328

9. Kahan BD. Cyclosporine. N Engl J Med 1989; 321: 1725-38. 10. Thomson AW, Neild GH. Cyclosporin: use outside transplantation. Br Med J 1991; 302: 4-5. 11. Ellis CN, Fradin MS, Messana JM, et al. Cyclosporine for plaque-type psoriasis. Results of a multidose, double-blind trial. N Engl J Med 1991; 324: 277-84. 12. Sowden JM, Berth-Jones J, Ross JS, et al. Double-blind, controlled,

study of cyclosporin in adults with severe refractory atopic dermatitis. Lancet 1991; 338: 137-40. 13. Brynskov J, Freund L, Rasmussen SN, et al. A placebo-controlled, double-blind, randomised trial of cyclosporine therapy in active chronic Crohn’s disease. N Engl J Med 1989; 321: 845-50. 14. Macleod MA, Sampson WF, Houston AS. Urinary clearance of 113In-DTPA and 99mTc-(Sn) DTPA measured by external arm counting. Urol Res 1977; 5: 71-74. 15. Fisher M, Veall N. Glomerular filtration rate estimation based on a single blood sample. Br Med J 1975; iii: 542. 16. Pocock SJ. Crossover trials. In: Pocock SJ. Clinical trials: a practical approach. Chichester: John Wiley, 1983: 116-17. crossover

Graffenried B, Friend D, Shand N, Schiess W, Timonen P. Cyclosporin A (SandimmunR) in autoimmune disorders. In: Thomson AW, ed. Cyclosporin. Mode of action and clinical applications. London: Kluwer, 1989: 213-50. 18. Mihatsch MJ, Thiel G, Ryffel B. Hazards of cyclosporine A therapy and recommendations for its use. J Autoimmunity 1988; 1: 533-43. 19. Cirillo R, Triggiani M, Siri L, et al. Cyclosporin A rapidly inhibits mediator release from human basophils presumably by interacting with cyclophilin. J Immunol 1990; 144: 3891-97. 20. Granelli-Piperno A, Keane M, Steinman RM. Evidence that cyclosporine inhibits cell mediated immunity primarily at the level of the T lymphocyte rather than the accessory cell. Transplantation 1988; 46 (suppl): 53-60S. 21. Frey FJ, Schnetzer A, Horber FF, Frey BM. Evidence that cyclosporine does not affect the metabolism of prednisolone after renal transplantation. Transplantation 1987; 43: 494-98. 22. Rocci ML, Tietze KJ, Lee J, Harris H, Danzeisen JR, Burke JF. The effect of cyclosporine on the pharmacokinetics of prednisolone in renal transplant patients. Transplantation 1988; 45: 656-60. 17.

von

Enzyme-linked immunosorbent assay for diagnosis of acute sporadic hepatitis E in Egyptian children

Hepatitis E virus (H EV) is thought to be a cause of enterically transmitted non-A, non-B (ET-NANB) hepatitis. Waterborne epidemics have been recorded in many developing countries, mainly affecting young-to-middle-aged adults; sporadic infection and overt illness in children are rare. However, a convenient and sensitive diagnostic test for HEV infection is not yet available. We now report the use of a solid-phase enzyme-linked immunoassay (ELISA) that detects IgM and IgG antibody to HEV. In a prospective study of endemic acute hepatitis during 1986 in rural Benha, Egypt, 15 (42%) of 36 children with NANB hepatitis (from whom convalescent-phase sera were available every 3 months to 9 or 12 months) were positive for anti-HEV-IgG by ELISA. Of 20 sera from healthy Benha children (controls), 5 (25%) were also positive for anti-HEV-IgG. When evaluated for anti-HEV-IgM, 6 of the 15 IgG-positive children, but none of the controls, were IgM positive and were thus regarded as having confirmed acute HEV infections. These 6 cases together with 2

(IgM negative, IgG seroconversion from positive to negative) presented sporadically over 9 months. presumptive

cases

This ELISA is a convenient method for the diagnosis of H EV infection; we have shown that the disease is present in Egypt, that it can occur endemically as sporadic cases, and that children do have overt infection.

Introduction E virus (HEV) and hepatitis C virus (HCV) be the principal causes of what was previously known as non-A, non-B (NANB) hepatitis.1-7 The two viruses have epidemiologically distinct modes of

Hepatitis

seem to

transmission, HCV being parenterally transmitted1,8 and HEV being transmitted by the faecal-oral route (enterically transmitted NANB or hepatitis E) .4-5,91-13 HEV is largely waterborne; epidemics have been reported from India, Pakistan, Bangladesh, Nepal, Burma, Borneo, Algeria, Somalia, Sudan, Ivory Coast, Mexico, China, and the Soviet Union,4,5,9,12,14,15 and the disease is also thought to be

Sporadic infection seems to be uncommon, and is mainly in travellers.14,16 Hepatitis E has its highest attack rate in young-to-middle-aged adults, with highest case-fatality rates in pregnant women.4,7,8,15 Overt infections endemic. seen

in children are rare.17,18 Until now, diagnosis of HEV infection has been based on detection by immune electronmicroscopy of virus-like particles (27-34 nm) in faecal specimens from acutely ill individuals, in the absence of (a) serological responses to hepatitis viruses A-D (HAV, HBV, HCV, HDV), EpsteinBarr virus (EBV), cytomegalovirus (CMV), and

hepatotropic arboviruses, or (b) exposure to hepatotoxic drugs. 2-5,12,13,19,20 However, immune electronmicroscopy is not a satisfactory test for broad surveys or for routine diagnosis because expensive equipment is required and many stool specimens from infected patients do not contain enough virus-like particles for detection.4,20 There is also an immunofluorescence method2l in which a fluoresceinconjugated IgG fraction from an anti-HEV serum with high titre is used to detect viral antigen in liver tissue sections of HEV-infected cynomolgus macaques .21 An enzyme-linked immunosorbent assay (ELISA) based on clonal recombinant HEV antigen has been developed ADDRESSES Department of Epidemiology and Biostatistics, University of California, San Francisco, California 94143 (Prof R. Goldsmith, MD); Molecular Virology Department, Genelabs Incorporated, Redwood City, California (P.O. Yarbough, PhD, G. R. Reyes, MD, K E. Fry, PhD, K. A. Gabor, BS) USA; Departments of Clinical Pathology (M. Kamel, MD) and Tropical Medicine (S Zakaria, MD), University of Cairo; Department of Pediatrics, Benha University, Benha (S. Amer, MD); and Department of Medicine, Ains Shams University, Cairo (Y. Gaffar, MD), Egypt. Correspondence to Prof Robert Goldsmith.

329

(Yarbough et al, unpublished) to detect HEV-IgM and HEV-IgG antibodies. We here report the use of this ELI SA (and an evaluation of several antigens) for the diagnosis of acute sporadic HEV infection among Egyptian children with NANB hepatitis.

TABLE I-ANTI-HEV TEST RESULTS FOR CHILDREN WITH NANB HEPATITIS FROM WHOM CONVALESCENT-PHASE SERA WERE AVAILABLE AND FOR CONTROL CHILDREN

I

Months after hosppitl addmission

Total

Patients and methods Patients 81 children (age range 2 months to 15 years; 54% boys) were admitted to hospital for acute hepatitis (jaundice of less than 4 weeks duration, moderate-to-severe illness, and at least one aminotransferase 2-5 times the upper limit of normal) at the Benha Fever Hospital, Kaliobiya Govemorate, Egypt, from December, 1985, to January, 1987. A few children were from Benha city, but most were residents of nearby rural communities. At admission, demographic information was obtained. When possible, venous blood was taken at admission and at 3, 6, 9, and 12 months. Sera were divided into aliquots and stored at - 20°C. Liver function tests were alanine (ALT) and aspartate (AST) aminotransferase (normal, 45 IU/1) and total bilirubin (5-1-17 umol/1). Control sera were obtained from 20 healthy children (age range 5 to 12 years; 65% boys) who lived in rural areas near Benha.

Serology HAV, HBV, HDV, CMV, EBV-Admission

sera were tested the Clinical Pathology Laboratory, Cairo, for anti-HAV-IgM and for HBV markers (HBsAg, anti-HBs, antiHBc[IgM], anti-HBc, HBeAg, anti-HBe) by ELISA. Sera positive for HBsAg or anti-HBc-IgM were tested for anti-HDV by

under code

at

competitive radioimmunoassay (Abbott, Wiesbaden-Delbenheim, Germany). The sera were also tested for anti-CMV(IgM and IgG) and anti-EBV-IgM by ELISA (Abbott). HBV tests were repeated at 6 months. Interpretation of the tests has been previously reported.22 Sera that were negative in these tests were then tested for HCV and HEV if convalescent-phase sera were available. HCV--Coded admission and follow-up sera were tested in duplicate at Genelabs and at the Cairo Pathology Laboratory by immunoassay (Abbott) for antibodies to C100-3. At Genelabs, sera were also tested by the ’Monolisa’ anti-HCV assay (SanofiDiagnostics Pasteur, Marnes-la Coquette, France) for antibodies against the nucleocapsid and non-structural regions (N S3) of HCV. The criterion

to

establish

current

infection

was

seroconversion

(negative to positive). HEV-Anti-HEV-IgG and anti-HEV-IgM tests by ELISA with four recombinant antigens23 were done twice under code at Genelabs. Serum was considered reactive for anti-HEV if the ELISA signal to HEV-glutathione-S-transferase24 fusion proteins was three times higher than that with non-recombinant glutathione-S-transferase proteins. A patient was regarded as and-HEV-IgG positive if a single serum was reactive to more than one antigen or if consecutive sera were reactive. Current infection was defined as IgM reactivity, in a single serum, with one or more of the recombinant antigens. HEV clones producing Mexico (M) antigen 406’3-2(M) and 406’4-2(M) were isolated from a lambda gtl 1 cDNA library.23 The library was constructed from an extract of a human stool sample collected during an epidemic of enterically transmitted NANB in Mexico in 1986.15 Two similar epitopes, HEV 3-2(B) and 4-2(B), were generated by polymerase chain reaction with HEV Burma (B)-derived primers followed by cloning into the expression vector lambda gtl11.23 The four HEV antigens were subsequently expressed in the pGEXl vector system24 as fusions with the C terminus of Sj26, a 26-kDa glutathione-Stransferase protein. HEV antigens (200 ng in 100 nl) were pipetted into polystyrene microtitre plates (Nunc-Immuno). After incubation for 1 h at 37°C followed by overnight incubation at 4°C, the wells were washed with phosphate-buffered saline (PBS)0 05% ’Tween’ and blocked with 200 ul 1 % bovine serum albumin in PBS for 2 h at room temperature. The plates were rewashed with PBS-tween; serum samples diluted 1/100 in antibody incubation buffer (10 mmol/1 "tris" pH 8-0, 150 mmol/1 NaCI, 1% gelatin) were then added to the wells and incubated for 2 h at room temperature. The wells were washed to remove any unbound

r

____.

_

..

i

..

i

..

..

i *No positive with one or more of four test antigens/total NAN B hepatitis tested tNo positive with antigen 3-2 (M) tested

..

i

i no

i

_,__ , ,

of the 36 children with

/total no of the 15 anti-HEV-tgG-positive children

primary antibody and incubated at room temperature with 100 µl of 1 I1gjml of alkaline phosphatase-conjugated goat anti-humanimmunoglobulin (either for 2 h with anti-IgG [gamma-chain specific] or for 1 h with anti-IgM [mu-chain specific]). After a fina! washing, 100 III of alkaline phosphate substrate (Sigma, UK) in diethanolamine buffer (pH 9-8) was added at room temperature for 30 min. Absorbance

was

read at 405

nm.

Results Of the 81 children tested, acute HAV infection was diagnosed in 5 (6-2%) and acute HBV infection in 11 (136%); none was positive for HDV, EBV, or CMV infection. By exclusion, therefore, NANB hepatitis was diagnosed in 65 (80-2%) children. HEV and HCV serology Convalescent-phase sera were available from 36 of the 65 children; the number of sera tested at each 3-month postadmission interval is shown in table 1. 20 children were aged 1-3 years, 10 were 4-6 years, 3 were 7-9 years, and 3 were 10-14 years; 60% were boys. Anti-HCV--Only 1 of the 36 children was positive for anti-HCV-at admission and at the 3-month intervals. Anti-HEV-lgG-Sera from 15 (41’7%) of the 36 children (67% boys) were positive with one or more of the four test antigens at various follow-up times (table I). 6 of the 15 were IgG/IgM positive (confirmed acute infection), 2 were IgG positive/IgM negative with seroconversion of IgG from positive to negative (presumptive infection), and 7 were IgG positive/IgM negative without seroconversion of IgG from positive to negative (present or postinfection). In the remaining 21 patients (58-3%) the cause of NANB hepatitis was unknown. On admission, 10 (39%) of the 26 available sera tested were seropositive; the proportion of patients with antibody gradually diminished until at 12 months, only 2 of the 12 tested were positive (table I). Seroconversion from

acute-phase-positive

to

convalescent-phase-negative

occurred between the 3rd and 6th month interval among 6 of the 15 IgG-positive patients (table n). Anti-HEV-lgM-Available admission sera were tested for IgM with three antigens; 3-2(B) was not included because of its low sensitivity for IgG testing. Of the 15 IgM-positive patients, admission sera from 11 were tested for IgM and 6 were positive only to antigen 3-2(M) (tablesI and II). The other 4 sets of sera were negative to the three antigens at all times after hospital admission. The 6 IgM-positive patients seroconverted at 3 months and remained negative thereafter.

330

TABLE II-FINDINGS FOR CHILDREN WHO SHOWED IgM OR IgG ANTIBODY SEROCONVERSION (POSITIVE TO NEGATIVE) I

I

.

I

NT not tested, + =positive.,

=

-

[I

negative

Control sera--Of the 20 control sera, 5 (25%) were positive for anti-HEV-IgG; none was positive for anti-

HEV-IgM (table I).

Comparative reactivity of four HEV antigens for anti-HEV-lgG The results of parallel assays to assess the relative reactivity of the four test antigens in detecting anti-HEVIgG are shown in table ill. Antibody to 3-2(M) was found in all 15 seropositive patients and was the most sensitive (ie, detected the highest proportion) and long-lasting. Although antibody to 4-2(B) was detected in only 12 of the 15 IgG-positive patients, positive sera that were negative with 3-2(M) were found in 2 patients at three test periods. Antibody to 3-2(B) and 4-2(M) was not detected in the absence of antibody to the other two antigens. The 8 confirmed or presumptive cases of hepatitis E presented sporadically from March to November, 1986. Among the 15 IgG-reactive patients, which included these 8 cases, there was no clustering by household or area, no secondary cases in households, and no association of cases with a common water source or with possible risk factors (eg, parenteral injections, surgery, transfusions, admission to hospital, or exposure to hepatitis cases during the 6 months before onset of acute symptoms). No patient had had a previous episode of jaundice or had used a hepatotoxic drug. During the study, there was no apparent increase in the usual endemic incidence of acute hepatitis in either children or adults or in mortality in pregnant women. The principal clinical findings on admission for the confirmed/presumptive cases were jaundice (8), fever (8), nausea and/or vomiting (4), and hepatomegaly (8); the mean concentration was bilirubin 74 jjmol/1, ALT 182 IU/1, and AST 145 IU/1. At the 6-month follow-up, 3 patients had slightly enlarged livers, but at 9 and 12 months the organs were no longer palpable and liver function had returned to normal. TABLE III-RELATIVE REACTIVITY OF FOUR TEST ANTIGENS TO DETECT ANTI-HEV-lgG

*Of the 15 IgG-posltlve children tWith one or more of the four antigens

Discussion The ELISA reported here detects both anti-HEV-IgM and anti-HEV-IgG and is a convenient method for the diagnosis of acute or past HEV infection. Of the children with acute NANB hepatitis tested for IgG antibody, 42% were HEV-IgG positive. But 25% of the controls also had anti-HEV-IgG. These findings could point to past infection in both groups; however, at admission 6 of 11 children were anti-HEV-IgM positive and thus were regarded as having acute hepatitis E, whereas none of the control sera was anti-HEV-IgM positive. Our lack of serum to test sera for IgM on admission for 4 of the 15 patients may have meant that no diagnosis could be made in some instances. It is noteworthy that IgM was found only in admission sera and in each instance was accompanied by IgG. We interpreted acute HEV infection by one of two serological patterns: first, the diagnosis was confirmed if both IgM and IgG were found in the acute-phase serum and if subsequently IgM disappeared in an early convalescentphase serum; and, second, diagnosis was presumptive in 2 patients who, although negative for acute-phase IgM, showed IgG seroconversion from positive in acute-phase to negative in convalescent-phase sera. This sera interpretation of presumptive infection is supported by the finding that during follow-up of the 6 IgM-confirmed infections, 4 patients lost detectable IgG antibody. The decrease in IgG reactivity over time confirms reports by others of diminishing reactivity or seroconversion from positive to negative when acute and serial convalescentphase human or monkey sera were tested by immune electronmicroscopy for agglutination of virus-like particles in stools from HEV-infected individuals.5,13,17 The four antigens used in our ELISA were known to react specifically with acute-phase and convalescent-phase sera from HEV-infected individuals and also with sera from experimentally infected cynomolgus macaques.23 In our study, 3-2(M) not only was the most sensitive antigen and detected IgG antibody for the longest time but also was the only antigen that detected IgM. We have also shown that the new ELISA is specific for anti-HEV (unpublished). To confirm the sensitivity of the ELISA, prospective studies that correlate autologous IgM and IgG with detection of HEV in stools by immune electronmicroscopy need to be done. Sporadic cases of HEV infection have been rarely reported, probably because diagnosis is not practicable with current methods. To our knowledge, this is the first report of HEV infection in Egypt. However, we do not know whether HEV infection is endemic in Egypt, either continuously or as a recurrent low-grade process, or whether outbreaks occur. Outbreaks of hepatitis E have been characterised by overt cases among young-to-middle-aged adults, with few overt cases in children.18 This finding has led to the proposal that, like HAV infections, most HEV infections in children are subclinical.’ Our results indicate that HEV infection in children can manifest as acute illness that may require admission to hospital, although the number of overt infections in children is probably small relative to the number of subclinical infections." We have also confirmed reports by other investigators that clinical features of acute hepatitis E do not distinguish it from other forms of acute viral hepatitis, with the possible exception of the occurrence of arthralgia;2,7 there have been no reports of progression to chronic liver disease.2,4

331

That we could not precisely diagnose 28 of the 36 cases of hepatitis suggests either that our testing methods for HEV or other known hepatitis viruses were inadequate or that there may be other viruses that cause some NANB hepatitis

infections.6,7 of

The study was done partly under the auspices University Linkage Project for Study of Liver Disease in Egypt, Maxy Project 842084. We thank Ms Karyn Yun, Genelabs; Ms Margaret Wright, Clinical Pathology Laboratory, Cairo; and Ms Rita Williams, University of California, San Francisco, for assistance in various phases of the study. Dr Sahen Dahborh kindly provided the sera used as control specimens.

REFERENCES 1. Alter HJ, Purcell RH, Shih JW, et al. Detection of antibody to hepatitis C virus in prospectively followed transfusion recipients with acute and chronic non-A, non-B hepatitis. N Ergl J Med 1989; 321: 1494-500.

2. Alter MJ. Non-A, non-B hepatitis: sorting through a diagnosis of exclusion. Ann Intern Med 1989; 110: 583-85. 3.Bradley DW, Maynard JE. Etiology and natural history of posttransfusion and enterically-transmitted non-A, non-B hepatitis. Seminar Liver Dis 1986; 6: 56-62. 4. Bradley DW. Enterically-transmitted non-A, non-B hepatitis. Br Med Bull 1990; 46: 442-61. 5. Purcell RH, Ticehurst JR. Enterically transmitted non-A, non-B hepatitis: epidemiology and clinical characteristics. In: Zuckerman AJ, ed. Viral hepatitis and liver disease. New York: Alan R. Liss, 1988: 131-37. 6. Editorial. The A to F of viral hepatitis. Lancet 1990; 336: 1158-59. 7. Tabor E. The three viruses of non-A, non-B hepatitis. Lancet 1985; i: 743-45. 8. Alter MJ, Sampliner RE. Hepatitis C. And miles to go before we sleep. N Engl J Med 1989; 321: 1538-40. 9. Wong DC, Purcell RH, Sreenivasan MA, Prasad SR, Pavri KM. Epidemic and endemic hepatitis in India: evidence for a non-A, non-B hepatitis virus aetiology. Lancet 1980; ii: 876-79. 10. Khuroo MS. Study of an epidemic of non-A, non-B hepatitis: possibility of another human hepatitis virus distinct from post-transfusion non-A, non-B type. Am J Med 1980; 68: 818-24. 11. Balayan MS, Andjaparidze AG, Savinskaya SS, et al. Evidence for a virus in non-A, non-B hepatitis transmitted via the fecal-oral route. Intervirology 1983; 20: 23-31. 12. Zuckerman AJ. Hepatitis E virus: the main cause of enterically transmitted non-A, non-B hepatitis. Br Med J 1990; 300: 1475-76. 13. Bradley DW, Krawczynski K, Cook EH, et al. Enterically transmitted non-A, non-B hepatitis: serial passage of disease in cynomolgus macaques and tamarins and recovery of disease-associated 27- to 34-nm viruslike particles. Proc Natl Acad Sci USA 1987; 84: 6277-81. 14. de Cock KM, Bradley DW, Sandford NL, Govindarajan S, Maynard JE, Redeker AG. Epidemic non-A, non-B hepatitis in patients from Pakistan. Ann Intern Med 1987; 106: 227-30. 15. Velazquez O, Stetler HC, Avila C, et al. Epidemic transmission of enterically transmitted non-A, non-B hepatitis in Mexico 1986-1097. JAMA 1990; 263: 3281-85. 16. Skidmore SJ, Yarbough PO, Gabor KA, Tam AW, Reyes GR, Flower AJE. Imported hepatitis E in UK. Lancet 1991; 337: 1541. 17. Panda SK, Datta R, Kaur J, Zuckerman AJ, Nayak NC. Enterically transmitted non-A, non-B hepatitis: recovery of virus-like particles from an epidemic in south Delhi and transmission studies in rhesus monkeys. Hepatology 1989; 10: 466-72. 18. Arankalle VA, Chadha MS, Mehendale SM, Banerjee K. Outbreak of enterically transmitted non-A, non-B hepatitis among schoolchildren. Lancet 1988; ii: 1199-1200. 19. Sreenivasan MA, Arankalle VA, Sehgal A, Pavri KM. Non-A, non-B epidemic hepatitis: visualization of virus-like particles in the stool by immune electron microscope. J Gen Virol 1984; 65: 1005-07. 20. Arankalle VA, Ticehurst J, Sreenivasan MA, et al. Aetiological association of a virus-like particle with enterically transmitted non-A, non-B hepatitis. Lancet 1988; i: 550-54. 21. Krawczynski K, Bradley DW. Enterically transmitted non-A, non-B hepatitis: identification of virus-associated antigen in experimentally infected cynomolgus macaques. J Infect Dis 1989; 159: 1042-49. 22. Zakaria S, Goldsmith RS, Kamel MA, El-Raziky EH. The etiology of acute hepatitis in adults in Egypt. Trop Geogr Med 1988; 40: 285-92. 23. Yarbough PO, Tam AW, Fry KE, et al. Hepatitis E virus: identification of type-common epitopes. J Virol 1991; 65: 5790-97. 24. Smith DB, Johnson KS. Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione-S-transferase. Gene 1988; 67: 31-40.

SHORT REPORTS Balloon dilatation of the arterial duct in congenital heart disease

The

systemic circulation of newborn infants with congenital left-heart obstruction is supplied from the right ventricle via a patent arterial duct between the pulmonary artery and descending aorta. The duct closes during the first few days of life, but infusion of prostaglandin E2 can prevent closure in some cases. We report four newborn infants (aged 3-8 days) with intractable heart failure due to severe obstruction of the left heart in the presence of a closing arterial duct. Infusion of prostaglandin E2 did not improve their clinical condition. Cardiac catheterisaton and balloon dilatation of their arterial ducts resulted in a dramatic improvement in the babies’ clinical condition; during subsequent surgical repair of the infants’ hearts, the arterial ducts were found to be widely patent. Balloon dilatation gives immediate and sustained wide patency of the arterial duct in infants who do not respond

adequately to prostaglandin E2. In newborn infants with severe congenital left-heart obstructions, such as an interrupted aortic arch, the right ventricle supplies systemic blood flow via a patent arterial duct between the pulmonary artery and descending aorta. As the duct closes within the first few days of life, hypoperfusion of the duct-dependent circulation results in metabolic acidosis and oliguria.1 At present, management of this defect depends on reopening the arterial duct with infusions of prostaglandin E! (PGE!) or PGEz to improve the baby’s condition before surgical repair. However, PGEz infusions produce clinical improvement in only 76% of patients with an interrupted aortic arch,3and surgical intervention during systemic hypoperfusion and severe metabolic acidosis has a high mortality.4 We report balloon dilatation of the arterial duct in four newborn infants with severe aortic obstructions and continuing clinical deterioration despite infusion of PGE2. The four babies (aged 3-8 days, weight 2-2-3-9 kg) presented with signs of severe heart failure, absent femoral pulses, and metabolic acidosis (mean pH 706). Cross-sectional echocardiography showed that patients 1 and 2 had interruption of the aortic arch between the left common carotid artery and left subclavian artery with a large ventricular septal defect, patient 3 had severe aortic stenosis with hypoplasia of the left ventricle and mitral valve, and patient 4 had a complete arterioventricular septal defect with aortic arch atresia distal to the left subclavian artery. In all the patients, the right ventricle was dilated and contracted poorly and the arterial duct was patent although constricted. Treatment with intravenous PGE2 (10-100 ng/kg per min), sodium bicarbonate, and dopamine did not lead to clinical improvement. Angioplasty of the arterial duct was done via a 1-6 mm diameter end-hole catheter (’Multipurpose A2’, Cordis, Brussels, Belgium)