparotomy, or at autopsy. Paraffin-embedded sections of liver were stained by the following techniques: hematoxylin and eosin, periodic acid-Schiff, PAS after diastase, and Masson and phosphotungstic acid hematoxylin. Since 1969, electron microscopy has been performed regularly, and in addition, quick frozen sections have been incubated with fluorescein isothiocyanate-labeled antibody against ~ A T (Behringwerke). Pulmonary function was assessed by spirometry, by measurements of lung volumes with the body plethysmograph, by determination of forced expiratory flow-volume curves, and by measurement by arterial pO.~?s In one patient who had clinically significant pulmonary disease, serial studies had been performed; in the others only one assessment of pulmonary function was made. RESULTS
Clinical and laboratory findings. Eighteen children, 10 boys and 8 girls, from 17 families were found to have cqAT deficiency and Pi type ZZ associated with liver disease. The mean • SD of serum chAT concentration determined by electroimmunoassay in 12 of these children was 28 • 7.6% of normal (range 21 to 42%). In four patients, the mean • SD of serum cqAT concentration measured by radial immunodiffusion was 0.3 • 0.11 mg/ ml (range 0.25 to 0.48). The age at which symptoms or physical signs were first observed and the mode of presentation of these children are shown in Table I. Fourteen presented with the neonatal hepatitis syndrome characterized by jaundice, dark urine, various degrees of failure to thrive, hepatomegaly and, in six patients, splenomegaly. All had eleva-
Volume 88 Number 1
Liver disease with ajA T deficiency
21
Table I. Onset and current status of liver disease of Pi type Z Z children
Current status
Onset
Case No.
Age
1 2 3 4
< 1 wk < l wk 2 wk 18 mo
5
1 wk
Neonatal hepatitis syndrome Neonatal hepatitis syndrome Neonatal hepatitis syndrome Hepatosplenomegaly, portal hypertension, bleeding Neonatal hepatitis syndrome
3 wk
Neonatal hepatitis syndrome
4 yr
7 8
6 mo < 1 wk
Hepatosplenomegaly Neonatal hepatitis syndrome
8 mo 1 yr
9 10 l1 12 13 14 15 16 17 18
1 wk 1 wk 3 wk 1 wk 4 mo 7 yr 10 wk 6 wk 2 wk I wk
I
Mode
Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis Hepatomegaly Hepatomegaly Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis
Age
syndrome syndrome syndrome syndrome
syndrome syndrome syndrome syndrome
tion of direct-reacting bilirubin in the serum and abnormalities of other tests of liver function. Two were sibs (Cases 1 and 17 in Table I). Three of four children who had no history of j a u n d i c e were noted to have an enlarged and firm liver on physical examination performed for other unrelated problems; the fourth was first seen for investigation of hematemesis. Other causes for the liver disease of these children could not be found. Specifically, none of the 12 patients tested had evidence of recent or past infection with hepatitis B virus. As shown in Table I, four patients (Cases 1 to 4) have died of liver failure or the complications of severe chronic liver disease at the ages indicated. Three others (Cases 5 to 7) are alive but have clinical evidence of severe liver disease and have persistent abnormalities of all liver function tests. Eleven (Cases 8 to 18) of our patient s are relatively well. Five of them have hepatomegaly with or without splenomega!y; the other six have no clinical evidence of liver disease. Some of these children, particularly the younger ones, have only mildly to moderately elevated SGOT values; the others have completely n o r m a l liver function. Four of the children who had had the neonatal hepatitis syndrome and now have no clinical or biochemical evidence of ongoing' liver disease are presently between 13 and 19 years of age.
12 10 10 5
yr yr yr yr
7 yr
4 3 7 7 2 7 13 13 16 19
Clinical
I
yr yr mo yr yr yr yr yr yr yr
]
Liver function
Died Died Died Died Hepatosplenomegaly, portal hypertension, clubbing Hepatosplenomegaly, portal hypertension, clubbing, poor growth Hepatosplenomegaly Hepatomegaly, poor growth Hepatomegaly Hepatomegaly Hepatosplenomgaly Well W ell Hepatomegaly Well Well Well Well
Abnormal Abnormal
Abnormal SGOT +, alkaline phosphatase SGOT § SGOT Normal SGOT Normal Normal Normal Normal Normal Normal
Fig. 1 illustrates the duration of jaundice in the patients who presented with the neonatal hepatitis syndrome. All patients were anicteric by 8 months of age. There was no correlation between the duration or the degree ofj aundice and the eventual outcome of the patient's disease. During the third year of life, the patients who developed severe progressive liver disease and some of those who are now relatively well had clinical evidence of liver disease consisting mainly of hepatosplenomegaly and firm consistency of the liver. These clinical signs were more pronounced and often associated with failure to thrive and clubbing of fingers in those whose liver disease advanced further during the ensuing years compared to those who are now relatively or entirely well. Between 2 and 3 years of age, the SGOT values of those who later developed severe liver disease were high, between 215 and 380 units with a m e a n of 300 units. At the same age the SGOT values o f five children who eventually did well were much lower, between 55 and 110 units with a mean of 80 units. At less than 6 months of age, there was a wider variation of SGOT activities in those who did well later, but we have an insufficient number of observations in the group that has not done well for an adequate comparison. Liver function tests other than SGOT were not found to be useful in predicting the outcome of the patient's disease. Serum
22
Moroz et aL
The Journal of Pediatrics January 1976
Table II. Clinical findings and liver function of Pi type ZZ sibs
Table IV. Hepatic histopathology of Pi type ZZ patients with liver disease and sibs and age (1) at which tissue was obtained
Sibs No.
Sib of Case(s) No. *
Age (yr)
Clinical findings
Liver function
Giant cell hepatitis
1 2 3
1, 17 5 7
14 21 3
Normal Normal Abnormal
1" (3 mo) 5 (3 too) 16 (3 too) 17 (3 too)
4
9
4
5
l0
3
6 7
11 14
2 5
Well Well Hepatosplenomegaiy Hepatomegaly Hepatomegaly Well Well
Normal
Hepatitis 8 (7 too) 10 (3 too) 11 (3 too) 13 (4 mo) 15 (3 mo) Sib 5t (3 yr)
Normal SGOT Normal
Portal fibrosis 12 (4 yr) 13 (4 mo) 14 (7 yr) 15 (12 yr) I8 (18 yr) Sib 3 (3 yr) Sib 5 (3 yr)
Cirrhosis 1 2 3 4 6 7
(7 yr, 12 yrA) (4 yr, 10 yrA) (8 yr, 10 yrA) (2 yr) (9 mo, 3 yr) (6 too)
A = autopsy. *Patient n u m b e r as in Table 1. tSib n u m b e r as in Table II.
*As in Table I.
Table IIL Incidence of Pi types
Patients with "idiopathic'" neonatal hepatitis Pi type
No.
ZZ MZ SZ MS MM Others Total
]
Normal population
%
No.
5 1 0 1 10 0
29.4 5.9 0 5.9 58.8 0
0 12 2 39 370 3
0 2.7 0.4 9.1 87.2 0.6
17
100.0
426
100.0
I
'%
alAT levels were similarly not helpful prognostically. Three of our 18 patients had clinically manifest pulmonary disease characterized by recurrent cough, wheezing, and infections. The details of our findings in two of these patients (Cases 1 and 17 in Table I) were published earlier, a Two of the three children with pulmonary involvement have died (Cases 1 and 3). Pulmonary function tests in one (Case 3) had been repeatedly normal in spite of recurrent respiratory symptoms, except for his last study at 9 years of age; it showed mild airway obstruction. One child (Case 17) appears to be well, but recent pulmonary function tests showed him to have airway obstruction. Pulmonary function was assessed in three others (Cases 15, 16, and 18). None of these had clinical evidence of pulmonary involvement. However, one had airway obstruction and reduced arterial pO~, another had moderate airway obstruction and gas trapping, and the third only reduced arterial pO2. The remaining 12 patients were too young to be tested. All three of our patients veho died with liver disease and
had an autopsy performed were found to have histologic evidence of membranoproliferative glomerulonephritis. All.three had had evidence of renal dysfunction during life, characterized by microscopic hematuria with or without elevated blood urea nitrogen, serum creatinine, or decreased creatinine clearance. Details of these findings are being published separately. 19 None of the 14 living children included in this study have shown similar renal problems as yet, except for one child who had obstructive uropathy and recurrent urinary tract infections which are unlikely to be related to her alAT deficiency. As part of this study, Pi typing was performed in sibs and parents of the Pi ZZ patients. For the 17 probands, there were 22 sibs. Of these, eight were Pi type ZZ, including one who was known to have disease (Case 17 in Table I), ten were Pi type MZ, and four were Pi type MM. Half-sibs as confirmed by blood grouping studies were excluded. The clinical and laboratory findings of the seven apparently healthy Pi ZZ sibs of our patients are shown in Table II. One (Sib 3) had previously unsuspected liver disease. Three others had minimal evidence of liver involvement; two had none. The eldest of the sibs (Sib 2) has had exertional dyspnea for several years, but recently performed pulmonary function tests and chest roentgenograms were normal. The remaining sibs had no clinical evidence of lung disease but were too young to have pulmonary function studies performed. The mean +_ SD of serum alAT concentration measured by electroimmunoassay in the Pi ZZ sibs was 22 _+ 3.6% of normal (range t8 to 20%). Two of the parents of the Pi type ZZ children with liver disease were Pi type SZ; the rest were Pi type MZ. The Pi SZ parents, eight MZ fathers, and seven MZ mothers had no clinical or biochemical evidence of liver disease. The
Volume 88 Number 1
remaining parents had no clinical evidence of liver disease, but liver function studies were not performed on them. From 1969 to the end of 1973, 17 infants with "obstructive" jaundice were studied by us. All other known causes of liver disease manifested by this form of jaundice were excluded17; these children could have been considered to have the "idiopathic" neonatal hepatitis syndrome. In Table III, the Pi types of these children are compared to those of a normal population of 176 adults and 248 children from the Toronto area studied by us. Five or 29.4% of our 17 patients were alAT deficient and were Pi type ZZ. Thus there was a significant association of Pi type ZZ with neonatal hepatitis of otherwise unknown cause. Although the number of patients was small, there appeared to be no significant association between other Pi type variants and idiopathic neonatal hepatitis. Histologie data. The hepatic histopathology of the Pi type ZZ patients with fiver disease and of their Pi type ZZ sibs from whom tissue was available and ages at which the tissue was obtained is summarized in Table IV. Case 9 did not have a biopsy performed; Sibs 4 and 6 had normal biopsies at 4 years and 2 years of age, respectively. Biopsies were not performed on Sibs 1,2, and 7. "Giant cell hepatitis" was defined as swelling of the hepatocytes, giant celt transformation, hepatic necrosis, mixed portal inflammatory cell infiltrate, and perilobular cholestasis. "Hepatitis" included similar changes but without giant cell transformation. "Cirrhosis" was defined as established portal fibrosis with nodular regeneration of fiver parenchyma. Mild mononuclear cell inflammatory infiltrate and decreased number of interlobular bile ducts were often noted in association with cirrhosis. T h e liver of patients who presented with the neonatal hepatitis syndrome in whom liver biopsy was performed early in the course of their disease showed the histologic characteristics of giant cell hepatitis or hepatitis. The severity of the pathologic changes in the patients who later developed severe, chronic liver disease was similar to those seen in patients who are now relatively well. In one of the children who developed progressive liver disease (Case 6), well=established cirrhosis was noted at 9 months of age. Recent liver biopsy in two children (Cases 15 and 18) who had had the neonatal hepatitis syndrome and are now I3 and 19 years of age, respectively, showed only mild to moderate periportal fibrosis. In one of them (Case 15), the liver biopsy in early life had shown hepatitis. Liver function is now normal in both of these patients. Of the four patients who had no history of jaundice, two had cirrhosis (Cases 4 and 7) that was noted as early as 6 months of age in one (Case 7). The other two (Cases 13
Liver disease with c~A T deficiency
23
and 14) had portal fibrosis, which was associated with hepatitis in one (Case 13). Liver biopsy was performed on four of the Pi type ZZ sibs. As already indicated, two were histologically normal, and two had portal fibrosis (Sibs 3 and 5) that was associated with hepatitis in one (Sib 5). In all cases eosinophilic inclusions were noted in the cytoplasm of hepatocytes primarily in the periportal area. The inclusions stained red with PAS-D, bright red with Masson, and dark purple with PTAH. Immunofluorescence and electron microscopy studies were performed on liver biopsy specimens of eight patients and three sibs. The immunofluorescence studies with FITC-labeled specific antibody against a~AT revealed positive reactions in each instance with material stored in the hepatocyte inclusions. The electron microscopic studies revealed that the storage of this material was within folds of rough endoplasmic reticulum. The degree of the storage of this abnormal material, as revealed by either light or electron microscopy or by the immunofluorescence studies, was not clearly related to the type or the extent of hepatic damage otherwise noted. Further details of the histopathology will be presented elsewhere. ~~ DISCUSSION The association of a~AT deficiency with liver disease was first observed in children who presented with neonatal hepatitis or cirrhosis. 2 Previously unsuspected hepatic involvement was later found in alAT-deficient adults with or without pulmonary emphysema ranging from mild portal fibrosis to cirrhosis or hepatoma. 4-~ Most of these patients have been Pi type Z, or ZZ when the genotype has been confirmed by family studies. ~1 A finding common to all individuals who carry the Z allele in combination with another Z, S, or the normal M allele is the storage of material that reacts with specific antibody against aIAT within the rough endoplasmic reticulum of primarily periportal hepatocytes21' ~ This material can be demonstrated by immunofluorescent techniques and by light and electron microscopy and was found in all our patients from whom hepatic tissue was available. Why alAT is stored in the hepatocytes of Pi type Z individuals is unknown. The Z type a~AT in serum appears to have less sialic acid than the normal type? 3-~s One of us has proposed that the Z protein lacks a carbohydrate side chain with two terminal sialic acid residues, a structural abnormality which could prevent its release from the hepatocytes because of adherence to membranes or unusual aggregation. 25 The alAT, isolated from the PAS-D staining inclusions from liver of a patient
24
Moroz etal.
of Pi type Z, is in the asialo form and has a pronounced tendency to aggregateY 6 The storage of alAT within hepatocytes of Pi type Z individuals does not fully explain the development of liver disease in them because large amounts of aIAT can be demonstrated in the otherwise normal livers of Pi type Z individuals. Lieberman and co-workers 27 suggested that some additional factor may be necessary to cause disease in a liver made vulnerable by the stored a~AT. Hepatitis B virus infection was suggested as One such factor, and Porter and associates9 found evidence of thisSnfection in several alAT-deficient children with neonatal hepatitis or in their parents. We, like others, 7, ~- 1o. 1~have been unable to demonstrate in our alAT-deficient patients evidence of recent or past infection with hepatitis B virus or other factors that can cause neonatal hepatitis or cirrhosis. Some other as yet undefined factors could be involved. As in other studies, the majority of our patients with fiver disease and a~AT deficiency presented with the neonatal hepatitis syndrome. Among those in whom liver tissue was available early in the course of the disease, giant cell hepatitis was usually observed. We found that since 1969, five of 17 children (29%) who were being investigated for the neonatal hepatitis syndrome were a~AT deficient and Pi type ZZ. Aagenes and associates ' ' have estimated that the ZZ genotype occurs in 40% of children who have intrahepatic cholestasis without septicemia or other causes of liver disease. Porter a n d associates~ found that 25% of their patients with neonatal hepatitis were Pi type ZZ. The overall frequency of neonatal hepatitis syndrome associated with a~AT deficiency and Pi type ZZ in these three studies was 14 of 55 cases, or 25.5%. Thus a~AT deficiency has emerged as the most frequently associated finding in children who, in the past, would have been considered to have "idiopathic" neonatal hepatitis. Furthermore, although the mechanism for the development of the liver disease is not yet clearly understood, aIAT deficiency appears to be causally related tO the neonatal hepatitis syndrome. While %AT deficiency is frequently associated with the neonatal hepatitis syndrome, a~AT-deficient individuals may develop other forms of hepatic involvement as already indicated. Thus anyone, particularly a child, who has clinical or laboratory evidence of liver disease should be tested for a~AT deficiency, the definitive test being Pi typing. Examination of the cellulose acetate protein electrophoresis tracing of serum provides a useful screening measure if the absence of the a~-globulin peak is observed. In our experience, the a,-globulin level obtained from the tracing may be misleading. Although 4 of our 18 patients have died and three others have severe chronic liver disease, the remaining l 1,
The Journal of Pediatrics Januarf 1976
of whom four have been re-examined periodically for over 13 years, have relatively little or no clinical, biochemical, or histologic evidence of continuing hepatic involvement. Thus contrary to most previous reports, our study suggests that early hepatic involvement in children with a1AT deficiency does not carry a uniformly grave prognosis. There were no clinical or biochemical findings early in the course of the disease of these children that foretold which would develop severe chronic liver disease. Hepatic histology was also not helpful prognostically. The amount of stored alAT as determined by light microscopy could not predict outcome, but immunofluorescence studies which are more reliable in defining the extent of alAT storage were not performed on tissues from patients seen prior to 1969. Patients who had cirrhosis early with or without a preceding history of jaundice had a poor outlook. There were few other factors that were helpful prognostically. Although liver function tests early in the course of the liver disease appeared to be of little value in predicting the outcome of the disease, persistently high SGOT values in the third year of life were associated with poor outcome. Three of the patients who died had other problems, such as pulmonary disease or membranoproliferative glomerulonephritis, 1~ which may have contributed to their deaths. The long-term prognosis for all of these children is unknown. Some will likely die from liver disease or its complications; others who appear to have recovered from their liver disease or have relatively mild abnormalities of liver function or hepatic histology may eventually develop pulmonary emphysema as young adults, may have further progression of their liver disease, or may develop hepatomata. Some of our children have already shown some evidence of pulmonary involvement, but further followup studies as they grow older are obviously necessary. The relative and composite risks of any Pi type Z individual developing various problems have not yet been dearly established. Aagenes and colleagues 11 have estimated that 20 to 30% develop cirrhosis in early life and die, 50 to 60% develop emphysema with or Without hepatic involvement as adults, and 10 to 20% have no clinical symptoms but frequently have subclinical lesions in the liver or lungs. Sveger and Eaurel128 recently found that 16% of Pi Z newborn infants d e t e c t e d by routine screening developed overt signs of liver disease during the early months of life. Neonatal hepatitis syndrome was the most common manifestation but, as in our experience, some presented with hepatosplenomegaly and abnormal liver function test results, but no jaundice. Another 50% o f these children had only mild to moderate elevations of
Volume 88 Number 1
SGPT without clinical evidence of disease. The remaining 33% had no clinical or biochemical evidence of hepatic dysfunction. The duration of follow-up of these children was short, but these data suggest that at least during the early months of life, only some of these Pi type Z infants will develop the type of hepatic involvement we have described. No specific effective treatment exists for the liver disease associated with a l A T deficiency. Our patients received only supportive therapy. Sharp 8 found no change in a l A T levels in the serum of some patients treated with phenobarbital. Liver transplantation in one patient resulted in normal alAT levels in his serum within two days after the operation, but he died of p u l m o n a r y complications. Although no specific treatment is available, alAT-deficient, Pi type Z children should be identified so that genetic counseling may be provided to their parents. Pi type Z siblings should also be identified and closely examined for evidence of liver disease. W h e n indicated, this examination should include a liver biopsy on which immunofluorescence studies should be performed. If an alAT-deficient child has evidence of liver disease, our studies suggest that his future is not completely bleak; recovery may occur. REFERENCES
1. Laurell C-B, and Eriksson S: The electrophoretic alglobulin pattern of serum in al-antitrypsin deficiency, Scand J Clin Lab Invest 15:132, 1963. 2. Sharp HL, Bridges RA, Krivit W, and Freier ER: Cirrhosis associated with alpha~-antitrypsin deficiency: a previously unrecognized inherited disorder, J Lab Clin Med 73:934, 1969. 3. Glasgow JFT, Lynch M J, Hercz A, Levison H, and SassKortsak A: Alphal-antitrypsin deficiency in association with both cirrhosis and chronic obstructive lung disease in two sibs, Am J Med 54:181, 1973. 4. Berg NO, and Eriksson S: Liver disease in adults with alpha~-antitrypsin deficiency, N Engl J Med 287:1264, 1972. 5. Babb RR, Lillington GA, and Kempson RL: Cirrhosis in an adult with emphysema and alphal-antitrypsin deficiency, Am J Dig Dis 18:803, 1973. 6. Palmer PE, Wolfe H J, and Gherardi GJ: Hepatic changes in adult alphai-antitrypsin deficiency, Gastroenterology 65"284, 1973. 7. Eriksson S, and Hfigerstrand I: Cirrhosis and malignant hepatoma in alpha~-antitrypsin deficiency, Acta Med Scand 195:451, 1974. 8. Sharp HL: Alpha,-antitrypsin deficiency, Hosp Prac 5:83, 1971.
Liver disease with a~A T deficiency
25
9. Porter CA, Mowat AP, Cook PJL, Hayner DWG, Shilkin KB, and Williams R: al-Antitrypsin deficiency and neonatal hepatitis, Br Med J 3:435, 1972. 10. ~stergaard PA: Hereditary alpha-l-antitrypsin deficiency and liver cirrhosis in children, Dan Med Bull 20:96, 1973. 11. Aagenes ~, Matlary A, EI~0 K, Munthe E, and Fagerhol M: Neonatal cholestasis in alpha-l-antitrypsin deficient children, Acta Pediatr Stand 61:632, 1972. 12. Castleman B, Scully RE, and McNeely BU: Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 50-1973, N Engl J Med 289:1301, 1973. 13. Talamo R C, and Feingold M: Infantile cirrhosis with hereditary alphal-antitrypsin deficiency, Am J Dis Child 125:845, 1973. 14. Fagerhol MK: The Pi-system. Genetic variants of serum a,antitrypsin, Ser Haematol 1:153, 1968. 15. Fagerhol MK, and Laurell C-B: The polymorphisrn of "prealbumins" and alpha~-antitrypsin in human sera, Ciin Chim Acta 16:199, 1967. 16. Laurell C-B: Electroimmuno assay, Scand J Clin Lab Invest 29:Suppl 21-23, 1972. 17. Sass-Kortsak A: Management of young infants presenting with direct-reacting hyperbilirubinemia, Pediatr Clin North Am 21:777, 1974. 18. Cooper DM, Hoeppner V, Cox D, Zamel N, Bryan AC, and Levison H: Lung function in alphal-antitrypsin heterozygotes (Pi type MZ), Am Rev Resp Dis 110:708, 1974. 19. Moroz SP, Cutz E, Balfe JW, and Sass-Kortsak A: Membranoproliferative glomerulonephritis in children with liver disease associated with al-antitrypsin deficiency, Pediatrics (February, 1976). 20. Cutz E, Moroz SP, Cox DW, and Sass-Kortsak A: Histology of the liver in al-antitrypsin deficiency. In preparation. 21. First International Workshop of the Pi System, XXII Colloquium, Protides of the Biological Fluids, Brugge, Belgium, May, 1974. Lancet 1:118, 1975. 22. Gordon HW, Dixon J, Rogers JC, Mittman C, and Lieberman J: Alphal-antitrypsin (A1AT) accumulation in the liver of emphysematous patients with A1AT deficiency, Hum Pathol 3:361, 1972. 23. Bell OF, and Carrell RW: Basis of the defect in c~-lantitrypsin deficiency, Nature 243:410, 1973. 24. Cox DW: Defect in alphal-antitrypsin deficiency, Lancet 2:844, 1973. 25. Cox DW: The effect of neuraminidase on genetic variants of alpha~-antitrypsin, Am J Hum Genet 27:165, 1975. 26. Eriksson S, and Larsson C: PAS-positive inclusion bodies from the fiver in alphal-antitrypsin deficiency, N Engl J Med 292:176, 1975. 27. Lieberman J, Mittman C, and Gordon HW: Alpha~antitrypsin in the livers of patients with emphysema, Science 175:63, 1972. 28. Sveger T, and Laurell C-B: Personal communication, 1975.
The J o u r n a l o f P E D I A T R I C S
19
Liver disease associated with alphal-antitrypsin deficiency in childhood Liver disease in children with alphal-antitrypsin deficiency and protease inhibitor type ZZ does not necessarily carry a bad prognosis. Fourteen of our 18 patients presented with the neonatal hepatitis syndrome and four had hepatomegaly without jaundice. Although four patients have died of cirrhosis and its complications, and three have severe liver disease, most of the 11 others, of whom four are over 13 years of age, have relatively little clinical, biochemical, or histologic evidence of liver disease. Persistent elevation of SGOT during the third year of life and renal or pulmonary problems were associated with a poor prognosis. Liver biopsy early in the course o f the disease was not helpful Frognostieally but was useful in assessment of the severity of Rver disease and demonstration of alA T storage, alA T deficiency was found in 29~ of our patients who presented with the neonatal hepatitis syndrome. One of seven apparently healthy Pi type ZZ sibs o f our patients had significant liver disease which had not been suspected previously.
Stanley P. Moroz, M.D., F.R.C.P. (C), Ernest Cutz, M.D., F.R.C.P. (C), Diane Wilson Cox, Ph.D., and Andrew Sass-Kortsak, M.D., F.R.C.P. (C),* Toronto, Ont., C a n a d a
ThE ASSOCIATIONS between alphal-antitrypsin deficiency and familial pulmonary emphysema in adults ~ and between alAT deficiency and neonatal hepatitis or cirrhosis in children 2 are well recognized. A recent report from our center described two alAT-deficient sibs who had both pulmonary and hepatic disease? Since then hepatic involvement has been reported in a~AT-deficient adults with or without pulmonary disease. 4-7 Most aiAT-deficient children with hepatic involvement have presented with the neonatal hepatitis syndrome. Some, however, had no history of early jaundice and presented with cirrhosis. Sharp ~ and others 9-12 noted a uniformly bad prognosis in these children. Although the jaundice cleared, the liver disease progressed to cirrhosis From the Departments of Pediatrics and Pathology, the University of Toronto, and the Research Institute, The Hospital for Sick Children. Supported by a National Health Grant (No. 605-7606)from the Department of Health and Welfare of Canada. *Reprint address: A. S.-K., The Hospitalfor Sick Children, 555 UniversityAve., Toronto, Ont., M5G 1X8, Canada.
and its complications. Because of the poor outcome of these patients, the factors that may affect prognosis have not been considered in previous reports. Two alAT-deficient sibs who had had the neonatal hepatitis syndrome were recently described as clinically Abbreviations used alAT: alpha,-antitrypsin Pi: protease inhibitor PAS-D: PAS after diastase PTAH: phosphotungstic acid hematoxylin FITC: fluorescein isothiocynanate well, 13 suggesting that the course of the liver disease associated with cqAT deficiency is not necessarily progressively downhill. These two children, however, had been followed only until age 18 months and 3 years, respectively, and their fiver function tests had remained abnormal. We had in our care 18 children who had liver disease in early life associated with a~AT deficiency and protease inhibitor type ZZ. Four of the surviving 14 patients have been followed by us for a period of 13 years or longer.
Vol. 88, No. 1, pp. 19-25
20
Moroz et al.
The Journal of Pediatrics January 1976
PATIENT t t t
1 2 3 5689I0-
111215161718-
,/ '
'
i
I
"
I
2
'
I
'
4
I 6
'
I
8
'
I 10
A G E (months)
Fig. 1. Duration of jaundice in 14 alAT-deficient children who had had the neonatal hepatitis syndrome. Patients who died (t) and those who have severe liver disease (*) are compared to those who are now relatively well. Seven seemingly healthy Pi type ZZ sibs of these patients were also examined. Hepatic histology of 21 of the 25 Pi type ZZ children were studied. Our experience with these 25 children is the subject of this report. MATERIALS AND METHODS Pi type was determined by starch gel electrophoresis at pH 4.9 a4 followed by crossed antigen-antibody electrophoresis? 5 Serum alAT concentration was determined in most Pi type ZZ individuals by electroimmunoassay?" Results were expressed as a percentage of normal with 100% representing the concentration of alAT in a pool of sera obtained from 1,000 normal donors by Dr. M. Fagerhol in Norway. Serum a~AT concentration in five patients early in the study was measured by radial immunodiffusion in antibody-impregnated agarose plates (Partigen, Behringwerke). Using this latter technique, the mean • SD for 30 healthy children was 2.80 +_ 0:18 rag/ ml? Since 1969, Pi typing has been performed regularly on the sera of all patients with liver disease who have been under the care of our gastroenterology service either as inor outpatients. In addition to this prospective study, we have screened all admissions to our hospital with liver disease during the period 1959 to 1969 by examining tracings of cellulose acetate paper electrophoresis of serum proteins for absence of the al-globulin peak and by reviewing fiver biopsies and liver tissue obtained at autopsy from patients who died with liver disease for the presence of cytoplasmic inclusions characteristic of a~AT deficiency? The suspicion of alAT deficiency raised by
these screening techniques was confirmed by determining the Pi type of the patient, whenever possible, and that of the parents and sibs. Liver function was assessed by routine biochemical techniques. Specifically, SGOT, alkaline phosphatase, serum proteins and their electrophoretic fractions, total and direct-reacting serum bilirubin and bromsulphalein retention were measured periodically in all patients. The patients with liver disease included in this study were carefully investigated for other possible causes of their liver disease. Those who presented with directreacting hyperbilirubinemia in early infancy were investigated as outlined in a recent publication. 17 Older children with liver disease were tested for the possibility of Wilson disease, cystic fibrosis, microcystic disease of the liver and kidney (renal tubular ectasia), toxic factors, etc. Investigation of both infants and older children included tests for hepatitis B~ antigen by crossed immunoelectrophoresis and hepatitis B~ antibody by passive he• Hepatic tissue was obtained by percutaneous needle biopsy, at laparotomy, or at autopsy. Paraffin-embedded sections of liver were stained by the following techniques: hematoxylin and eosin, periodic acid-Schiff, PAS after diastase, and Masson and phosphotungstic acid hematoxylin. Since 1969, electron microscopy has been performed regularly, and in addition, quick frozen sections have been incubated with fluorescein isothiocyanate-labeled antibody against ~ A T (Behringwerke). Pulmonary function was assessed by spirometry, by measurements of lung volumes with the body plethysmograph, by determination of forced expiratory flow-volume curves, and by measurement by arterial pO.~?s In one patient who had clinically significant pulmonary disease, serial studies had been performed; in the others only one assessment of pulmonary function was made. RESULTS
Clinical and laboratory findings. Eighteen children, 10 boys and 8 girls, from 17 families were found to have cqAT deficiency and Pi type ZZ associated with liver disease. The mean • SD of serum chAT concentration determined by electroimmunoassay in 12 of these children was 28 • 7.6% of normal (range 21 to 42%). In four patients, the mean • SD of serum cqAT concentration measured by radial immunodiffusion was 0.3 • 0.11 mg/ ml (range 0.25 to 0.48). The age at which symptoms or physical signs were first observed and the mode of presentation of these children are shown in Table I. Fourteen presented with the neonatal hepatitis syndrome characterized by jaundice, dark urine, various degrees of failure to thrive, hepatomegaly and, in six patients, splenomegaly. All had eleva-
Volume 88 Number 1
Liver disease with ajA T deficiency
21
Table I. Onset and current status of liver disease of Pi type Z Z children
Current status
Onset
Case No.
Age
1 2 3 4
< 1 wk < l wk 2 wk 18 mo
5
1 wk
Neonatal hepatitis syndrome Neonatal hepatitis syndrome Neonatal hepatitis syndrome Hepatosplenomegaly, portal hypertension, bleeding Neonatal hepatitis syndrome
3 wk
Neonatal hepatitis syndrome
4 yr
7 8
6 mo < 1 wk
Hepatosplenomegaly Neonatal hepatitis syndrome
8 mo 1 yr
9 10 l1 12 13 14 15 16 17 18
1 wk 1 wk 3 wk 1 wk 4 mo 7 yr 10 wk 6 wk 2 wk I wk
I
Mode
Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis Hepatomegaly Hepatomegaly Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis Neonatal hepatitis
Age
syndrome syndrome syndrome syndrome
syndrome syndrome syndrome syndrome
tion of direct-reacting bilirubin in the serum and abnormalities of other tests of liver function. Two were sibs (Cases 1 and 17 in Table I). Three of four children who had no history of j a u n d i c e were noted to have an enlarged and firm liver on physical examination performed for other unrelated problems; the fourth was first seen for investigation of hematemesis. Other causes for the liver disease of these children could not be found. Specifically, none of the 12 patients tested had evidence of recent or past infection with hepatitis B virus. As shown in Table I, four patients (Cases 1 to 4) have died of liver failure or the complications of severe chronic liver disease at the ages indicated. Three others (Cases 5 to 7) are alive but have clinical evidence of severe liver disease and have persistent abnormalities of all liver function tests. Eleven (Cases 8 to 18) of our patient s are relatively well. Five of them have hepatomegaly with or without splenomega!y; the other six have no clinical evidence of liver disease. Some of these children, particularly the younger ones, have only mildly to moderately elevated SGOT values; the others have completely n o r m a l liver function. Four of the children who had had the neonatal hepatitis syndrome and now have no clinical or biochemical evidence of ongoing' liver disease are presently between 13 and 19 years of age.
12 10 10 5
yr yr yr yr
7 yr
4 3 7 7 2 7 13 13 16 19
Clinical
I
yr yr mo yr yr yr yr yr yr yr
]
Liver function
Died Died Died Died Hepatosplenomegaly, portal hypertension, clubbing Hepatosplenomegaly, portal hypertension, clubbing, poor growth Hepatosplenomegaly Hepatomegaly, poor growth Hepatomegaly Hepatomegaly Hepatosplenomgaly Well W ell Hepatomegaly Well Well Well Well
Abnormal Abnormal
Abnormal SGOT +, alkaline phosphatase SGOT § SGOT Normal SGOT Normal Normal Normal Normal Normal Normal
Fig. 1 illustrates the duration of jaundice in the patients who presented with the neonatal hepatitis syndrome. All patients were anicteric by 8 months of age. There was no correlation between the duration or the degree ofj aundice and the eventual outcome of the patient's disease. During the third year of life, the patients who developed severe progressive liver disease and some of those who are now relatively well had clinical evidence of liver disease consisting mainly of hepatosplenomegaly and firm consistency of the liver. These clinical signs were more pronounced and often associated with failure to thrive and clubbing of fingers in those whose liver disease advanced further during the ensuing years compared to those who are now relatively or entirely well. Between 2 and 3 years of age, the SGOT values of those who later developed severe liver disease were high, between 215 and 380 units with a m e a n of 300 units. At the same age the SGOT values o f five children who eventually did well were much lower, between 55 and 110 units with a mean of 80 units. At less than 6 months of age, there was a wider variation of SGOT activities in those who did well later, but we have an insufficient number of observations in the group that has not done well for an adequate comparison. Liver function tests other than SGOT were not found to be useful in predicting the outcome of the patient's disease. Serum
22
Moroz et aL
The Journal of Pediatrics January 1976
Table II. Clinical findings and liver function of Pi type ZZ sibs
Table IV. Hepatic histopathology of Pi type ZZ patients with liver disease and sibs and age (1) at which tissue was obtained
Sibs No.
Sib of Case(s) No. *
Age (yr)
Clinical findings
Liver function
Giant cell hepatitis
1 2 3
1, 17 5 7
14 21 3
Normal Normal Abnormal
1" (3 mo) 5 (3 too) 16 (3 too) 17 (3 too)
4
9
4
5
l0
3
6 7
11 14
2 5
Well Well Hepatosplenomegaiy Hepatomegaly Hepatomegaly Well Well
Normal
Hepatitis 8 (7 too) 10 (3 too) 11 (3 too) 13 (4 mo) 15 (3 mo) Sib 5t (3 yr)
Normal SGOT Normal
Portal fibrosis 12 (4 yr) 13 (4 mo) 14 (7 yr) 15 (12 yr) I8 (18 yr) Sib 3 (3 yr) Sib 5 (3 yr)
Cirrhosis 1 2 3 4 6 7
(7 yr, 12 yrA) (4 yr, 10 yrA) (8 yr, 10 yrA) (2 yr) (9 mo, 3 yr) (6 too)
A = autopsy. *Patient n u m b e r as in Table 1. tSib n u m b e r as in Table II.
*As in Table I.
Table IIL Incidence of Pi types
Patients with "idiopathic'" neonatal hepatitis Pi type
No.
ZZ MZ SZ MS MM Others Total
]
Normal population
%
No.
5 1 0 1 10 0
29.4 5.9 0 5.9 58.8 0
0 12 2 39 370 3
0 2.7 0.4 9.1 87.2 0.6
17
100.0
426
100.0
I
'%
alAT levels were similarly not helpful prognostically. Three of our 18 patients had clinically manifest pulmonary disease characterized by recurrent cough, wheezing, and infections. The details of our findings in two of these patients (Cases 1 and 17 in Table I) were published earlier, a Two of the three children with pulmonary involvement have died (Cases 1 and 3). Pulmonary function tests in one (Case 3) had been repeatedly normal in spite of recurrent respiratory symptoms, except for his last study at 9 years of age; it showed mild airway obstruction. One child (Case 17) appears to be well, but recent pulmonary function tests showed him to have airway obstruction. Pulmonary function was assessed in three others (Cases 15, 16, and 18). None of these had clinical evidence of pulmonary involvement. However, one had airway obstruction and reduced arterial pO~, another had moderate airway obstruction and gas trapping, and the third only reduced arterial pO2. The remaining 12 patients were too young to be tested. All three of our patients veho died with liver disease and
had an autopsy performed were found to have histologic evidence of membranoproliferative glomerulonephritis. All.three had had evidence of renal dysfunction during life, characterized by microscopic hematuria with or without elevated blood urea nitrogen, serum creatinine, or decreased creatinine clearance. Details of these findings are being published separately. 19 None of the 14 living children included in this study have shown similar renal problems as yet, except for one child who had obstructive uropathy and recurrent urinary tract infections which are unlikely to be related to her alAT deficiency. As part of this study, Pi typing was performed in sibs and parents of the Pi ZZ patients. For the 17 probands, there were 22 sibs. Of these, eight were Pi type ZZ, including one who was known to have disease (Case 17 in Table I), ten were Pi type MZ, and four were Pi type MM. Half-sibs as confirmed by blood grouping studies were excluded. The clinical and laboratory findings of the seven apparently healthy Pi ZZ sibs of our patients are shown in Table II. One (Sib 3) had previously unsuspected liver disease. Three others had minimal evidence of liver involvement; two had none. The eldest of the sibs (Sib 2) has had exertional dyspnea for several years, but recently performed pulmonary function tests and chest roentgenograms were normal. The remaining sibs had no clinical evidence of lung disease but were too young to have pulmonary function studies performed. The mean +_ SD of serum alAT concentration measured by electroimmunoassay in the Pi ZZ sibs was 22 _+ 3.6% of normal (range t8 to 20%). Two of the parents of the Pi type ZZ children with liver disease were Pi type SZ; the rest were Pi type MZ. The Pi SZ parents, eight MZ fathers, and seven MZ mothers had no clinical or biochemical evidence of liver disease. The
Volume 88 Number 1
remaining parents had no clinical evidence of liver disease, but liver function studies were not performed on them. From 1969 to the end of 1973, 17 infants with "obstructive" jaundice were studied by us. All other known causes of liver disease manifested by this form of jaundice were excluded17; these children could have been considered to have the "idiopathic" neonatal hepatitis syndrome. In Table III, the Pi types of these children are compared to those of a normal population of 176 adults and 248 children from the Toronto area studied by us. Five or 29.4% of our 17 patients were alAT deficient and were Pi type ZZ. Thus there was a significant association of Pi type ZZ with neonatal hepatitis of otherwise unknown cause. Although the number of patients was small, there appeared to be no significant association between other Pi type variants and idiopathic neonatal hepatitis. Histologie data. The hepatic histopathology of the Pi type ZZ patients with fiver disease and of their Pi type ZZ sibs from whom tissue was available and ages at which the tissue was obtained is summarized in Table IV. Case 9 did not have a biopsy performed; Sibs 4 and 6 had normal biopsies at 4 years and 2 years of age, respectively. Biopsies were not performed on Sibs 1,2, and 7. "Giant cell hepatitis" was defined as swelling of the hepatocytes, giant celt transformation, hepatic necrosis, mixed portal inflammatory cell infiltrate, and perilobular cholestasis. "Hepatitis" included similar changes but without giant cell transformation. "Cirrhosis" was defined as established portal fibrosis with nodular regeneration of fiver parenchyma. Mild mononuclear cell inflammatory infiltrate and decreased number of interlobular bile ducts were often noted in association with cirrhosis. T h e liver of patients who presented with the neonatal hepatitis syndrome in whom liver biopsy was performed early in the course of their disease showed the histologic characteristics of giant cell hepatitis or hepatitis. The severity of the pathologic changes in the patients who later developed severe, chronic liver disease was similar to those seen in patients who are now relatively well. In one of the children who developed progressive liver disease (Case 6), well=established cirrhosis was noted at 9 months of age. Recent liver biopsy in two children (Cases 15 and 18) who had had the neonatal hepatitis syndrome and are now I3 and 19 years of age, respectively, showed only mild to moderate periportal fibrosis. In one of them (Case 15), the liver biopsy in early life had shown hepatitis. Liver function is now normal in both of these patients. Of the four patients who had no history of jaundice, two had cirrhosis (Cases 4 and 7) that was noted as early as 6 months of age in one (Case 7). The other two (Cases 13
Liver disease with c~A T deficiency
23
and 14) had portal fibrosis, which was associated with hepatitis in one (Case 13). Liver biopsy was performed on four of the Pi type ZZ sibs. As already indicated, two were histologically normal, and two had portal fibrosis (Sibs 3 and 5) that was associated with hepatitis in one (Sib 5). In all cases eosinophilic inclusions were noted in the cytoplasm of hepatocytes primarily in the periportal area. The inclusions stained red with PAS-D, bright red with Masson, and dark purple with PTAH. Immunofluorescence and electron microscopy studies were performed on liver biopsy specimens of eight patients and three sibs. The immunofluorescence studies with FITC-labeled specific antibody against a~AT revealed positive reactions in each instance with material stored in the hepatocyte inclusions. The electron microscopic studies revealed that the storage of this material was within folds of rough endoplasmic reticulum. The degree of the storage of this abnormal material, as revealed by either light or electron microscopy or by the immunofluorescence studies, was not clearly related to the type or the extent of hepatic damage otherwise noted. Further details of the histopathology will be presented elsewhere. ~~ DISCUSSION The association of a~AT deficiency with liver disease was first observed in children who presented with neonatal hepatitis or cirrhosis. 2 Previously unsuspected hepatic involvement was later found in alAT-deficient adults with or without pulmonary emphysema ranging from mild portal fibrosis to cirrhosis or hepatoma. 4-~ Most of these patients have been Pi type Z, or ZZ when the genotype has been confirmed by family studies. ~1 A finding common to all individuals who carry the Z allele in combination with another Z, S, or the normal M allele is the storage of material that reacts with specific antibody against aIAT within the rough endoplasmic reticulum of primarily periportal hepatocytes21' ~ This material can be demonstrated by immunofluorescent techniques and by light and electron microscopy and was found in all our patients from whom hepatic tissue was available. Why alAT is stored in the hepatocytes of Pi type Z individuals is unknown. The Z type a~AT in serum appears to have less sialic acid than the normal type? 3-~s One of us has proposed that the Z protein lacks a carbohydrate side chain with two terminal sialic acid residues, a structural abnormality which could prevent its release from the hepatocytes because of adherence to membranes or unusual aggregation. 25 The alAT, isolated from the PAS-D staining inclusions from liver of a patient
24
Moroz etal.
of Pi type Z, is in the asialo form and has a pronounced tendency to aggregateY 6 The storage of alAT within hepatocytes of Pi type Z individuals does not fully explain the development of liver disease in them because large amounts of aIAT can be demonstrated in the otherwise normal livers of Pi type Z individuals. Lieberman and co-workers 27 suggested that some additional factor may be necessary to cause disease in a liver made vulnerable by the stored a~AT. Hepatitis B virus infection was suggested as One such factor, and Porter and associates9 found evidence of thisSnfection in several alAT-deficient children with neonatal hepatitis or in their parents. We, like others, 7, ~- 1o. 1~have been unable to demonstrate in our alAT-deficient patients evidence of recent or past infection with hepatitis B virus or other factors that can cause neonatal hepatitis or cirrhosis. Some other as yet undefined factors could be involved. As in other studies, the majority of our patients with fiver disease and a~AT deficiency presented with the neonatal hepatitis syndrome. Among those in whom liver tissue was available early in the course of the disease, giant cell hepatitis was usually observed. We found that since 1969, five of 17 children (29%) who were being investigated for the neonatal hepatitis syndrome were a~AT deficient and Pi type ZZ. Aagenes and associates ' ' have estimated that the ZZ genotype occurs in 40% of children who have intrahepatic cholestasis without septicemia or other causes of liver disease. Porter a n d associates~ found that 25% of their patients with neonatal hepatitis were Pi type ZZ. The overall frequency of neonatal hepatitis syndrome associated with a~AT deficiency and Pi type ZZ in these three studies was 14 of 55 cases, or 25.5%. Thus a~AT deficiency has emerged as the most frequently associated finding in children who, in the past, would have been considered to have "idiopathic" neonatal hepatitis. Furthermore, although the mechanism for the development of the liver disease is not yet clearly understood, aIAT deficiency appears to be causally related tO the neonatal hepatitis syndrome. While %AT deficiency is frequently associated with the neonatal hepatitis syndrome, a~AT-deficient individuals may develop other forms of hepatic involvement as already indicated. Thus anyone, particularly a child, who has clinical or laboratory evidence of liver disease should be tested for a~AT deficiency, the definitive test being Pi typing. Examination of the cellulose acetate protein electrophoresis tracing of serum provides a useful screening measure if the absence of the a~-globulin peak is observed. In our experience, the a,-globulin level obtained from the tracing may be misleading. Although 4 of our 18 patients have died and three others have severe chronic liver disease, the remaining l 1,
The Journal of Pediatrics Januarf 1976
of whom four have been re-examined periodically for over 13 years, have relatively little or no clinical, biochemical, or histologic evidence of continuing hepatic involvement. Thus contrary to most previous reports, our study suggests that early hepatic involvement in children with a1AT deficiency does not carry a uniformly grave prognosis. There were no clinical or biochemical findings early in the course of the disease of these children that foretold which would develop severe chronic liver disease. Hepatic histology was also not helpful prognostically. The amount of stored alAT as determined by light microscopy could not predict outcome, but immunofluorescence studies which are more reliable in defining the extent of alAT storage were not performed on tissues from patients seen prior to 1969. Patients who had cirrhosis early with or without a preceding history of jaundice had a poor outlook. There were few other factors that were helpful prognostically. Although liver function tests early in the course of the liver disease appeared to be of little value in predicting the outcome of the disease, persistently high SGOT values in the third year of life were associated with poor outcome. Three of the patients who died had other problems, such as pulmonary disease or membranoproliferative glomerulonephritis, 1~ which may have contributed to their deaths. The long-term prognosis for all of these children is unknown. Some will likely die from liver disease or its complications; others who appear to have recovered from their liver disease or have relatively mild abnormalities of liver function or hepatic histology may eventually develop pulmonary emphysema as young adults, may have further progression of their liver disease, or may develop hepatomata. Some of our children have already shown some evidence of pulmonary involvement, but further followup studies as they grow older are obviously necessary. The relative and composite risks of any Pi type Z individual developing various problems have not yet been dearly established. Aagenes and colleagues 11 have estimated that 20 to 30% develop cirrhosis in early life and die, 50 to 60% develop emphysema with or Without hepatic involvement as adults, and 10 to 20% have no clinical symptoms but frequently have subclinical lesions in the liver or lungs. Sveger and Eaurel128 recently found that 16% of Pi Z newborn infants d e t e c t e d by routine screening developed overt signs of liver disease during the early months of life. Neonatal hepatitis syndrome was the most common manifestation but, as in our experience, some presented with hepatosplenomegaly and abnormal liver function test results, but no jaundice. Another 50% o f these children had only mild to moderate elevations of
Volume 88 Number 1
SGPT without clinical evidence of disease. The remaining 33% had no clinical or biochemical evidence of hepatic dysfunction. The duration of follow-up of these children was short, but these data suggest that at least during the early months of life, only some of these Pi type Z infants will develop the type of hepatic involvement we have described. No specific effective treatment exists for the liver disease associated with a l A T deficiency. Our patients received only supportive therapy. Sharp 8 found no change in a l A T levels in the serum of some patients treated with phenobarbital. Liver transplantation in one patient resulted in normal alAT levels in his serum within two days after the operation, but he died of p u l m o n a r y complications. Although no specific treatment is available, alAT-deficient, Pi type Z children should be identified so that genetic counseling may be provided to their parents. Pi type Z siblings should also be identified and closely examined for evidence of liver disease. W h e n indicated, this examination should include a liver biopsy on which immunofluorescence studies should be performed. If an alAT-deficient child has evidence of liver disease, our studies suggest that his future is not completely bleak; recovery may occur. REFERENCES
1. Laurell C-B, and Eriksson S: The electrophoretic alglobulin pattern of serum in al-antitrypsin deficiency, Scand J Clin Lab Invest 15:132, 1963. 2. Sharp HL, Bridges RA, Krivit W, and Freier ER: Cirrhosis associated with alpha~-antitrypsin deficiency: a previously unrecognized inherited disorder, J Lab Clin Med 73:934, 1969. 3. Glasgow JFT, Lynch M J, Hercz A, Levison H, and SassKortsak A: Alphal-antitrypsin deficiency in association with both cirrhosis and chronic obstructive lung disease in two sibs, Am J Med 54:181, 1973. 4. Berg NO, and Eriksson S: Liver disease in adults with alpha~-antitrypsin deficiency, N Engl J Med 287:1264, 1972. 5. Babb RR, Lillington GA, and Kempson RL: Cirrhosis in an adult with emphysema and alphal-antitrypsin deficiency, Am J Dig Dis 18:803, 1973. 6. Palmer PE, Wolfe H J, and Gherardi GJ: Hepatic changes in adult alphai-antitrypsin deficiency, Gastroenterology 65"284, 1973. 7. Eriksson S, and Hfigerstrand I: Cirrhosis and malignant hepatoma in alpha~-antitrypsin deficiency, Acta Med Scand 195:451, 1974. 8. Sharp HL: Alpha,-antitrypsin deficiency, Hosp Prac 5:83, 1971.
Liver disease with a~A T deficiency
25
9. Porter CA, Mowat AP, Cook PJL, Hayner DWG, Shilkin KB, and Williams R: al-Antitrypsin deficiency and neonatal hepatitis, Br Med J 3:435, 1972. 10. ~stergaard PA: Hereditary alpha-l-antitrypsin deficiency and liver cirrhosis in children, Dan Med Bull 20:96, 1973. 11. Aagenes ~, Matlary A, EI~0 K, Munthe E, and Fagerhol M: Neonatal cholestasis in alpha-l-antitrypsin deficient children, Acta Pediatr Stand 61:632, 1972. 12. Castleman B, Scully RE, and McNeely BU: Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Case 50-1973, N Engl J Med 289:1301, 1973. 13. Talamo R C, and Feingold M: Infantile cirrhosis with hereditary alphal-antitrypsin deficiency, Am J Dis Child 125:845, 1973. 14. Fagerhol MK: The Pi-system. Genetic variants of serum a,antitrypsin, Ser Haematol 1:153, 1968. 15. Fagerhol MK, and Laurell C-B: The polymorphisrn of "prealbumins" and alpha~-antitrypsin in human sera, Ciin Chim Acta 16:199, 1967. 16. Laurell C-B: Electroimmuno assay, Scand J Clin Lab Invest 29:Suppl 21-23, 1972. 17. Sass-Kortsak A: Management of young infants presenting with direct-reacting hyperbilirubinemia, Pediatr Clin North Am 21:777, 1974. 18. Cooper DM, Hoeppner V, Cox D, Zamel N, Bryan AC, and Levison H: Lung function in alphal-antitrypsin heterozygotes (Pi type MZ), Am Rev Resp Dis 110:708, 1974. 19. Moroz SP, Cutz E, Balfe JW, and Sass-Kortsak A: Membranoproliferative glomerulonephritis in children with liver disease associated with al-antitrypsin deficiency, Pediatrics (February, 1976). 20. Cutz E, Moroz SP, Cox DW, and Sass-Kortsak A: Histology of the liver in al-antitrypsin deficiency. In preparation. 21. First International Workshop of the Pi System, XXII Colloquium, Protides of the Biological Fluids, Brugge, Belgium, May, 1974. Lancet 1:118, 1975. 22. Gordon HW, Dixon J, Rogers JC, Mittman C, and Lieberman J: Alphal-antitrypsin (A1AT) accumulation in the liver of emphysematous patients with A1AT deficiency, Hum Pathol 3:361, 1972. 23. Bell OF, and Carrell RW: Basis of the defect in c~-lantitrypsin deficiency, Nature 243:410, 1973. 24. Cox DW: Defect in alphal-antitrypsin deficiency, Lancet 2:844, 1973. 25. Cox DW: The effect of neuraminidase on genetic variants of alpha~-antitrypsin, Am J Hum Genet 27:165, 1975. 26. Eriksson S, and Larsson C: PAS-positive inclusion bodies from the fiver in alphal-antitrypsin deficiency, N Engl J Med 292:176, 1975. 27. Lieberman J, Mittman C, and Gordon HW: Alpha~antitrypsin in the livers of patients with emphysema, Science 175:63, 1972. 28. Sveger T, and Laurell C-B: Personal communication, 1975.