Prosthetic-Joint Infections
http://www.100md.com
《新英格兰医药杂志》
To the Editor: Zimmerli et al. (Oct. 14 issue),1 in their review article, include a table outlining the treatment of infections associated with prosthetic joints. I would like to comment on some of the choices of antibiotics suggested in the table. The short half-life of nafcillin supports a dose of 2 g every four hours (not every six hours) to better obtain persistent, adequate levels of the drug. There is essentially no difference in toxicity at this dose, but one receives 50 percent more drug at very little additional cost, potentially with far better results. Similarly, the doses of penicillin and ampicillin are best given every four hours, or even possibly by continuous infusion for optimal results.
Although quinolones have antistaphylococcal activity, there is great potential for resistance. Other, possibly superior antibiotics that should have been included in the table are linezolid (oral and intravenous formulations) for both patients with methicillin-resistant Staphylococcus aureus infection and those who are severely allergic to beta-lactam antibiotics, clindamycin as another oral alternative, and a single daily high dose of ceftriaxone as an option for selected intravenous outpatient regimens.
Aaron E. Glatt, M.D.
Our Lady of Mercy Medical Center
Bronx, NY 10466
aglatt@olmhs.org
References
Zimmerli W, Trampuz A, Ochsner PE. Prosthetic-joint infections. N Engl J Med 2004;351:1645-1654.
To the Editor: Zimmerli et al. state that technetium-99m–labeled monoclonal anti-NCA-90 antibody Fab' fragments (LeukoScan) have an accuracy of 81 percent for detecting arthroplasty-associated infection, based on a single study.1 However, comparison with previous studies suggests that this test does not have as accurate results as was suggested. A study from the United Kingdom reviewed the results of this test in 55 patients with infected prosthetic joints, infected internal-fixation devices, or septic arthritis.2 The positive predictive value in the prosthetic-joint group was 67 percent (58 percent in all patients). Another study from the United Kingdom showed a positive predictive value of 20 percent.3 A study from France with 32 possible foci of infection in 23 patients showed a positive predictive accuracy of 78 percent.4 These studies confirm our impression — it is difficult to see how the test could be relied on to diagnose infection, owing to its low positive predictive accuracy. A positive scan is insufficient for diagnosis and requires further correlation with gallium or white-cell imaging.
Eitan Melamed, M.D.
Ilan Cohen, M.D.
Dror Robinson, M.D., Ph.D.
Rabin Medical Center
49372 Petach-Tikva, Israel
eitanm@clalit.org.il
References
Ivancevic V, Perka C, Hasart O, Sandrock D, Munz DL, Ivaneeviae V. Imaging of low-grade bone infection with a technetium-99m labelled monoclonal anti-NCA-90 Fab' fragment in patients with previous joint surgery. Eur J Nucl Med Mol Imaging 2002;29:547-551. abstract.
Ryan PJ. Leukoscan for orthopaedic imaging in clinical practice. Nucl Med Commun 2002;23:707-714.
Cosgriff PS, Aslam M, Minhas THA. Tc-99m leukoscan in suspected orthopaedic infection. Nucl Med Commun 1999;20:478-478. abstract.
Devillers A, Garin E, Polard JL, et al. Comparison of Tc-99m-labelled antileukocyte fragment Fab' and Tc-99m-HMPAO leukocyte scintigraphy in the diagnosis of bone and joint infections: a prospective study. Nucl Med Commun 2000;21:747-753.
The authors reply: Dr. Glatt suggests a shorter dose interval, or a continuous infusion, for penicillins rather than the one stated in our article. This seems reasonable on the basis of pharmacokinetic and pharmacodynamic considerations. However, our dose intervals were based on the European standard interval (i.e., the six-hour interval), as used for floxacillin in our randomized, controlled trial.1 In addition, the clinical advantage of shorter dose intervals, with more laborious treatment, has never been proved. In Table 1 of our article, we included only antimicrobial agents for which published data were available for the treatment of staphylococcal infections related to orthopedic devices. This was not the case for linezolid, clindamycin, or ceftriaxone. Since our article was published, data from a study of therapy with linezolid have been reported.2 However, in view of the high cumulative toxicity of linezolid in this study, with 40 percent myelosuppression within 16 weeks, our conservative approach seems to be justified.
Melamed et al. point out that several studies using antigranulocyte labeling show a lower accuracy in diagnosing device-associated infection than the study referred to in our review. However, they cite not the accuracy but the positive predictive value. In our review, we chose a representative study with results exactly in the range of those in the reports cited by Melamed et al. In the study by Ivancevic et al.,3 the sensitivity was 100 percent, the specificity 69 percent, the accuracy 81 percent, the positive predictive value 67 percent, and the negative predictive value 100 percent in patients with hip arthroplasty. Thus, we fully agree that the technique involving technetium-99m–labeled anti-NCA-90 antibody Fab' fragments is not ideal for diagnosing infection, owing to insufficient specificity. Indeed, imaging with labeled leukocytes combined with technetium-99m sulphur colloid bone marrow scintigraphy has a higher accuracy, 90 percent.4 However, leukocyte scintigraphy is cumbersome and requires skill in order to get granulocytes with intact chemotaxis. We previously showed large differences of chemotactic potency between technetium-99m-hexamethylpropylene-amino-oxime–labeled and indium-111-oxine–labeled granulocytes.5 For these reasons, most European centers switched to the in vivo technique with labeled monoclonal antibodies. Taken together, nuclear-imaging methods are not ideal or decisive tests for diagnosis or follow-up of infections associated with prosthetic joints.
Werner Zimmerli, M.D.
Basel University Medical Clinic
CH-4410 Liestal, Switzerland
werner.zimmerli@unibas.ch
Andrej Trampuz, M.D.
University Hospital Basel
CH-4031 Basel, Switzerland
References
Zimmerli W, Widmer AF, Blatter M, Frei R, Ochsner PE. Role of rifampin for treatment of orthopedic implant-related staphylococcal infections: a randomized controlled trial. JAMA 1998;279:1537-1541.
Razonable RR, Osmon DR, Steckelberg JM. Linezolid therapy for orthopedic infections. Mayo Clin Proc 2004;79:1137-1144.
Ivancevic V, Perka C, Hasart O, Sandrock D, Munz DL, Ivaneeviae V. Imaging of low-grade bone infection with a technetium-99m labelled monoclonal anti-NCA-90 Fab' fragment in patients with previous joint surgery. Eur J Nucl Med Mol Imaging 2002;29:547-551.
Love C, Tomas MB, Marwin SE, Pugliese PV, Palestro CJ. Role of nuclear medicine in diagnosis of the infected joint replacement. Radiographics 2001;21:1229-1238.
Stoeckli TC, Zimmerli W, Maecke HR, Fridrich R. Comparison of chemotaxis and superoxide generation of indium-111-oxine- and technetium-99m-HMPAO-labelled granulocytes. Scand J Clin Lab Invest 1996;56:305-309.
Although quinolones have antistaphylococcal activity, there is great potential for resistance. Other, possibly superior antibiotics that should have been included in the table are linezolid (oral and intravenous formulations) for both patients with methicillin-resistant Staphylococcus aureus infection and those who are severely allergic to beta-lactam antibiotics, clindamycin as another oral alternative, and a single daily high dose of ceftriaxone as an option for selected intravenous outpatient regimens.
Aaron E. Glatt, M.D.
Our Lady of Mercy Medical Center
Bronx, NY 10466
aglatt@olmhs.org
References
Zimmerli W, Trampuz A, Ochsner PE. Prosthetic-joint infections. N Engl J Med 2004;351:1645-1654.
To the Editor: Zimmerli et al. state that technetium-99m–labeled monoclonal anti-NCA-90 antibody Fab' fragments (LeukoScan) have an accuracy of 81 percent for detecting arthroplasty-associated infection, based on a single study.1 However, comparison with previous studies suggests that this test does not have as accurate results as was suggested. A study from the United Kingdom reviewed the results of this test in 55 patients with infected prosthetic joints, infected internal-fixation devices, or septic arthritis.2 The positive predictive value in the prosthetic-joint group was 67 percent (58 percent in all patients). Another study from the United Kingdom showed a positive predictive value of 20 percent.3 A study from France with 32 possible foci of infection in 23 patients showed a positive predictive accuracy of 78 percent.4 These studies confirm our impression — it is difficult to see how the test could be relied on to diagnose infection, owing to its low positive predictive accuracy. A positive scan is insufficient for diagnosis and requires further correlation with gallium or white-cell imaging.
Eitan Melamed, M.D.
Ilan Cohen, M.D.
Dror Robinson, M.D., Ph.D.
Rabin Medical Center
49372 Petach-Tikva, Israel
eitanm@clalit.org.il
References
Ivancevic V, Perka C, Hasart O, Sandrock D, Munz DL, Ivaneeviae V. Imaging of low-grade bone infection with a technetium-99m labelled monoclonal anti-NCA-90 Fab' fragment in patients with previous joint surgery. Eur J Nucl Med Mol Imaging 2002;29:547-551. abstract.
Ryan PJ. Leukoscan for orthopaedic imaging in clinical practice. Nucl Med Commun 2002;23:707-714.
Cosgriff PS, Aslam M, Minhas THA. Tc-99m leukoscan in suspected orthopaedic infection. Nucl Med Commun 1999;20:478-478. abstract.
Devillers A, Garin E, Polard JL, et al. Comparison of Tc-99m-labelled antileukocyte fragment Fab' and Tc-99m-HMPAO leukocyte scintigraphy in the diagnosis of bone and joint infections: a prospective study. Nucl Med Commun 2000;21:747-753.
The authors reply: Dr. Glatt suggests a shorter dose interval, or a continuous infusion, for penicillins rather than the one stated in our article. This seems reasonable on the basis of pharmacokinetic and pharmacodynamic considerations. However, our dose intervals were based on the European standard interval (i.e., the six-hour interval), as used for floxacillin in our randomized, controlled trial.1 In addition, the clinical advantage of shorter dose intervals, with more laborious treatment, has never been proved. In Table 1 of our article, we included only antimicrobial agents for which published data were available for the treatment of staphylococcal infections related to orthopedic devices. This was not the case for linezolid, clindamycin, or ceftriaxone. Since our article was published, data from a study of therapy with linezolid have been reported.2 However, in view of the high cumulative toxicity of linezolid in this study, with 40 percent myelosuppression within 16 weeks, our conservative approach seems to be justified.
Melamed et al. point out that several studies using antigranulocyte labeling show a lower accuracy in diagnosing device-associated infection than the study referred to in our review. However, they cite not the accuracy but the positive predictive value. In our review, we chose a representative study with results exactly in the range of those in the reports cited by Melamed et al. In the study by Ivancevic et al.,3 the sensitivity was 100 percent, the specificity 69 percent, the accuracy 81 percent, the positive predictive value 67 percent, and the negative predictive value 100 percent in patients with hip arthroplasty. Thus, we fully agree that the technique involving technetium-99m–labeled anti-NCA-90 antibody Fab' fragments is not ideal for diagnosing infection, owing to insufficient specificity. Indeed, imaging with labeled leukocytes combined with technetium-99m sulphur colloid bone marrow scintigraphy has a higher accuracy, 90 percent.4 However, leukocyte scintigraphy is cumbersome and requires skill in order to get granulocytes with intact chemotaxis. We previously showed large differences of chemotactic potency between technetium-99m-hexamethylpropylene-amino-oxime–labeled and indium-111-oxine–labeled granulocytes.5 For these reasons, most European centers switched to the in vivo technique with labeled monoclonal antibodies. Taken together, nuclear-imaging methods are not ideal or decisive tests for diagnosis or follow-up of infections associated with prosthetic joints.
Werner Zimmerli, M.D.
Basel University Medical Clinic
CH-4410 Liestal, Switzerland
werner.zimmerli@unibas.ch
Andrej Trampuz, M.D.
University Hospital Basel
CH-4031 Basel, Switzerland
References
Zimmerli W, Widmer AF, Blatter M, Frei R, Ochsner PE. Role of rifampin for treatment of orthopedic implant-related staphylococcal infections: a randomized controlled trial. JAMA 1998;279:1537-1541.
Razonable RR, Osmon DR, Steckelberg JM. Linezolid therapy for orthopedic infections. Mayo Clin Proc 2004;79:1137-1144.
Ivancevic V, Perka C, Hasart O, Sandrock D, Munz DL, Ivaneeviae V. Imaging of low-grade bone infection with a technetium-99m labelled monoclonal anti-NCA-90 Fab' fragment in patients with previous joint surgery. Eur J Nucl Med Mol Imaging 2002;29:547-551.
Love C, Tomas MB, Marwin SE, Pugliese PV, Palestro CJ. Role of nuclear medicine in diagnosis of the infected joint replacement. Radiographics 2001;21:1229-1238.
Stoeckli TC, Zimmerli W, Maecke HR, Fridrich R. Comparison of chemotaxis and superoxide generation of indium-111-oxine- and technetium-99m-HMPAO-labelled granulocytes. Scand J Clin Lab Invest 1996;56:305-309.