Вы находитесь на странице: 1из 7

Drug-induced skin disease

Allen P. Kaplan, M.D. Stony Brook, N.Y.

Drug-induced cutaneous reactions encompas.s a wide variee of rashes that depend in purt on
route of administration (e.g., contact ver.sus systemic,) as well as type qf cutaneous response and
molecular mechanism underlying the reaction. One such reaction is a type IV immunologic
reaction (delayed hypersensitivity) manifest as contact dermatitis and commonly elicited lq
drugs such as antihistamines, antibiotic ointments. local anesthetics, and paraben ester.s in
cosmetic creams and lotions. A generalized eruption of this sort will occasionally occur with
systemic administration of a drug to someone previously sensitized by topical application.
Systemic administration qf agents can cause non.spec$c pruritus or maculopapular eruptions
that resemble visual exanthems. The pathogenesis is unclear and no immune mechanism has
been demonstrated. If the drug is continued, exfoliative dermatitis can result. Other types qf
reactions are urticarial in nature and include acute urticarialangioedema. erythema malt[forme
(bullous and nonbullous), Stevens-Johnson syndrome, urticaria in association with serum
sickness-like reactions, and urticaria associated with anaphylactoid reactions. In man) of these,
an allergic reaction in which there is an immunoglobulin (Ig) E-dependent release qf mediators
in the skin causes hives or swelling. In oi’hers, circulating immune complexes may be presrnt,
qften involving IgG antibody complexed with drug and complement,fixation: hives may then be
caused by anaphylatoxin release or a concomitant IgE-mediated reaction. In some in.sttrnces ( a
cellular reaction may augment the aforementioned i@ammatorv reuctions, perhap.s as part (/‘a
late-phase reaction or a true delayed hypersensitivin component. Mnjor offenders involved in
such reactions are antibiotics such as penicillin and sulfa, barbiturates, anticonvu1sant.s. and
phenothiazines. Urticarial reactions to aspirin and other nonsteroidal anti-inj?ammatop drugs
are in general nonimmunologic. Other types qf cutaneous reactions to drugs include erytht>ma
nodosum. fixed drug eruptions, purpura (thromboqtopenic or nonthrombocytopenic), phototo.wic
and photoallergic reactions, and exfoliative dermatitis or toxic epidermal necrolysis. Phototoxic,
reactions are nonimmunologic, occur on first exposure, and depend on the light absorption
spectrum of the drug. Photoallergic reacTions resemble contact dermatitis but involve a delayed
hypersensitivity reaction to an antigen formed by interaction of the drug with light and binding
of the altered drug (hapten) to a cutane0u.s protein to Jbrm a complete antigen. Exfoliative
dermatitis or toxic epidermal necrolysis can be severe sequalae of maculopapular eruptions in
which the drug is continued or can be severe erythema multij~orme-like reactions and are
potentially fatal. Superinfection with group 2 staphy1ococcu.c mq be contributory in Some cases.
Finally, cases of physically induced allergy such as dermatographism have been reported Gftcr
drug reactions. (J ALLERGY CLINIMMUNOL74:573, 1984.)

The most common allergic drug reactions are der- are contact dermatitis, urticaria/angioedema, and
matologic; thus the skin is of particular importance maculopapular (exanthematous) eruptions. Such re-
when adverse effects of drugs are considered. Almost actions not only are diverse in terms of clinical man-
any type of skin eruption can be secondary to medi- ifestation, but also involve differing pathogenic mech-
cation However, the three most common reactions anisms. It is also important to note that for many types
of skin reactions, including some common ones, the
From the Division of Allergy, Rheumatology, and Clinical Im- precise mechanism by which cutaneous inflammation
munology, Department of Medicine, State University of New occurs is unknown. Table I lists the major cutaneous
York at Stony Brook, Health Sciences Center, Stony Brook, N.Y.
reactions seen with drug therapy. In this article I will
Reprint requests: Dr. Allen P. Kaplan, Division of Allergy, Rheu-
matology, and Clinical Immunology, Department of Medicine, outline the scope of the problem and, for selected
State University of New York at Stony Brook, Health Sciences types of cutaneous reactions, discuss pathogenesis,
Center. Stony Brook, NY 11794. diagnostic testing, and/or therapy.
574 Kapian

TA@E I. Clinical cutaneous manifestations oi

drug allergy*

Contact dtmtatitiq
Allergic eczematouscontact dermatitis
Systemiceczematous“contact-type” dermatitis
Dermatitis medicamentosa
U&aria and angjoedema
Exanthematic eruptions
Exfoliative dermatitis
Erythema multiforme-like eruptions
Erylhema nodosum-like eruptions
Fixed drug eruptions
Purpuric eruptions
Henoch-Schiinlein(anaphylactoid) purpura
Other drug-induced dermatoses(possibly allergic)
Eczematous eruptions
Toxic epidermal necrolysis
Lichen planus-like eruptions
FIG. 1. Acute facial contact dermatitis. Other drug-induced dermatoses (not allergic)
.\cneiform eruptions
CONTACT DERMATtTtS Pigmcntary changes
It is known that topical application of medicine and ‘Alt~peCiil
intermittent or repetitive exposure to a drug is asso- .l-.-.__ll_ - -
*TBken from DeSwarte RD: Drug allergy. In Pattersqn R. editor:
ciated with the highest sensitization risk. Contact der- Mcrgic diseases, Philadelphia, 1972. JB Lippincott Co, p 408.
matitis is the major disorder resulting from topical
application and must be differentiated from primary phages are activated by lymphocyte products, ag-
irritant contact dermatitis, in which the pathogenic gregated by macrophage aggregation factor (a se-
mechanism does not involve an immune response, as creted fibronectin),’ and immobilized by antigen-
well as other similar dermatoses such as seborrheic dependent migration inhibition factor.’ At least during
dermatitis and atopic dermatitis. Many substances ca- the early phase, the skin reaction appears to also in-
pable of inducing such reactions are of low molecular volve basophil accumulation, perhaps secondary to
weight and act as haptens. i.e.. they can combine with release of one or more chemotactic factors. Thus con-
preformed antibody but do not elicit an immune rc- tact dermatitis seems to be an example of cutaneous
sponse. Some substances are thought to combine with basophil hypersensitivity in man.‘. ’ Degranulation of
skin proteins to form a complete antigen, while others such basophils as well as cutaneous mast cells may
(e.g., nickel or chromates) that are not reactive with contribute to the marked pnui~us associated with con-
proteins can combine with membrane protems of an- tact dermatitis; however. the mechanism by which
tigen-presenting cells. ’ Contact allergens are taken up such degranulation-might occur is unknown.
by the Langerhans cell, a macrophage-like cell of skin. In sensitized persons, the reaction usually becomes
and the complex (altered metabolically) is presented evident 6 to 48 hours after contact, but later reactions
to lymphocytes in the skin and then in regional lymph are occasionally seen. A larger interval suggests that
nodes. The reaction is then disseminated at a cellular the exposure that caused the reaction may have been
level to involve macrophages and lymphocytes that the sensitization event. Severity depends on the quan-
are not reactive with antigen. particularty when a sec- tit! of allergen applied and the degree of existing
ondary response is elicited on subsequent exposure to hypersensitivity. The reaction initially consists of red-
the contactant. Lymphocytes are activated by release ness, pruritus. and a papular eruption. It tken becomes
of macrophage interleukin 1’ plus lymphocyte (helper edematous with vesiculation and, when sevece, can be
and/or effector cell) interleukin 2. lnterleukin 2 also hullous ( Fig. 1) HistologicaHy, one seesspongiosis (in-
activates cytotoxicisuppressor lymphocytes.” Macro- tercellular edema., microvesiculation) and infiltration
VOLUME 74 Drug-induced skin disease 575

FIG. 2. Urticarial lesion caused by reaction to penicillin.

with lymphocytes and other mononuclear cells. as 40 to 60 mgiday for an adult), but can then be rapidly
well as basophils and a variable number of eosino- tapered.
phils. When the dermatitis is chronic because of con-
tinuous or repetitive exposure, lichenification of skin URTICARIA AND ANGIOEDEMA
is seen and on biopsy one sees acanthosis, localized Drug reactions are among the most commonly rec-
areas of parakeratosis, and thinning between the rete ognized causes of urticaria and angioedema. Most ap-
pegs. Systemic exposure to an allergen that previously pear to be immunoglobulin (Ig) E mediated. For such
caused contact dermatitis can cause flare-ups at sites reactions, combination of the drug or drug metabolite
of previous reactivity, but there is also danger of in- with IgE bound to the surfaces of cutaneous mast cells
ducing a systemic generalized eruption. leads to cell activation, degranulation, and release of
Drugs important in the elicitation of contact der- mast cell vasoactive mediators such as histamine,
matitis when applied to the skin include local anes- prostaglandin DZ, leukotrienes, and platelet activating
thetics (particularly those containing paraben esters), factor (1 -O-alkyl-2-O-acetyl-sn-glyceryl-3-phosphoryl-
antihistamines applied in topical medications, anti- choline).’ All of these vasoactive factors can contrib-
biotics such as penicillin, sulfa, or neomycin, and ute to the vasodilatation and increased vascular per-
corticosteroid creams containing paraben esters, as meability associated with hive formation; however,
well as cosmetic creams and lotions. histamine is clearly one major factor, while the spe-
The causal agent of contact dermatitis is often rec- cific contribution made by the other moieties is pres-
ognized by the patient or easily elicited when the his- ently unknown. Angioedema can occur in the absence
tory is taken. However, it is often necessary to do of urticaria, although drug reactions are frequently
skin testing when the agent or agents suspected are associated with both urticaria and angioedema. The
uncertain. Patch testing is a procedure in which sus- pathogenesis is essentially the same, the difference
pected materials are applied to a small area of skin being the depth of skin at which the reaction occurs;
and patches are left in place 24 to 48 hours. A lesion urticaria reflects inflammation in the superficial der-
with the features of contact dermatitis (erythema, pap- mis, while angioedema involves the deep dermis and
ules, edema, vesicles) is considered positive and in- subcutaneous tissue. Urticarial lesions can also be
dicates prior sensitization. Like all skin tests, the re- seen with serum sickness-like reactions. These are
action must be correlated with the history, localization associated with circulating immune complexes con-
of the eruption, clinical course, and other variables to sisting of the agent plus IgG antibody. There is fixation
be meaningful in terms of etiology. of the classic complement pathway and histamine re-
Treatment consists of cold wet dressings (e.g., Bur- lease that is presumed to be caused by the anaphy-
row’s solution in diluted saline or tap water) and anti- latoxins C3a, C5a, and perhaps C4a. The syndrome,
histamines taken orally. When the reaction is severe, typified by serum sickness reactions caused by peni-
involving large portions of the body, or when prom- cillin, consists of fever, lymphadenopathy, urticaria,
inent on the face, a course of systemic corticosteroid arthralgias, periarticular swelling, occasionally bona
therapy is indicated. High doses for 7 to 10 days may fide arthritis, and in about 20% of the occurrences,
be needed to control symptoms (e.g., prednisone at glomerulonephritis.
FIG. 3. Erythema muitiforme caused by adverse reaction to sulfa.

The major causes of drug-induced urticaria and an- tion of penicillin and related drugs, IS usually of httle
gioedema are penicillin and related compounds, sulfa value. Generally, no routine tests are avaifable that
drugs, barbiturates (particularly phenobarbital). anti- reliably confirm or refute the diagnosis of drug-in-
convulsants, salicylates, foreign antisera. allergy ex- duced urticaria or angioedema; thus an empiric ap-
tracts, radiocontrast materials, and opiate analgesics proach is indicated. Radioallergosorbent testing, if
(Fig. 2). Some of these agents, such as salicylates, available, may be helpful to establish the IgE depen-
opiates, and radiocontrast agents, cause urticaria by dence of the reaction. Prausnitz-Kustner testing of the
nonimmune mechanisms. One must also be aware of ability to passively transfer the wheal-and-flare re-
potential cross-reactions such as reactivity to thiazide action by serum from an allergic to nonallergic in-
dhrretics, furosemide, sulfonylurea hypoglycemic dividual may be helpful in an investigative laboratory
agents, carbonic anhydrase inhibitors, and procaine- situation.
type local anesthetics in individuals sensitive to sulfa. Recently, much mformation has accumulated on
Urticarial penicillin reactions occur most commonly urticaria secondary to ingestion of aspirin. ” The bulk
to the penicilloyl major determinant or one of the of evidence suggests that these reactions are not im-
minor determinants.’ Chronic reactions can occasion- munologic. Patients who are sensitive to aspirin do
ally be traced to small amounts of contaminating pen- not tolerate indomethacin, pyrazones, mefenamic
icillin in food, drink, or other medicines and to viral acid, and coal tar dye derivatives such as tartrazine
vaccines. Since the semisynthetic penicillins are yellow dye No. 5 (used to color foods and medica-
formed by combination of various side chains with tions). Although salicylates per se are generally not
the 6-amino-penicillenic acid nucleus. cross-reaction thought to cause hives, at least one investigator has
occurs between all penicillins, such as methicillin. reported that sodium salicylate in very large doses can
cloxaciliin, carbenicillin, and ampicillin. However. cause urticaria. All the above drugs have the property
the common morbilliform rash associated with am- of being inhibitors of prostaglandin synthetase,‘! al-
picillin therapy does not appear to be an immunologic though the most recent data concerning tartrazine do
reaction. Cephalosporins, which are derivatives of 7- not con&m this mechanism. Aspirin may actsby di-
aminocephalosporanic acid, have been shown to ex- rectly acetylating and inactivating the synthetase. tt
hibit cross-reactivity with the penicillins to a greater is also possible that a subpopulation of patients may
degree than appreciated by most people. react to fhe contaminant aspirin anhydride by a true
Although a variety of in vitro and in vivo methods immunologic reaction. !’
for the assessment of drug allergies have been dc-
scribed, such as observing migration inhibition factor
release when leukocytes are challenged with the drug, Erythema multiforme can be considered a more se-
most of these tests offer little advantage to complete vere variant of the basic underlying process that results
abstinence from the suspecfed agent. Skin testing of in urticaria. Lesions resemble chmnic urticaria more
compounds of low molecular weight. with the excep- than acute urticaria in that they are clearly intiltrative
VOLUME 74 Drug-induced skin disease 577

and are raised beyond the lesions caused by wheal

formation. In this instance the lesions can vary in size
and shape, may have a half moon-type configuration,
or shovv central clearing, sometimes with a red central
spot producing the iris or target lesion (Fig. 3). I,gE-
mediated reactions may be responsible for erythema
multifcmme-like reactions, but the pathogenesis may
also involve elements of a late-phase reaction” or
immune complex-mediated inflammation. As an ex-
ample of the latter mechanism, some forms of idio-
pathic erythema multiforme have been shown to be
vasculitic and are associated with immune complexes
of IgG antibody with herpes simplex type I virus.li
Thus erythema multiforme may or may not have a
clear vasculitic component analogous to comparisons
that have been made concerning typical chronic ur-
ticaria and urticarial vasculitis.‘5 These are distin-
guishable by biopsy but may or may not be distin-
guishable by observation. Lesions of erythcma
multiiorme may become bullous and are often accom-
panied by constitutional symptoms including fever,
malaise, gastrointestinal upset, arthralgias, and per-
iarticular inflammation. Drugs commonly responsible
includr sulfa, penicillin, barbiturates, salicylates, an-
ticonvulsants, phenothiazines, and phenylbutazone.
The Stevens-Johnson syndrome (Fig. 4) is a ful-
minant, primarily bullous form of the disease that
occurs most often in children and involves mucosal FIG. 4. Stevens-Johnson syndrome demonstrating bul-
Ious facial lesions in healing phase.
surfaces including the mouth, eyes, and genital tract.
Before the availability of steroids, the mortality rate
approximated 30%; therapy now consists of high-dose pruritic, although this is variable, and it may be pre-
steroids until clear defervescence occurs. Occular dominantly erythematous, macular, papular, or mor-
damaglz can cause blindness, while visceral organ in- billiform. Virtually any drug can produce it, although
volvement including bronchitis, pneumonia, and ne- common agents are the same ones that cause urticaria,
phritis may be seen. As sulfa drugs are being used angioedema, and erythema multiforme. Ampicillin is
more commonly, the incidence of erythema multi- a particularly common offender in the penicillin drug
forme and its sequelae appears to be increasing sig- category.
nificantly. Repeated administration of the same drug at a future
Taxi c epidermal necrolysis, or Lyell’s syndrome, time will most likely cause recurrence of the same
consists of generalized erythema of the skin followed rash with the same time interval, rather than an ac-
byextensivedesquamationresembling “scaldedskin.” celerated, more severe reaction. When such a rash is
It appears to result from a group 2 staphylococcus encountered, it is best to discontinue causative agents
infection in infants and young children; drug reactions or switch drugs when continued therapy is needed.
have been implicated in older children and adults and Continued administration can in ‘some cases lead to
it may represent an often fatal variant of erythema confluence of lesions, generalized erythroderma, and/
multiforme. The high mortality rate results from fluid or exfoliative dermatitis. Although drugs such as pen-
imbalance and sepsis; the latter usually is due to see- icillin and sulfa can cause so many different forms of
ondary infection including staphylococcus, so that it cutaneous reactions, it is the host response to them
then overlaps with the infectious form. that determines the outcome and we cannot predict or
explain why some individuals react in one way rather
The:<e are among the most common drug reactions
and can be difficult to distinguish from viral exan- PHOTOSENSITIVE REACTIONS6
thems, particularly when common offending drugs are Photosensitive reactions are produced by interaction
used to treat such infections. The eruption may be of a drug with light energy that, depending on the
578 Kaplan

cause sensitization, followed by reactions with sub-

sequent exposure to drug plus light. Compared to pho-
totoxic reactions, there is cross-reactivity between im-
munochemically related antigens, reactivity can be
transferred with mononuclear cells in animal mod&.
and patch testing can be performed by application of
the suspected agent for 24 hours followed by exposure
to a light source. Reported agents are primarily ones
taken systemically, e.g., phenothiazines. sulfa drugs
and related compounds, and griseofulvin.

Drug-induced cutaneous vasculitis can consist of a
combination of papules, palpable purpura, urticaria,
nodules, and ulceration. Drugs implicated include
penicillin, sulfa. phenylbutazone, thiouracil, iproni-
azid, and busulfan. The reaction is presumed to result
from immune complexes of drug and IgG antibodies
causing complement-mediated inflammation of smalf
cutaneous vessels. Rigorous demonstration of drug-
antibody complexes in these instances is usually lack-
ing, and there is often uncertainty regarding a drug
as the cause. There may be an underlying disorder
that can cause a similar rash and there is usually no
contirmatory test available to distinguish which is the
FIG.5. Fixed drug eruption of hand from phenolphthalein cause. while subsequent rechallenge with any sub-
Exposure. (From DeSwarte RD: Drug allergy. ln Patterson petted drug is contraindicated.
R, editor: Allergic diseases. Philadelphia, 1972, JB Lippin- Erythema nodosum-like lesions can also be seen
:ott co, pp 414-5.) after drug reactions: however, this lesion is more fre-
quently associated with diseases such as sarcoidosih,
particular interaction, can range from the ultraviolet mycobacterial diseases. fungal diseases. or connective
spectrum to the visible light spectrum. The drug can tissue diseases. These lesions are typically bilateral,
be administered topically, orally. or systemically. The warm, and tender nodular lesions over &he anterior
Eruption is thus limited to light-exposed areas such as aspects of the legs. Suspected agents have been io.-
the face, neck, arms, back of the hands. the V area dides, bromides, sulfa, penicillin, and salicylates.
af the chest, and the anterior aspect of the legs. In contrast to the above entities, fixed drug erup.-
Two mechanisms appear to be operative and are tions are virtually pathognomonic of drug hypersen-
defined as phototoxic reactions or photoallergic re- i;itivity ( Fig. 51. These lesions recur at the same site
actions. Phototoxic reactions are not immunologic and each time the drug is encountered. Typically, one sees
occur on first exposure to the drug when adequate a rounded lesion of 2 to 3 inches that is edematous.
light is present. The reaction resembles a severe sun- then turning erythematoub, often with a purple hue.
bum, some vesiculation may occur, and the area may Lesions may be mildly pruritic or burning, rna:i hub.
become hyperpigmented. Phototoxic reactions inctude side within 2 to 3 weeks of drug discontinuation, and
those caused by coal tar derivatives (used to treat may leave a residual hyperpigmentation. C’ommon
psoriasis) or by systemic administration of dimeth- causative agents Include phenophthalein (red dye).
ylchlorotetracycline (Dectomycin), chlorpromazine. sulfa drugs. barbiturates, tetracycline, aspirin. and
Dr sulfa. Dimethylchlorotetracycline exposed to tight phenylbutazone. The mechanism is unknown. One
has recently been shown capable of activating the must postulate some permanently altered reactivity of
:lassic complement pathway. I7Thus it is possible that the skin site. since recurrences occur at the same lo-
:omplement plays a role in the cutaneous inflamma- cation regardless of route of administration.
:ion seen
Photoallergic reactions are a variant of contact der- CONCLUDING (=OMIVIENTS
matitis. Radiant energy alters the drug to alfow com- Although drug reactions are extremely common and
Jlex formation with a cutaneous protein to become many are immunologically mediated, our diagnostic
Immunogenic. There is usually an initial exposure to abilities are actually quite limited. For example. it’ a
VOLUME 74 Drug-induced skin disease 579

patient has urticaria resulting from an antibiotic, pen- macrophage agglutination factor and fibronectin. Immunology
icillin and related compounds remain the only drugs 46:115, 1982
5. Weiser WY, Greineder DK, Remold HG, David JR: Studies
for which there are skin testing reagents that can be
on human migration inhibition factor: charactertzation of three
used in a diagnostic or predictive fashion. We are molecular species. J Immunol 128: 1958, 198 I
unable I:Otest for even so common an offender as sulfa 6. Dvorak HF, Dvorak AM, Mihm MC: Morphologic \tudies in
drugs by in vitro or in vivo methods. Of course, pre- cellular hypersensitivity in guinea pig and man. I/r Contact
vention of drug reactions is critical and available skin hypersensitivity in animals. Basel, 1974, S Karger, pp 54-65
7. Askenase PW: Role of basophils. masl cells, and va\oamines
testing reagents can be utilized to assist in achieving
in hypersensitivity reactions with a delayed time course. Prog
this goal. It is generally important to utilize medication Allergy 23: 199, 1977
for proper indications. Avoidance of potent sensitizers 8. Kaplan AP: Urticaria and angioedema. ~‘n Middleton E Jr. Reed
whenever possible is a good principle to follow, par- CE, Ellis EF, editors: Allergy principles and practice, ed 2.
ticularly with topical application. We should also re- St. Louis, 1983, The CV Mosby Co, vol 2, pp 1341-60
9. Parker CW: Drug allergy. In Parker CW, (editor: Clinical im-
frain from the use of drug mixtures when a single
munology. Philadelphia, 1980. WB Saunders Co, vol 2. pp
agent can be equally effective. Future investigation 1219-60
should deal with studies of the biochemical or im- 10. Yurchak AM, Wichner K. Arbesman CE: Immunologic studies
munochemical mechanisms by which each form of on aspirin. J ALLERGY 46:245, 1970
inflammation is initiated. We need to obtain data re- 1 I. Szczeklik P, Gryglewski RJ, Czemiewska-Mysik G: Relation-
ship of inhibition of prostdglandin biosynthesis by analgesics
garding the mechanism by which commonly used
to asthma attacks in aspirin sensitive patients. Br Med J l:67,
drugs complex with proteins. Also, drug metabolism 1975
should be studied and various metabolites assessedas 12. DeWeck AL: Immunological effects of aspirin anhydride. a
possible toxic or immunogenic agents. Perhaps we contaminant of commercial acetyl salicylic acid preparation.
Int Arch Allergy Appl Immunol 41:393, 1971
will one day have information regarding patient sus-
13. Solley GO, Gleich AJ, Jordan RE, Schroeter AL: The late
ceptibility to drug reactions in general and to particular phase of the immediate wheal and flare skin reaction. J Clin
kinds of adverse reactions. ’ Invest 58:408, 1976
14. Kazmierowski JA, Wuepper KD: Erythema multiforme: Im-
mune complex vasculitis of the superlicial cutaneous micro-
I. Shelley WB. Juhlin L: Langerhans cells form a reticuloen- vasculature. J Invest Dermatol 71:366, 197X
dothelial trap for external contact allergens. Nature 261:46. IS. Zeiss CR, Burch FX, Marder RJ, Furey NL, Schmid FR,
I910 Gewurz H: A hypocomplementemic vascuhtic urticarial syn-
2. Luger TA, Oppenheim JJ: Characteristics of interleukin I and drome. Report of four new cases and definition of the disease.
epidcrmal cell-derived thymocyte activating factor. In Weiss- Am J Med 68:867. 1980
mann G. editor: Advances in inflammation research. vol 5. 16. DeSwarte RD: Drug allergy. In Patterson R, editor: Allergic
New York, 1983, Raven Press, pp 1-25 diseases. Philadelphia, 1972. JB Lippincott Co. pp 393-493
3. Luger TA, Smolen JS, Chused TM, Steinberg AD, Oppenheim 17. Lim HW, Novotny H, Gigli I: Role of complement and poly-
JJ: Human lymphocytes with either OKT 4 or OKT 8 phenotype morphonuclear cells in dimethylchlorotetracycline induced
produce interleukin 2 in culture. J Clin Invest 70:470, 1982 phototoxicity in guinea pigs. Inhibition by decomplementation
4. Godfrey HP, Pirohit A: Reversible binding of a guinea pig in vivo. 3 Clin Invest 71:1326, 1983
Iymphokine to gelatin and fibrinogen: possible relationship of