30

congenita and myotonic dystrophy in the same study.10 Furthermore, efficacy of treatment is difficult to measure objectively.3 We were aware that objective measurement would not be possible in our index case, but also thought that placebo effect was likely to be negligible. Trimeprazine is an aminated phenothiazine compound with 1- receptor blocking activity, and it is commonly used for its sedative effects in childhood. It has antimuscarinic activities with some quinidine-like effects, which may explain its beneficial effect in myotonia. Although trimeprazine has been well tolerated and effective in this family, there are reservations about the long-term use of any phenothiazine compound because of the association with extrapyramidal side-effects, and in the short term acute dystonias or dyskinesias may occur, especially in children. Our observations suggest that in adults, antazoline merits further investigation as a treatment for myotonia.

developments in electromyography and clinical neurophysiology. Basel: Karger, 1973; vol 1: 413-19. 3. Durrelli L, Mutani R, Piredda S, Fassio F, Delsedime M. The quantification of myotonia: a problem in the evaluation of new antimyotonic drugs. J Neurol Sci 1983; 59: 167-73. 4. Bryant SH. The electrophysiology of myotonia with a review of congenital myotonia of goats. In: Desmedt JE, ed. New developments in electromyography and clinical neurophysiology. Basel: Karger, 1973; vol 1: 420-50. 5. Peter JB, Dromgoole SHM, Campion DS, et al. Experimental myotonia and hypocholesterolaemic agents. Exp Neurol 1975; 49: 115. 6. Rowland LP. Diseases of the motor unit, the motor neuron, peripheral nerve and muscle. In: Kandel ER, Schwartz JH, eds. Principles of neural science. Amsterdam: Elsevier, 1985: 196-208. 7. Adrian RH, Marshall MW. Action potentials reconstructed in normal and myotonic muscle fibres. J Physiol (Lond) 1976; 258: 125-43.

8. Lipicky RJ, Bryant SH, Salmon JH. Cable parameters, sodium, potassium, chloride and water content and potassium efflux in isolated external intercostal muscle of normal volunteers and patients with myotonia congenita. J Clin Invest 1973; 50: 2091-102. 9. Munsat TL. Therapy in myotonia. A double blind evaluation of diphenylhydantoin, procaineamide, and placebo. Neurology 1967; 17:

We thank Dr J. Payan for his help with electrophysiological studies; Dr M. Baraitser, Department of Clinical Genetics, Hospital for Sick Children, London, for his advice; Ciba Geigy Pharmaceuticals for supplies of ’Antistin’;

359-67. 10. Streib W. Successful

and members of the

11.

family for their co-operation and support. REFERENCES

12. 1. Becker PE. Generalised

non-dystrophic myotonia. In: Desmedt JE, ed. New developments in electromyography and clinical neurophysiology. Basel: Karger, 1973; vol 1: 407-12. 2. Kuhn E. Myotonia—the clinical evidence. In: Desmedt JE, ed. New

treatment with tocainide of recessive generalised congenital myotonia. Ann Neurol 1986; 19: 501-04. Cook JD, Henderson-Tilton AC. Beneficial responses to a calcium channel antagonist in myotonic syndromes. Neurology 1984; 34 (suppl 1): 193-94. Grant R, Sutton DL, Behan PO, Ballantyre JP. Nifedipine in the treatment of myotonic dystrophy. J Neurol Neurosurg Psychiatry 1987;

50: 199-206. 13. Gerst JW. Brumback RA, Staton RD. Lithium induced improvement of myotonia. J Neurol Neurosurg Psychiatry 1984; 47: 1044-45.

Evaluation of enzyme immunoassays for antibody to human T-lymphotropic viruses type I/II

To evaluate the sensitivity and specificity of HTLV-I/ II assays, serum from 1100 pregnant Haitian women was tested with seven commercially available HTLV I/II assays. Serum that was found to be reactive in any assay was analysed by western blot and all indeterminate samples were further characterised by radioimmunoprecipitation assays (RIPA). 59 (5·4%) samples were HTLV I/I I antibody positive by western blot and/or RIPA. The sensitivity of these seven assays ranged from 93·2% to 100%, with the ’Recombinant HTLV-I’ (Cambridge Bioscience) and ’Serodia HTLV-I’ (Fujirebio) assays having the highest sensitivity (100%). The specificity of these assays ranged from 98·4% to 100%, with the Abbott assay having the highest specificity (99·5%, 100%) according to two different methods of evaluation. Whether the antigens used in any assay were whole disrupted virus or recombinant gene products made no difference. The low positive predictive values of some of these assays (71·8-91·7%), even in a high prevalence population, and the need for RIPA to test indeterminate sera, indicate that for routine screening of blood donors there is still room for

improvement assays for

both in

screening and confirmatory

HTLV-I/II. Introduction

Human T-lymphotropic virus type I (HTLV-I) has been identified as a cause of adult T-cell leukaemia/lymphoma (A TL)I-3 and of a neurological disorder called tropical spastic paraparesis (TSP) or HTLV-1-associated myelopathy (HAM).4-7 HTLV-I infection is endemic primarily in southwestern Japan, the Caribbean basin, and selected areas of Africa, and it has recently been identified in intravenous drug users and blood transfusion recipients in ADDRESSES Laboratory of Immunoregulation, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Maryland (R L Kline, MS, T. C Quinn, MD); Division of Infectious Diseases, Johns Hopkins University School of Medicine, Baltimore (T. Brothers, BS); Department of International Health, Johns Hopkins University School of Hygiene and Public Health, Baltimore, Maryland (N Halsey, MD); Centers for Development and Health, Port-au-Prince, Haiti (R Boulos, MD); and Retrovirus Diseases Branch, Centers for Disease Control, Atlanta, Georgia (M D. Lairmore, DVM), USA. Correspondence to Dr T C Quinn, Division of Infectious Diseases, Johns Hopkins Hospital, Blalock 1111, 600 N Wolfe Street, Baltimore, Maryland 21205, USA.

31

Europe, and Latin American HTLV-I can be transmitted perinatally (especially postnatally by breastfeeding), sexually, or parenterally by sharing of blood contaminated needles and syringes or by transfusion of infected blood. 10,13 -1’ Transmission of HTLV-1 infection by blood transfusion has been described in Japan and the US, with a 48% to 82% seroconversion rate in recipients of HTLV-I contaminated blood.12,14 The finding that the USA,

HTLV-1 infection has been linked to blood transfusions has raised suspicion about the safety of the nation’s blood supply and prompted the recommendation that all blood donations and cellular components be screened for HTLV-1. In a recent study of 39 898 blood donors, 10 individuals (0-025%) were found to be seropositive for HTLV-Llz A closely related human retrovirus, HTLV-II, originally isolated from 2 patients with hairy cell leukaemia, has also been found in intravenous drug users. It is not known whether HTLV-11 is associated with any clinical disease; a study of 21 patients with hairy cell leukaemia did not reveal any association between the disorder and HTL V - I.17 HTLV-II infection was reported in an isolated symptomfree population of Guaymi Indians in Panama.18 Because of antibody cross-reactivity between HTLV-1 and HTLV-II, results of serological surveys for either virus are really a reflection of the seroprevalence of both, hence the designation HTLV-1/11. Differentiation between the viruses requires specific peptide tests, culture, and/or tests based on polymerase chain reaction. Our study compares the reliability, sensitivity, and specificity of seven

commercially available enzyme immunoassays (EIAs) for antibody to HTLV-1/11, as assessed by western blot (WB) and radioimmunoprecipitation assays (RIPA). Material and Methods Serum was collected from 1100 pregnant Haitian women as part of a prospective perinatal study on HIV-I and HTLV-1.19 Samples that had been frozen at - 70°C were thawed and kept at 4°C for the duration of the study. The assays used were:

Code Kit A ’Recombinant HTLV-I

B

Antigen source

EIA’ Genetically

(Cambridge Bioscience, Worcester, Massachusetts, USA) ’Synth EIA HTLV-I EIA’ (UBI-Olympus, Lake Success, New York, USA)

engineered HTLV-I (gp21) genederived peptide Whole disrupted virus plus synthetic peptide corresponding to

Results Of the 1100

samples tested,

59 (5-4%) were positive by and RIPA. blot, EIA, Using method X, 5 samples 1036 were negative. Using were indeterminate and method Y, 25 samples were indeterminate and 1016 were western

Whole virus

disrupted

TABLE II-TEST PERFORMANCE WHEN ALL INDETERMINATE SAMPLES ARE EXCLUDED FROM ANALYSIS (METHOD Y)

Whole ’Vironostika HTLV-1 Microelisa system’ (Organon virus Teknika, Durham, North

disrupted

viral envelope ’Retro-tek HTLV-I ELISA’ Whole disrupted (Cellular Products, Buffalo, virus

D

New York, USA) Whole ’Abbott HTLV-I EIA’ (Abbott Laboratories, North virus

Chicago, Illinois, USA) ’DuPont HTLV-I ELISA’

(DuPont, Wilmington, Delaware, USA)

Carolina, USA) G

labelled HTLV-I infected MT-2 cells. The RIPA tested for antibodies to gag (p24,p28) and env (gp68) proteins. Results from the western blot and RIPA were combined and any indeterminate sample that had antibodies to gag and env by either was considered positive.16.20 Samples that were still indeterminate were analyzed in two different ways. In method X, samples that were western blot indeterminate and RIPA negative were counted as negative samples. Any sample that was western blot indeterminate and RIPA indeterminate was designated as indeterminate and excluded from analysis. In method Y, all western blot indeterminate and RIPA negative or indeterminate samples were considered unresolvable and were excluded from analysis. All samples that showed no reactivity by western blot and RIPA or were negative on all HTLV-I EIAs were considered negative. Sensitivity, specificity, and positive and negative predictive values were calculated with methods described by Griner et al. 21

disrupted

C

F

was considered positive by the assay. All samples that were initially reactive by any assay were further tested by western blot (DuPont) and judged positive if antibodies to gag (pl9, p24) and env (gp46) were detected.*" Samples that showed reactivity to HTLV-1 gene products but did not fulfil the requirements listed above were deemed indeterminate. All indeterminate were further tested samples by radioimmunoprecipitation assay (RIPA). RIPA made use of

reactive, the sample

negative. The only EIA assays that detected all 59 HTLV-I positive samples were A and G (table 1). B did not detect 1 sample as positive, F 2, C and E 3 each, and D 4. Sensitivities ranged from 100% (A and G) to 93-2% (D) and were not affected by the method of analysis. Specificity depended on the method used for analysis. D had the highest specificity with either method X or method Y (tables I and II). With method X specificities ranged from 99 5% (D) to 97-8% (E); with method Y they ranged from 100% (D) to 98-4% (A).

env

segment of HTLV-1

E

TABLE I-TEST PERFORMANCE WHEN ALL WESTERN BLOT INDETERMINATE AND RIPA NEGATIVE SAMPLES ARE DEEMED NEGATIVE BY METHOD X

’Serodia HTLV-I’

(Fujirebio Whole disrupted

Inc, Tokyo, Japan)

virus

Every serum sample was tested with all seven assays according to the manufacturer’s specifications. Samples that were initially reactive were re-tested in duplicate. If either duplicate retest was

32

TABLE III-WESTERN BLOT AND RIPA CHARACTERISTICS OF EIA-NEGATIVE SAMPLES

Positive predictive values were low for most assays whether analysed by method X or method Y (tables I and n). With method X, they ranged from91-7% (D) to 71-8% (E) and with method Y the range was 100% (D) to 78-7% (A). Negative predictive values were the same whether method X or Y was used and ranged from 100% (A and G) to 99-6%

(D). Samples that were negative by at least one EIA but positive by western blot and RIPA all showed reactivity to gag on western blot, and to env on RIPA (table in). 1 positive sample with weak gag reactivity by western blot and env reactivity by RIPA was detected as positive only by A and G.

Discussion EIA methods for screening sera for antibody, the most important factors in determining how well an assay performs are the sensitivity and specificity of the assays, and the positive and negative predictive values of the assay, which are dependent on the prevalence of the infection in the population. The Cambridge Bioscience ’Recombinant HTLV-I EIA’ (A) and Fujirebio ’Serodia HTLV-I’assay (G) were the only assays that detected all 59 samples judged positive by our criteria. However, their specificities were among the lowest. The Abbott HTLV-1I EIA (D) showed the highest specificity, but did not detect 4 positive samples. All of the assays detected all strong positive samples. The 4 positive samples that were detected by only some of the 7 assays were deemed indeterminate until RIPA confirmed the presence of antibody to env proteins. There are some difficulties in relating the sensitivities and specificities of HTLV-1 EIA kits to results by "confirmatory assays" such as western blot or RIPA. Other investigators have found that the western blot assay seems less sensitive than RIPA for the detection of env antibodies, and RIPA seems less sensitive than western blot for the detection of gag antibodies.22 Hence some investigators are recommending that western blot and RIPA be used together as confirmatory testing for HTLV-1.’° In our study 18 (30-5%) of the 59 HTLV-1/11 positive samples required RIPA for confirmation since the western blots were indeterminate. However, confirmation by RIPA would be nearly impossible in areas where HTLV-1 is endemic because of the complexity and prohibitive cost of RIPA. Thus, it is very important that a serological confirmatory assay be developed that is as sensitive for HTLV-Igag as for HTLV-I env. Another difficulty is that some samples remain indeterminate after confirmatory testing. Reasons for indeterminate findings include the presence of non-specific When

evaluating

antibodies in serum, the presence of viral cross-reacting antibodies such as HTLV-11, or early seroconversion.11,11 The only methods for resolving the reaction in an indeterminate sample reaction are to re-test the patient to look for seroconversion, or to do more specific and complicated tests such as viral culture or polymerase chain reaction. We excluded indeterminate samples in calculating sensitivity and specificity since follow-up serum samples were not available. The cost of a test is related to volume and factors such as packaging with other tests (HIV-1, hepatitis B virus) and equipment. In general, those assays that use whole disrupted virus are cheaper than those that use recombinant and synthetic peptides. It has been suggested that the specificity of recombinant and synthetic peptide assays should be better than those that use whole disrupted virus. However, our data show that the specificity of the whole disrupted virus assays was as good and in some cases better than recombinant and synthetic peptide assays. A possible explanation for the lack of increased specificity is that recombinant assays may contain bacterial (Escherichia coli) antigens that are recognised by antibodies in non-infected individuals. In a low prevalence population such as routine blood donors, the low positive predictive values of some assays may result in the discarding of uninfected blood. When confirmatory assays are insensitive for antibody to envelope proteins, many samples will be labelled as indeterminate and complicate counselling of the individual about infection status. These issues, added to the low sensitivity of some assays, emphasise the urgency of the need to improve the current screening assays and to develop more practical and sensitive confirmatory assays. In view of the clinical importance of HTLV-1 and the high prevalence of HTLVII in selected populations,25 improved assays that discriminate between HTLV-I and HTLV-11 should also be developed. This study was supported Institutes of Health.

in part by grant RO IA126521 from the National

REFERENCES 1. Poiesz BJ, Ruscetti FW, Gazdar AF, Bunn PA, Minna JD, Gallo RC. Detection and isolation of type C retrovirus partciles from fresh and

cultured lymphocytes of a patient with cutaneous T-cell lymphoma. Proc Natl Acad Sci USA 1980; 77: 7415-19. 2. Catovsky D, Greaves MF, Rose M, et al. Adult T-cell lymphomaleukaemia in blacks from the West Indies. Lancet 1982; i: 639-43. 3. Minamoto GY, Gold JW, Scheinberg DA, et al. Infection with human T-cell leukemia virus type I in patients with leukemia. N Engl JMed 1988; 318: 219-22. 4. Gessain A, Barin F, Vernant JC, et al. Antibodies to human Tlymphotropic virus type-1 in patients with tropical spastic paraparesis. Lancet 1985; ii: 407-10. 5. Osame M, Usuku K, Izumo S, et al. HTLV-I Associated myelopathy, a new clinical entity (letter). Lancet 1986; i: 1031-32. 6. Bhagavati S, Ehrlich G, Kula RW, et al. Detection of human T-cell lymphoma/leukemia virus type 1 DNA and antigen in spinal fluid and blood of patients with chronic progressive myelopathy. N Engl J Med

1988; 318: 1141-47. 7. Brew BJ, Price RW. Another retroviral disease of the nervous system. N Engl J Med 1988; 318: 1195-96. 8. Hinuma Y, Komoda H, Chosa T, et al. Antibodies to adult T-cell leukemia-virus-associated antigen (ATLA) sera from patients with ATL and controls in Japan: a nation-wide seroepidemiologic study Int J Cancer 1982; 29: 631-35. 9. Blattner WA, Blayney DW, Robert-Guroff M, et al. Epidemiology of human T-cell leukemia/lymphoma virus. JInfect Dis 1983; 147: 406-16. 10. Saxinger W, Blattner WA, Levine PH, et al. Human T-cell leukemia virus (HTLV-I) antibodies in Africa. Science 1984; 225: 1473-76.

33

11. Clark J, Saxinger C, Gibbs WN, et al. Seroepidemiologic studies of human T-cell leukemia/lymphoma virus type I in Jamaica. Int J Cancer

1985; 36: 37-41. AE, Fang CT, Slamon DJ,

12. Williams

et al. Seroprevalence and epidemiological correlates of HTLV-I infection in U.S. blood donors.

Science 1988; 36: 37-41. 13.

Murphy EL, Figueroa JP, Gibbs WN, et al. Sexual transmission of human T-lymphotropic virus type I. Ann Intern Med 1989; 111: 555-60.

19.

Boulos R, Halsey NA, Holt E, et al HIV-1 in Haitian women 1982-1988. J Acquir Immune Defic Syndr 1990; 3: 721-28. 20. Anderson DW, Epstein JS, Lee TH, et al. Serologic confirmation of human T-lymphotropic virus type I infection in healthy blood and plasma donors. Blood 1989; 74: 2585-91. 21. Griner PF, Mayewski RJ, Mushlin AI, Greenland P. Selection and interpretation of diagnostic tests and procedures. Ann Intern Med 1981; 94: 553-600.

Hartley TM, Khabbaz RF, Cannon RO, Kaplan JE, Lairmore MD. Characterization of antibody reactivity to human T-cell lymphotropic virus types I and II using immunoblot and radioimmunoprecipitation assays. J Clin Microbiol 1990; 28: 646-50. 23. Khabbaz RF, Hartley TM, Lairmore MD, Kaplan JE. Epidemiologic assessment of screening tests for antibody to human T-cell lymphotropic virus type I (HTLV-I). Am J Public Health (in press). 24. Khabbaz RF, Hartel D, Lairmore M, et al. HTLV-II infection in a cohort of New York intravenous drug users: evidence for an old infection. J Infect Dis (in press). 25. Lee H, Swanson P, Shorty J, Zack J, Rosenblatt J, Chen I. High rate of HTLV-II infection in seropositive I.V. drug abusers in New Orleans. Science 1990; 244: 471-75. 22.

14. Okochi K, Sato H, Hinuma Y. A retrospective study on transmission of adult T-cell leukemia virus by blood transfusion: seroconversion in recipients. Vox Sang 1984; 46: 245-53. 15. Quinn TC, Zacarias FRK, St John RK. HIV and HTLV-I infections in the Americas: a regional perspective. Medicine 1989; 68: 189-209. 16. Centers for Disease Control. Licensure of screening tests for antibody to human T-lymphotropic virus type 1. MMWR 1988; 37: 736-47. 17. Rosenblatt JD, Gasson JC, Glaspy J, et al. Relationship between T-cell leukemia virus-I I and atypical hairy cell leukemia: a serologic study of hairy cell leukemia patients. Leukemia 1987; 1: 397-401. 18. Lairmore MD, Jacobson S, Gracia F, et al. Isolation of human T-lymphotropic virus type 2 from Guaymi Indians in Panama. Proc Natl Acad Sci USA (in press).

VIEWPOINT

Helping a child to understand her own testicular feminisation

Children do not think as adults do. They would therefore be less worried than adults are about a diagnosis with serious or ominous implications, yet they are commonly left uninformed until someone judges that they are old enough to understand. For most, this means delivery of painful information during the very vulnerable teenage years. A better approach is to unfold the truth stage by stage, matching simple statements to the child’s conceptual until the personal implications are finally realised as part of a maturing process. Use of this approach for a child with testicular feminisation is described.

growth

Introduction We have over the years come to a deeper understanding of how children think.12 Whereas at one time it was customary to withhold distressing information until a child was old enough to understand it, it is now recognised as more helpful to tell the truth stage by stage as understanding unfolds. This approach is recognised as the best for matters such as adoption, conception through a donoror the diagnosis of a malignant disorder.4 It should also be used in any chronic disorder of early onset, including a chromosomal abnormality such as Turner’s or Kleinfelter’s syndrome. I describe here how it relates to the condition of persistent testicular feminisation.

Case-report The child was the firstborn of healthy parents and weighed kg at birth. Abnormalities of the genitalia (hyper-

2-82

trophied clitoris, prominent labia, and closely approximated vaginal introitus and urethral orifice) were noted at birth. Buccal smear showed no Barr bodies but a Y chromosome. At 9 days, normal male chromosomes were reported. At 2 weeks, examination under anaesthetic showed a moderate phallus, a female urethra with no verumontanum, and a normal vagina with a cervix. Laparotomy at 3 weeks of age showed a normal uterus with bilateral fallopian tubes and abnormal gonads. A biopsy sample showed cryptorchidism as in testicular feminisation. After discussion, the parents, paediatric surgeon, and clinical geneticist agreed that the child should be raised as a girl. Phallidectomy and excision of the gonads were carried out when she was 17 months old. By then she had a normal 6-month-old sister. Paediatric follow-up was arranged.

Subsequent developments When the child was 2 years 9 months, her mother mentioned that she had been told to withhold the truth about the child’s condition. Advice from other senior experienced doctors was that this was generally the correct ADDRESS

Paediatric Department, North Staffordshire Hospital Centre, Stoke-on-Trent, UK (J Goodall, FRCPE). Correspondence to Dr Goodall at Melton, Burrington Drive, Trentham, Stoke-on-Trent ST4 8SP, UK