260 Archives ofDisease in Childhood 1994; 70: 260-263

ORIGINAL ARTICLES Arch Dis Child: first published as 10.1136/adc.70.4.260 on 1 April 1994. Downloaded from

Early diagnosis of severe combined syndrome

R A Hague, S Rassam, G Morgan A J Cant

Abstract treatment die within the first year of life from Infants with severe combined immuno- overwhelming infection.' The prognosis of deficiency syndrome (SCIDS) have a children with SCIDS has been transformed by greatly improved prognosis if diagnosed transplantation and data from a and treated before they develop over- European collaborative study of affected whelming infection. Clinical and labora- infants (in which Newcastle upon Tyne and tory data on 45 patients with SCIDS were London are the participating centres in the retrospectively reviewed to assess the UK) have shown 97°/0 long term survival for value of absolute counts in recipients ofHIA identical sibling marrow and making an early diagnosis. Ninety infants 52% survival for recipients of HLA mis- matched for age, sex, and presenting matched (haploidentical) marrow from a symptoms were used as controls. Thirteen parent.2 The survival rate for infants trans- (29%) infants with SCIDS were diagnosed planted after 6 months of age is only 45%/o, at birth as previous siblings had been however, compared with 70°/ in younger affected; 32 (71%) were diagnosed after children. Although facilities for bone marrow the development of symptoms. Eighteen transplantation for the treatment of SCIDS (56%) of these remained undiagnosed are available within the UK, many affected until after 6 months of age. The first children without a family history are older than symptoms occurred at a median of 5 6 months of age at referral, or are terminally ill weeks 1 to 8 and the with (range day months) overwhelming infection, making trans- http://adc.bmj.com/ first admission to hospital was at 4 months plantation a risky procedure. (range 1 week to 16 months). Symptoms The incidence of SCIDS is one in 66 000 live included respiratory infection (91%), births in most parts of the world,3 6 although it vomiting and diarrhoea (81%), failure to is higher in some populations.7 Thus in thrive (88%), candidiasis (50°/0), and skin paediatric practice in the UK a case will be lesions (28%). The mean lymphocyte encountered only rarely. Early symptoms and count was 171X109/1 compared with signs can be non-specific and only after a 792X iO1/l in controls. Excluding one child number of infective episodes, or an infection of on September 25, 2021 by guest. Protected copyright. with Omenn's syndrome (lymphocyte unusual severity, may immunodeficiency be count 23-3X 109I), all symptomatic infants suspected. A full blood count with differential with SCIDS had lymphocyte counts less count is usually performed in than 2-8X 1o9/ at presentation. The infants with an infection severe enough to merit median delay between the first abnormal admission to hospital. Previous workers have lymphocyte count and diagnosis was seven reported conflicting data on the proportion of weeks (range one day to 13 months). infants with SCIDS who are lymphopenic.81l Twenty eight (88%) of 32 infants would We have therefore documented the presenting have been diagnosed before 6 months of clinical features of such infants and studied the Department of Paediatric age if investigated after the first low usefulness oftheir absolute lymphocyte count in Inmmunology, lymphocyte count. These data indicate alerting a paediatrician to the possibility of Newcastle General that low lymphocyte counts are predictive SCIDS at an early stage when immediate Hospital, Newcastle upon Tyne ofSCIDS. Paediatricians are urged to pay treatment would greatly improve the prognosis. R A Hague attention to the absolute lymphocyte A J Cant counts in all infants in whom a full blood count is performed. Those with Methods Department of lympho- , Institute cyte counts persistently less than 2-8X 109/1 Patients with SCIDS seen in Newcastle upon ofChild Health, should be investigated for SCIDS. Tyne, London, and Edinburgh were identified London (Arch Dis Child 1994; 70: 260-263) from hospital records. Clinical and laboratory S Rassam G Morgan data on each child were reviewed retrospec- tively and for those referred from elsewhere Correspondence to: Dr A J Cant, Department of Severe combined immunodeficiency syndrome additional data about the initial presentation Paediatric Immunology, (SCIDS) is a heterogeneous group of inherited were sought from the referring hospital. Two Newcastle General Hospital, Westgate Road, Newcastle disorders characterised by the failure ofcellular control groups were identified among infants upon Tyne NE4 6BE. and humoral immunity. Children present with presenting to Newcastle General Hospital. Accepted 5 December 1993 recurrent or persistent infections and without The first group consisted of infants matched Early diagnosis ofsevere combined immunodeficiency syndrome 261

14 ment of symptoms was 5 weeks, the earliest 12 being at 1 day of age. Only one child, who 10 first became unwell at 8 months, had been Arch Dis Child: first published as 10.1136/adc.70.4.260 on 1 April 1994. Downloaded from cn C.) 8 asymptomatic at 6 months of age.

0 6 The median age on first admission for 0 symptomatic infants was 4 months (range 1 z week to 16 months) (fig 2). Respiratory infec- 2 tion (cough and wheeze, or less commonly 0 2 3 4 5 6 7 8 pneumonia), seen in 22 infants, was the most Age (months) common presenting symptom. Nine presented with and with Figure 1 in index vomiting diarrhoea, eight children. candidal infection, six were failing to thrive, and two had skin sepsis. Six children had only one of these symptoms by the time they were for age and first presenting symptom (for first admitted to hospital, eight had two, 10 example, chest infection, wheeze, diarrhoea, had three, eight had four, and one had all five failure to thrive) with the infants with SCIDS (one was admitted for an unrelated cause). presentinig with infection. The second control The diagnosis of SCIDS was not made until a group cc)nsisted of uninfected term infants median age of 7 months (range 4 weeks to 16 undergoi.ng blood counts either before an months). By the time all infants had at least operatiori or for the investigation of jaundice, two symptoms and most were failing to thrive. and was imatched with the infants with SCIDS The age of onset of the commonly observed who wer*e diagnosed at birth because of a symptoms and their frequency among the family hiEstory of the syndrome. group is shown in the table. White cell counts in index cases and controls Twenty eight (88%) of the 32 infants had a were me-asured by an automated Coulter 'routine' blood count checked at the time of method and blood was also filmed for a the first hospital admission, the count not differentiial white cell count. Statistical analysis being checked until later in the other four of the two groups used Student's t test. infants. The median age for the first blood count was 4 months (range 1 week to 11 months). The mean total white cell count in Results these infants was in the normal range at Forty fivee children with SCIDS were identified 8 29X 109/l and did not differ significantly and each child was matched with two control from the controls (9 25 X 109/1): only five were children from the appropriate group (90 leucopenic (white cell count <4 OX 109/1) and

children) Thirteen (29%) infants were seven had a white cell count greater than http://adc.bmj.com/ diagnosed at or around the time of birth and 11 X 109/l. In 44 infants with SCIDS, excluding were asyimptomatic (all had a family history of the one child with Omenn's syndrome a previously affected sibling). Thirty two (lymphocyte count 23 4X 109/l), the mean (71%/o) c:hildren with no family history of lymphocyte count was 1 72X 109/1 compared SCIDS Nwere diagnosed once symptoms had with 6e02 x 109/1 in controls (p

0 in whom there is a clinical suspicion of immunodeficiency should also be investigated.

t= Arch Dis Child: first published as 10.1136/adc.70.4.260 on 1 April 1994. Downloaded from moi As Gossage and Buckley point out,10 the reference range for 5 lower limit of the 950/o a) lymphocyte counts in infants is much higher 0C-' than in older children or adults, in whom <10 x 1 This C lymphopenia is defined as 09/l.11 10 has been overlooked by reviewers of SCIDS9 0 and by Hosking and Roberton when formulat- 0) ing a score for when to investigate a child for -0 15 immunodeficiency.14 The insensitivity of this scoring system has already been shown by E Lyall et al,'5 in whose series this system did not 20 identify two children with SCIDS, perhaps because the wrong criteria for lymphopenia were used. It is hardly surprising therefore gr, _- Il 2 that a lymphocyte count of between 1 and 0 6 8 10 1 2 12 10 8 6 4 2 2 4 2 8X 109/1 should escape recognition when it is Figure 3 Histogram of lymphocyte counts from index and control infants in increments of more usual to pay attention to the total white 0-5X 10 94. cell count and neutrophil count than to work out the absolute lymphocyte count from the Discussion differential percentages. Our data indicate that It has long been recognised that infants wit]h this calculation is of vital importance. SCIDS may be lymphopenic, but the propor The pattern of symptoms and signs which tion of infants reported to have subnormaLl we observed among infants with SCIDS is 5-8 16 lymphocyte counts has varied from 10-20% tio similar to those reported by other groups.2 96%./o10 The 95% reference range for lympho Although some features may be highly cyte counts in the first year of life varies fronn suggestive of SCIDS, such as the cutaneous 2 0-11 -OX 109/1 at birth to 4-0-13.5x 109/1 at i6 manifestations in a child with maternofetal months to 1 year. 1 We chose a cut offpoint feir engraftment,17 most could also be seen in lymphopenia of 2 8X 109/1 as the value unde.r children with other which counts from all the symptomatic infantts including HIV infection, and also in nornal with SCIDS lay. Five asymptomatic infantts children, particularly those from poorer with SCIDS had higher lymphocyte counts aLt socioeconomic backgrounds.'8 The control birth, and this value will therefore be less diss- data from our study shows that even during viral in criminatory for infants in the first month oflife-. acute infections, many of which will be http://adc.bmj.com/ If a persistently low lymphocyte count is noted origin, the mean lymphocyte count usually in a child free of infection, however, investigai- remains within the normal range. The control tion is probably still justified. It should be infants who were lymphopenic on initial blood emphasised that a normal lymphocyte courit count were easily distinguished from those does not preclude the diagnosis of SCIDS,'2' 13 with SCIDS as their lymphocyte count and occasionally even lymphocyte subsets caIn returned to normal within a few days. counits On the basis of our findings we propose the be normal, so that infants with higher on September 25, 2021 by guest. Protected copyright. algorithm outlined in fig 4 for the further investigation of children under 1 year of age Is total lymphocyte presenting to hospital who have full blood count 2.8 x 109/1 ? counts checked and are found to be lymphopenic. Lymphocyte subset analysis should be performed on all those whose lymphopenia persists, as well as those in whom Yes No there is a clinical suspicion of immuno- deficiency. Immunoglobulin is more often measured as a screening test, but in the early months of life is less useful and may be mis- transfer Repeat count No further action iif leading because of the transplacental infant clinically we1I1 of maternal IgG and physiologically low concentrations of IgA and IgM at this age. Infants with SCIDS admitted to hospital a is made often acquire noso- Is total lymphocyte before diagnosis count 2.8 x 109/l ? comial infections such as respiratory syncytial and rotavirus while being nursed in an unprotected environment; these infections are often is therefore considerable Yes No fatal. There urgency in the recognition of the low lympho- cyte count and further investigation, so that an affected child can be nursed in aseptic isolation as soon as Investigate No further action if and definitive treatments started (lymphocyte subsets) infant clinically well possible. The additional cost of a repeat full count is cheap at about £3 50, and Figure 4 Protocolfor selection ofinfants with low lymphocyte counts for investigation foir blood SCIDS. only a small number of infants should need Early diagnosis ofsevere combined immunodeficiency syndrome 263

lymphocyte subset which costs about WHO scientific group. Clin Immunol Immunopathol 1979; analysis, 13: 296-359. £75. This is surely justified when compared 5 Fasth A. disorders in Sweden: cases among children 1974-9. J Clin Immunol 1982; 2: with the enormous cost, both human and Arch Dis Child: first published as 10.1136/adc.70.4.260 on 1 April 1994. Downloaded from 86-92. financial, of the unsuccessful treatment of an 6 Hayakawa H, Iwata T, Yata J, Kobayashi N. Primary infant with SCIDS diagnosed (and so treated) immunodeficiency in Japan: overview of a nationwide survey on primary immunodeficiency syndromes. J Clin too late. We have initiated a prospective study Immunol 1981; 1: 31-9. of the protocol to evaluate further the implica- 7 Jones JF, Rittenbaugh CK, Spence MA, Hayward A. Severe combined immunodeficiency among the Navajo. I. tions of its adoption. At present, survival after Characterisation of phenotypes, epidemiology, and haploidentical bone marrow transplantation in population genetics. Hum Biol 1991; 63: 669-82. 8 Bortin MM, Rimm AA. Severe combined immuno- Europe is poorer than that in some American deficiency disease. Characterization of the disease and centres, one of the chief differences being age results of transplantation. JAMA 1977; 238: 591-600. 9 Gelfand EW. SCID continues to point the way. N Engl J at diagnosis,2 and results also vary between Med 1990; 322: 1741-3. European centres. We hope that a greater 10 Gossage DL, Buckley RH. Prevalence of lymphocytopenia in severe combined immunodeficiency. N Engl _J Med awareness of the importance of absolute 1990; 322: 1422-3. lymphocyte counts among paediatricians 11 Lubin BH. Reference values in infancy and childhood. In: Nathan DG, Oski FA, eds. Hematology of infancy and will lead to earlier diagnosis and therefore childhood. Philadelphia: WB Saunders, 1987. improved survival for infants with SCIDS born 12 Peter HH, Freidrich W, Dopfer R, et al. NK function in severe combined immunodeficiency (SCID): evidence of in the UK. a common T and NK cell defect in some but not all SCID patients._ Immunol 1983; 131: 2332-9. We thank the referring paediatricians, records departments, and 13 Fontan G, Garcia Rodriguez MC, Carrasco S, et al. Severe haematology laboratories of all hospitals from which the combined immunodeficiency with T retain- children were referred for their assistance in providing data. We ing functional activity. Clin Immunol Immunopathol 1988; thank Dr 0 B Eden, then consultant in haematology, Royal 46: 432-41. Hospital for Sick Children, Edinburgh for allowing us to 14 Hosking CS, Roberton DM. The diagnostic approach to include infants under his care in the series. We also thank Dr M recurrent infections in childhood. Clin Immunol Allergy Reid and Dr P Saunders and the staff of the departments of 1981; 1: 631-9. haematology ofNewcastle General Hospital and Royal Victoria 15 Lyall EGH, Eden OB, Dixon R, et al. Assessment of a Infirmary, Newcastle upon Tyne, for their help in obtaining clinical scoring system for detection of immunodeficiency data from control children, and for giving advice. in children with recurrent infections. Pediatr Infect Dis J 1991; 10: 673-6. 16 Patel MS, Hosking CS. Severe combined immuno- 1 Stiehm ER, ed. Immunologic disorders of infants and children. deficiency - the experience of the Royal Children's 3rd Ed. Philadelphia: W B Saunders, 1989: 166. Hospital, Melbourne, 1969-1979. Australian Paediatric 2 Fischer A, Landais P, Freidrich W, et al. European Joumnal 1982; 18: 169-76. experience of bone-marrow transplantation for severe 17 de Raeve L, Song M, Levy J, Mascart-Lemone F. combined immunodeficiency. Lancet 1990; 336: 850-4. Cutaneous lesions as a clue to severe combined immuno- 3 Hosking CS, Roberton DM. Epidemiology and treatment deficiency. Pediatr Dermatol 1992; 9: 49-51. of . Birth Defects 1983; 19: 18 Mok JYQ, Hague RA, Yap PL, et al. Vertical transmission 223-77. of HIV - the Edinburgh experience. Arch Dis Child 1989; 4 World Health Organisation. Immunodeficiency. Report of a 64: 1140-5. http://adc.bmj.com/ on September 25, 2021 by guest. Protected copyright.