Interleukin 1 Receptor Antagonist Deficiency Presenting as Infantile Pustulosis Mimicking Infantile Pustular Psoriasis

COMMENTS AND OPINIONS

Antipruritic Potency of Serotonin Type 3(5-HT3) Receptor Antagonists—a Reply

W e read with great interest the case report ofDowns and Kennedy1 about the successfultreatment of intractable palmoplantar pru-

ritus with ondansetron. Because of our continuous re-search on pruritus and the action of serotonin type 3 (5-HT3) receptor antagonists, we think that there are majorpoints that should be clarified.

First of all, palmoplantar pruritus is an unusual lo-cation and a descriptive term but is not a diagnostic anddisease entity. This raises the question on the etiologyof palmoplantar pruritus. According to the report, neu-rological examinations, including electroneurography andupper spine radiography, had not been performed. It ispossible that the palmoplantar pruritus was caused bynerve compression, nerve alteration, or neuropathy. Thereport further lacks important information such as theoccupation and hobbies of the patient, to rule out exter-nal factors. The medical history does not mention the pa-tient’s prior external medications. Elderly patients oftenuse vasoactive emollients and ointments, especially af-ter having experienced vein thrombosis. Such therapeu-tics, applied by the patients themselves, sometimes causecontact dermatitis, which the authors might see ex-cluded by the lack of skin lesions and the correspond-ing histological findings. Furthermore, topical agents suchas those containing, for example, capsaicin, can lead tovarious skin sensations without specific skin lesions andhistological findings. Unfortunately, there are no data onallergological examinations such as atopy screening (ex-cept for total IgE levels) and epicutaneous tests. Neitherhistamine, eosinophilic cationic protein, serotonin, nor5-hydroxyindolacetic acid had been measured as a stan-dard parameter in the peripheral blood. It seems that theauthors did not consider an adverse drug reaction in theirdifferential diagnosis. Drug-induced itching is usually gen-eralized, but localized varieties occur.2 The authors men-tioned that the patient’s hands and feet appeared “healthy”so that “pruritus sine materia” would be the more ap-propriate term to characterize this type of pruritus. Al-together, we think that there are important data missingin this case report.

The nature of substances that mediate cholestaticand uremic pruritus are not completely known. It needsto be clarified that the mechanisms of action leading tothe clinically observed antipruritic effect of 5-HT3 re-ceptor antagonists in these types of pruritus are not un-

derstood. The first case reports on the antipruritic effectof 5-HT3 receptor antagonists3 encouraged us to inves-tigate the antipruritic potency of a 5-HT3 receptor an-tagonist (tropisetron) under experimental conditions. Inhealthy volunteers, we could not verify an antipruriticeffect of tropisetron on histamine- and serotonin-induced itch under experimental conditions.4 As our firstresults also pointed toward a major role of skin mast cells(which cannot be explained in more detail in this letter)we performed the same experimental design when skinmast cells were depleted before.5 Our study demon-strated that tropisetron does not affect histamine-induced itch but has a measurable effect on serotonin-induced reactions in mast cell–depleted skin.5 It wouldbe interesting to know whether the histological exami-nation of the patient’s skin showed any specific findingson skin mast cell number and function. Though we didnot investigate the agent ondansetron in our studies, ourdata are comparable because pharmacokinetic differ-ences among these drugs are unlikely to contribute toclinical differences in activity. Downs and Kennedy shouldconsider that ondansetron seems to have weak antago-nistic activities on 5-HT1b, 5-HT1c, and adrenergic andopioid receptors.

According to our results, there is no definitive ex-planation for the effectiveness of ondansetron in this typeof pruritus. Besides the medication’s effects on the skin’smast cells, antiserotoninergic activity and influence oncentral nervous processing may be possible.

The most interesting aspect of this publication is thatthe patient has remained free of pruritus for 1 year. It re-mains a question whether it is sensible to apply ondanse-tron on a regular basis, considering the high cost and pos-sible adverse effects of long-term medication.

Elke Weisshaar, MDHarald Gollnick, MDDepartment of Dermatology and VenereologyOtto von Guericke UniversityLeipziger Strasse 4439120 MagdeburgGermany(e-mail: [email protected])

1. Downs AMR, Kennedy CTC. Successful treatment of intractable palmoplan-tar pruritus with ondansetron. Arch Dermatol. 1998;134:925-926.

2. Sarno AM, Bernhard JD. Drug-induced pruritus without rash. In: BernhardJD, ed. Itch-Mechanisms and Management of Pruritus. New York, NY: McGraw-Hill Book Co; 1994:329-335.

3. Schworer H, Hartmann H, Ramadori G. Relief of cholestatic pruritus by a novelclass of drugs: 5-hydroxytryptamine type 3 (5-HT3) receptor antagonists: ef-fectiveness of ondansetron. Pain. 1995;61:33-37.

4. Weisshaar E, Ziethen B, Gollnick H. Can a serotonin type 3 (5-HT3) recep-tor antagonist reduce experimentally-induced itch? Inflamm Res. 1997;46:412-416.

5. Weisshaar E, Ziethen B, Gollnick H. The antipruritic effect of a 5-HT3 recep-tor antagonist (tropisetron) is dependent on mast cell depletion: an experi-mental study. Exp Dermatol. In press.

CORRESPONDENCE

ARCH DERMATOL / VOL 135, MAY 1999599

©1999 American Medical Association. All rights reserved.Downloaded From: http://archderm.jamanetwork.com/ by a University of Alabama-Birmingham User on 01/01/2013

In reply

We appreciate the interest of Weisshaar et al in our casereport.1 The answers to points they raise are as follows:

1. They request clarification of our designation “pal-moplantar pruritus.” This is purely descriptive of itch lim-ited to the volar surfaces of hands and feet, and was not in-tended to convey a disease entity.

2. Concerning neurology, the patient was thoroughlyevaluated by a neurologist, and the results of a computedtomographic scan of the upper spine and brain were nor-mal. Sensory evoked potentials were not done, however, andcan be a useful investigation in a patient with a clinicallynormal nervous system.

3. Concerning factors that may indicate a contact der-matitis, her occupation was housewife, and she had no hob-bies. There were no prior medications other than those statedin the case report. Without further evidence for contact der-matitis or urticaria, patch or prick tests were not consid-ered relevant.

4. Concerning extensive allergological investiga-tions, because the patient had no symptoms or signs or fam-ily history of atopic allergy and no elevated total IgE lev-els, these investigations were deemed unnecessary.

5. Measurements for histamine, eosinophilic cationicprotein, serotonin, and 5-hydroxyindolacetic acid were notperformed.

6. We are unaware of localized pruritus being a po-tential adverse effect of any of the patient’s preexisting sys-temic medication, and also point out that these medicationswere taken only during the winter. Her pruritic symptomspersisted all year.

7. The number and morphological appearances of mastcells in the skin biopsy specimen were not abnormal; no func-tional studies (in vitro or in vivo) were undertaken.

We found the account of the commentators’ previousexperimental work and complementary case report very in-teresting. Our case report demonstrates that ondansetronhas provided a sustained clinical benefit without adverse ef-fects for a difficult management problem of localized pru-ritus and is worth considering in such situations.

Anthony M. R. Downs, MRCPDepartment of DermatologyBristol Royal InfirmaryBristol BS2 8HWEnglandCameron T. C. Kennedy, FRCPBristol

1. Downs AMR, Kennedy CTC. Successful treatment of intractable palmoplan-tar pruritus with ondansetron. Arch Dermatol. 1998;134:925-926.

Sezary Syndrome, Cutaneous T-CellLymphoma, and ExtracorporealPhotopheresis

T he report by Fraser-Andrews et al1 of theirexperience with extracorporeal photopheresis(ECP) treatment of patients with leukemic

cutaneous T-cell lymphoma (CTCL) who had previ-

ously failed treatment with systemic chemotherapy pro-vides valuable information. Since most of the previouslyreported patients received ECP as an initial systemictherapy, the study patients of Fraser-Andrews et al dif-fer in this quite important way from those whom weand other investigators have treated with this type ofactive immunotherapy. The median survival of 39months for the patients in the study of Fraser-Andrewset al did not differ in a statistically significant mannerfrom the median survival of 22 months for their histori-cal control group, and was shorter than the 60 to 100months in prior reports in which many of the bestresponders also had clear evidence of clonal malignantcells. The relatively poorer survival data of Fraser-Andrews and colleagues may result from their ECPrecipients having been selected from those for whomsystemic chemotherapy had not been adequate manage-ment.

The reason that we and others have long advisedthat ECP therapy for patients with CTCL precedeimmunosuppressive chemotherapy, rather than viceversa, is that the immunocompetence of the patient isprerequisite to the ECP response.2 In experimentalmodels of ECP therapy, immunosuppression with cor-ticosteroids completely prevented otherwise potentresponses.3

Extracorporeal exposure of malignant cells to pho-toactivated methoxsalen massively increases their dis-play of tumor antigens by increasing peptide processingand transport.4 The best CTCL responses to ECP occurin patients whose CD8 (cytotoxic) T-cell counts have notbeen suppressed by either prior chemotherapy or dis-ease progression, probably because the CTCL-specific an-tigens5 are weakly immunogenic.

We enthusiastically concur with the suggestion ofFraser-Andrews et al that a multi-institutional, prospec-tive, controlled study be undertaken of the efficacy ofECP for CTCL. Yet it would make little sense to spendseveral years and great expense conducting a trial of atherapy that would be outmoded by the time the trialconcluded. The understanding of the mechanism ofECP has now matured to the point of permitting itslogical improvements. At this time, ECP is a rapidlyevolving technology, comprehension of its scientificbasis finally driving its evolution. For the past 9months, our group has been conducting a phase I clini-cal trial of an altered ECP approach that enhances theimmunogenicity of the CTCL cells, while simulta-neously augmenting numbers of dendritic antigen-presenting cells. If our results remain encouraging, wewould be more interested in testing this method inphase III trials than the standard ECP method that itwill hopefully replace.

The principal lesson learned from clinical re-sponses to ECP is that selective immunotherapy for pa-tients with CTCL is possible when the capacity of the pa-tient to respond to reinfused or “vaccinating” alteredmalignant cells remains intact. Sixteen years after the in-troduction of this treatment, scientifically based improve-ments of the immunotherapeutic method have finally be-come possible. It is those potentially improveddescendants of ECP that will have sufficient promise to



merit the type of multi-institutional trials that Fraser-Andrews and colleagues propose.

Richard L. Edelson, MDDepartment of DermatologyYale University School of Medicine333 Cedar St, 500 LCIPO Box 208059New Haven, CT 06520

The Department of Dermatology at Yale University, whichI chair, has received research support from Therakos Inc,the company that manufactures the photopheresis appara-tus. In addition, I have been a consultant for Therakos andalso have devised the clinical protocol that is, in part, dis-cussed herein.

1. Fraser-Andrews E, Seed P, Whittaker S, Russell-Jones R. Extracorporeal pho-topheresis in Sezary syndrome: no significant effect in the survival of 44 pa-tientswithaperipheralbloodT-cell clone. Arch Dermatol. 1998;134:1001-1005.

2. Wolfe JT, Lessin SR, Singh AH, Rook AH. Review of immunomodulation byphotopheresis: treatment of cutaneous T-cell lymphoma, autoimmune dis-ease, and allograft rejection. Artif Organs. 1994;12:888-897.

3. Yamane Y, Lobo FM, John LA, Edelson RL, Perez MI. Suppression of anti-skin-allograft responsebyphotodamagedeffectorcells: themodulatingeffectsofpred-nisolone and cyclophosphamide. Transplantation. 1992;54:119-124.

4. Hanlon D, Berger C, Edelson R. Photoactivated 8-methoxypsoralen treat-ment causes a peptide dependent increase in antigen display by transformedlymphocytes. Int J Cancer. 1998;78:70-75.

5. Berger CL, Longley BJ, Imaeda S, Christensen I, Heald P, Edelson RL. Tumor-specific peptides in cutaneous T-cell lymphoma: association with class I ma-jor histocompatibility complex and possible derivation from the clonotypicT-cell receptor. Int J Cancer. 1998;76:304-311.

In reply

Edelson suggests that prior treatment with chemotherapymight explain the failure of ECP treatment to affect sur-vival in our patients with Sezary syndrome (SS). In fact,only 4 patients had been treated with high-dose chemo-therapy such as fludarabine or pentostatin prior to startingECP therapy. Most had received phototherapy (16 pa-tients) or low-dose oral chemotherapy with chlorambucil(16 patients). This is still a popular first-line therapy withmany dermatologists, and it seems highly likely that otherstudies of ECP also included patients who had been simi-larly pretreated. Indeed, in Edelson’s original study,1 28 of37 patients had received systemic chemotherapy prior to ECP.

Edelson’sgrouphaspreviouslyclaimedthatpatientswithCD8 counts below 0.15 ×109/L do not respond well to ECP.ThemeanCD8countofourcohortpriortoECPwas0.25 ×109/L(range, 0.01-0.56 ×109/L), which is only slightly lower thanthe mean CD8 count of our patients with SS at diagnosis(0.32 ×109/L).2 Thus, there is littleevidence tosupport theviewthat chemotherapy has immunosuppressed our patients.

We do not deny that ECP may benefit patients with SSclinically, and in a minority of cases there may be a com-plete response.3 However, in our own studies we have not con-firmed the observation that this depends on the initial CD8count. However, our studies were confined to patients with aT-cell clone confirmed by genotypic analysis of the periph-eral blood. We have observed that the CD8 count in patientswith clonal SS is lower than in patients who fulfill the clini-cal criteria for a diagnosis of SS but do not exhibit a periph-eral blood T-cell clone.2 It is this nonclonal group who mightaccount for the predictive power of the initial CD8 count and

the prolonged survival of patients in Edelson’s cohort.Edelsonalsoproposes selectionbiasasanexplanation for

our results, in that only those patients who failed other treat-ments were referred for ECP. This is a possibility, but selec-tion bias could apply equally in the opposite direction, as dis-cussed in our article. The only conclusion one can draw withcertaintyfromthisdiscussionistheneedforarandomizedstudy.

Inhis letter,Edelsonseems toquestion theneed foraran-domizedstudyofECP,butenthusiasmforaparticular therapyis no substitute for scientific rigor. As Zackheim and coau-thors4 pointed out in a recent review of cutaneous T-cell lym-phoma,“Randomisedcontrol trials(RCT)areneeded inCTCLas they are in many other disease areas. If one wishes to havean unambiguous result in the evaluation of therapeutic effi-cacy, there is no substitute for a carefully performed RCT.”

In a review of 6 medical therapies, each subject to mul-tiple therapeutic trials with historical as well as randomizedcontrols, 79% of the 56 studies with historical controls foundthe experimental therapy to be superior, whereas only 20% ofthe50studieswithrandomizedcontrolsagreed.5 In thecontextofCTCL, someof the studies reportingprolongedsurvivalwithECP have not even confined their analysis to SS, but have in-cludedpatientswithcutaneouspatch-orplaque-stagedisease.

Investigators faced with a lack of response to ECP inpatients with CTCL now advocate the use of ECP plus in-terferon alfa or, as in Edelson’s case, immune enhancementusing tumor-specific peptides and dendritic antigen-presenting cells. With the advent of cytokine therapy suchas interleukin 2 or interleukin 12, the number of combina-tions could be extended indefinitely, but does not overridethe need to test scientifically the efficacy of ECP itself againsta standard chemotherapy regimen such as methotrexate.

Over the past 2 years, we have tried to establish such astudybuthavebeenunsuccessfuldespite thesupportof thepresi-dentof the InternationalSociety forCutaneousLymphomaandthe chairman of the European Organization for Research andTreatmentofCancerCutaneousLymphomaGroup.Therakos,the company that provides ECP equipment, has little interestin such a trial because it already has Food and Drug Admin-istration approval for the use of ECP in CTCL treatment.

Over a decade has passed since Edelson’s seminal ar-ticle was published in the New England Journal of Medi-cine,1 and the lack of any randomized study of ECP therapyfor CTCL is increasingly difficult to justify.

R. Russell-Jones, MA, FRCPSean Whittaker, MD, FRCPE. Fraser-Andrews, MA, MRCPSt John’s Institute of DermatologySt Thomas’ HospitalLambeth Palace RdLondon SE1 7EH, England

1. Edelson R, Berger C, Basparro F, et al. Treatment of cutaneous T-cell lymphomaby extracorporeal photochemotherapy. N Engl J Med. 1987;316:297-303.

2. Fraser-Andrews E, Wolstencroft RA, Dean AJ, Russell-Jones R, Whittaker SJ.TCR gene analysis of peripheral blood defines two distinct subsets in Sezarysyndrome [abstract]. Br J Dermatol. 1998;139:(suppl 51):29.

3. Russell-Jones R, Fraser-Andrews E, Spittle M, Whittaker S. Extracorporeal pho-topheresis in Sezary syndrome [letter]. Lancet. 1997;350:886.

4. Zackheim HS, Koh HK, Weinstock MA. Assessing clinical outcomes in cuta-neous T-cell lymphoma: haematology/oncology clinics of North America. 1995;9:1021-1029.

5. Sacks H, Chalmers T, Smith H. Randomized versus historical controls for clini-cal trials. Am J Med. 1982;72:233-240.



Randomized Trials and Scientific Methods

W e enjoyed the article by Smolle et al1 that dis-cussed a randomized clinical trial of home-opathy for warts. We strongly agree with the

statement in their introduction that “. . . scientific meth-ods are valuable tools for distinguishing helpful alterna-tive medical methods from superstition and quackery.”

Scientific methods do not necessarily require a ran-domized trial. We can argue that sometimes a random-ized trial is harmful. When a trial is performed, a test ofstatistical significance is done, such as a t test. The stan-dard in clinical medicine for a statistically significant re-sult is a P value less than or equal to .05. Accepting a Pvalue of .05 means that there is a 1 in 20 chance that atherapy without merit could be shown to be significant.One way to decrease the likelihood of getting a spuriousresult is to limit trials to topics that have a potentiallymedically explainable result. Is there even 1 scientifi-cally defensible hypothesis for the therapeutic basis ofhomeopathy? Have any basic science experiments evershown that the chemically pure water used in homeopa-thy is different from normal water?

A trial of a therapy with no medical basis is danger-ous because by chance alone the trial could suggest thatthe therapy is effective. A potentially false-positive trialof a novel therapy may be followed by a second trial us-ing a large number of patients. The likelihood that thesecond, larger trial would also give a false-positive re-sult is low. But what about the interval between the tri-als? How many patients would spend their money andtime chasing an ineffective therapy? Further, some al-ternative medicine practitioners and customers maychoose to look at only “positive” studies. These prob-lems could be avoided by limiting randomized trials totherapies that have a rational physiological explanation.

David A. Wrone, MDArthur J. Sober, MDDepartment of DermatologyMassachusetts General Hospital55 Fruit St, Bartlett Hall 622Boston, MA 02114

1. Smolle J, Prause G, Kerl H. A double-blind, controlled clinical trial of home-opathy and an analysis of lunar phases and postoperative outcome. Arch Der-matol. 1988;134:1369-1370.

Careful Evaluation Neededof “Alternative” Claims

T he November 1998 issue of the ARCHIVES on al-ternative medicine and dermatology (Novem-ber 1998) certainly is thought provoking. I would

like to add my opinion to the discussion. In the enthu-siasm and excitement over alternative approaches to medi-cal care, I believe it is important to maintain the perspec-tive of the past. In the prescientific era of medicine,treatment was frequently ineffective if not harmful. In-fections, injuries, periodic epidemics of communicable

disease, and other medical maladies, with few practicaltreatments, contributed greatly to human suffering anddeath. Adoption of the scientific method and the bio-logical model of disease has led to the great triumphs ofmedicine in combating disease, promoting wellness, andextending the human life span. We must guard againstthe adoption of unproven remedies based solely on folk-lore, tradition, or anecdotal reports of success. Unfortu-nately, some of the articles within the issue on alterna-tive medicine present anecdotes, opinions, andunreferenced statements and conclusions without chal-lenge or editorial comment.

The discovery of new approaches to promoting well-ness and combating disease is among our highest goals.So-called alternative treatments must be tested in well-controlled, properly designed studies. Those that con-vincingly demonstrate benefit are perhaps best not thoughtof as alternative therapies but rather simply as therapiesthat add to the armamentarium of medicine. There shouldbe no bias against unconventional thinking, but ratheran expectation that the utility of new techniques can bevalidated. Several of the articles within the special issueon alternative medicine do exactly that.

When new therapies appear to work without an un-derstanding of their biological or physical mechanismsof effect, research to uncover these mechanisms of ac-tion may be very helpful in designing even more effec-tive techniques. When it is difficult or impossible to con-struct a hypothesis for a mechanism of action for a therapythat is consistent with the current understanding of thephysical laws by which the universe is governed, one mustexpect a particularly high level of proof. Marcello Truzzihas written, “ . . . when such claims are extraordi-nary . . . we must demand extraordinary proof. . . . ”

Richard D. Granstein, MDDepartment of DermatologyNew York Presbyterian Hospital525 E 68th St, Room F-340New York, NY 10021(e-mail: [email protected])

1. Truzzi M. Editorial. The Zetetic. 1976;1:3-6.

Randomized Trial of Aromatherapy:Successful Treatment for Alopecia Areata

T he randomized trial of “aromatherapy” for alo-pecia areata by Hay et al1 in the “Alternative Medi-cine” theme issue in the ARCHIVES is illustrative

of several basic principles. Although the treatment re-sulted in a statistically significant improvement, I havequestions about their initial randomization. The only state-ment on randomization procedure is, “Eighty-four pa-tients were randomized into 2 groups by the aromathera-pist.”1 There is mention neither of the technique used torandomize the patients nor of disease duration, which isa major predictive factor in alopecia areata.2 It must beremembered that approximately half of adult patients withrecent onset of patchy alopecia areata have spontaneousremission within the first year. Thus, the reported re-



sults could easily be accounted for by differences in dis-ease duration.

Assuming the treatment is effective, what conclu-sions does one make? No doubt some will take this asevidence that aromatherapy is valid, and “alternative medi-cine” is both a valid and important tool for the completeclinician. However, the logical response is to determinewhich of the botanicals is biologically active, isolate theactive material, and standardize it. A large percentage ofour medications come from botanicals, and it is no sur-prise that essential oils contain active moieties. The find-ing of active materials in essential oils does not justifyrandom application of good smells based on unprovenclaims. Derivation of pharmaceuticals from botanicals ishard science and has no room for the vague, irrationalclaims of practitioners of “alternative medicine.”

Many other subjects touched by alternative medi-cine are amenable to scientific analysis, without validat-ing the alternative medicine approach. Mind-body con-nections are real, and a subject for neurology,neuroimmunology, neurochemistry, and psychiatry.Much of alternative medicine is distinguished only byits lack of basis in rational thought, gullibility, and alack of skepticism.

Richard S. Kalish, MD, PhDDepartment of DermatologyState Universitiy of New York, Stony BrookHealth Sciences Center, T-16, 060Stony Brook, NY 11794-8165

1. Hay IC, Jamieson M, Ormerod AD. Randomized trial of aromatherapy: suc-cessful treatment for alopecia areata. Arch Dermatol. 1998;134:1349-1352.

2. Gordon PM, Aldrige RD, McVittie E, Hunter JA. Topical diphencyprone foralopecia areata: evaluation of 48 cases after 30 months’ follow-up. Br J Der-matol. 1996;134:869-871.

In reply

In reply to Professor Kalish’s questions, patients were in-terviewed in a dedicated clinic for the study. After consent,evaluation and photography were performed, identical in-structions were given by the aromatherapist, and random-ization was carried out by giving every alternate subject theactive treatment (alternating with placebo base oil/no es-sential oils). Thus there could be no selection bias; treat-ment was both random and blind. Table 1 details diseaseduration in active and control groups.

Disease severity was classified by the McDonald-Hulland Norris1 4-point scale, as described in the article. Theinitial assessment of the disease severity in the 2 groups isoutlined in Table 2.

We do not deny the frustrating unpredicability of thisdisease and constructed the study with statistical advice onpower, numbers, and methodology anticipating the skepti-cism that would be met if we did find a significant effect.Any study can produce a false-positive conclusion and veri-fication in further studies is desirable; but our results dem-onstrate that this therapy was successful. In Britain, pa-tients are only referred to a dermatologist after seeing ageneralist. Generally, British dermatologists are likely toreview more severe cases than those of our colleagues in theUnited States. The photographic illustration we showed was

one of several showing a degree of improvement we wouldnot have expected by chance.

We agree entirely that this does not validate claims thataromatherapy is effective in other therapeutic areas. As withChinese herbalism, some alternative therapies have a his-tory and tradition. Over hundreds of years practitioners mayhave gained wisdom from trial and error in identifying ac-tive botanicals. We have found references from 100 yearsago documenting the use of oils similar to those used in thisstudy for hair loss. The task of identifying the active con-stituent would be fascinating but likely too complex, andwould require support from the pharmaceutical industry.

Isabelle C. Hay, MRCPA. D. Ormerod, FRCPAberdeen Royal InfirmaryDepartment of DermatologyWard 48, ForesterhillAberdeen AB25 2ZNScotland

1. MacDonald-Hull S, Norris JF. Diphencyprone in the treatment of alopecia ar-eata. Br J Dermatol. 1988;119:367-374.

VIGNETTES

Oral Psoralen–UV-A forSystemic Scleroderma

S ystemic sclerosis is an autoimmune connective tis-sue disorder that is characterized by massive depo-sition of collagen in the skin and/or visceral or-

gans. Attempts to treat the disease have included therapywith anti-inflammatory drugs, immunosuppressive agents,penicillamine, colchicine, calcitonin, retinoids, inter-feron gamma, and extracorporeal photochemotherapy.1

Recently, bath psoralen UV-A (PUVA) therapy was ad-ministered successfully to patients with circumscribedscleroderma and to a few patients with systemic sclero-sis.2,3 Encouraged by this handful of positive reports, weexamined whether standard oral PUVA therapy would

Table 1. Disease Duration for Both Activeand Control Groups

DiseaseDuration, y

ActiveGroup, No.

ControlGroup, No. Total

0-4 10 19 295-9 15 9 24.9 18 13 31

Table 2. Disease Severity Measured on the 4-Point Scale1

PatientGroups, No.

Disease Severity Score

Total1 2 3 4

Active 11 9 22 0 42Control 9 6 26 0 41



be an effective treatment for systemic sclerosis. Since therewould be no need for special bath facilities, oral PUVAwould be much easier to administer and offer to pa-tients on an outpatient basis.

Patients, Therapy, and Methods. Four women (mean age,58 years; age range, 44-73 years) with systemic sclerosis(mean disease duration, 3 years; range, 1-6 years) whohad undergone progressive skin changes during the pre-vious 6 months were treated with oral PUVA therapy af-ter they had previously not responded to various treat-ment agents, including systemic steroids, calcitonin, andinterferon gamma. Before oral PUVA therapy was be-gun, a 1-month washout treatment was administeredwhere the patients had been treated with other therapyfor systemic sclerosis. Each patient received 0.6 mg of8-methoxypsoralen (8-MOP, Oxsoralen) per kilogram ofbody weight 1 hour prior to UV-A irradiation. The ini-tial UV-A dose was 50% of the individual minimal pho-totoxic dose (range, 0.5-4 J/cm2). Psoralen UV-A was given3 times weekly for 10 weeks, resulting in a mean cumu-lative dose of 70.5 J/cm2 (range, 50.5-92.0 J/cm2). Twopatients had to be switched from 8-methoxypsoralen to5-methoxypsoralen (5-MOP, Geralen) because the 8-me-thoxypsoralen caused nausea. To evaluate the responseto therapy, clinical scores were obtained for skin sever-ity in 22 skin regions, joint motility, grip strength, andskin thickness (via 20-MHz ultrasonography). Also, skinbiopsy specimens were taken immediately before andwithin 7 days after oral PUVA treatment. The biopsy speci-mens were processed, sectioned, stained with hematoxy-lin-eosin, and finally scored on a scale from 0 to 3 by ablinded observer (H.P.S.) for 6 parameters (thickness ofdermis, thickness of collagen bundles, width of spacesbetween collagen bundles, thickness of subcutaneoussepta, presence of inflammatory infiltrate, and entrap-ment of eccrine glands), resulting in an overall histo-logical skin assessment with a possible score ranging from0 to 18.

Results. Whereas 3 of 4 patients reported an improve-ment of their subjective overall status and 1 patient ex-hibited stabilization of disease, the mean objective scoresfor skin severity, grip strength, joint motility, and skinthickness (via 20-MHz ultrasonography) before and af-ter therapy did not differ significantly. However, the his-tological analysis revealed that oral PUVA treatment hadameliorated the histological skin score of all patients. Themean histological skin score was 6.3 before therapy vs3.0 after therapy (Wilcoxon test, P = .07).

Comment. The possible mechanism of action of PUVAin systemic sclerosis remains unclear at present. Psor-alen UV-A may influence the course of disease by itslocal and/or systemic immunosuppressive effects. Alter-natively, PUVA treatment, like UV-A treatment, maydecrease the amount of collagen by directly inhibitingits synthesis and/or stimulating collagenase activity.The UV-induced production and release of cytokinessuch as tumor necrosis factor a or interleukin 6 may beresponsible for decreased collagen synthesis.4 Perhapssinglet oxygen molecules induce the activity of matrix-

metalloproteinases such as collagenase.5 Whatever themechanism, our observation suggests that oral PUVAtherapy may be an effective treatment for patients withsystemic sclerosis who are undergoing skin changes,that further studies are clearly necessary to determinethe significance of PUVA-induced amelioration of histo-logical changes in the long-term clinical follow-up ofpatients with systemic sclerosis, and that more work ina greater number of patients with systemic sclerosisseems justified and desirable.

Angelika Hofer, MDDepartment of DermatologyUniversity of GrazAuenbruggerplatz 8A-8036 Graz, AustriaH. Peter Soyer, MDWolfgang Salmhofer, MDHelmut Kerl, MDPeter Wolf, MDGraz

1. Perez MI, Kohn SR. Systemic sclerosis. J Am Acad Dermatol. 1993;28:525-547.

2. Luftl M, Degitz K, Plewig G, Rocken M. Psoralen bath plus UV-A therapy:possibilities and limitations. Arch Dermatol. 1997;133:1597-1603.

3. Kanekura T, Fukumaru S, Matsushita S, Terasaki K, Mizoguchi S, Kanzaki T.Successful treatment of scleroderma with PUVA therapy. J Dermatol. 1996;23:455-459.

4. Vowels BR, Cassin M, Boufal MH, Walsh LJ, Rook AH. Photochemotherapyinduces the production of tumor necrosis factor-alpha by monocytes: impli-cation for the treatment of cutaneous T-cell lymphoma and systemic sclero-sis. J Invest Dermatol. 1992;98:686-692.

5. Wlaschek M, Briviba K, Stricklin P, Sies H, Scharffetter-Kochanek K. Singletoxygen may mediate the ultraviolet A–induced synthesis of interstitial colla-genase. J Invest Dermatol. 1995;104:194-198.

Low and Irreproducible Methoxsalen Levelsin Patients Receiving Photochemotherapy

P soralen UV-A therapy (PUVA) is used for the treat-ment of psoriasis1 and cutaneous T-cell lym-phoma.2 The efficacy may be related to the amount

of drug present at the time of irradiation. From 1988 to1996, 5274 blood specimens were collected from pa-tients receiving photochemotherapy and analyzed by high-pressure liquid chromotograpy.3 Patients for whom at least10 determinations were made (1029 determinations in68 patients) were selected for detailed analysis. We showthat it is not uncommon for patients to have very lowlevels of methoxsalen.

Patients and Results. Patients with similar parameterswere categorized into 3 subgroups (Figure 1) corre-sponding to high (group 1), average (group 2), and low(group 3) methoxsalen levels. For group 1, the average± SD methoxsalen level was 311 ± 218 ng/mL; for group2, 201 ± 152 ng/mL; and for group 3, 118 ± 86 ng/mL.Figure 1 most clearly illustrates the differences betweenthe 3 cluster groups because it can be seen that eventhough the patients in group 3 ingested some of the high-est doses of methoxsalen, their blood levels were the low-est. The average ingested doses were 44, 46, and 50 mg,respectively. In group 1, only 3 of the measured levelscorresponded to ingested doses greater than 60 mg; for



group 2, 41 mg; and for group 3, 83 mg. When the av-erage plasma level–ingested drug ratio is computed, thebioavailability in group 1 was found to be more than 3times that for group 3 (7.5 compared with 2.4).

Figure 2 shows chronological patterns ofmethoxsalen levels for representative patients from eachof these groups. Although the average methoxsalenlevel in group 1 is much higher than for groups 2 and 3,there were still many instances of low or 0 values. Fore-most among our observations is that there is a high fre-quency of low and perhaps suboptimal levels of thedrug at the time of therapy: 1369 (31.7%) bloodmethoxsalen levels were lower than 50 ng/mL, and ofthese, 976 were lower than 20 ng/mL. Second, we haveobserved substantial variations in patient chronologicalprofiles over months and years (Figure 2). An analysisof the pattern of methoxsalen levels of 1 patient (datanot shown) placed that patient in group 3. However, acloser inspection showed that the 21 analyses could bedivided into 2 groups (specimens 1-7 and 8-21). Twovery different averages resulted (211 ± 117 and 22 ± 32,respectively). This kind of result indicates the value ofperiodically determining methoxsalen levels rather thanrelying on a single random measurement that may notbe representative of typical methoxsalen bioavailabilityin a particular patient.

Comment. Familiarity with the patient’s methoxsalenplasma profile may provide an explanation for the lackof response to therapy. These variations have been at-

tributed to food ingestion, metabolic disposition, and va-sodilating drugs (for a discussion, see Gasparro,2 chap-ter 2 and references therein). Attempts to eliminate orminimize variations in psoralen blood levels have had lim-ited success. One method of ensuring the delivery ofmethoxsalen to the target tissue in patients with psoria-sis has been the implementation of bath delivery.4 Whilelogistically more involved and less convenient than oralingestion, this approach has been widely implementedin Europe. Other approaches have included supposito-

700

600

500

400

300

200

100

0Patient Groups

Met

hoxs

alen

Lev

el, n

g/m

L

Group 1

Group 2

Group 3

∗

∗

∗

Figure 1. Analysis of methoxsalen levels: group 1, high (n = 11), group 2, average (n = 18), and group 3, low (n = 29). Only patients for whom at least 10determinations were made are included in this plot. The average methoxsalen level and SD are shown for each patient. A linear inverse correlation betweenmethoxsalen plasma levels and ingested dose was observed for this limited group of patients (r2 = 0.93). The asterisks indicate those patients whose levels arefurther analyzed in Figure 2.

800

700

600

500

400

300

200

100

01 8 9 10 11 12 13 14 15765432 16 17 18 19 20 21 22 23 24 25 26 27

Chronological Sample Order, d

Met

hoxs

alen

Lev

el, n

g/m

L

Group 1 PatientGroup 2 PatientGroup 3 Patient

Figure 2. The chronological pattern of methoxsalen levels in 3 of the 68patients from the groups shown in Figure 1 (see the asterisks in Figure 1):patient 6 from the high group (group 1), patient 48 from the average group(group 2), and patient 13 from the low group (group 3). The methoxsalenplasma concentration in nanograms per milliliter is plotted as a function ofthe analysis number (in chronological order).



ries, topical application, and intravenous infusion (for adiscussion of these issues see Gasparro,2 chapter 2 andreferences therein). Knobler et al5 formulated a sterileaqueous solution of methoxsalen for direct injection intothe leukocyte/plasma fraction collected during the ini-tial phase of photopheresis.5

Variations in methoxsalen levels highlight theneed to establish a patient’s pattern of methoxsalen lev-els. While bath PUVA has been used in Europe, logisti-cal issues have limited its implementation in the UnitedStates. Thus, for patients with psoriasis there seems tobe no alternative to oral ingestion. For photopheresis,however, of these alternate routes, the direct injectionof a methoxsalen solution into the collected lympho-cytes is feasible. Finally, it is important to note thatwhile no level of methoxsalen has been related to thera-peutic efficacy, it would seem prudent to aim for repro-ducibility.

Peter Donath, MDVienna, AustriaDeidre Bethea, BSLouis Amici, MSKier DeLeo, MSXiao-Mei Wang, MDJohn Battista, BSPhiladelphia, PaRobert M. Knobler, MDViennaFrancis P. Gasparro, PhDDepartment of DermatologyThomas Jefferson UniversityPhiladelphia, Pa(e-mail: [email protected])

1. Parrish JA, Fitzpatrick TB, Tannenbaum L, et al. Photochemotherapy of pso-riasis with oral methoxsalen and long wave ultraviolet light. N Engl J Med.1974;291:1207-1212.

2. Gasparro FP, ed. Extracorporeal Photochemotherapy. Georgetown, Tex: RG Lan-des Co, Medical Intelligence Unit; 1994:chap 1.

3. Gasparro FP, Battista J, Song J, et al. Rapid and sensitive analysis of 8-me-thoxypsoralen in plasma. J Invest Dermatol. 1988;90:234-236.

4. Fischer T, Alsins J. Treatment of psoriasis with trioxsalen baths and dyspro-sium lamps. Acta Derm Venereol. 1976;56:383-390.

5. Knobler RM, Trautinger F, Graninger W, et al. Parenteral administration of8-methoxypsoralen in photopheresis. J Am Acad Dermatol. 1993;28:580-584.

Bath Psoralen–UV-A Therapyfor Persistent Grover Disease

I n 1970, Grover described a novel disease character-ized by the sudden onset of intense pruritus and dis-seminated papules on the trunk. It has a typical his-

tological picture, manifests predominantly in white menolder than 40, and is frequently self-limited. However, insomepatientsthediseasepersistsforseveralmonthsoryears,and therefore theoriginalname, transientacantholyticder-matosis, has been abandoned in favor of Grover disease.1

Although a benign course of Grover disease re-sponds to appropriate skin care and topical corticoste-roids, treatment of persistent disease frequently remainsunsatisfactory, and a standard therapy is still missing. Re-lief may be obtained by systemic treatments with eitherisotretinoin, corticosteroids, or relatively high doses ofmethotrexate.1,2 The etiology remains unclear, but it is wellestablished that Grover disease may be provoked by UVexposure, ionizing radiation, heat, and sweating.3 Most in-terestingly, Paul and Arndt4 reported provocation of Groverdisease during oral psoralen–UV-A (PUVA) therapy andclearing following continuation of oral PUVA, suggest-ing that Grover disease may be a photosensitive derma-tosis responsive to photochemotherapy. Bath PUVA therapyis an important alternative to oral PUVA. For most indi-cations it is at least as effective as oral PUVA but avoidsmost of the adverse effects.5

Report of Cases. Three patients with persistent Groverdisease received bath PUVA therapy (Table). Despite thelong-standing disease, all patients’ symptoms cleared com-pletely with 10 to 37 treatments (Figure 1 andFigure 2). Bath PUVA therapy was performed as de-scribed5 and complete remission of pruritus and skin erup-tions was achieved in all 3 patients (Figures 1 and 2).No relapse was observed within more than 1 year. His-tological samples from a previously involved skin areawere reassessed 6 months after bath PUVA therapy withno pathologic findings.

Comment. Despite careful treatment, Grover disease doesnot resolve in some patients.1 Systemic corticosteroids,

Clinical Presentation, History, and Treatment of 3 Consecutive Men With Persistent Grover Disease

Patient Age, y(Fitzpatrick Skin Type) Clinical Presentation

HistologicalCharacteristics Prior Therapy

Psoralen–UV-ATreatments, No.

CumulativeUV-A Dose, cm2

58 (I) Widespread, symmetric,papular eruptions,extensive itchingfor 8 mo

Parakeratosis, focalacantholysis,dyskeratosis

Clobetasol-dipropionateointment for 2 mo

37 54.4

70 (I) Disseminated, partlycrusted seropapules onthe trunk; severe pruritusover 2 mo

Suprabasal cleft withacantholytic anddyskeratotic cells

16 mg oral prednisoloneand betamethasone17,21-dipropionate ointmentfor 1 mo

10 9.8

73 (II) Keratotic, partly excoriatedand crusted papules onthe back, disseminatedeczema for 3 y; severepruritus

Epidermal spongiosis withfocal acantholysis anddyskeratosis

Various external corticosteroidsand oral antihistaminesfor 2 y

20 49.7



vitamin A derivatives, or other immunosuppressive agentsmay be effective in these patients, but relapses are fre-quent after the drug therapy is discontinued. As oral PUVAtherapy achieved complete clearance of skin eruptionsand pruritus in a single patient, and its usefulness hasbeen suggested in reviews,4,6 we investigated the effi-cacy of bath PUVA therapy. The results suggest that thismode of PUVA treatment, which is almost devoid of theadverse effects associated with oral PUVA and which in-duces sensitivity to UV-A for less than 1 hour,5 is an im-portant therapy for persistent Grover disease.

Matthias Luftl, MDKlaus Degitz, MDGerd Plewig, MDMartin Rocken, MDDepartment of DermatologyLudwig-Maximilians-University MunichFrauenlobstrasse 9-11D-80337 MunichGermany(e-mail: [email protected])

1. Parsons JM. Transient acantholytic dermatosis (Grover’s disease): a global per-spective. J Am Acad Dermatol. 1996;35:653-666.

2. Helfman RJ. Grover’s disease treated with isotretinoin: report of four cases.J Am Acad Dermatol. 1985;12:981-984.

3. Fawcett HA, Miller JA. Persistent acantholytic dermatosis related to actinicdamage. Br J Dermatol. 1983;109:349-354.

4. Paul BS, Arndt KA. Response of transient acantholytic dermatosis to photo-chemotherapy. Arch Dermatol. 1984;120:121-122.

5. Luftl M, Degitz K, Plewig G, Rocken M. Psoralen bath plus UV-A therapy:possibilities and limitations. Arch Dermatol. 1997;133:1597-1603.

6. Honig B, Morison WL, Karp D. Photochemotherapy beyond psoriasis. J AmAcad Dermatol. 1994;31:775-790.

Figure 1. Grover disease prior to (left) and after (right) bath psoralen–UV-A therapy.

Figure 2. Histological specimen of a representative papule(hematoxylin-losin) original magnification 3 40.



Effects on Human Epidermisof Chronic Suberythemal Exposureto Pure Infrared Radiation

W e have characterized the chronic effects ofnear infrared radiation on human epider-mis in vivo and want to compare this with

the alterations reported to occur after UV irradiation.

Patients and Methods. We studied 8 normal healthy vol-unteers (4 men, 4 women, age range 28-39 years; skinphototypes I through III) with no history of skin disor-ders and minimal solar damage.

An infrared lamp (Hydrosun 500; Hydrosun Mediz-intechnik GmBh, Germany) with single-output band (620-1370 nm) and an irradiance of 440 mW · cm−2 was used.Irradiance between 250 nm and 400 nm is 0.0034mW · cm−2, but from 250 nm to 340 nm is only 0.00045mW · cm−2, 106 less than the main infrared band; 9.5 min-utes’ irradiation at 30 cm is equivalent to a 1-hour ex-posure to a typical solar spectrum (ASTM).1

Irradiation of buttock skin for 9.5 minutes pro-duced no pain and was administered randomly to eitherthe left or the right side daily for 6 weeks, excluding week-ends. The 10 3 10-cm area irradiated was masked usinga 15-mm-thick wooden board. A skin temperature in-crease of 2.6°C was measured during irradiation, result-ing in a typical skin temperature of 33.4°C.

Asessments were made 1 day after the last irradia-tion. Skin roughness was assessed by stylus profilome-try of silicone rubber “negative” replicas of the skin. DINstandard roughness parameters Ra (arithmetic meanheight of surface profile) and Rz (mean of maximum peakto valley height from 5 equal sections of the profile) weremeasured in triplicate using 4.8-mm scans.

Four-millimeter punch biopsy specimens weretaken from irradiated and unirradiated sites. One half ofeach biopsy specimen was for standard (hematoxylin-eosin) histology, the other for frozen section and histo-chemical studies. Using image analysis under lightmicroscopy, the mean epidermal thickness (MET) wasdetermined from a minimum of 5 fields across eachspecimen. Using cryostat sections, activity of glucose-6

Characteristics of Nonirradiated and Irradiated Skin Samples

Characteristic*

Sites, Mean (SD)

Difference, Mean (SD)

Within Subject Differences, P

Nonirradiated Irradiated t Test Wilcoxon Signed Rank Test

Ra, µm 13.2 (3.2) 15.5 (4.9) 2.35 (2.5) .015 .007Rz, µm 61.1 (11.9) 69.7 (19.9) 8.61 (10.2) .024 .015MET, µm 57.82 (14.0) 63.7 (12.1) 5.83 (8.69) .049 .040G6PEAK 1.29 (0.5) 1.62 (0.4) 0.334 (0.46) .04 .053G6AUC 19.61 (6.6) 24.53 (6.2) 4.92 (6.4) .033 .030

*Ra indicates arithmetic mean height of surface profile; Rz, mean peak to valley height; MET, mean epidural thickness; G6PEAK, density of glucose-6-phosphatedehydrogenase (G6PD) enzyme staining in granular layer (optical density units); and G6AUC, area under curve of G6PD epidermal distribution (arbitrary units).

Buttock skin sample from volunteer 6, stained for glucose-6 phosphate dehydrogenase enzyme. Left, Irradiated site. Right, Nonirradiated site.



phosphate dehydrogenase (G6PD) enzyme was mea-sured.2 Using a suitable narrow bandwidth filter (597nm, 43.5-nm bandwidth), the activity of the enzymewas assessed by taking the peak optical density (a mea-sure of granular layer staining), and by integrating thestain density over the whole of the epidermis. Within-subject differences were tested using the parametricpaired t test and the nonparametric Wilcoxon matchedpairs signed rank test.

Results. Summary results are given in the Table. Themean epidermal thickness from irradiated sites was 5.8µm (SD, 8.7 µm) thicker than the nonirradiated sites. Boththe peak (granular layer) and the total (area under op-tical density curve) staining of G6PD enzyme were sig-nificantly higher in irradiated sites, with an average dif-ference of 0.33 (SD, 0.46) optical density units and 4.92(SD, 6.4) arbitrary units, respectively. The Figure showstypical sections from irradiated (left) and nonirradiated(right) specimens. Irradiated sites had significantly greaterroughness than the control sites by both the Ra and theRz parameters.

Comment. A significantly thicker epidermis was foundin irradiated skin than in nonirradiated skin in thisstudy. Glucose-6-phosphate dehydrogenase is anenzyme that is found in the cytoplasm of cells, and anincrease in its activity has been found to be a markerof UV damage.2 It is found in nonirradiated epidermisas a band of reaction product in the granular layerwith only a small amount of activity in the malphigianand basal cell zones. In irradiated epidermis, this bandspreads throughout the epidermis showing an increasedactivity down as far as the basal layer (Figure, left).

Cytochemical changes in G6PD activity have alsobeen reported in solar keratoses and in preneoplastic andneoplastic epithelium in both rat and human tissue. Itseems possible therefore that metabolic changes, such asthe increase in G6PD activity seen here after infrared ir-radiation, are sensitive markers of early epidermal dam-age and are not specifically related to UV radiation.

The higher roughness parameters suggest that somedamage has been inflicted on the irradiated sites. Thiscould be caused by the increased production of epider-mis, or a disruption in normal desquamation

Definite signs of epidermal response to damage havebeen detected, so it seems that the pure infrared irradia-tion regimen used is capable of inducing skin changes,even in the absence of a notable increase in skin tissuetemperature. In particular, the level of the enzyme G6PDwas considerably elevated after the mild but pure infra-red irradiation regimen. It is believed that this is a newfinding and merits further investigation.

Christopher Edwards, PhDStephen A. GaskellStephanie A. Hill, DSR(T)Robert Heggie, BScAnthony D. Pearse, MScRonald Marks, FRCP, FRCPathDepartment of DermatologyUniversity of Wales College of MedicineHeath ParkCardiff CF4 4XN, England

1. American Society for Testing and Materials.Terrestial Solar Spectral Irradianceat Air Mass 1.5 for a 37 (Tilted Surface). West Conshohocken, Pa; 1996:E892-87.

2. Pearse AD, Marks R. Response of human skin to ultraviolet radiation: disso-ciation of erythema and metabolic changes following sunscreen protection.J Invest Dermatol. 1983;80:191-194.



Documents

Interleukin 1 Receptor Antagonist Deficiency Presenting as Infantile Pustulosis Mimicking Infantile Pustular Psoriasis