Skeletal Radiol (1992) 21:297-299
Skeletal Radiology
Ultrasonography after hip arthroplasty K. F61des, M.D. 1, M. Gaal, M.D. 1, P. Balint, M.D. t, K. Nemenyi, M.D. 1, C. Kiss, M.D. 1, G.P. Balint, M.D. 1, and W.W. Buchanan, M.D. 2 1 National Institute of Rheumatology and Physiotherapy, H-1025 Budapest, Hungary 2 Department of Medicine, McMaster University, Hamilton, Ontario, Canada
Abstract. U l t r a s o n o g r a p h y was p e r f o r m e d in 55 patients w h o h a d total C h a r n l e y hip arthroplasties. Effusions were identified in 19 patients a n d confirmed in all but 3 by arthrocentesis or at surgery. Aspirations were perf o r m e d in 5 and d e m o n s t r a t e d infection in 2. It is concluded that u l t r a s o u n d is a valuable noninvasive m e t h o d for assessing painful hip arthroplasty. It can d c m o n strate the presence o f effusion, which should be aspirated to exclude infection. Key words: Hip joint a r t h r o p l a s t y - U l t r a s o n o g r a p h y
Infection and aseptic loosening are the two m o s t c o m m o n complications o f total hip a r t h r o p l a s t y [2, 4, 6, 10-14, 16]. Early diagnosis is especially important. After physical examination, the following procedures m a y be ordered: plain r a d i o g r a p h y , radionuclide b o n e scan, art h r o g r a p h y , a n d aspiration [6, 14]. There is scant inform a t i o n on u l t r a s o u n d as a m e a n s o f investigating surgically treated hip joints [1, 7]. L o o s e n i n g o f hip prostheses m a y be the consequence o f a synovitis [2, 4, 6]. In this study we have a t t e m p t e d to answer the following questions: (1) C a n an effusion in a prosthetic hip joint be detected by ultrasound, independent o f etiology? (2) W h a t is the a c c u r a c y o f u l t r a s o u n d o f the hip joint in patients with total hip replacement? (3) W h a t are the causes o f n o n m e c h a n i c a l and aseptic loosening, and h o w can u l t r a s o u n d help to identify t h e m ? (4) C a n early diagnosis aid therapeutic intervention? and (5) W h a t is the incidencc o f effusion detected by u l t r a s o u n d in the contralateral, painless, untreated hip?
Materials and methods Fifty-five patients (37 women and 18 men) who had Charnley hip arthroplasties carried out 6 months to 8 years previously were Address reprint requests to: K. F61des, M.D., Department of Rheumatology, Frankel-Leo u 17-19 H-1025, National Institute of Rheumatology & Physiotherapy, Budapest, Hungary
studied. The mean age was 65 years, with a range of 57 68 years. None of the patients had complications within 3 months of surgery. Twenty-five patients complained of hip pain. All had undergone plain radiographs of the hip, and 10 had technetium-99m pertechnetate (99mTcO4) bone scans. Twenty of the patients subsequently had second operations. Fluid was obtained from 3 of 5 patients who had arthrocentesis. Arthrocentesis was performed with a 20 gauge needle using the freehand ultrasound-guided technique [3]. The needle was directed in the sagittal plane toward the neck of the prosthesis and was maintained caudal to the transducer head. The needle was then guided into the effusion under sonographic control [9]. The transducer was covered with a sterile glove, and sterile gel was used. Routine hematology, blood biochemistry, and cultures were obtained in all patients. Synovial fluid obtained in 3 of the 5 patients who underwent arthrocentesis was examined for total and differential cell counts and lactic dehydrogenase and glucose levels, and crystals, and was cultured. Ultrasound was performed in 55 patients. Of the 25 patients who had pain following arthroplasty, 20 had second operations, and 5 had arthrocentesis. Control ultrasound was only performed in the latter group. Ultrasound was also performed on the untreated contralateral hip of the 20 patients who had second operations. Radiographs of these painless hips showed only mild osteoarthritic changes. Thirty patients who had successful Charnley hip arthroplasties were studied as controls. Five of these patients had bilateral arthroplasties. Ultrasound examinations were performed with a Hitachi EUB 450 real-time scanner with a 5-MHz transducer. Scanning was donc in the sagittal plane, parallel to the neck of the prosthesis. An effusion was defined as fluid within the capsule [1, 7] (Fig. 1).
Results Table i summarizes the findings in 20 patients with hip pain due to loosening o f the prosthesis; this g r o u p underwent second operations. O f the patients w h o were considered to have noninfective loosening o n clinical grounds, 4 had s o n o g r a p h i c evidence o f effusion. Effusion was surgically confirmed in 3, as well as in 1 o f the remaining 6 patients w h o had negative u l t r a s o u n d examinations. One o f the 3 patients w h o had a positive u l t r a s o u n d rcsult and an effusion confirmed at the second o p e r a t i o n was s h o w n to have a Pseudomonas aeruginosa infection (Fig. 2). Infec9 1992 International Skeletal Society
298
K. F61des et al. : US after hip arthroplasty
Fig. 1. Ultrasonogram of hip joint showing large amount of effusion (effusio in Hungarian). The capsule is outlined by arrows. The horizontal lines below indicate echoes from the metallic neck of the prosthesis. The effusion was shown to be noninfective
Fig. 2. A further example on the right of effusion following Charnley arthroplasty. This was proven to be infected at operation. Note the small echoes in the effusion, which are more frequent in infected effusions. The operated hip on the left shows no evidence of effusion
Table 1. Results of radiographs, radionuclide bone scans, and ultrasonography in 20 patients who had second operations because of painful loosening of the prosthesis following Charnley arthroplasties
and responded to antibiotic therapy. In the remaining 2 patients, it was not possible to aspirate any fluid from the affected hip joints. Ultrasound demonstrated an effusion in only 1 of 30 painless hip joints which had been surgically treated. It was not considered ethical to attempt aspiration in an asymptomatic patient with an effusion demonstrated on ultrasound examination. However, it is of interest that the effusion could not be demonstrated 2 weeks after treatment with a nonsteroidal anti-inflammatory analgesic. In the group consisting of 20 subjects who had second operations and 5 who had painful hips after joint surgery, 2 were shown to have an effusion of the contralateral untreated hip. These effusions were confirmed by aspiration. Both patients had mild pain due to mild osteoarthritis. D a t a in the literature [8] suggest a high sensitivity and lower specificity for sonographic detection of effusions. The sensitivity calculated for our study in patients treated with second operations was 92.8% ; the specificity was 83.3% (Table 2).
X-radiography evidence of loosening
Radionuclide bone scan
Ultrasonography
Fluid present at surgery
Not done Not done
4 6
3 1
10 0
10 0
Noninf ective a
Positive Negative
7 3
Possibly infective a
Positive Negative
10 0
9 1
" As determined by clinical laboratory findings prior to second operation Table 2. Data base for calculation of sensitivity and specificity of ultrasound Number
True positive
False positive
True negative
False negative
13
1
5
1
3
2
0
0
Reoperated
20 Arthrocentesis 5
Sensitivity: 92.8% ; specificity: 83% tion was surgically confirmed in only 4 of the 10 patients suspected of having joint infection before surgery who had positive ultrasound examinations. Five patients did not have second operations but had effusions demonstrated by ultrasound. Joint aspiration confirmed the effusion in 3. Two proved to be infected
Discussion The purpose of this study was to determine the value of ultrasound in assessing the failed hip arthroplasty. In m a n y cases, the postoperative pain is due to mechanical loosening of the prosthesis. This m a y give rise to a secondary inflammation of the synovium, probably as a result of a foreign body reaction [2, 6, 15]. In addition to this aseptic process, infection m a y supervene. In both instances, effusions m a y develop. We hoped that ultrasound might prove a useful noninvasive procedure to confirm the presence of such an effusion. In 13 of 14 patients, it was possible to obtain surgical confirmation of the presence of a hip joint effusion identified by ultrasound. Thus, the incidence of false posi-
K. F61des et al. : US after hip arthroplasty rives was 1 in 14. Only 1 of the 6 patients with a negative preoperative ultrasound result was found to have fluid. Three of the 5 patients who had painful hips after surgery but did not undergo second operations had effusions identified at sonography and confirmed by arthrocentesis. The remaining 2 patients had effusion shown by ultrasonography, but attempts at aspiration were unsuccessful. It is possible that these patients may have had effusions since " d r y taps ", even with clearly demonstrable effusions, are not u n c o m m o n in the knee. Ultrasound demonstrated only 12 effusions in the 25 contralateral hips of the patients who had hip arthroplasties and the 30 painless hips in the 30 control subjects. Because it would have been unethical to attempt to confirm the presence of an effusion by joint aspiration in asymptomatic patients, we were unable to determine the reliability of the sonographic finding. However, 2 other patients in this group had mild pain as a result of osteoarthritis, and effusions were confirmed by aspiration in both. Ultrasound therefore appears to have potential in identifying the presence of hip joint effusion, but it cannot be expected to determine whether infection is present. This is best done by a labelled leukocyte scan (ind i u m - l l l - W B C ) . Ultrasonography m a y therefore alert the clinician to the presence of fluid, which can then be aspirated in order to exclude infection. Ultrasound provides a simple method of visualizing the needle location during aspiration and requires less manipulation of the needle. Ultrasound has many advantages over conventional fluoroscopic techniques: lack of ionizing radiation, superior visualization of the needle, relatively low cost, and short examination time [5].
References
1. Baratelli M, Cabitza P, Parini L (1986) Ultrasonography in the investigation of loose hip prosthesis. Ital J Orthop Traumatol 12:77
299 2. Deutman R, Mulder THJ, Brian R, et al (1977) Metal sensitivity before and after total hip arthroplasty. J Bone Joint Surg [Am] 59 : 862 3. Fornage BD (1989) Simple phantom for training in US-guided needle biopsy using the freehand technique. J Ultrasound Med 8:701 4. Goldring SR, Bringhurst FR, Roelke M, et al (1984) Loosening of prosthetic components after total hip replacement (THR): presence of a synovial-iike membrane and its role in bone lysis (abstract). Arthritis Rheum 27 (Suppl 4) : S 41 5. Hankey S, McCall WJ, Park MW, O'Connor TB (1979) Technical problems in arthrography of the painful hip arthroplasty. Clin Radiol 30:653 6. Kaufman RL, Tong I, Beardmore DT (1985) Prosthetic synovitis: clinical and histologic characteristics. J Rheumatol 12:1066 7. Komppa GH, Northern JR, Haas DK, et al (1985) Ultrasound guidance for needle aspiration of the hip in patients with painful hip prosthesis. J Clin Ultrasound 13:433 8. Koski JM, Antilla P, Haalainen M, Isomaki H (1990) Hip joint ultrasonography: correlation with intraarticular effusion and synovitis. Br J Rheumatol 29 : 189 9. Mayekawa DS, Rails PW, Kerr RM, Lee KP, et al (1989) Sonographically guided arthrocentesis of the hip. J Ultrasound Med 8 : 665 10. McLaughlin ER, Whitehill R (1977) Evaluation of the painful hip by aspiration and arthrography. Surg Gynecol Obstet 144:381 1t. Smith RE, Turner RJ (1973) Total hip replacement using methylmethacrylate cement: an analysis of data from 3, 482 cases. Clin Orthop 95:231 12. Spinelli R (1976) The role of scintigraphy in the diagnosis of the late complications of total prosthesis of the hip. Infection, loosening, ossification. Ital J Orthop Traumatol 2 : 79 13. Stoker DJ (1980) A simple technique of joint puncture following hip arthroplasty. Radiology 136:234 14. Tehranzadeh J, Schneider R, Freiberger HR (1981) Radiological evaluation of painful total hip replacement. Radiology 141:355 15. Vernon-Roberts B, Freeman MAR (1976) Morphological and analytical studies of the tissues adjacent to joint prosthesis: investigations into the causes of loosening of prosthesis. In: Engineering in medicine. Springer, New York, 2:162 16. Wilson PD Jr, Aglietti P, Salvati EA (1974) Subacute sepsis of the hip treated by antibiotics and cemented prosthesis. J Bone Joint Surg [Am] 56:879
Skeletal Radiology
Ultrasonography after hip arthroplasty K. F61des, M.D. 1, M. Gaal, M.D. 1, P. Balint, M.D. t, K. Nemenyi, M.D. 1, C. Kiss, M.D. 1, G.P. Balint, M.D. 1, and W.W. Buchanan, M.D. 2 1 National Institute of Rheumatology and Physiotherapy, H-1025 Budapest, Hungary 2 Department of Medicine, McMaster University, Hamilton, Ontario, Canada
Abstract. U l t r a s o n o g r a p h y was p e r f o r m e d in 55 patients w h o h a d total C h a r n l e y hip arthroplasties. Effusions were identified in 19 patients a n d confirmed in all but 3 by arthrocentesis or at surgery. Aspirations were perf o r m e d in 5 and d e m o n s t r a t e d infection in 2. It is concluded that u l t r a s o u n d is a valuable noninvasive m e t h o d for assessing painful hip arthroplasty. It can d c m o n strate the presence o f effusion, which should be aspirated to exclude infection. Key words: Hip joint a r t h r o p l a s t y - U l t r a s o n o g r a p h y
Infection and aseptic loosening are the two m o s t c o m m o n complications o f total hip a r t h r o p l a s t y [2, 4, 6, 10-14, 16]. Early diagnosis is especially important. After physical examination, the following procedures m a y be ordered: plain r a d i o g r a p h y , radionuclide b o n e scan, art h r o g r a p h y , a n d aspiration [6, 14]. There is scant inform a t i o n on u l t r a s o u n d as a m e a n s o f investigating surgically treated hip joints [1, 7]. L o o s e n i n g o f hip prostheses m a y be the consequence o f a synovitis [2, 4, 6]. In this study we have a t t e m p t e d to answer the following questions: (1) C a n an effusion in a prosthetic hip joint be detected by ultrasound, independent o f etiology? (2) W h a t is the a c c u r a c y o f u l t r a s o u n d o f the hip joint in patients with total hip replacement? (3) W h a t are the causes o f n o n m e c h a n i c a l and aseptic loosening, and h o w can u l t r a s o u n d help to identify t h e m ? (4) C a n early diagnosis aid therapeutic intervention? and (5) W h a t is the incidencc o f effusion detected by u l t r a s o u n d in the contralateral, painless, untreated hip?
Materials and methods Fifty-five patients (37 women and 18 men) who had Charnley hip arthroplasties carried out 6 months to 8 years previously were Address reprint requests to: K. F61des, M.D., Department of Rheumatology, Frankel-Leo u 17-19 H-1025, National Institute of Rheumatology & Physiotherapy, Budapest, Hungary
studied. The mean age was 65 years, with a range of 57 68 years. None of the patients had complications within 3 months of surgery. Twenty-five patients complained of hip pain. All had undergone plain radiographs of the hip, and 10 had technetium-99m pertechnetate (99mTcO4) bone scans. Twenty of the patients subsequently had second operations. Fluid was obtained from 3 of 5 patients who had arthrocentesis. Arthrocentesis was performed with a 20 gauge needle using the freehand ultrasound-guided technique [3]. The needle was directed in the sagittal plane toward the neck of the prosthesis and was maintained caudal to the transducer head. The needle was then guided into the effusion under sonographic control [9]. The transducer was covered with a sterile glove, and sterile gel was used. Routine hematology, blood biochemistry, and cultures were obtained in all patients. Synovial fluid obtained in 3 of the 5 patients who underwent arthrocentesis was examined for total and differential cell counts and lactic dehydrogenase and glucose levels, and crystals, and was cultured. Ultrasound was performed in 55 patients. Of the 25 patients who had pain following arthroplasty, 20 had second operations, and 5 had arthrocentesis. Control ultrasound was only performed in the latter group. Ultrasound was also performed on the untreated contralateral hip of the 20 patients who had second operations. Radiographs of these painless hips showed only mild osteoarthritic changes. Thirty patients who had successful Charnley hip arthroplasties were studied as controls. Five of these patients had bilateral arthroplasties. Ultrasound examinations were performed with a Hitachi EUB 450 real-time scanner with a 5-MHz transducer. Scanning was donc in the sagittal plane, parallel to the neck of the prosthesis. An effusion was defined as fluid within the capsule [1, 7] (Fig. 1).
Results Table i summarizes the findings in 20 patients with hip pain due to loosening o f the prosthesis; this g r o u p underwent second operations. O f the patients w h o were considered to have noninfective loosening o n clinical grounds, 4 had s o n o g r a p h i c evidence o f effusion. Effusion was surgically confirmed in 3, as well as in 1 o f the remaining 6 patients w h o had negative u l t r a s o u n d examinations. One o f the 3 patients w h o had a positive u l t r a s o u n d rcsult and an effusion confirmed at the second o p e r a t i o n was s h o w n to have a Pseudomonas aeruginosa infection (Fig. 2). Infec9 1992 International Skeletal Society
298
K. F61des et al. : US after hip arthroplasty
Fig. 1. Ultrasonogram of hip joint showing large amount of effusion (effusio in Hungarian). The capsule is outlined by arrows. The horizontal lines below indicate echoes from the metallic neck of the prosthesis. The effusion was shown to be noninfective
Fig. 2. A further example on the right of effusion following Charnley arthroplasty. This was proven to be infected at operation. Note the small echoes in the effusion, which are more frequent in infected effusions. The operated hip on the left shows no evidence of effusion
Table 1. Results of radiographs, radionuclide bone scans, and ultrasonography in 20 patients who had second operations because of painful loosening of the prosthesis following Charnley arthroplasties
and responded to antibiotic therapy. In the remaining 2 patients, it was not possible to aspirate any fluid from the affected hip joints. Ultrasound demonstrated an effusion in only 1 of 30 painless hip joints which had been surgically treated. It was not considered ethical to attempt aspiration in an asymptomatic patient with an effusion demonstrated on ultrasound examination. However, it is of interest that the effusion could not be demonstrated 2 weeks after treatment with a nonsteroidal anti-inflammatory analgesic. In the group consisting of 20 subjects who had second operations and 5 who had painful hips after joint surgery, 2 were shown to have an effusion of the contralateral untreated hip. These effusions were confirmed by aspiration. Both patients had mild pain due to mild osteoarthritis. D a t a in the literature [8] suggest a high sensitivity and lower specificity for sonographic detection of effusions. The sensitivity calculated for our study in patients treated with second operations was 92.8% ; the specificity was 83.3% (Table 2).
X-radiography evidence of loosening
Radionuclide bone scan
Ultrasonography
Fluid present at surgery
Not done Not done
4 6
3 1
10 0
10 0
Noninf ective a
Positive Negative
7 3
Possibly infective a
Positive Negative
10 0
9 1
" As determined by clinical laboratory findings prior to second operation Table 2. Data base for calculation of sensitivity and specificity of ultrasound Number
True positive
False positive
True negative
False negative
13
1
5
1
3
2
0
0
Reoperated
20 Arthrocentesis 5
Sensitivity: 92.8% ; specificity: 83% tion was surgically confirmed in only 4 of the 10 patients suspected of having joint infection before surgery who had positive ultrasound examinations. Five patients did not have second operations but had effusions demonstrated by ultrasound. Joint aspiration confirmed the effusion in 3. Two proved to be infected
Discussion The purpose of this study was to determine the value of ultrasound in assessing the failed hip arthroplasty. In m a n y cases, the postoperative pain is due to mechanical loosening of the prosthesis. This m a y give rise to a secondary inflammation of the synovium, probably as a result of a foreign body reaction [2, 6, 15]. In addition to this aseptic process, infection m a y supervene. In both instances, effusions m a y develop. We hoped that ultrasound might prove a useful noninvasive procedure to confirm the presence of such an effusion. In 13 of 14 patients, it was possible to obtain surgical confirmation of the presence of a hip joint effusion identified by ultrasound. Thus, the incidence of false posi-
K. F61des et al. : US after hip arthroplasty rives was 1 in 14. Only 1 of the 6 patients with a negative preoperative ultrasound result was found to have fluid. Three of the 5 patients who had painful hips after surgery but did not undergo second operations had effusions identified at sonography and confirmed by arthrocentesis. The remaining 2 patients had effusion shown by ultrasonography, but attempts at aspiration were unsuccessful. It is possible that these patients may have had effusions since " d r y taps ", even with clearly demonstrable effusions, are not u n c o m m o n in the knee. Ultrasound demonstrated only 12 effusions in the 25 contralateral hips of the patients who had hip arthroplasties and the 30 painless hips in the 30 control subjects. Because it would have been unethical to attempt to confirm the presence of an effusion by joint aspiration in asymptomatic patients, we were unable to determine the reliability of the sonographic finding. However, 2 other patients in this group had mild pain as a result of osteoarthritis, and effusions were confirmed by aspiration in both. Ultrasound therefore appears to have potential in identifying the presence of hip joint effusion, but it cannot be expected to determine whether infection is present. This is best done by a labelled leukocyte scan (ind i u m - l l l - W B C ) . Ultrasonography m a y therefore alert the clinician to the presence of fluid, which can then be aspirated in order to exclude infection. Ultrasound provides a simple method of visualizing the needle location during aspiration and requires less manipulation of the needle. Ultrasound has many advantages over conventional fluoroscopic techniques: lack of ionizing radiation, superior visualization of the needle, relatively low cost, and short examination time [5].
References
1. Baratelli M, Cabitza P, Parini L (1986) Ultrasonography in the investigation of loose hip prosthesis. Ital J Orthop Traumatol 12:77
299 2. Deutman R, Mulder THJ, Brian R, et al (1977) Metal sensitivity before and after total hip arthroplasty. J Bone Joint Surg [Am] 59 : 862 3. Fornage BD (1989) Simple phantom for training in US-guided needle biopsy using the freehand technique. J Ultrasound Med 8:701 4. Goldring SR, Bringhurst FR, Roelke M, et al (1984) Loosening of prosthetic components after total hip replacement (THR): presence of a synovial-iike membrane and its role in bone lysis (abstract). Arthritis Rheum 27 (Suppl 4) : S 41 5. Hankey S, McCall WJ, Park MW, O'Connor TB (1979) Technical problems in arthrography of the painful hip arthroplasty. Clin Radiol 30:653 6. Kaufman RL, Tong I, Beardmore DT (1985) Prosthetic synovitis: clinical and histologic characteristics. J Rheumatol 12:1066 7. Komppa GH, Northern JR, Haas DK, et al (1985) Ultrasound guidance for needle aspiration of the hip in patients with painful hip prosthesis. J Clin Ultrasound 13:433 8. Koski JM, Antilla P, Haalainen M, Isomaki H (1990) Hip joint ultrasonography: correlation with intraarticular effusion and synovitis. Br J Rheumatol 29 : 189 9. Mayekawa DS, Rails PW, Kerr RM, Lee KP, et al (1989) Sonographically guided arthrocentesis of the hip. J Ultrasound Med 8 : 665 10. McLaughlin ER, Whitehill R (1977) Evaluation of the painful hip by aspiration and arthrography. Surg Gynecol Obstet 144:381 1t. Smith RE, Turner RJ (1973) Total hip replacement using methylmethacrylate cement: an analysis of data from 3, 482 cases. Clin Orthop 95:231 12. Spinelli R (1976) The role of scintigraphy in the diagnosis of the late complications of total prosthesis of the hip. Infection, loosening, ossification. Ital J Orthop Traumatol 2 : 79 13. Stoker DJ (1980) A simple technique of joint puncture following hip arthroplasty. Radiology 136:234 14. Tehranzadeh J, Schneider R, Freiberger HR (1981) Radiological evaluation of painful total hip replacement. Radiology 141:355 15. Vernon-Roberts B, Freeman MAR (1976) Morphological and analytical studies of the tissues adjacent to joint prosthesis: investigations into the causes of loosening of prosthesis. In: Engineering in medicine. Springer, New York, 2:162 16. Wilson PD Jr, Aglietti P, Salvati EA (1974) Subacute sepsis of the hip treated by antibiotics and cemented prosthesis. J Bone Joint Surg [Am] 56:879
