In their article, Carrol and Coakley thoroughly analysed 21 patients with transient hyperphosphatasaemia (TH) of infancy and early childhood. 1 We would like to add our experience with TH. Between 1987 and 2001, we have diagnosed TH in 52 patients (33 boys and 19 girls), using the criteria devised by Kraut et al . 2 Transient hyperphosphatasaemia was detected in patients with a wide variety of clinical disorders, gastrointestinal infections being the most frequent (18 patients), followed by bronchitis/bronchopneumonia (nine patients), anaemia (six patients) and history of prematurity (five patients). In the remaining 14 patients, the following diagnoses were established: urinary tract infection, history of rickets, cervical lymphadenitis, hypercalciuria, breath-holding spells, epilepsy, Quincke's oedema, breast milk jaundice, juvenile idiopathic arthritis, hyperinsulinaemia, coeliac disease. There was no relationship between the course and severity of these disease states and alkaline phosphatase (ALP) activity. The mean age of the patients at the time of TH was 17 months ± 14 SD (range 1-60 months; median 13 months). The ALP values ranged between 15 and 127 µ kat/L (mean 36 µ kat/L ± 23 SD; median 31 µ kat/L), which was 2.5-20-fold above the age-related upper reference range. The ALP values returned to within normal limits in 12-343 days (mean 71 days ± 69 SD; median 52 days). However, exact maximal ALP levels and the duration of TH are difficult to establish because of the accidental detection of this state. The median age in our patients supports previous observations of TH preponderance between the first and second year of life. 1,3 We observed a marked seasonal clustering in September through November (Fig. 1), which is in accordance with previous reports. 1,3,4 As TH is presumed to be of infectious origin and infections tend to be more frequent during the autumn months, this could explain the seasonal occurrence. 1 As we detected TH in two pairs of twins, this might further support the hypothesis regarding the infectious origin of TH. However, we didn't find a common infectious agent in our patients, as antibodies to the following viruses were detected in 10 of 32 examined patients: cytomegalovirus in six, EpsteinBarr virus in three and adenovirus in one. In regard to bone metabolism, an increased urinary hydroxyproline/creatinine ratio was found in three children of five examined, as has been reported previously. 5 In our opinion, this does not solely support the hypothesis of increased bone turnover in TH, as hydroxyproline is currently considered the least sensitive marker of bone resorption. 6 In conclusion, our observations are in accordance with previous reports on TH and we advise that children with TH should be spared from excessive diagnostic procedures. 1,3,7
REFERENCES