ARTICLE
Auteur(s) : Laurence
Valeyrie-Allanore, Jean-Claude Roujeau
Department of Dermatology, Hôpital Henri Mondor, Université
Paris XII, Créteil, 94010 France
As often with dermatologic diseases, there has been a long lasting
debate between “mergers” and “splitters” concerning the clinical
manifestations of “drug allergy”. Some consider drug rashes as a
spectrum, while others try to separate distinct entities, with more
or less success and “overlaps”.In this issue of the European
Journal of Dermatology, Petkov et al. report 3 well documented
cases of severe cutaneous adverse drug reactions (SCARs)
characterized by a cutaneous phenotype of epidermal necrolysis but
also associated with eosinophilia and systemic manifestations that
are usually considered characteristics of Drug Hypersensitivity
Syndrome (DHS) [1]. At first glance this report strongly challenges
the splitters point of view.From the detailed clinical data and
figures in Petkov’s paper, one has to accept without doubt the 3
cases as definite TEN, based on the clinical pattern of dusky spots
and blisters, epidermal detachment on a large part of the body
surface area, full-thickness necrosis of the epidermis on biopsy,
erosions of mucous membranes, high fever. We are more reluctant,
however, concerning the diagnosis of Drug Hypersensitivity
Syndrome. The authors define DHS as “a triad of fever, skin rash
and symptomatic or asymptomatic internal organ involvement”. This
definition is so large that it includes most drug eruptions, from
anaphylaxis to severe photosensitivity, from drug-induced lupus to
pustular drug rashes (AGEP) and obviously also SJS/TEN. This is in
line with a merger conception, including under the denomination of
DHS the majority of serious cutaneous adverse reactions to drugs.
This choice does not allow for individualisation of the originality
of the drug reactions that have been reported under the
denomination of DHS or a variety of other names.From our
experience, from the literature and from discussion with many other
experts, we consider that this syndrome has the following
characteristics features:
- 1. Rash developing late (usually > 3 weeks)
after onset of treatment.
- 2. Long lasting symptoms (usually > 2
weeks) after discontinuation of the causative drug.
- 3. Fever (> 38 ˚C).
- 4. Multi-organ involvement.
- 5. Eosinophilia (> 1.5
× 109/L).
- 6. Lymphocyte activation (node enlargement,
lymphocytosis, atypical lymphocytes).
- 7. Frequent virus reactivation.
Japanese Dermatologists call this syndrome DIHS (Drug Induced
Hypersensitivity Syndrome) [2], while we have suggested the acronym
of DRESS (Drug Reaction with Eosinophilia and Systemic Symptoms)
[3].Whatever the name, DHS, DIHS or DRESS, the important point is
to agree on the main characteristics of the syndrome, on the number
of criteria needed for establishing the diagnosis and to check the
sensitivity and specificity of such criteria for discriminating
difficult cases such as those reported by Petkov et al.Concerning
disease characteristics, most experts agree on the criteria listed
above. Further progress is needed, however, before reaching a
consensus on the practical use of this list. Recommendations in
Japan are to diagnose “typical DIHS” if all 7 criteria are present,
and “atypical DIHS” if at least 5 are present [2]. In Europe, the
RegiSCAR group has proposed a different approach, with calculation
of a numerical score giving different weights to each item [4]. The
score value results in a diagnosis of DRESS being excluded,
possible, probable or definite. With this scoring system cases
could be classified as definite DRESS without fulfilling all 7
criteria. Important considerations in both systems are 1) to
exclude any clinical or laboratory abnormality that pre-existed the
onset of the drug reaction or had an obvious other cause, and 2) to
define a minimal level of alteration that can be accepted as
indicating organ involvement. As an example, both the RegiSCAR
group and the Japanese group define liver abnormalities as ALT
levels above twice the upper limit of normal values. Using these
criteria, none of the 3 cases presented in Petkov’s paper would be
classified as DIHS and all 3 would be only possible DRESS.These
“rules” are not engraved in marble. They are only preliminary tools
that need validation and improvement. For instance, the fact that
neither high eosinophilia nor lymphadenopathy are usual
manifestations of epidermal necrolysis is not taken in account in
our validation rules for SJS or TEN. We thank Petkov et al. for
pointing to these alterations as a possible indicator of
overlapping mechanisms in their patients.The question here is to
figure whether these overlapping cases are frequent enough to ruin
the splitters theories. We do not have any good answers yet.
Obviously we will not be able to decide who has been right (or more
probably, to what extent everyone has been wrong…) before knowing
much more about the basic molecular mechanisms of “drug
allergy”.Recent results suggest that it is probably more fruitful
to investigate series of cases as homogeneous as possible, to
decipher the molecular mechanisms of drug reactions. A group of
geneticians and dermatologists in Taiwan has evidenced a very
strong association of SJS and TEN with HLA B*1502 and HLA B*5801
when induced by carbamazepine and by allopurinol respectively [5,
6]. Concerning allopurinol, the same association with B*5801 was
also observed in patients suffering from DHS/DIHS/DRESS [6]. But
carbamazepine association to HLA B*1502 was only observed for SJS
and TEN and not for DHS/DIHS/DRESS, nor for milder drug eruptions
[7]. Behind all possible theoretical considerations on the
mechanisms, one may suspect that if a large series of cases with
“allergy” to carbamazepine had been first investigated, the
specific link with SJS would probably had been missed.We therefore
consider that the “splitter” attitude is more efficient for
research by allowing the detection of any difference between
subgroups, whether this difference affects the nature or the
intensity of mechanisms. It would then be easier to merge all
subgroups if they appear to be artificially split than to
understand what happens in a mixture of “drug hypersensitivity” of
different mechanisms and poorly defined phenotypes.This is the main
reason why we strongly suggest that all reports of severe
dermatological manifestations of hypersensitivity to drugs include
all the relevant information allowing the use of diagnosis criteria
as soon as validated.
References
1 Petkov T, Pehlivanov G, Grozdev I,
Kavaklieva S, Tsankov N. Toxic epidermal necrolysis as a
dermatological manifestation of drug hypersensitivity syndrome. Eur
J Dermatol 2007; 17(5): 422-7.
2 Shiohara T, Inaoka M, Kano Y. Drug-induced
hypersensitivity syndrome (DIHS): A reaction induced by a complex
interplay among herpesviruses and antiviral and anti drug immune
responses. Allergol Int 2006; 55: 1-8.
3 Bocquet H, Bagot M, Roujeau JC. Drug-induced
pseudolymphoma and drug hypersensitivity syndrome (drug rash with
eosinophilia and systemic symptoms: DRESS). Semin Cutan Med Surg
1996; 15: 250-7.
4 Kardaun SH, Sidoroff A, Valeyrie-Allanore L,
et al. Variability in the clinical pattern of cutaneous
side-effects of drugs with systemic symptoms: does a DRESS
synbdrome really exists? Br J Dermatol 2007; 156: 609-11.
5 Chung WH, Hung SL, Hong HS, et al. A
marker for Stevens-Johnson syndrome. Nature 2004; 428: 486.
6 Hung SL, Chung WH, Liou LB, et al.
HLAB*5801 allele as a genetic marker for severe cutaneous reactions
caused by allopurinol. Proc Natl Acad Sci USA 2005; 102: 4134.
7 Hung SI, Chung WH, Jee SH, et al. Genetic
susceptibility to carbamazepine-induced cutaneous adverse drug
reactions. Pharmacogenet Genomics 2006; 16: 297-306.
|
Printable version
Version PDF