Dr Hadjivassiliou and colleagues (1) (June 26th) are not correct in
invoking dermatitis herpetiformis (DH) as being analogous to the
neurological disorders which they attribute to gluten sensitivity.
Firstly, all patients with DH do in fact have demonstrable gluten
sensitivity of the small intestine. Thus, although only two-thirds of
patients with DH have a degree of villous-atrophy on a routine diagnostic
single biopsy, if appropriate criteria are used, nearly 100% have
demonstrable gluten sensitivity. Brow et al. (2) showed that if multiple
biopsies are taken, the incidence of villous atrophy rises to 95%. Also,
lymphocytic infiltration of the epithelium occurs in 90% of patients with
DH irrespective of villous architecture, and this abnormal lymphocytic
infiltration resolves with gluten withdrawal (3). Further, it has been
shown that in DH patients without villous atrophy, histological damage is
seen on gluten challenge (4). Thus, all patients with DH do have gluten
sensitivity of the intestine. Second, in DH, a clear and a direct
relationship between the intestine and skin has been demonstrated. Dermal
IgA deposits are the hallmark and do not occur in its absence. After many
years of gluten withdrawal, the IgA will disappear from the skin but
returns with re-introduction of gluten in the diet (4).
By contrast, the putative role for gluten in the cases of ataxia
/neuropathy described by Hadjiivassiliou et al, remains speculative. The
acid test, namely clear and unequiviocal clinical and or imaging /
neurophysiological improvement following gluten free diet, is still
lacking. In the editorial, the authors quote their own previous work
despite the fact that their previous paper actually provided no evidence
for effectiveness of diet. We accept that an unusually high proportion of
their patient are HLA DQ2 and this may well point to an
immunopathogenesis. But, remembering that HLA DQ2 is also associated
with a range of non-gluten sensitive autoimmune conditions, it need not be
gluten related. Neurological case selection rested largely on the finding
of IgG anti-gliadin antibody. But IgG anti-gliadin is extremely common
even in normal individuals. If the hypothesis is correct, what
explanation is given for the absence of neurological defects in the
overwhelming majority of individuals, including the many with coeliac
disease or DH, despite their HLA DQ2 and IgG gliadin antibody positive
status ? Whatever disease process underlies these neurological cases, it
would appear to be different to that leading to skin lesions in DH.
Dr D J Unsworth PhD.MRCPath. FRCP
Southmead Hospital, Bristol. BS10 5ND.
Professor L Fry MD FRCP
EMERITUS PROFESSOR OF DERMATOLOGY
C/O St. Mary's Hospital.
Paddington. London W2 1NY.
1 Hadjivassiliou M, Grunewald RA, Davies-Jones GAB. Gluten
sensitivity: a many headed hydra BMJ 1999; 318: 1710-1.
2 Brow J, Parker F, Weinstein W, Rube CE. The small intestinal
mucosa in dermatitis herpetiformis: severity and distributions of the
small intestinal lesion and associated malabsorption: Gastroenterol 1972;
3 Fry L, Seah PP, Hayer PG, Hoffrand AV, McMinn RMH. The small
intestine in dermatitis herpetiformis J.Clin. Path. 1974; 27: 814-27
4 Weinstein WM Latent coeliac sprue. Gastroenterol. 1973 ; 64 : 489-
5 Leonard JN, Haldenton GP, Tucker W, Unsworth DJ, Swain AF, McMinn
RMH. Holborrow EJ, Fry L. Gluten challenge in dermatitis herpetiformis
NEJM. 1983: 308: 816-19.
Competing interests: No competing interests