IJCRI – International Journal of Case Reports and Images, Vol. 4 No. 6, June 201 3. ISSN – [0976-31 98]

IJCRI 201 3;4(6):31 6–320.www.ijcasereportsandimages.com

Postoperative Stevens–Johnson syndrome secondary topiperacillin–tazobactam: A case reportMihir M Shah, Kelsey Larson, Neilendu Kundu, R Matthew Walsh

ABSTRACTIntroduction: Toxic epidermal necrolysis (TEN)and Stevens–Johnson syndrome (SJS) aresevere, potentially fatal cutaneous reactions tomedications affecting only 1–2 per millionindividuals each year. They present astenderness, erythema and blistering of both theskin and mucus membranes. We describe a caseof Stevens–Johnson syndrome linked topiperacillin–tazobactam treatment, which is apreviously unreported cause and a rareoccurrence in postoperative patients. CaseReport: A 37­year­old female with chronicpancreatitis post­total pancreatectomy withauto­islet transplant was re­admitted onpostoperative day 12 with abdominal pain andleukocytosis. Her workup revealed a small intra­abdominal fluid collection suspicious for anabscess and she was empirically started onpiperacillin–tazobactam. She developed hiveson her extremities with associated itching afterfour days of treatment. Piperacillin–tazobactamwas discontinued, diphenhydramine wasadministered for symptomatic relief, and shewas discharged on an oral regimen ofciprofloxacin and metronidazole. She returnedthe following day with a rash extending onto herchest, abdomen, feet, face, lips and mouth, with

newly appeared bullae covering approximately35–50% of body surface. A skin biopsy wasconsistent with SJS, with piperacillin–tazobactamidentified as the likely causative agent.Conclusion: The most common medicationslinked to TEN/SJS are sulfonamide antibiotics(especially sulfamethoxazole/trimethoprim),cephalosporins, quinolones, allopurinol,carbamazepine, phenytoin, phenobarbital, andNSAIDs (oxicam­type). While medications themost common cause, mycoplasma pneumoniaeand herpes simplex virus (HSV) have also beenlinked to SJS. The mortality rate of biopsy­confirmed SJS ranges from 1–5%, whereas themortality of TEN ranges from 25–35%. Thoughmany patients experience pruritis and/or rashesas side­effects to medications given post­operatively, it is important to keep in mind moreserious reactions such as TEN and SJS as theseconditions as they have a substantial associatedmorbidity and even mortality.Keywords: Stevens–Johnson syndrome, Toxicepidermal necrolysis, Piperacillin–tazobactam

*********Shah MM, Larson K, Kundu N, Walsh RM. PostoperativeStevens–Johnson syndrome secondary topiperacillin–tazobactam: A case report. InternationalJournal of Case Reports and Images 2013;4(6):316–320.

*********doi:10.5348/ijcri­2013­06­321­CR­6

INTRODUCTIONStevens­Johnson syndrome (SJS) and toxicepidermal necrolysis (TEN) are rare, but severe,potentially fatal exfoliating dermatitis conditions

CASE REPORT OPEN ACCESS

Mihir M Shah1 , Kelsey Larson1 , Neilendu Kundu1 , RMatthew Walsh2

Affi l iations: 1MD, Resident, General Surgery, ClevelandClinic Foundation, Cleveland, Ohio, USA; 2MD, Faculty,General Surgery, Cleveland Clinic Foundation, Cleveland,Ohio, USACorresponding Author: Mihir M Shah, 35 Severance Circle,Apt 311 , Cleveland Heights, Ohio, USA, 4411 8; Ph: (+1 )21 5-200-6256; Fax: (+1 ) 21 6-445-7653; Email :mihir253@yahoo.co. in

Received: 21 August 201 2Accepted: 25 October 201 2Published: 01 June 201 3

Shah et al. 31 6

IJCRI – International Journal of Case Reports and Images, Vol. 4 No. 6, June 201 3. ISSN – [0976-31 98]

IJCRI 201 3;4(6):31 6–320.www.ijcasereportsandimages.com Shah et al. 31 7

affecting only 1–2 per million individuals annually [1].They typically present as tenderness, erythema, andblistering of both skin and mucus membranes [1]. Wereport a case of SJS linked to piperacillin­tazobactamtreatment, which is a previously unreported cause and arare occurrence in postoperative patients.

CASE REPORTA 37­year­old female with a past medical historysignificant for Turner’s syndrome, type­II diabetesmellitus and chronic pancreatitis, had undergonepancreatic duct drainage and partial pancreatic headresection (Frey procedure) 18 months prior to thisadmission. She had persistent pancreatic pain requiringnarcotics and repeated hospitalizations. She presentedto the clinic with chronic, epigastric abdominal pain. Adiscussion was held and she elected to proceed withtotal pancreatectomy with auto­islet transplant, whichwas subsequently performed at our institution. Herpostoperative course was unremarkable and she wasdischarged on postoperative day (POD) 11 insatisfactory condition.The patient returned the day after discharge withworsening abdominal pain, mild distention, andworsened leukocytosis. A computed tomography (CT)scan was performed demonstrating a 4x2 cm fluidcollection in the mesentery. She was empirically startedon piperacillin­tazobactam 3.37 mg IV Q6 hrs afterobtaining samples for blood and urine cultures. Herblood and urine cultures were negative, she remainedafebrile, and her white blood cell count continued todecrease. Prior to discharge four days later, shedeveloped hives and itching on her extremities.Piperacillin­tazobactam was stopped and she wasadministered diphenhydramine for relief of symptoms.Her symptoms were controlled and she was dischargedhome in stable condition on an oral regimen ofciprofloxacin 250 mg BID and metronidazole 500 mgdaily.On the evening of discharge (POD 16), the patientnoted spreading of her rash onto her chest, abdomen,feet, mouth and lips with the appearance of bullae. Shedeveloped fever, chills, rigors as well as dysuria andoropharyngeal burning. She returned to the emergencydepartment the next morning. Upon evaluation, shewas noted to have 35–50% of her body surface areacovered with a rash and bullous lesions (Figure 1A–D).Initial concern was for TEN versus SJS. Dermatologyand ophthalmology were promptly consulted, and a skinbiopsy was obtained. IgA levels and blood cultures werenegative. Her skin lesions were treated with Vaselinegauze (Covidien Mansfield, MA) and Kerlix (CovidienMansfield, MA), and she was started on solumedrol 60mg IV BID.Skin biopsy sections demonstrated interfacedermatitis characterized by detached epidermis withnumerous dyskeratotic keratinocytes and scantperivascular infiltrate within the dermis, consistent withSJS (Figure 2A–C). Piperacillin­tazobactam was

suspected as the likely causative agent after a thoroughreview by all teams involved. She did not have anyophthalmologic issues. After stabilization and initialtreatment, she was transferred to the regional burncenter for continuation of management. She wasinitially managed with local skin care and IV antibioticswith significant improvement. She was subsequentlylost to follow­up after discharge from the burn center.Unfortunately she is now listed as deceased in electronicmedical record system, with unknown date and cause ofdeath.

DISCUSSIONEpidemiology: TEN and SJS occur inapproximately 1–2 per million individuals each year,with relatively similar rates reported in multiplecountries around the world [1, 2]. Small differences inthe rates of SJS and TEN can likely be attributed togenetics, medication availability, prescribing practicesand other concurrent conditions, including cancers andinfection. One notable population with a significantlyincreased risk for SJS and TEN includes those infectedwith HIV, as the annual incidence is 1000 times higherthan the general population (1/1000) [1, 3, 4].Clinical Evaluation: Initial presentation of thedisease typically begins several days prior to cutaneousmanifestations. Symptoms include fevers, keratitis sicca,odynophagia and malaise making it difficult todifferentiate from routine drug reactions [1, 2, 5]. Whenskin lesions first appear, they often begin on the trunk

Figure 1: (A) Right hand and forearm with flaccid bullae,superficial desquamation and a solitary denuded bulla, (B)Right medial thigh with tense and flaccid vesicles and bullaeupon a background of erythema, (C) Upper cutaneous lip,vermillion border and oral aperatures with superficialdesquamation. Tongue without vesicles, bullae ordesquamation, (D) Right anterior thigh with solitary serous­filled bulla within a field of superficial desquamation.

IJCRI – International Journal of Case Reports and Images, Vol. 4 No. 6, June 201 3. ISSN – [0976-31 98]

IJCRI 201 3;4(6):31 6–320.www.ijcasereportsandimages.com Shah et al. 31 8

and face in addition to the palms and soles and appearas dusky red, flat, poorly delineated areas. They thencontinue to spread outward [1, 5]. Evidence ofinvolvement of the buccal, genital, and/or ocularmucosa should raise suspicion of SJS or TEN and occursin >90% of patients [1]. Ocular involvement early iscommon, and can include conjunctivitis, edema,erythema, crusting, discharge erosions and cornealulcerations [1, 6, 7].Epidermal detachment usually occurs as a slightlylater secondary manifestation. The patent can be testedby applying lateral mechanical pressure and observingfor detachment (Nikolsky’s sign) or visualizing bullaeextending to surrounding skin when vertical pressure isapplied (Asbor–Hansen sign). While notpathognomonic for SJS or TEN, these exam findings arehelpful for ascertaining the danger of the reactions. Inaddition to epidermal involvement, the respiratory,gastrointestinal (including the liver), and urinarysystems can be involved [8–10].The TEN survivors have a 50% risk of long­termsequelae including changes in pigmentation of the skinand ocular complications ranging from simple (keratitissicca, 46%) to severe (visual loss 5%, corneal ulceration2%) [1, 11].Etiology: The pathophysiology of SJS and TENremains unclear. Current theories support immune­mediated phenomena, resulting in keratinocyteapoptosis followed by necrosis. The generalizedmechanism for this reaction is MHC­1 drug presentationleading to clonal expansion of cytotic T­cells, found inhigh numbers in the blisters of SJS/TEN patients [1].However, additional mediators including FasL andgranulysin (cytotoxic molecule), are likely to playprimary roles in the actual generalized apoptosis of thekeratinocytes [1, 5]. The Fas­FasL pathway has been

linked to apoptosis in SJS/TEN via triggeredintracellular DNA degradation [5]. The use of IVIG astreatment for SJS/TEN is based on this link between thedisease and FasL [1, 5]. The exact pathway by which amedication leads to secretion or activation of Fas­FasLand/or ganulysin and thus keratinocyte apoptosisremains to be explained.Genetics: The link between SJS and genetic HLAclass was first reported in the Han Chinese populationwith a strong association between HLA­B*1502 andcarbamazepine [1, 12]. This association was confirmedin other Asian populations. However, this was not validin European populations, leading to the conclusion thatgenetic markers are not specific in predicting SJS/TENsusceptibility in the general population [1].Drugs: Numerous medications have been identifiedas causative agents for SJS and TEN. The most commonmedications linked to SJS/TEN over short­term useinclude trimethoprim­sulfamethoxazole and othersulfonamides, aminopenillins, cephalosporins,quinolones, and chlormezaone. Some medications arelinked to SJS/TEN following long­term use, such ascarbamazepine, phenytoin, phenobarbital, valproic acid,NSAIDs (oxicam­type), nevirapine, lamotrigine,sertraline, and allopurinol. With these medications,SJS/TEN usually occurs within two months afterbeginning treatment [1, 3].Disease: While medications, the primary cause ofSJS and TEN, mycoplasma pneumoniae and HSV havealso been linked to development of these diseaseswithout medication exposure [1]. HIV patients are atsignificantly higher risk for development of SJS/TENcompared to the general population.Diagnosis: The diagnosis of SJS versus TEN relieson clinical vigilance and pathologic confirmation. SJS isdefined by mainly isolated lesions with <10% bodysurface area skin detachment, while TEN is a moresevere and wide­spread form with widely confluentlesions and skin detachment covering >30% of the bodysurface area [1]. Skin biopsy is the only way to confirmdefinitive diagnosis and rule­out other autoimmuneblistering diseases, fixed drug reactions, andstaphylococcal scaled skin syndrome. Histopathologicanalysis reveals necrosis of the epidermis involving alllayers, with evidence of keratinocyte apoptosis. Directimmunofluorescent staining is also carried out to rule­out autoimmune causes [1].Given the severe nature of SJS/TEN and its initialpresentation, a rapid diagnostic test to predict SJSversus benign drug reaction is being developed [13].Patients who are diagnosed with SJS have increasedsystemic levels of FasL as well as increased granulysinlevels. While the FasL serum levels are too low tomeasure, the granulysin serum levels can be measuredin rapid fashion to yield positive versus negative resultsin 15 minutes, similar to a rapid strep test. This methodhad sensitivity of 80% and specificity of 95.8% forSJS/TEN versus benign drug reaction. While not widelyavailable for clinical use, further development of fast,efficient, early diagnostic tools is important forimproved patient outcomes in the future.

Figure 2: (A) Low power image demonstrating subepidermalblister, (B) Within the dermis there is a sparse perivascularlymphocytic infiltrate with interface change and numerousdyskeratotic cells, (C) High power image of the epidermishighlighting the numerous dyskeratotic keratinocytes.

IJCRI – International Journal of Case Reports and Images, Vol. 4 No. 6, June 201 3. ISSN – [0976-31 98]

IJCRI 201 3;4(6):31 6–320.www.ijcasereportsandimages.com Shah et al. 31 9

Management: Initial management when adiagnosis of SJS/TEN is suspected begins withimmediate cessation of all new medications as well asurgent consultation to dermatology and ophthalmology.Universally accepted treatment includes monitoringvolume status, evidence of sepsis, and conservativewound treatment without debridement [1]. The role ofIVIG in treatment in addition to corticosteroids isgenerally accepted, although controversy remainsregarding the exact details of the treatment [14].Prognosis: The SCORTEN scoring system wasdeveloped as a validated means for predicting diseaseseverity and mortality associated with SJS/TEN (Table 1and Table 2) [1]. This score is useful for determinationof appropriate treatment (all scores >3 should betreated in ICU and transferred to a regional burn centeras soon as possible) and patient/family counseling. Theleading cause of death in most patients is sepsis leadingto acute renal failure [1]. The patient in this case reporthad a SCORTEN score of 4 (tachycardia + >10% bodyarea involvement + glucose elevation + bicarbonate>20), corresponding to a 58.3% mortality.Table 1: SCORTEN Scoring System Each “yes” answer to theparameter yields one point, each “no” answer is worth zeropoints.Age >40 years old?Presence of malignancy?Tachycardia (>120 beats per minute)?Initial surface of epidermal detachment >10%?Serum urea >10mmol/L?Serum glucose >14mmol/L?Bicarbonate >20mmol/L?

Table 2: The sum of the SCORTEN criteria yields a scorewhich is correlated with a predicted mortality.SCORTEN (sum of 1A items) Predicted % mortality

0­1 3.2%2 12.1%3 35.8%

4 58.3%>5 90%

CONCLUSIONWhile many patients experience pruritis and/orrashes as side­effects to medications given post­operatively, it is important to keep in mind more seriousreactions such as toxic epidermal necrolysis andStevens–Johnson syndrome as these conditions can be

rare causes of post­operative morbidity and mortality. Ifthere is any concern for possible toxic epidermalnecrolysis or Stevens–Johnson syndrome, promptconsultations to dermatology and ophthalmology shouldbe obtained, possible causative agents should beimmediately discontinued, and early transfer to a burncenter should be initiated following confirmeddiagnosis.*********

Author ContributionsMihir Shah – Conception and design, Acquisition ofdata, Analysis and interpretation of data, Drafting thearticle, Critical revision of the article, Final approval ofthe version to be publishedKelsey Larson – Conception and design, Acquisition ofdata, Analysis and interpretation of data, Drafting thearticle, Critical revision of the article, Final approval ofthe version to be publishedNeilendu Kundu – Acquisition of data, Analysis andinterpretation of data, Drafting the article, Criticalrevision of the article, Final approval of the version to bepublishedR Matthew Walsh – Conception and design, Acquisitionof data, Analysis and interpretation of data, Drafting thearticle, Critical revision of the article, Final approval ofthe version to be publishedGuarantorThe corresponding author is the guarantor ofsubmission.Conflict of InterestAuthors declare no conflict of interest.Copyright© Mihir Shah et al. 2013; This article is distributedunder the terms of Creative Commons Attribution 3.0License which permits unrestricted use, distribution andreproduction in any means provided the original authorsand original publisher are properly credited. (Please seewww.ijcasereportsandimages.com/copyright­policy.phpfor more information.)

REFERENCES1. Harr T, French LE. Toxic epidermal necrolysis andStevens­Johnson Syndrome. Orphanet Journal ofRare Diseases 2010;5:39.2. Sanmarkan AD, Sori T, Thappa DM, Jaisankar TJ.Retrospective analysis of Stevens­JohnsonSyndrome and Toxic Epidermal Necrolysis over aperiod of 10 years. Indian Journal of Dermatology2011 Jan­Feb;56(1):25–9.3. Roujeau JC, Kelly JP, Naldi L, et al. Medication useand the risk of Stevens­Johnson syndrome or toxicepidermal necrolysis. N Engl J Med1995;333(24):1600–7.4. Saiag P, Caumes E, Chosidow O, Revuz J, RoujeauJC. Drug­induced toxic epidermal necrolysis (Lyellsyndrome) in patients infected with the human

IJCRI – International Journal of Case Reports and Images, Vol. 4 No. 6, June 201 3. ISSN – [0976-31 98]

IJCRI 201 3;4(6):31 6–320.www.ijcasereportsandimages.com Shah et al. 320

immunodeficiency virus. Journal of the AmericanAcademy of Dermatology 1992;26(4):567–74.5. Chung WH, Hung SI. Genetic Markers and DangerSignals in Stevens­Johnson Syndrome and ToxicEpidermal Necrolysis. Allergology International2010;59(4):325–32.6. Chang YS, Huang FC, Tsent SH, Hsu CK, Ho CL,Sheu MH. Erythema multiforme, Stevens JohnsonSyndrome, and Toxic Epidermal Necrolysis: Acuteocular manifestations, causes and management.Cornea 2007;26(2):123–9.7. Sotozono C, Ueta M, Koizumi N, et al. Diagnosis andtreatment of Stevens­Johnson Syndrome and ToxicEpidermal Necrolysis with ocular complications.Opthamology 2009;116(4):685–90.8. Lebargy F, Wolkenstein P, Gisselbrecht M, et al.Pulmonary Complications in toxic epidermalnecrolysis: A prospective clinical study. IntensiveCare Medicine 1997;23(12):1237–44.9. Sugimoto Y, Mizutani H, Sato T, Kawamura N,Ohkouchi K, Shimizu M. Toxic epidermal necrolysiswith severe gastrointestinal mucosal cell death: apatient who excreted long tubes of dead intestinalepithelium. Journal of Dermatology1998;25(8):533–8.

10. Ducic I, Shalom A, Rising W, Nagamoto K, MunsterAM. Outcome of patients with toxic epidermalnecrolysis syndrome revisited. Plastic andReconstructive Surgery 2002;110(3):768–3.11. Yip LW, Thong BY, Lim J, et al. Ocularmanifestations and complications of StevensJohnson Syndrome and Toxic Epidermal Necrolysis:an Asian series. Allergy 2007;62(5):527–31.12. Chung WH, Hung SI, Hong HS, et al. MedicalGenetics: a marker for Stevens­Johnson Syndrome.Nature 2004;428(6982):486.13. Fujita Y, Yoshioka N, Abe R, et al. Rapidimmunochromatographic test for serum granulysinis useful for the prediction of Stevens JohnsonSyndrome and Toxic Epidermal Necrolysis. Journalof the American Academy of Dermatology 2011July;61(1):65–8.14. Worswick S, Cotliar J. Stevens­Johnson Syndromeand Toxic Epidermal Necrolysis: a review oftreatment options. Dermatologic Therapy 2011 Mar­April;24(2):207–18.

Access full text article onother devices Access PDF of article onother devices