Genitourin Med 1993;69:208-212

Paranasal sinus disease in HIV antibody positivepatients

Alison Grant, Marcelle von Schoenberg, H R Grant, R F Miller

AbstractObjective-To investigate the prevalenceof radiologically-diagnosed paranasalsinus disease in HIV-1 seropositivepatients.Subjects and setting-476 patientsadmitted to a dedicated inpatient unit forHIV and AIDS at the MiddlesexHospital, London, between September1988 and February 1992.Design-Retrospective review ofpatients' case notes and radiologicalrecords.Results-30 patients (6.3%) had radiolog-ical evidence of paranasal sinus disease.At the time of admission, sinusitis was inthe differential diagnosis in only 12 of the30 patients; 13 patients were initiallydiagnosed as having meningitis. Pseudo-monas aeruginosa was the causativeorganism in four patients, all of whomhad advanced HIV disease. All fourresponded to appropriate antibiotics buthad early relapse of infection.Conclusions-Sinusitis is an importantand under-recognised cause of morbidityin patients with HIV disease.

(Genitourin Med 1993;69:208-212)

sinuses on the basis of sinus radiographs, CTor magnetic resonance imaging of the brain.

Data analysisWe examined the corresponding case notesand recorded the following information: age,sex and risk factors for HIV infection; riskfactors for sinus disease; presenting symp-toms; laboratory investigations and outcome.

Clinical classificationWe defined sinusitis as one or more of thefollowing symptoms: facial pain, rhinorrhoea,post-nasal drip or nasal obstruction due tounderlying inflammation of the nasal orparanasal sinus mucosa. We classified theclinical severity of sinusitis as follows: (i)Asymptomatic: radiological evidence of sinusdisease but either asymptomatic or symptomsattributable to another identified pathology,(ii) Mild: symptomatic sinus disease whichwas self-linting and did not require any spe-cific intervention with antibiotics or antihista-mines, (iii) Moderate: symptomatic diseasewhich was not self-limiting and requiredtreatment, but responded to conventionalmeasures, (iv) Severe: symptomatic diseasethat either persisted after a standard course oftreatment or required operative intervention.

Department ofMedicineA GrantR F MillerDepartment ofOtolaryngology,Ferens InstituteUniversity College andMiddlesex School ofMedicine MiddlesexHospital, MortimerStreet, LondonW1N 8AAM von SchoenbergH R GrantAddress for correspondence:Dr R F Miller, Departmentof Medicine, UniversityCollege and MiddlesexSchool of Medicine,Middlesex Hospital,Mortimer Street, LondonWIN 8AA UK.Accepted for publication18 December 1992

IntroductionAlthough an increased susceptibility to bacte-rial infections has been well described inpatients with HIV disease there have beenfew reports of sinusitis.' Recently we haveseen several HIV seropositive patients pre-senting with severe sinus disease, in four ofwhom Pseudomonas aeruginosa, a rare cause ofsinusitis in the immunocompetent, was thecausative organism.

In this paper we describe our experiencesof managing sinusitis in a cohort of HIVseropositive patients admitted to theMiddlesex Hospital over a 42 month periodand suggest a policy for the management ofthis condition.

MethodsCasefindingAll HIV positive patients at the MiddlesexHospital are admitted to a dedicated unitunder a single consultant. We reviewed theradiological records of all these patients over

a 42 month period (September 1988 toFebruary 1992) and identified all patientswho had radiological abnormalities of the

Radiological classificationWe devised a system for classifying the radio-logical severity of sinus disease by assessingfour radiological parameters in ascendingorder of clinical significance: a higher scoreindicates more severe disease (table 1).

ResultsBetween September 1988 and February1992, 476 HIV seropositive patients wereadmitted to the Middlesex Hospital: of these,30 had radiological evidence of sinus disease.

Table 1 Radiological classification ofsinus disease

Radiological abnormality Score

Mucosal thickening 1 pointFluid level 4 pointsOpaque sinus 8 pointsComplications:Orbital abscess 16 pointsOsteomyelitis 16 pointsIntracranial abscess 32 points

Radiological severity was calculated from the sum of the scorefor each abnormality for each sinus (right and left maxillary,ethmoid, sphenoid and frontal sinuses respectively). A totalscore of 1-7 = mild, 8-15 = moderate and > 16 = severeradiological disease.

208

on February 11, 2020 by guest. P

rotected by copyright.http://sti.bm

j.com/

Genitourin M

ed: first published as 10.1136/sti.69.3.208 on 1 June 1993. Dow

nloaded from

Paranasal sinus disease in HIVantibody positive patients

Sinus radiographs were performed on anothernine patients who presented with headache,and were normal. Two further patients hadskull radiographs performed following a headinjury, and these showed no evidence of sinusdisease. These 11 patients were excludedfrom the subsequent analysis.

Patient characteristicsThe median age was 34 years (range 22 to56). Twenty-seven were male, of whom 21were homosexual, four were intravenous drugusers, one was African and three had noapparent risk factor. Two men were bothhomosexual and intravenous drug users. Ofthe three women, one was an intravenousdrug user and two were African. Twenty-threepatients had a previous AIDS-defining diagnosis.

Risk factors for sinus diseaseSix patients had a past history of sinusitis.Four patients had a history of asthma, onewith coincident bronchiectasis, and one (anintravenous drug user) had bronchiectasiswithout a history of atopy. Ten patients werecigarette smokers and 20 were non-smokersor ex-smokers.

Drug therapyFive patients were taking long term oralsteroids (for immune thrombocytopaenia,oedema due to cutaneous Kaposi's sarcoma,chronic asthma, myositis and to improvewell-being respectively). Twelve patients weretaking continuous antibiotics orally: co-tri-moxazole (9) and dapsone (1) as prophylaxisagainst 3Pneumocystis carinii pneumonia;rifampicin, isoniazid and pyrazinamide fortreatment of tuberculosis (2); erythromycin(1) and doxycycline (1) for recurrent bronchi-tis. Nine patients were receiving monthlynebulised pentamidine. Four patients weretaking zidovudine and two ganciclovir, but allsix had neutrophil counts greater than 0-6 x109 per litre.

Laboratory investigationsOne patient had severe neutropaenia; his onlymedication was co-timoxazole 960mg daily.The median white cell count was 3.3 (range0-6-15.7) x 109 per litre and the medianlymphocyte count was 06 (range 0.28-2.1)x 109 per litre. A CD4 count was available in19 patients: the median was 0.02 (range0-0.78) x 109 per litre (normal range =0.35-2.2, x 109 per litre).

Clinical presentationOf the 30 patients, 16 had specific symptomsof sinus disease (usually unilateral facialpain). 11 others had non-specific symptoms,usually symmetrical headache, and theremaining three patients were asymptomatic.

In 13 patients a clinical diagnosis ofmeningitis was made initially, and of these 10had a lumber puncture performed to excludethis diagnosis. Of the three patients in whomlumbar puncture was not carried out, the rea-sons were the presence of a mass lesion oncranial imaging, spontaneous resolution of

the headache and refusal to undergo the pro-cedure respectively. Two patients were diag-nosed as having meningitis on the basis ofcerebrospinal fluid examination, but in theremaining eight sinusitis was the most likelycause of the presenting symptoms.

Twelve of the 30 patients had a clinicaldiagnosis of sinusitis made at the time ofadmission.We present two case histories to illustrate

some of the problems caused by severe sinusdisease.

Case 1A 56 year old Caucasian homosexual manpresented with a 5 day history of frontalheadache, fever and nasal congestion. Tenmonths previously he had had Pneumocystiscarinii pneumonia: at this time his CD4 countwas 0.06 (normal range 0.35-2.2) x 109 perlitre. Subsequently zidovudine 250 mg qdsand nebulised pentamidine 300 mg monthlywere commenced, and he had remained wellin the intervening period. On examination hewas febrile at 38°C and neurological exami-nation was normal. Cryptococcal meningitiswas suspected. Prior to lumbar puncture, CTof the brain was performed, which showeddiffuse cortical atrophy. Microscopy of thecerebrospinal fluid obtained at lumbar punc-ture was normal and culture was negative. Atthis point sinusitis was suspected, and sosinus radiographs were peformed, revealingbilateral maxillary fluid levels. He was treatedwith oral amoxycillin and ephedrine nosedrops but his symptoms persisted: a secondlumbar puncture was performed and cere-brospinal fluid examination was again nor-mal. In view of his persisting symptoms ofsinusitis, a right antral wash-out was per-formed. Culture of lavage fluid revealed amixed growth of organisms including Pseudo-monas aeruginosa, sensitive to polymixin,amikacin, ciprofloxacin, azlocillin, cefta-zidime and gentamicin. Follow-up sinusradiographs showed a pansinusitis (fig 1).Chloramphenicol was commenced. Thepatient improved but his fever returned whenchloramphenicol was stopped. He was there-fore given oral ciprofloxacin, but after com-pleting this second course of antibiotics hissymptoms of headache and nasal congestionrecurred, and so he was maintained on con-tinuous antibiotics after discharge.

Six months later the patient was readmit-ted with a 10 day history of severe frontalheadache and retro-orbital pain. A CT scanof the sinuses showed right ethmoidal sinusi-tis with erosion of the medial orbital wall (fig2). Sputum culture revealed Pseudomonasaeruginosa with the same sensitivity pattern asfound previously. After seven days of intra-venous azlocillin and gentamicin there wasclinical improvement, but when antibioticswere stopped he deteriorated rapidly. To pro-mote resolution of infection his ethmoid aircells were drained under direct transnasalendoscopic vision with good results. In thepostoperative period he went into renal fail-ure and he died 3 weeks later.

209 on F

ebruary 11, 2020 by guest. Protected by copyright.

http://sti.bmj.com

/G

enitourin Med: first published as 10.1136/sti.69.3.208 on 1 June 1993. D

ownloaded from

Grant, von Schoenberg, Grant, Miller

Figure 1 Case 1: Sinusradiographs at originalpresentation showingbilateral maxillary fluidlevels and opaque rightfrontal sinus.

Case 2A 35 year old Caucasian homosexual manpresented with a two day history of left peri-orbital pain, purulent nasal discharge, post-nasal drip, fever and pain on swallowing. Healso described transient sensory symptoms inhis left arm. In the six weeks prior to thisadmission he had been hospitalised severaltimes because of recurrent bacterial chestinfections due to Pseudomonas aeruginosa.Treatment had been complicated by thedevelopment of hypersensitivity reactions toseveral antibiotics.On examination he was unwell and febrile

at 37 5°C. There were no abnormal neuro-logical signs. In view of his sensory symp-toms, magnetic resonance imaging of thehead was performed: this showed no intrac-erebral abnormality but showed mucosalthickening and a fluid level in the left maxil-lary antrum, mucosal thickening of both eth-moid sinuses and the right sphenoid sinusand abnormal signal from the left maxillasuggestive of osteomyelitis (figs 3 and 4). Achest radiograph showed a small cavitating

lesion in the right upper zone. The CD4count was severely reduced at 0-01 x 109 perlitre. A left antral wash-out revealed mucopu-rulent fluid which on culture yielded a heavygrowth of Pseudomonas aeruginosa, as did sam-ples of spontaneously expectorated sputum.These isolates were sensitive to polymixin,amikacin, ciprofloxacin, azlocillin, cef-tazidime and gentamicin. Bronchoalveolarlavage was performed in order to exclude pul-monary tuberculosis, and lavage fluid alsoyielded Pseudomonas aeruginosa on culture,with the same pattern of antibiotic sensitivityas the previous isolates. Oesophageal candidi-asis was found on endoscopy. The patientimproved on intravenous azlocillin and gen-tamicin, but he developed a diffuse drug rashafter nine days and so his antibiotics werestopped. Four days later his symptomsreturned, and he was treated with intravenousaztreonam and tobramycin to which heresponded rapidly. After ten days of treat-ment he again developed a drug rash andantibiotics were stopped. There was no recur-rence of his symptoms and he was subse-quently discharged home.

Clinical severitySix patients were asymptomatic, eight hadmild disease, 10 had moderate disease and sixhad severe disease. Of those with severe dis-ease, four patients required operative inter-vention. One patient had bilateral intranasalantrostomies to facilitate sinus drainage andresolution of infection. Three patients under-went antral wash-outs, one requiring the pro-cedure twice because of persisting symptoms:this latter patient ultimately underwent func-tional endoscopic sinus surgery.

ImagingSixteen patients had plain sinus radiographs,11 patients had CT of the head or the sinusesand 16 patients had magnetic resonance

Figure 2 Case 1: CTscan on re-presentationshowing right ethmoidalsinusitis with erosion of themedial orbital wall(arrows).

Figure 3 Case 2: T2-weighted magnetic resonance imageshowing mucosal thickening and a fluid level maxillaryantrum, and abnormal signalfrom the left maxilla(arrows).

210 on F

ebruary 11, 2020 by guest. Protected by copyright.

http://sti.bmj.com

/G

enitourin Med: first published as 10.1136/sti.69.3.208 on 1 June 1993. D

ownloaded from

Paranasal sinus disease in HIV antibody positive patients

Figure 4 Case 2:T2-weighted magneticresonance image showingbilateral ethmoidal andright sphenoidal mucosalthickening.

imaging studies. Using the scoring systemdescribed, 19 patients had mild disease, sixhad moderate disease and five had severeradiological disease. Table 2 shows the corre-lation between the clinical and the radiologi-cally-defined severity of sinusitis.

BacteriologySix patients had positive microbiological cul-tures relevant to their sinus disease.Pseudomonas aeruginosa was isolated in fourpatients, all of whom had advanced HIV dis-ease. In all four cases the infection was com-munity-acquired. These isolates were fromsputum (1 patient); sputum, post-nasal spacebiopsy and blood culture (1 patient); sputum,bronchoalveolar lavage fluid and sinuswashings (1 patient) and sinus washings(1 patient). The sensitivity pattern of the

Table 2 Correlation of clinical and radiological severtyofsinusitis

Radiological seventy

Clinical severity Mild Moderate Severe

Asymptomatic 4 2 0Mild 6 2 0Moderate 9 0 1Severe 0 2 4

Table 3 Outcome of ten patients specifically treatedfor sinusitis

Patient Organism Initial therapy Outcome

1 AB Improved, no recurrence2 AB and Improved, no recurrence

bilateral AWO3 Ps. aeruginosa AB Persistent symptoms, required

AWO twice. Further recurrenceat 6 months.

4 H. influenzae AB Improved, no recurrence5 S. aureus AB and BINA Improved, no recurrence6 Ps. aeruginosa AB and left AWO Relapsed after 1 course of

antibiotics, better aftersecond course

7 AB Partial improvement of symptoms8 Ps. aeruginosa AB Partial improvement of symptoms,

recurrence at 2 months9 AB Improved, no recurrence10 AB Improved, recurrence at 3 months

AB-antibiotics, AWO-antral washout, BINA-bilateral intranasal antrostomies.

organisms was the same in all four patients,as described in the case histories.Two further patients had positive microbi-

ological cultures: one had Haemophilusinfluenzae isolated from the sputum and onehad Staphylococcus aureus isolated from sinuswashings.

InterventionTen patients received specific treatment forsinusitis: all 10 were given 0 5% ephedrinenose drops for symptom relief, and two alsoreceived systemic antihistamines. All 10 werecommenced on antimicrobial therapy, five bythe intravenous route and five orally. Fourpatients required operative intervention.Thirteen patients received broad spectrumantibiotics primarily with the intention oftreating infection at another site which is like-ly to have had an additional beneficial effecton the observed sinus disease. In sevenpatients there was no specific intervention.

OutcomeThe outcome in those patients treated specifi-cally for sinusitis is shown in table 3. Therewas recurrent disease in one of the six asymp-tomatic patients, two of eight with mild dis-ease, 3 of 10 with moderate disease and fourof six who had severe disease.

DiscussionOnly four patients underwent surgical proce-dures, and those procedures that were per-formed were relatively minor. This compareswith intervention rates reported by othergroups of four out of 752 and 18 out of 68.As a definite bacteriological diagnosis wasmade in only a few patients, we suggest thatearly surgical intervention might both providethe diagnosis and alleviate symptoms. Thebenefit of early intervention is illustrated bythe case histories.

Previous authors have differed in theirstrategy for the management of sinusitis inpatients with AIDS, particularly with respectto operative intervention.6 We concur withthe view that if sinusitis does not respond toinitial treatment with standard antibioticssuch as doxycycline then a sinus wash-outshould be performed in order to make a bac-teriological diagnosis and enable rationalantibiotic therapy.4 This is particularly impor-tant because in addition to Pseudomonasaeruginosa, a wide variety of unusual organ-isms have been described as a cause of sinusi-tis in patients with AIDS, includingCryptococcus neoformans and cytomegalovirus,7infections which would not respond to stan-dard broad-spectrum antibiotics.

In an HIV seropositive patient who pre-sents with headache meningitis must beexcluded, and cranial imaging (with CT ormagnetic resonance imaging) and a lumbarpuncture are needed to rule out this diagno-sis. Once meningitis has been excluded,sinusitis should be considered in the differen-tial diagnosis. Plain sinus radiographs shouldbe performed to confirm the diagnosis: if

21 on F

ebruary 11, 2020 by guest. Protected by copyright.

http://sti.bmj.com

/G

enitourin Med: first published as 10.1136/sti.69.3.208 on 1 June 1993. D

ownloaded from

Grant, von Schoenberg, Grant, Miller

these are normal, sinusitis is unlikely. In astudy of 54 immunocompetant patients pre-sentmg with symptoms and signs of sinusitis,only 9% of patients with normal sinus radi-ographs had bacteriological evidence ofsinusitis on antral wash-out.8 Conventionalaxial CT may not detect the presence of sinusdisease; coronal scans may be required. If apatient with sinusitis is not systemicallyunwell and there is no clinical evidence ofcomplications such as orbital cellulitis orintracranial infection, a 10-day course of oralantibiotics (for example doxycycline) shouldbe given. Failure to respond to this regime isan indication for coronal CT of the sinuses toassess disease extent and look for evidence ofosteomyelitis of the ethmoids, orbital bonesor mai.ilae, or intracranial sepsis, particularlya frontal lobe abscess; at this stage sinuswash-outs should be performed to make abacteriological diagnosis.

If the patient is systemically unwell, admis-sion is recommended for a sinus wash-outand intravenous antibiotics. If there is noimprovement after 48 hours of treatment, orif there are complications of sinus disease(orbital cellulitis or intracranial infection),coronal CT of the sinuses should be per-formed to identify the extent and severity ofthe disease and assess the need for furthersurgical intervention.

Sinusitis due to Pseudomonas aeruginosamay require a prolonged course of antibiotics,because of the tendency to relapse aftershorter courses of antibiotics and thepotential for osteomyelitis. This is analagousto the treatment of malignant otitis externa, achronic osteomyelitis of the temporal bone inimmunocompromised patients due to

infection of the external ear canal byPseudomonas aeruginosa.9

Sinusitis is rarely included in the differen-tial diagnosis of headache in HIV seropositivepatients. Surgical intervention may berequired either to make a microbiologicaldiagnosis or as a therapeutic procedure torelieve symptoms. An urgent ENT opinionshould be sought in patients who are systemi-cally unwell in order that sinus wash-outs orother surgical intervention may be carriedout. Sinusitis due to Pseudomonas aeruginosaoccurs in the context of advanced HIV dis-ease, and responds to specific antimicrobialtherapy, but the course of treatment mayneed to be prolonged in order to prevent therelapse of symptoms and signs or the devel-opment of chronic infection.

1 Rubin JS, Honigberg R Sinusitis in patients with theacquired immunodeficiency syndrome. Ear Nose ThroatJf 1990;69:460-3.

2 Zurlo JJ, Feuerstein IM, Lebovics R, Lane HC. Sinusitisin HIV-1 infection. Amj?,Med 1992;93:157-62.

3 Godofsky EW, Zinreich J, Armstrong M, Leslie JM,Weikel CS. Sinusitis in HIV-infected patients: a clinicaland radiographic review. AmJ Med 1992;93:163-70.

4 Poole MD, Postma D, Cohen MS. Pyogenic otorhinologicinfections in acquired immune deficiency syndrome.Arch Otolaryngol 1984;110:130-1.

5 Schrager LK. Bacterial infections in AIDS patients. AIDS1988;2 (Suppl 1):S183-9.

6 Lamprecht J, Wiedbrauk C. Sinusitis und andere typischeErkrankungun im HNO-Bereich im Rahmen des erwor-benen Immundefekt-Syndroms (AIDS). HNO 1988;36:489-92.

7 Meiteles LZ, Lucente FE. Sinus and nasal manifestationsof the acquired immunodeficiency syndrome. Ear NoseThroatJ 1990;69:454-9.

8 Pfieiderer AG, Drake Lee AB, Lowe D. Ultrasound of thesinuses; a worthwhile procedure? A comparison of ultra-sound and radiography in predicting the findings. ClinOtolaryngol 1984;9:335-41.

9 Hammond V. Diseases of the external ear. In: Kerr hA(ied. Scott-Brown's Otolaryngology. London: Butterworths,1987;3:168-9.

212 on F

ebruary 11, 2020 by guest. Protected by copyright.

http://sti.bmj.com

/G

enitourin Med: first published as 10.1136/sti.69.3.208 on 1 June 1993. D

ownloaded from