- Email: [email protected]
Infectious complications of old nonfunctioning arteriovenous grafts in renal transplant recipients: A case series
Infectious Complications of Old Nonfunctioning Arteriovenous Grafts in Renal Transplant Recipients: A Case Series George M. Nassar, MD, and Juan Carlos Ayus, MD ● Occult infection of old nonfunctioning arteriovenous grafts (AVGs) is frequent among hemodialysis patients. It is a risk factor for bacteremia and serious AVG-related infection. Immunocompromised patients are at increased risk of dissemination of occult AVG infection. We present a series of five renal transplant recipients who developed acute life-threatening infections that originated in their old nonfunctioning AVGs. Their presenting symptoms were noticeably varied. In two patients, infection of the AVG was characterized by local physical signs of infection around the AVG. In three patients, no physical signs of AVG infection were detected by physical examination. Among these, two presented with bacteremia, and one presented with failure to thrive. Detection of AVG infection in the absence of local signs of infection requires a high index of suspicion. Surgical resection and antimicrobial treatment led to a complete cure in four of these patients. One patient developed recurrent bacterial endocarditis and died. Old nonfunctioning AVGs are potential sources of serious infection in renal transplant recipients. Renal transplant recipients with old nonfunctioning AVGs who present with unexplained bacteremia, fever of unknown origin, or failure to thrive should be investigated for occult AVG infection. Screening for occult infection of the old nonfunctioning AVG may be considered before kidney transplantation, especially if the candidate gives a history of previous bacteremia or fever of unknown origin. © 2002 by the National Kidney Foundation, Inc. INDEX WORDS: Kidney transplantation; arteriovenous graft (AVG); infection.
I
MMUNOSUPPRESSED kidney transplant recipients are at increased risk of infections resulting from myriad causes.1,2 Before kidney transplantation, prospective recipients undergo rigorous evaluation to detect and treat potential sources of infection.3 Clinically silent urinary tract infection, symptomatic cholelithiasis, periodontal abscess, positive skin test for tuberculosis, and positive serologies for hepatitis viruses are a few examples of potential sources of infection that may become life-threatening with immunosuppression. Every effort is made to treat or eliminate such sources of infection before transplantation. The prosthetic arteriovenous graft (AVG), composed of polytetrafluoroethylene, which many patients on hemodialysis have in their extremities, is not investigated routinely for the possibility of occult infection before transplantation.3 From the Nephrology, Dialysis and Transplantation Associates and Renal Research Inc; and Department of Medicine, Baylor College of Medicine, Houston; and Department of Medicine, University of Texas Health Science Center at San Antonio, San Antonio, TX. Received March 22, 2002; accepted in revised form June 12, 2002. Address reprint requests to Juan Carlos Ayus, MD, 1967 Haddon Street, Houston, TX 77019. E-mail: [email protected] © 2002 by the National Kidney Foundation, Inc. 0272-6386/02/4004-0021$35.00/0 doi:10.1053/ajkd.2002.35696 832
There is lack of a high index of suspicion of AVG infection post-transplantation in patients presenting with bacteremia or fever of unknown origin.1,2 To highlight the threat posed by occult AVG infection, we present a series of five renal transplant recipients who developed an acute febrile illness resulting from infection of their old nonfunctioning AVGs (Table 1). CASE REPORTS
Case 1 A 64-year-old man with end-stage renal disease (ESRD) secondary to IgA nephropathy was begun on hemodialysis in 1992. His initial vascular access was a synthetic AVG placed in the left forearm. This AVG was complicated by several episodes of thrombosis and one episode of local infection managed by antimicrobial therapy. Three months after the episode of infection, this AVG was abandoned, and a second AVG was placed in the right forearm. In 1995, the patient underwent a cadaveric renal transplant and was placed on cyclosporine, prednisone, and azathioprine. Four months later, he developed fever, tenderness, and erythema surrounding the nonfunctioning left forearm AVG. Surgical resection of this AVG revealed purulent material inside and surrounding the graft that grew Staphylococcus aureus. After surgical excision and antimicrobial therapy, the patient recovered completely.
Case 2 A 46-year-old white woman developed ESRD secondary to diabetic nephropathy and was placed on hemodialysis via an AVG. After 2 years of hemodialysis punctuated by recurrent AVG thrombotic complications, she received a kidney/ pancreas transplant in December 1999 and was placed on immunosuppression with tacrolimus, mycophenolate, and
American Journal of Kidney Diseases, Vol 40, No 4 (October), 2002: pp 832-836
AVG INFECTIONS IN RENAL TRANSPLANT RECIPIENTS Table 1.
Case No.
Clinical Presentations, Mode of Diagnosis, Operative Findings, and Outcome of five Renal Transplant Recipients With Infection of Old Nonfunctioning Arteriovenous Grafts Clinical Presentation
Method of Diagnosis
1
Fever, erythema around AVG
Physical examination
2
Fever, erythema around AVG
Physical examination
3
Bacteremia
Indium scan
4
Bacteremia Recurrent endocarditis Failure to thrive Weight loss
Indium scan
5
833
Indium scan
prednisone. In February 2000, she suffered from an episode of acute transplant rejection, which was managed by antithymoglobulin, and required the addition of rapamycin to her immunosuppression regimen. In March 2000, her AVG clotted spontaneously. In May 2000, the patient started having recurrent episodes of fever of unknown origin managed by empirical antibiotic coverage. In October 2000, blood cultures grew Staphylococcus epidermidis, and she was given 4 weeks of intravenous vancomycin. In December 2000, she presented to the emergency department with 3-day history of high-grade fever and erythema around the clotted AVG in the right forearm. She was started on intravenous empirical antimicrobial therapy with vancomycin and ceftazidime pending cultures. Her AVG was surgically removed 24 hours after admission. Purulent material inside and surrounding the AVG was seen during surgery, but bacterial pathogens failed to grow on culture.
Case 3 A 37-year-old black Nigerian woman developed ESRD secondary to hypertensive nephrosclerosis. After 7 years of hemodialysis, she received a cadaveric kidney transplant and was started on cyclosporine, prednisone, and azathioprine. One year after transplantation, her right forearm AVG spontaneously clotted. Four months later, she was admitted to the hospital with fever and malaise. Physical examination failed to reveal any evidence of localized infection, and blood cultures were positive for S aureus. As part of a diagnostic workup, an indium-labeled white blood cell scan was done. The scan revealed significant diffuse uptake in the area of the old clotted AVG. The patient underwent surgical excision of the graft, which revealed purulent material inside the AVG that grew S arueus. The patient was treated with vancomycin and recovered without complications.
Case 4 A 35-year-old woman of Spanish descent developed ESRD secondary to glomerulonephritis. After 5 years of hemodialysis via an AVG, she received a cadaveric kidney transplant.
Operative Findings
Purulent material inside and outside the AVG Purulent material inside and outside the AVG Purulent material inside the AVG Purulent material inside the AVG Purulent material inside the AVG
Bacterial Culture
Outcome
S aureus
Cured
No growth
Cured
S aureus
Cured
S aureus
Death
S aureus
Cured
Her immunosuppression protocol consisted of cyclosporine, prednisone, and azathioprine. Her AVG spontaneously clotted 18 months after transplantation. Her renal allograft functioned well until 30 months after transplantation, when creatinine increased suddenly from 1.2 mg/dL (106 mol/L) to 1.8 mg/dL (159 mol/L). Soon afterward, the patient became febrile, and blood cultures grew S aureus. She was treated with antibiotics for 3 weeks, during which she became afebrile and showed clinical improvement. One month after discontinuation of antibiotic therapy, she had recurrence of bacteremia with S aureus. At this time, she developed bacterial endocarditis with severe aortic regurgitation. The patient was started on intravenous antimicrobial therapy and underwent aortic valve replacement when her blood cultures became negative for bacterial growth. Two weeks later, she developed recurrence of S aureus bacteremia. Intravenous antimicrobial therapy was resumed, and an indium scan was done. The indium scan showed intense diffuse uptake of tracer in the region of the old clotted AVG. Surgical excision of the AVG was performed and revealed purulent material inside the AVG, which grew S aureus. The patient was continued on intravenous antibiotics, but she soon died as a result of multiple complications.
Case 5 A 47-year-old white man with ESRD secondary to diabetic nephropathy was on hemodialysis for 7 years when he received a kidney transplant from his brother in September 2001. At the time of transplantation, he had an old clotted AVG in the left forearm and was receiving hemodialysis via a central venous catheter. He was placed on rapamycin, cyclosporine, and prednisone. Despite an uneventful postoperative course and successful kidney transplantation, he continued to suffer from constitutional symptoms and failure to thrive. At 1 month after transplantation, his weight had decreased from 64.4 kg postoperatively to 58.2 kg, and he had persistent relative hypoalbuminemia (serum albumin, 3.4 g/dL). An indium scan was performed as part of the diagnostic workup. It showed increased tracer uptake in the
834
NASSAR AND AYUS
area corresponding to the old clotted AVG in the left forearm, raising the suspicion of AVG-related infection. The AVG subsequently was resected. Intraoperatively, purulent material was found inside the AVG that grew S aureus. Four weeks later, the patient’s weight increased to 65.3 kg, and serum albumin rose to 3.9 g/dL (39 g/L).
DISCUSSION
The natural history of the prosthetic AVG is fraught with occlusive and thrombotic events, which, among other factors, lead to loss of graft function.4 It is common practice to keep these nonfunctioning AVGs in the extremities of hemodialysis patients. In a survey of 743 hemodialysis patients, we found 35% to have at least one old clotted AVG in their extremities.5 Renal transplant recipients who had been on hemodialysis frequently have an old clotted AVG before transplantation. In addition, a high percentage of functional AVGs thrombose after transplantation. Although old nonfunctioning AVGs are considered by many health care providers to be innocuous, we have found them to be common sites of clinically silent infection.6-9 Such an infection frequently is undetectable by physical examination but is likely to be apparent on indiumlabeled white blood cell scan.6-8 Occult AVG infection may be caused by direct needle puncture of the AVG or by hematogenous spread during periods of previous bacteremia.8 Acute dissemination of infection from the AVG into the bloodstream has devastating consequences, such as bacterial endocarditis, septic arthritis, and epidural abscess.9 An immunocompromised host, such as a renal transplant recipient, is at increased risk of infectious complications from occult AVG infection. The patients presented demonstrate several clinical vignettes associated with infection of old nonfunctioning AVGs. In the first two patients, infection of the AVG was detectable by physical examination, allowing prompt surgical resection of infected graft material. In both cases, purulent material was found in association with the AVG, but in patient 2, it showed no bacterial growth, most likely because antimicrobial agents had been started 24 hours before resection of the AVG. Patient 2 had had bacteremia with S epidermidis about 1 month before presenting with acute AVG infection. This fact may be subject to several possible interpretations, as follows: (1)
Occult infection of the AVG was the cause of the previous bacteremia. Antimicrobial treatment cleared the bacteremia but failed to eradicate the AVG infection, which subsequently manifested as local acute AVG infection. (2) The old clotted AVG was seeded during the previous bacteremia and went on to develop locally detectable infection. Cultures of the purulent material failed to show bacterial pathogens, limiting our ability to confirm a relationship between the previous episode of bacteremia and the acute local AVG infection. (3) The two events are unrelated. In our experience, all these clinical scenarios are possible.8 The clinical scenarios of the third and fourth patients show that acute bacteremia can be the only presenting sign of AVG infection. In these cases, the absence of local signs of AVG infection may lead to delay in proper diagnosis. A high index of suspicion is required to pursue the possibility of occult AVG infection. In both patients, the indium scan was useful in detecting occult AVG infection, which subsequently was confirmed by surgical excision and culture of the AVG. In patient 3, a high index of suspicion of AVG infection led to an early diagnosis and prompt surgical resection of an AVG harboring occult infection. In patient 4, the lack of early suspicion of AVG infection markedly delayed detection and resection of infected AVG material. The latter continued to be a source of recurrent bacteremia resistant to eradication by antimicrobials. A totally curable infection escaped detection and evolved to serious metastatic complications and ultimately death of the patient. The case of the fifth patient shows a completely different clinical presentation—the development of a chronic inflammatory state–like illness resulting from occult AVG infection. In this patient, failure to thrive, weight loss, and hypoalbuminemia raised the suspicion of occult AVG infection. Occult AVG infection as a cause of chronic inflammatory state is supported by data that show high prevalence of erythropoietinresistant anemia, hypoalbuminemia, and elevated C-reactive protein levels among hemodialysis patients with occult infection of old nonfunctioning AVGs.10 In these patients, resection of AVGs harboring occult infection led to resolution of laboratory and clinical parameters of chronic inflammatory state. Patient 5 illus-
AVG INFECTIONS IN RENAL TRANSPLANT RECIPIENTS
trates this phenomenon, and a high index of suspicion of occult AVG infection facilitated early diagnosis and proper treatment. This led to clinical improvement, weight gain, and normalization of serum albumin. Kidney transplant patients are immunosuppressed and prone to a multitude of infectious complications.1,2 Among these, bacteremia continues to be a major source of morbidity and mortality. This fact is shown best by a report by Abbott et al,11 who reviewed the U.S. Renal Data System database between 1994 and 1997. They found 1,447 episodes of septicemia among 33,479 renal transplant patients. This translated to 1,896 hospitalizations and amounted to a gigantic adjusted incidence ratio of hospitalizations for septicemia of 41.5 compared with the general population. An infected vascular device was present in 10.6% of hospitalizations for septicemia. This dismal rate of hospitalizations for septicemia among renal transplant patients should be fought with every effort to identify and eliminate its causes. The five patients presented here highlight that occult infection of old clotted AVGs is a potential source of bacteremia and life-threatening illness in renal transplant patients. It is curable if identified without delay and treated without hesitation. The approach we advise is as follows: In patients with obvious signs of local AVG infection, prompt surgical resection and antimicrobial therapy are required. A high index of suspicion for occult AVG infection should be exercised in all renal transplant patients with old nonfunctioning AVGs and presenting with unexplained bacteremia, fever of unknown origin, or failure to thrive. In such cases, we advocate performing an indium-labeled white blood cell scan, which we previously showed to have a high degree of sensitivity and specificity for detecting occult AVG infection.6-8 Surgical resection of all AVGs that show local uptake by indium scan is advised.6-8 Infectious complications of the old nonfunctioning AVG in renal transplant patients have not been the subject of a formal study. To our knowledge, our patients represent the first report on this subject in renal transplant patients to date. We are aware that this report does not provide data on how the AVGs became infected in the first place or on the incidence and magnitude of
835
infections caused by occult AVG infections in renal transplant patients. Nevertheless, this report does convey the message to the renal transplant community about the potential threats of occult AVG infection in everyday clinical practice. Finally, comment in regard to renal transplant candidates who have asymptomatic old nonfunctioning AVGs in their extremities is warranted. Should these AVGs be left alone? Should they undergo evaluation to exclude occult AVG infection before transplantation? Should they undergo routine resection? Definite answers cannot be offered at present without data addressing this particular issue, and different arguments can be made. We argue against the current prevailing practice of leaving these AVGs in place without any evaluation. Screening these AVGs with indium scan is indicated based on data that show a high prevalence (20% to 70%) of occult infection of old nonfunctioning AVGs among asymptomatic hemodialysis patients.5 The risk of occult AVG infection may be higher if the patient gives a history of previous bacteremia or fever of unknown origin. As stated earlier, the indium scan is a reliable tool for detecting occult AVG infection. If the indium scan shows positive uptake around an old nonfunctioning AVG, the AVG should be removed before kidney transplantation. The value of indium scan in patients undergoing evaluation for renal transplantation also extends to evaluation of their functioning AVGs. Functional AVGs also are prone to occult infection, and if they show increased uptake by indium scan, these functional accesses should be removed surgically before transplantation. ACKNOWLEDGMENT We thank Dr Mario Rubin and Dr Roberto Alcazar for providing data on two patients (cases 1 and 5).
REFERENCES 1. Rubin RH: Infectious disease complications of renal transplantation. Kidney Int 44:221-236, 1993 2. Fishman JA, Rubin RH: Infection in organ transplant recipients. N Engl J Med 338:1741-1751, 1998 3. McKay DB, Milford EL, Sayegh MH: Clinical aspects of renal transplantation, in Brenner BM and Rector FC (eds): The Kidney (ed 5). Philadelphia, PA, Saunders, 1995, pp 2602-2652
836
4. Feldman HI, Kobrin S, Wasserstein A: HD vascular access morbidity. J Am Soc Nephrol 7:523-535, 1996 5. Nassar GM, Quinibi W, Aboujaoude W, Ayus JC: Prevalence and complications of old clotted arteriovenous grafts in HD patients. J Am Soc Nephrol 12:298A, 2001 (abstract) 6. Ayus JC, Sheikh-Hamad D: Silent infection in clotted HD access grafts. J Am Soc Nephrol 9:1314-1317, 1998 7. Sheikh-Hamad D, Ayus JC: The patient with a clotted PTFE graft developing fever. Nephrol Dial Transplant 13: 2392-2393, 1998
NASSAR AND AYUS
8. Nassar GM, Ayus JC: Clotted arteriovenous grafts: A silent source of infection. Semin Dial 13:1-3, 2000 9. Nassar GM, Ayus JC: Infectious complications of the HD access. Kidney Int 60:1-13, 2001 10. Nassar GM, Fishbane S, Ayus JC: Occult infection of old nonfunctioning arteriovenous grafts: A novel cause of erythropoietin resistance and chronic inflammation in hemodialysis patients. Kidney Int 61:S49-S54, 2002 (suppl 80) 11. Abbott KC, Oliver JD 3rd, Hypolite I, et al: Hospitalizations for bacterial septicemia after renal transplantation in the United States. Am J Nephrol 21:120-127, 2001
I
MMUNOSUPPRESSED kidney transplant recipients are at increased risk of infections resulting from myriad causes.1,2 Before kidney transplantation, prospective recipients undergo rigorous evaluation to detect and treat potential sources of infection.3 Clinically silent urinary tract infection, symptomatic cholelithiasis, periodontal abscess, positive skin test for tuberculosis, and positive serologies for hepatitis viruses are a few examples of potential sources of infection that may become life-threatening with immunosuppression. Every effort is made to treat or eliminate such sources of infection before transplantation. The prosthetic arteriovenous graft (AVG), composed of polytetrafluoroethylene, which many patients on hemodialysis have in their extremities, is not investigated routinely for the possibility of occult infection before transplantation.3 From the Nephrology, Dialysis and Transplantation Associates and Renal Research Inc; and Department of Medicine, Baylor College of Medicine, Houston; and Department of Medicine, University of Texas Health Science Center at San Antonio, San Antonio, TX. Received March 22, 2002; accepted in revised form June 12, 2002. Address reprint requests to Juan Carlos Ayus, MD, 1967 Haddon Street, Houston, TX 77019. E-mail: [email protected] © 2002 by the National Kidney Foundation, Inc. 0272-6386/02/4004-0021$35.00/0 doi:10.1053/ajkd.2002.35696 832
There is lack of a high index of suspicion of AVG infection post-transplantation in patients presenting with bacteremia or fever of unknown origin.1,2 To highlight the threat posed by occult AVG infection, we present a series of five renal transplant recipients who developed an acute febrile illness resulting from infection of their old nonfunctioning AVGs (Table 1). CASE REPORTS
Case 1 A 64-year-old man with end-stage renal disease (ESRD) secondary to IgA nephropathy was begun on hemodialysis in 1992. His initial vascular access was a synthetic AVG placed in the left forearm. This AVG was complicated by several episodes of thrombosis and one episode of local infection managed by antimicrobial therapy. Three months after the episode of infection, this AVG was abandoned, and a second AVG was placed in the right forearm. In 1995, the patient underwent a cadaveric renal transplant and was placed on cyclosporine, prednisone, and azathioprine. Four months later, he developed fever, tenderness, and erythema surrounding the nonfunctioning left forearm AVG. Surgical resection of this AVG revealed purulent material inside and surrounding the graft that grew Staphylococcus aureus. After surgical excision and antimicrobial therapy, the patient recovered completely.
Case 2 A 46-year-old white woman developed ESRD secondary to diabetic nephropathy and was placed on hemodialysis via an AVG. After 2 years of hemodialysis punctuated by recurrent AVG thrombotic complications, she received a kidney/ pancreas transplant in December 1999 and was placed on immunosuppression with tacrolimus, mycophenolate, and
American Journal of Kidney Diseases, Vol 40, No 4 (October), 2002: pp 832-836
AVG INFECTIONS IN RENAL TRANSPLANT RECIPIENTS Table 1.
Case No.
Clinical Presentations, Mode of Diagnosis, Operative Findings, and Outcome of five Renal Transplant Recipients With Infection of Old Nonfunctioning Arteriovenous Grafts Clinical Presentation
Method of Diagnosis
1
Fever, erythema around AVG
Physical examination
2
Fever, erythema around AVG
Physical examination
3
Bacteremia
Indium scan
4
Bacteremia Recurrent endocarditis Failure to thrive Weight loss
Indium scan
5
833
Indium scan
prednisone. In February 2000, she suffered from an episode of acute transplant rejection, which was managed by antithymoglobulin, and required the addition of rapamycin to her immunosuppression regimen. In March 2000, her AVG clotted spontaneously. In May 2000, the patient started having recurrent episodes of fever of unknown origin managed by empirical antibiotic coverage. In October 2000, blood cultures grew Staphylococcus epidermidis, and she was given 4 weeks of intravenous vancomycin. In December 2000, she presented to the emergency department with 3-day history of high-grade fever and erythema around the clotted AVG in the right forearm. She was started on intravenous empirical antimicrobial therapy with vancomycin and ceftazidime pending cultures. Her AVG was surgically removed 24 hours after admission. Purulent material inside and surrounding the AVG was seen during surgery, but bacterial pathogens failed to grow on culture.
Case 3 A 37-year-old black Nigerian woman developed ESRD secondary to hypertensive nephrosclerosis. After 7 years of hemodialysis, she received a cadaveric kidney transplant and was started on cyclosporine, prednisone, and azathioprine. One year after transplantation, her right forearm AVG spontaneously clotted. Four months later, she was admitted to the hospital with fever and malaise. Physical examination failed to reveal any evidence of localized infection, and blood cultures were positive for S aureus. As part of a diagnostic workup, an indium-labeled white blood cell scan was done. The scan revealed significant diffuse uptake in the area of the old clotted AVG. The patient underwent surgical excision of the graft, which revealed purulent material inside the AVG that grew S arueus. The patient was treated with vancomycin and recovered without complications.
Case 4 A 35-year-old woman of Spanish descent developed ESRD secondary to glomerulonephritis. After 5 years of hemodialysis via an AVG, she received a cadaveric kidney transplant.
Operative Findings
Purulent material inside and outside the AVG Purulent material inside and outside the AVG Purulent material inside the AVG Purulent material inside the AVG Purulent material inside the AVG
Bacterial Culture
Outcome
S aureus
Cured
No growth
Cured
S aureus
Cured
S aureus
Death
S aureus
Cured
Her immunosuppression protocol consisted of cyclosporine, prednisone, and azathioprine. Her AVG spontaneously clotted 18 months after transplantation. Her renal allograft functioned well until 30 months after transplantation, when creatinine increased suddenly from 1.2 mg/dL (106 mol/L) to 1.8 mg/dL (159 mol/L). Soon afterward, the patient became febrile, and blood cultures grew S aureus. She was treated with antibiotics for 3 weeks, during which she became afebrile and showed clinical improvement. One month after discontinuation of antibiotic therapy, she had recurrence of bacteremia with S aureus. At this time, she developed bacterial endocarditis with severe aortic regurgitation. The patient was started on intravenous antimicrobial therapy and underwent aortic valve replacement when her blood cultures became negative for bacterial growth. Two weeks later, she developed recurrence of S aureus bacteremia. Intravenous antimicrobial therapy was resumed, and an indium scan was done. The indium scan showed intense diffuse uptake of tracer in the region of the old clotted AVG. Surgical excision of the AVG was performed and revealed purulent material inside the AVG, which grew S aureus. The patient was continued on intravenous antibiotics, but she soon died as a result of multiple complications.
Case 5 A 47-year-old white man with ESRD secondary to diabetic nephropathy was on hemodialysis for 7 years when he received a kidney transplant from his brother in September 2001. At the time of transplantation, he had an old clotted AVG in the left forearm and was receiving hemodialysis via a central venous catheter. He was placed on rapamycin, cyclosporine, and prednisone. Despite an uneventful postoperative course and successful kidney transplantation, he continued to suffer from constitutional symptoms and failure to thrive. At 1 month after transplantation, his weight had decreased from 64.4 kg postoperatively to 58.2 kg, and he had persistent relative hypoalbuminemia (serum albumin, 3.4 g/dL). An indium scan was performed as part of the diagnostic workup. It showed increased tracer uptake in the
834
NASSAR AND AYUS
area corresponding to the old clotted AVG in the left forearm, raising the suspicion of AVG-related infection. The AVG subsequently was resected. Intraoperatively, purulent material was found inside the AVG that grew S aureus. Four weeks later, the patient’s weight increased to 65.3 kg, and serum albumin rose to 3.9 g/dL (39 g/L).
DISCUSSION
The natural history of the prosthetic AVG is fraught with occlusive and thrombotic events, which, among other factors, lead to loss of graft function.4 It is common practice to keep these nonfunctioning AVGs in the extremities of hemodialysis patients. In a survey of 743 hemodialysis patients, we found 35% to have at least one old clotted AVG in their extremities.5 Renal transplant recipients who had been on hemodialysis frequently have an old clotted AVG before transplantation. In addition, a high percentage of functional AVGs thrombose after transplantation. Although old nonfunctioning AVGs are considered by many health care providers to be innocuous, we have found them to be common sites of clinically silent infection.6-9 Such an infection frequently is undetectable by physical examination but is likely to be apparent on indiumlabeled white blood cell scan.6-8 Occult AVG infection may be caused by direct needle puncture of the AVG or by hematogenous spread during periods of previous bacteremia.8 Acute dissemination of infection from the AVG into the bloodstream has devastating consequences, such as bacterial endocarditis, septic arthritis, and epidural abscess.9 An immunocompromised host, such as a renal transplant recipient, is at increased risk of infectious complications from occult AVG infection. The patients presented demonstrate several clinical vignettes associated with infection of old nonfunctioning AVGs. In the first two patients, infection of the AVG was detectable by physical examination, allowing prompt surgical resection of infected graft material. In both cases, purulent material was found in association with the AVG, but in patient 2, it showed no bacterial growth, most likely because antimicrobial agents had been started 24 hours before resection of the AVG. Patient 2 had had bacteremia with S epidermidis about 1 month before presenting with acute AVG infection. This fact may be subject to several possible interpretations, as follows: (1)
Occult infection of the AVG was the cause of the previous bacteremia. Antimicrobial treatment cleared the bacteremia but failed to eradicate the AVG infection, which subsequently manifested as local acute AVG infection. (2) The old clotted AVG was seeded during the previous bacteremia and went on to develop locally detectable infection. Cultures of the purulent material failed to show bacterial pathogens, limiting our ability to confirm a relationship between the previous episode of bacteremia and the acute local AVG infection. (3) The two events are unrelated. In our experience, all these clinical scenarios are possible.8 The clinical scenarios of the third and fourth patients show that acute bacteremia can be the only presenting sign of AVG infection. In these cases, the absence of local signs of AVG infection may lead to delay in proper diagnosis. A high index of suspicion is required to pursue the possibility of occult AVG infection. In both patients, the indium scan was useful in detecting occult AVG infection, which subsequently was confirmed by surgical excision and culture of the AVG. In patient 3, a high index of suspicion of AVG infection led to an early diagnosis and prompt surgical resection of an AVG harboring occult infection. In patient 4, the lack of early suspicion of AVG infection markedly delayed detection and resection of infected AVG material. The latter continued to be a source of recurrent bacteremia resistant to eradication by antimicrobials. A totally curable infection escaped detection and evolved to serious metastatic complications and ultimately death of the patient. The case of the fifth patient shows a completely different clinical presentation—the development of a chronic inflammatory state–like illness resulting from occult AVG infection. In this patient, failure to thrive, weight loss, and hypoalbuminemia raised the suspicion of occult AVG infection. Occult AVG infection as a cause of chronic inflammatory state is supported by data that show high prevalence of erythropoietinresistant anemia, hypoalbuminemia, and elevated C-reactive protein levels among hemodialysis patients with occult infection of old nonfunctioning AVGs.10 In these patients, resection of AVGs harboring occult infection led to resolution of laboratory and clinical parameters of chronic inflammatory state. Patient 5 illus-
AVG INFECTIONS IN RENAL TRANSPLANT RECIPIENTS
trates this phenomenon, and a high index of suspicion of occult AVG infection facilitated early diagnosis and proper treatment. This led to clinical improvement, weight gain, and normalization of serum albumin. Kidney transplant patients are immunosuppressed and prone to a multitude of infectious complications.1,2 Among these, bacteremia continues to be a major source of morbidity and mortality. This fact is shown best by a report by Abbott et al,11 who reviewed the U.S. Renal Data System database between 1994 and 1997. They found 1,447 episodes of septicemia among 33,479 renal transplant patients. This translated to 1,896 hospitalizations and amounted to a gigantic adjusted incidence ratio of hospitalizations for septicemia of 41.5 compared with the general population. An infected vascular device was present in 10.6% of hospitalizations for septicemia. This dismal rate of hospitalizations for septicemia among renal transplant patients should be fought with every effort to identify and eliminate its causes. The five patients presented here highlight that occult infection of old clotted AVGs is a potential source of bacteremia and life-threatening illness in renal transplant patients. It is curable if identified without delay and treated without hesitation. The approach we advise is as follows: In patients with obvious signs of local AVG infection, prompt surgical resection and antimicrobial therapy are required. A high index of suspicion for occult AVG infection should be exercised in all renal transplant patients with old nonfunctioning AVGs and presenting with unexplained bacteremia, fever of unknown origin, or failure to thrive. In such cases, we advocate performing an indium-labeled white blood cell scan, which we previously showed to have a high degree of sensitivity and specificity for detecting occult AVG infection.6-8 Surgical resection of all AVGs that show local uptake by indium scan is advised.6-8 Infectious complications of the old nonfunctioning AVG in renal transplant patients have not been the subject of a formal study. To our knowledge, our patients represent the first report on this subject in renal transplant patients to date. We are aware that this report does not provide data on how the AVGs became infected in the first place or on the incidence and magnitude of
835
infections caused by occult AVG infections in renal transplant patients. Nevertheless, this report does convey the message to the renal transplant community about the potential threats of occult AVG infection in everyday clinical practice. Finally, comment in regard to renal transplant candidates who have asymptomatic old nonfunctioning AVGs in their extremities is warranted. Should these AVGs be left alone? Should they undergo evaluation to exclude occult AVG infection before transplantation? Should they undergo routine resection? Definite answers cannot be offered at present without data addressing this particular issue, and different arguments can be made. We argue against the current prevailing practice of leaving these AVGs in place without any evaluation. Screening these AVGs with indium scan is indicated based on data that show a high prevalence (20% to 70%) of occult infection of old nonfunctioning AVGs among asymptomatic hemodialysis patients.5 The risk of occult AVG infection may be higher if the patient gives a history of previous bacteremia or fever of unknown origin. As stated earlier, the indium scan is a reliable tool for detecting occult AVG infection. If the indium scan shows positive uptake around an old nonfunctioning AVG, the AVG should be removed before kidney transplantation. The value of indium scan in patients undergoing evaluation for renal transplantation also extends to evaluation of their functioning AVGs. Functional AVGs also are prone to occult infection, and if they show increased uptake by indium scan, these functional accesses should be removed surgically before transplantation. ACKNOWLEDGMENT We thank Dr Mario Rubin and Dr Roberto Alcazar for providing data on two patients (cases 1 and 5).
REFERENCES 1. Rubin RH: Infectious disease complications of renal transplantation. Kidney Int 44:221-236, 1993 2. Fishman JA, Rubin RH: Infection in organ transplant recipients. N Engl J Med 338:1741-1751, 1998 3. McKay DB, Milford EL, Sayegh MH: Clinical aspects of renal transplantation, in Brenner BM and Rector FC (eds): The Kidney (ed 5). Philadelphia, PA, Saunders, 1995, pp 2602-2652
836
4. Feldman HI, Kobrin S, Wasserstein A: HD vascular access morbidity. J Am Soc Nephrol 7:523-535, 1996 5. Nassar GM, Quinibi W, Aboujaoude W, Ayus JC: Prevalence and complications of old clotted arteriovenous grafts in HD patients. J Am Soc Nephrol 12:298A, 2001 (abstract) 6. Ayus JC, Sheikh-Hamad D: Silent infection in clotted HD access grafts. J Am Soc Nephrol 9:1314-1317, 1998 7. Sheikh-Hamad D, Ayus JC: The patient with a clotted PTFE graft developing fever. Nephrol Dial Transplant 13: 2392-2393, 1998
NASSAR AND AYUS
8. Nassar GM, Ayus JC: Clotted arteriovenous grafts: A silent source of infection. Semin Dial 13:1-3, 2000 9. Nassar GM, Ayus JC: Infectious complications of the HD access. Kidney Int 60:1-13, 2001 10. Nassar GM, Fishbane S, Ayus JC: Occult infection of old nonfunctioning arteriovenous grafts: A novel cause of erythropoietin resistance and chronic inflammation in hemodialysis patients. Kidney Int 61:S49-S54, 2002 (suppl 80) 11. Abbott KC, Oliver JD 3rd, Hypolite I, et al: Hospitalizations for bacterial septicemia after renal transplantation in the United States. Am J Nephrol 21:120-127, 2001