Journnl oj Infih~n (2000) 40, 184-204doi:lO. 105 3/jinf.l998.0480, available online at http://www.idealibrary.com on IDEhl@

CASE REPORTS

i

Efficacy of Long-term Antibiotic Suppressive Therapy in Provenor Suspected Infected Abdominal Aortic Grafts

Debajyoti Roy and David I. Grove*

Department of Clinical Microbiology and Infectious Diseases The Queen Elizabeth HospitaI, Woodville,South Australia

We have reviewed our experience of long-term antibiotic suppressive therapy in patients who underwent repair of anabdominal aortic aneurysm (AAA) and developed proven or strongly suspected infection of a graft. Five patients withabdominal aortic repair complicated by proven or suspected graft infections were treated with continuing antibioticsuppressive therapy based on microbiology culture reports. Two patients developed infection of an established graft,two patients had a graft inserted into an infected area and one patient was thought to be at high risk of developinginfection of a recently placed graft. All patients had severe co-existent medical problems and were considered too ill totolerate further definitive surgery. Response to therapy was monitored by absence of symptoms, fever, inflammatorymarkers and survival. All patients are alive with a median survival of 32 months, the longest having survived for6 years. In selected patients with abdominal aortic graft infections, indefinite antibiotic suppressive therapy may be anacceptable alternative to further surgery. 0 2000 The British Infection Society

IntroductionInfection of an arterial graft is one of the most feared complica-tions of vascular surgery. Patients with an infected aortic grafthave a mortality rate of 15-75X and a leg amputation rate ofg-75%-+ Complete excision of the infected aortic graft withextra-anatomical bypass has been the benchmark against whichother therapies have been compared. This procedure is associ-ated with serious morbidity in 75% of patients and more thanone-third of patients require amputation.5,h Other techniqueshave included graft excision with in situ replacement and directirrigation of the infected aortic graft with antibiotics.‘,iml 1

Cases

Case 1

Some patients are poor candidates for complete excision ofthe graft because of underlying severe medical problems such ascongestive heart failure, severe ischaemic heart disease or pul-monary insufficiency and would be expected to have a highoperative mortality. In some patients a graft may have to beplaced in an infected area resulting in inevitable infection of thegraft. Finally, another group of patients who might benefit fromprolonged antibiotic therapy are those in whom a diagnosis ofgraft infection is strongly suspected but not proven. This paperreports the outcome of long-term antibiotic suppressive therapyin five such patients.

*Address correspondence to: LX Grove.Accepted for publication 8 October 1998.

A 77-year-old male underwent emergency repair for a leakingabdominal aortic anearysm (AAA) with the insertion of a bifur-cated aortic graft. Five years later he developed ischaemic bowelwhich was resected. The thrombosed left femoral artery wasbypassed with a right to left femoral cross-over graft. Five yearslater he complained of severe claudication in the right leg forwhich he underwent femoro-popliteal bypass grafting from theright limb of the aorto-bifemoral graft. Four months later hedeveloped a left inguinal sinus. Fluid cultures from the sinusgrew mixed aerobic bacteria. Clinical examination suggestedand an ultrasound study confirmed the presence of a falseaneursym near the junction of the cross-over graft and the leftlimb of the aorto-bifemoral graft. The erythrocyte sedimenta-tion rate (ESR) was 107 mm/h. The false aneurysm wasrepaired and the femoro-femoral interposition and the left limbof the aorto-bifemoral graft were both resected but the aorticgraft and its right limb were left intact leaving some prostheticmaterial in an infected area. Staph$ococcus auricularis was iso-lated from cultures of the false aneurysm as well as from boththe cross-over graft and the left prosthetic limb. He was treatedwith intravenous vancomycin for 2 weeks and subsequentlybegun on oral flucloxacillin for life, initially in a dose of 500 mgfour times daily then reduced over 12 months to 500 mg twicedaily: 3 5 months later he is well.

Olh3-44j3/001020184 + 21 $35.00/O 0 2000 The British Infection Society

Case 2

A 77-year-old male developed severe abdominal pain andhypotension. A computed tomography (CT) scan of theabdomen was consistent with a large leaking infrarenal AAA.The aneurysm was repaired and the post-operative period wasuneventful. Ten years later he presented with fever and abdom-inal pain of 1 weeks duration. A CT scan of the abdomenrevealed gas locules adjacent to the graft (Fig. 1). Enterococcusfuecalis was isolated on blood cultures. The ESR and C-reactiveprotein (CRP) were 63 mm/h and 72 units/ml respectively.No surgical intervention was made in view of his age andpoor general condition. He was treated initially with van-comycin for 10 days intravenously then changed to long-term oral antibiotic therapy with amoxycillin/clavulanic acid500 mg/12 5 mg three t imes dai ly: 32 months later he remainswell.

Case 3 Figure 1. Computerized tomography scan of the abdomen showingtwo collections of gas between the aortic graft and the calcified native

A 56-year-old male presented to the emergency service with left aorta, together with evidence of small bowel obstruction.

hip pain for the preceding 2 weeks and a fever of 1 week’sduration. Staphylococcus aureus was isolated on blood cultures.Flucloxacillin therapy was initiated and the fever settled within48 h. He was discharged from hospital and advised to takeflucloxacillin for a further 2 weeks. Six months later he pre-sented to the emergency service with fever, hypotension andabdominal pain. A CT scan of the abdomen revealed a largepara-aortic collection surrounding a 7 cm infrarenal aorticfalse aneurysm (Fig. 2). He underwent emergency AAA repairwith a dacron graft at which time a diagnosis of a mycoticaneurysm with posterior erosion was made. Histopathologicalevaluation of the aort ic wall indicated a chronic inflammatoryreaction. The para-aortic fluid grew S. aureus on culture. Hewas treated with flucloxacillin together with gentamicinintravenously for 2 weeks. Since the risk of recurrence ofinfection in the graft was considered high he was treatedindefinitely with oral flucloxacillin in a dose reducing over 2years from 1 g three t imes daily to 0.5 g twice daily. Six yearslater he is well and his ESR has fallen from 150 to 23 mm in thefirst hour.

Figure 2. Computerized tomography scan of the abdomen showing aCase4 mass 7 cm in diameter thought to be a false aneursym arising from the

aorta, the calcified rim of which is disrupted posteriorly and on the left.

A 64-year-old male underwent an elective AAA repair and anSurrounding this is a large area of low density extending through the

aorto-bifemoral graft was placed. His other significant medicalretroperitoneal soft tissues and into both psoas muscles which is consis-

problems included severe ischaemic heart disease. He presentedtent with an inflammatory mass or pus.

8 years later with a large right-sided femoral aneurysm at thejunction with the graft . He had been febrile for the preceding 2weeks. The aneurysm was resected and a dacron graft inserted.Perigraft fluid collected at surgery grew Staphylococcus

Case 5

haemolyticus sensitive to amoxycillin. His ESR and CRP were 69 A 77-year-old male developed fever 10 days after an electivemm/h and 12 units/ml respectively, He was treated initially with AAA repair in which a straight Gore-tex graft was emplaced.amoxycil l in intravenously for 2 weeks then was placed on oral Invest igat ions included a white cel l count of 40.4x10y/1 wi thamoxycillin as life-long suppressive therapy life in a dose reduc- 95% polymorphonuclear cells, ESR of 71 mm/h and a CRP ofing over 18 months from 1 g three times daily to 0.5 g twice 156 uni ts /ml . Staphylococcus aureus and Bacteroides stercoris_ . .~daily: 30 months later he is well. were isolated from three of four sets of blood cultures.

Case Reports 185

186 Case Reports

St~phyhoccus aureus was also isolated from a resected mycoticaneurysm where a radial arterial l ine had been inserted. The tei-choic acid antibody level was greatly raised at > 1: 16 suggestiveof deep-seated infection, He was treated with intravenousflucloxacillin for 2 weeks. Since the risk of seeding of the graftwas considered to be high he was treated with oral flucloxacillinlong-term in a dose reducing over 12 months from 1 g threetimes daily to 0.5 g twice daily. Thirty-two months later heis well with an ESR and CRP of 7 mm/h and 3 units/mlrespectively.

Summary of Cases

Five patients with aortic graft infections have been treated withlong-term antibiotic suppressive therapy, The diagnosis of graftinfection was based on culture of perigraft fluid taken at surgery(three patients), or the presence of positive blood cultures in onepatient with CT scan evidence of graft infection and in onepatient with a recent graft emplacement who was considered athigh risk of graft infection. The first two patients developedinfection of established aortic grafts and were treated only withlong-term antibiotics. The next two patients had a graft insertedinto an infected area. The final patient had bacteraemia and wasconsidered at high risk of developing infection of a recentlyplaced aortic graft . The choice of antibiotic therapy was basedon microbiology results. Response to therapy was monitored byobservation of symptoms, temperature, inflammatory markersand survival. All patients are currently alive and are on long-term antibiotic therapy. The median follow-up period is 32months (range: 30-72 months) .

Discussion

Prosthetic graft infection is one of the most feared complicationsof vascular surgery. Foreign material such as vascular graftsenhances the pathogenici ty of known virulent organisms byreducing the inoculum necessary to ini t iate infection. I t alsoallows relatively avirulent organisms like Staphylococcus epider-midis to establish infection.

Aortic graft infection following aneurysm resection has beenreported to develop in around 2% of patients8 The usual ther-apy for vascular graft infection is surgical excision with in situplacement of a graft or extra-anatomical bypass. These proce-dures are associated with high morbidity and mortality. Forpatients with anticipated high surgical mortality, prolongedantibiotic therapy may be a better option than surgicalexcision.“,14 Similarly, antibiotic therapy is the only therapeuticmodality left when a graft has to be inserted or re-emplaced intoan infected site or when the graft is in a location that cannot beexcised.

Bouhoutsos et al.‘” reviewed 532 patients with syntheticarterial grafts and found 14 patients with infected grafts. Tenpatients had the graft removed and were given antibiotics. Fourpatients in whom infect ions developed within 3 weeks of sur-gery were treated with antibiotics alone and all recovered.However, no details were provided as to the method of diagnosis,infecting organism or the antibiotics used.

Identification of the organism involved is critical in the man-agement of these patients with suppressive antibiotics,Individual ant ibiot ic suscept ibi l i ty pat terns need to be detcr-mined so that the most appropriate antibiotics are chosen. Thereis little information on which antibiotics are the most appropri-ate to use. Some antibiot ics such as r i fampin and cl indamycinare concentrated in leucocytes and may be carried to graft inter-slices where bacteria often reside.‘” When infection is recog-nized at operation, soaking the graft in an antibiotic such asrifampicin may be of benefit as this procedure has been shownin prel iminary studies to be of benefi t in experimental animalsand in humans. Ij,lh

All of our patients were treated with beta-lactam antibioticswith good responses. Once long-term therapy has been initiated,it is hard to know when to stop as overt infection may recur oncessation. Clearly, this may have devastating consequences in anaortic graft and most of our patients are loath to take the risk ofs topping a l l t reatment .

In conclusion, our experience of five patients treated withappropriate long-term antibiot ics suggests that this approachshould be strongly considered when the replacement of aninfected graft is associated with an unacceptably high risk ofmortality or when a graft has to be inserted into an infectedsite.

References

1 Hoffert PWK, Gensler S. Haimovici H. Infection complicating arterialgrafts. Personal experience with 12 cases and review of literature.Arch Surg 1965; 90: 427-431.

2 Conn JH. Hardy JD. Chavez CM. Infected arterial grafts:experience in 22 cases with emphasis on unusual bacteria and tech-niques. Ann Surg 1980; 171: 704-709.

3 Fry WJ, Lindenauer SM. Infections complicating the use ofplastic arterial implants. Arch Surg 1967: 94: 600-609.

4 Goldstone J. Moore WS. Infections in vascular prostheses:clinical manifestations and surgical management. Am J Surg 19 74:128: 225-231.

5 Perler BA, Van der Kolk CA, Dufresene CR, Williams GM. Caninfected prosthetic grafts be salvaged with rotational muscle flap?Surgery 1991: 110: 30-34.

6 Quinones-Baldrich WJ. Hernandez JJ, Moore WS. Long-term resultsfollowing surgical management of aortic graft infections. Arch Surg1991: 126: 507-11.

i Plate G. Hollier LA, O’Brien P Pairuiero PC, Cherry KJ. Kazmier FJ.Recurrent aneurysms and late vascular complications following repairof abdominal aortic aneurysms. Arch Surg 1985: 120: 590-594.

8 Lawrence PF. Management of infected aortic grafts. Surg Clin NorthAm 1995; 75: 783-797.

9 Crawford ES. Manning LG, Kelly TF. Redo surgery after operations foraneurysm and occlusion of the abdominal aorta. Surger;/ 19 77: 8 1:41-52.

10 Strachan CL, Newsom SB. AshtonTR. The clinical use of an antibiotic-bonded graft. EurJVasc Surg 1991: 5: 627-632.

11 Calligaro KD, Veith FJ, Schwartz ML rt al. Are gram negative bacteriaa contraindication to selective preservation of infected arterialgrafts? JVusc Surg 1992; 16: 337-346.

12 McCann RI,. Schwartz LB, Georgiade GS. Management of abdominalaortic graft complications. Ann Surg 1993: 217: 729-734.

13 Morris GE, Friend PJ, Vassallo DJ. Antibiotic irrigation and conserva-tive surgery for major aortic graft infection. I Vast Surg 1994; 20:X8-95.

Case Reports 187

14 Bouhoutsos J, Chavatza D. Martin P, Morris T. Infected synthetic arte- 16 Gupta AK, Bandyk DF, Johnson BL. In situ repair of mycotic abdomi-rialgrafts. BrJSurg 1974; 61: 108-111. nal aortic aneurysms with rifampicin-bonded gelatin-impregnated

15 Vicaretti M, Hawthorne WJ. Ao PY, Fletcher JP. An increased concen- Dacron grafts: a preliminary case report. 1 Vusc Surg 1996: 24:tration of rifampicin bonded to gelatin-sealed Dacron reduces the 472-476.incidence of subsequent graft infections following a staphylococcalchallenge. Cardiovmc Surg 1998; 6: 268-273.

doi:10.1053/jinf,1999.0572. available online at http://www.idealibrary.com on IDEhl@

Superior Vena Cava Syndrome During the Treatment ofPulmonary Tuberculosis in an HIV-1 Infected Patient

C. Minguez*l, B. Rocal, C. Gonztilez-Mifio2 and E. Simhl

‘Infectious Diseases Unit, Department of Medicine, Hospital General de Castelldn, 2Primary Care Centre “El Barranquet”,Castelldn, Spain

Tuberculosis is nowadays an uncommon cause of superior vena cava syndrome (SVCS). We report the case of an HIV-infected patient who presented with respiratory symptoms accompanied by cervical and mediastinal lymphadeno-pathy. Sputum examination showed acid-fast bacilli, and treatment was instituted with isoniazid, rifampicin,pyrazinamide and ethambutol. A few days later SVCS developed, presumably as a consequence of inflammatory lym-phadenitis. With corticosteroids, all symptoms disappeared. To our knowledge, no cases of SVCS provoked by this kindof paradoxical reaction have been described previously, 0 2000 The British Infection Society

Introduction

Obstruct ion of venous flow through the superior vena cava pro-vokes distension of upper body veins, facial swelling, dyspnoea,and cyanosis, known as superior vena cava syndrome (SVCS).Before the arrival of antibiotics, infectious diseases such assyphilis or tuberculosis produced most cases of SVCS,‘,2 but nowcancer is the main cause of this syndrome.3 In recent years, theAIDS epidemic has led to an increased incidence of tuberculosisand its complications.4-h We report the case of a patient withAIDS who developed SVCS after starting treatment for tubercu-losis, presumably as a consequence of a paradoxical reaction totherapy. In a review of the li terature (Medline 1980-1999), wehave found anecdotal reports of SVCS related to tuberculosis.“’but none in HIV-l infected patients.

*Address correspondence to: Dr Carlos Minguez, Department ofMedicine (Infectious Diseases Unit), Hospital Genera1 de Castelk%. AvBenicassim s/n, 12004 Castelldn, Spain.Accepted for publication: 25 June 1999.

Case ReportA 34-year-old woman with a history of intravenous drug use,and HIV-1 and hepatitis C virus infection, was diagnosed of pul-monary tuberculosis in September 1996. Isoniazid, r i fampicin,pyraxinamide and ethambutol were instituted, and initially agood response was observed. A few months later, the patientstopped therapy on her own, and in March 199 7 she was read-mitted because of fever, weight loss and breathlessness. Onexamination. prominent tender lymphadenopathies were notedin her neck and axillary regions. Blood analysis showed ahaemoglobin of 8.8 g/d1 with normal mean cell volume andCD4+ lymphocyte count 87 Z 106/l. Chest radiographsdisclosed a diffuse interstitial infiltrate and a left upper lobeopacity. An acid-fast stain of the sputum was positive, andisoniazid, rifampicin, pyrazinamide and ethambutol werereintroduced. No antiretroviral drugs were prescribed. Twentydays later the woman experienced headache, swelling of the faceand neck, distension of the neck veins, and cyanosis. A com-puted tomography scan (CT) of the chest d isc losed mult iplemediastinal and axil lary lymphadenopathies (Fig. la) , a part ialcollapse of the left upper lobe, and bilateral pleural effusion. Twolymph node biopsies showed granuloma formation and necrosis,

0 2000 The British Infection Society