Vo1.11,1992
23. Biaser M J, Smith PF, Kohler PF: Susceptibility of Campylobacter isolates to the bactericidal activity of human serum. Journal of Infectious Diseases 1985,151: 227-235. 24. Endtz HP, van der Meet JWM, Mouton RP: Campylobacter jejuni bacteraemia in hypogammaglobulinaemia: emergence of ciprofloxacin resistance and results of serum bactericidal assays. In: Kaijser B, Falsen E (ed): Proceedings of the Fourth International Workshop on Campylobacter Infections. University of G6teburg, Ktingalve, Sweden, 1988, p. 150--151. 25. Perlman DM, Ampel NM, Schifman RB, Cohn DL, Patton CM, Aguirre ML, Wang WL, Blaser M J: Persistent Campylobacterjejuni infections in patients infected with human immunodefieieney virus (HIV). Annals of Internal Medicine 1988, 108: 540-546. 26. Endtz HP, Mouton RP, van der Reyden T, Ruijs G J, Biever M, van Klingereu B: Fluoroquinolone resistance in Campylobacter spp. isolated from human stools and poultry products. Lancet 1990, 335: 787.
Prosthetic Joint Infection due to Candida parapsilosis in the UK: Case Report and Literature Review J. Paul 1, S.H. White 2, K.M. Nicholls 1, D.W. C r o o k 1.
Candida infection of joint replacements is a rare but increasingly reported phenomenon. A case of Candida parapsilasis prosthetic knee joint infection occurring in the UK is described. Cure followed removal of the prosthesis and treatment, first with a combination of amphotericin and 5fluorocytosine, then ketoconazole.
The overwhelming majority of cases of infection of total joint arthroplasties are ascribed to bacterial pathogens (1). However, there is a growing North American literature referring to cases of prosthetic joint candidosis. A favourable outcome of these infections is seen after removal of the prosthesis and administration of specific antifungal treatment. However, such successful treatment depends on isolating Candida and recognising it as the aetiologic agent. We describe a case in 1Public Health Laboratory, Levels 6/7, John Radcliffe Hospital, Oxford OX3 9DU, UK. 2Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, UK.
847
the UK of infection of a total knee arthroplasty with Candida parapsilosis where our culturebased diagnosis was supported by histological and serological evidence, and where serial serological tests for candida allowed monitoring of the therapeutic response. The literature on candida joint infection, its diagnosis and treatment is reviewed.
Case Report. A 63-year-old man underwent a bicompartmental Oxford knee replacement for severe osteoarthritis. Continued pain led to revision to an Insall Burstein total condylar replacement 20 months later. Loosening of the prosthesis but no evidence of infection was noted at the time of the revision. Three months after revision, the joint became hot and red suggesting infection. Exploration showed inflamed granulations, but routine microbiological cultures were negative. Four months later persistent pain resulted in removal of all foreign material and the knee was fused using compressive biplanar external fixation. Tissue was taken for histological and microbiological investigation. Culture of debrided cancellous bone, joint capsule and periarticular tissue revealed Candida parapsilosis in pure growth. Histological examination of debrided bone showed yeasts and chronic, non-specific inflammatory changes. Sera taken 8 and 105 days before the time of removal of the prosthesis were sent to the Central Public Health Laboratory, London, UK, to test for the presence of candidal precipitins and agglutinins. The results proved to be highly supportive of a diagnosis of infection with Candida parapsilosis. A culture of the organism sent to the Central Public Health Laboratory, London, was found to be sensitive to amphotericin, 5-fluorocytosine and ketaconazole and the identification was confirmed as Candida parapsilosis. One week after removal of the prosthesis, treatment with intravenous amphotericin, 50 mg daily, and 5fluorocytosine, 2500 mg 6-hourly, was started. This regimen was well tolerated and continued for 40 days, and was followed by treatment for 24 days with ketoconazole, 200 mg 12-hourly. The knee became painless and results of serial serological tests for candida indicated a therapeutic response. At follow-up two years after removal of the prosthesis, the patient's fused knee was trouble-free and serological tests for candida were negative.
Discussion. Fourteen cases of prosthetic joint candidosis including our own case have been reported in the world literature. Apart from the
848
case reported here, all were from the USA. As we are unaware of any major differences in the incidence of systemic candidal infections between the USA and other developed regions of the world, there is no apparent reason to explain this bias in cases reported. Recognition of this infectious complication in the UK should alert practitioners to the potential of further cases in European countries. The Candida species most commonly observed in clinical practise is Candida albicans, which occurs in all types of candidosis (2). Compared with candida infection as a whole, a disproportionate number of cases of prosthetic joint infection have been attributed to Candidaparapsilosis. Six of the 14 infections reviewed here including the present case involved Candida parapsilosis (3-7), four Candida albicans (7-10), one Candida (Torulopsis) glabrata (11), and three Candida tropicalis (11, 12). The clinical manifestations of infection is reported to vary for the different candidal species. Lambertus et al. (12) showed that there is a more indolent course in Candida parapsilosis infections of arthroplasties than in cases caused by the more invasi'~e species Candida albicans and Candida tropicalis, which present with unequivocal signs of infection much sooner after surgery. As Candida parapsilosis is a member of the human skin flora, direct inoculation into a joint or haematogenous spread are considered the likely modes of prosthetic joint infection. As an example of direct inoculation, MacGregor et al. (3) describe a case where arthrocentesis led to joint infection. Haematogenous seeding of the joint was the likely route of Candida albicans infection in two cases with long-term indwelling intravenous catheters (8, 9, 13) and also in an intravenous drug abuser with a Candida parapsilosis shoulder joint infection (4). Patients with rheumatoid arthritis are perceived to be more likely to develop fungal infections of their joints than patients with other types of joint disease, although the numbers are too small to establish a definitive association. Of the 14 cases reported here, four received a joint replacement because of rheumatoid disease (3, 9, 8, 10), three because of degenerative changes associated with trauma (12, 4, 7), one because of avascular necrosis (5), and six because of mechanical wearing of the joint (12, 7, 11, 6) including the present case. Diagnosis of candida joint infection depends almost entirely on culture of the organism from joint fluid or tissue and visualisation of it in his-
Eur. J. Clin. Microbiol. Infect. Dis.
tological sections. Unfortunately, the serodiagnosis of eandida infection is rendered difficult due to the confusion generated by false positive reactions seen in uninfected individuals whose immune systems have been exposed to the antigens of these ubiquitous commensals and false negative reactions seen in infected immunodeficient patients (2). However, Murray et al. (14) demonstrated the value of serial serological tests (immunodiffusion and latex agglutination) for confirming the diagnosis and for monitoring the therapeutic response in a case of candida arthritis and bursitis. Similarly, when serology was applied to the management of the present ease to detect candida, it was found to be of assistance for the same reasons. All the cases reviewed here were treated by removing the prosthesis and by giving antifungal therapy, mostly amphotericin supplemented by 5fiuorocytosine or ketoconazole, which was reported to result in eventual cure. As experience accumulates with the newer antifungal agents, such as fluconazole, these may prove to be excellent alternative agents for treatment of serious candidosis. References 1. Giilespie WJ: Infection in total joint replacement. Infectious Diseases Clinics of North America 1990, 4:
465-484, 2. Odds FC: Candida and candidosis. Balli6re Tindatl, London, 1988. 3. MacGregor RR, Schimmer BM, Steinberg ME:
Results of combined amphotericinB/5-fluorocytosine therapy for prosthetickneejoint infectedwith Candida parapsilosis. Journal of Rheumatology 1979, 6: 451455. 4. IJchtman EA: Candida infection of prosthetic shoulder joint. Skeletal Radiology1983, 10: 176--177. 5. Morley DC, Patlerson A: Candida parapsilosis infection of total hip replacement: a ease. Orthopaedic Reviews 1983, 12: 61--64. 6. Younkln S, Evarts CMcC, Ste|gbigel RT: Candida parapsilosis infection of a total hip-joint replacement: successful re-implantation after treatment with amphotericin B and 5-Iluorocytosine.Journal of Bone and Joint Surgery 1984, 66A: 142-143. 7. Lira EVA, Stern PJ: Candida infection after implant arthroplasty: a case report. Journal of Bone and Joint Surgery 1986, 68A: 143-145. 8. lskander MK, Khan M: Candida albicans infectionof a prosthetic knee replacement.Journal of Rheumatology 1988, 15: 1594-1595. 9. Koch AE: Candida albicans infection of a prosthetic knee replacement:a report and reviewof the literature. Journal of Rheumatology 1988, 15: 362-365. 10. Levine M, Rehm SJ, Wilde AH: Infection with Candick: albicans of a total knee arthroplasty.ClinicalOrthopaedics 1988, 226: 235-239.
Vol. 11, 1992
849
11. Goodman JS, Seeibert DG, Reahl GE, Geclder RW: Fungal infection of prosthetic joints: a report of two eases. Journal of Rheumatology 1983, 10: 494-495. 12. Lambertus M, Thordarson D, Goetz MB: Fungal prosthetic arthritis: presentation of two eases and review of the literature. Reviews of Infectious Diseases 1988, 11: 1038-1043. 13. Koch AE: Reply to the editor. Journal of Rheumatology 1988, 15: 1595. I4. Murray HW, Fialk MA, Roberts RB: Candidaarthritis: a manifestation of disseminated candidiasis. American Journal of Medicine 1976, 60:. 587-595.
furthermore be a useful tool for monitoring antiviral drug therapy (3-5). To improve the sensitivity of the test it may be useful to separate the immuno-complexes without damaging the binding capacity of the free antigen. Nishanian et al. (6) have developed a simple method based on low-pH treatment of the test samples before performing a standard enzyme immunoassay (EIA) to detect HIV p24 antigen, which has little or no effect on HIV antigen reactivity. We evaluated the use of acid treatment to improve p24 antigen detection in HIV-1 seropositive subjects, in particular asymptomatic individuals.
Improved Detection of HIV p24 Antigen in Serum after Acid Pretreatment
Materials and Methods. Five hundred and fiftyfour serum samples were used, 509 of which were from HIV seropositive individuals belonging to the following high risk groups: intravenous drug users (n = 478), homosexuals (n = 5), heterosexuals (n = 11), prostitutes (n = 9), and pediatric cases (n = 6). Clinically the status of the subjects was classified as asymptomatic (n = 390), A R C (n = 49) and AIDS (n = 70). Sera from a control group of 45 healthy blood donors not belonging to high risk groups were also tested. Sera were tested for the presence of antibodies to HIV-1 using a commercial E I A (Rapid Elavia Mixt, Pasteur, France) and results confirmed by Western blot analysis (Bioblot, Biokit, Spain). HIV-1 p24 antigen was detected using a commercial kit (HIV p24 Ag Assay, Coulter, USA) following the manufacturer's instructions. Although the sensitivity level of the technique is 5 pg/ml, we preferred to use a cut-off value of 10 pg/ml owing to the fact that the acid treated samples were diluted 1:2. The results were registered quantitatively. Positive samples were confirmed by a neutralization assay using the HIV p24 antigen (Coulter). The specimen was confirmed as positive for HIV-1 p24 antigen if the reduction of absorbance of the aliquot containing confirmatory reagent was at least 50 % compared with that containing the control reagent. The samples were treated according to the method proposed by Nishanian et ai. (6). In this technique two volumes of test serum were mixed with one volume of 0.5 N HCI (pH 2.5-3.0), the mixture incubated for 60 rain at 37 *C and then neutralized with one volume of 0.5 N N a O H to about pH 7. The treated samples were tested parallel to nontreated sera as described in detail above to detect p24 antigen. The controls and standards were not acid pretreated.
M.A. Rodriguez-Iglesias 1., J.R, Alvarez 1, A. Ver~ara 2, M.S. Garcia-Valdivia 1, I. Jesfis 1, J. Mira I
HIV-1 p24 antigen was detected in 554 sera (509 from HIV-1 seropositive individuals and 45 sera from seronegative controls) using a conventional method with acid pretreatment of the sample in order to separate the p24 antigen/anti-p24 antibody immune complexes. In asymptomatic individuals there was a substantial increase in antigen detection (48.2 % vs 8.4 %). Similar results were also observed in A R C (59.1% vs 12.2 %) and AIDS patients (85.7 % vs 37.1%). It can be concluded that the acid treatment improves the sensitivity of conventional techniques to detect HIV-1 p24 antigen.
Following seroconversion the HIV-1 p24 antigen diminishes to usually undetectable levels due to formation of immune complexes with the antibody and remains thus for lengthy periods of time in clinically asymptomatic seropositive individuals. The onset of AIDS related complex (ARC) or AIDS may be associated with detectable antigenemia. A steadily increasing antigen concentration and decrease in anti-p24 antibody titre correlates strongly with progression of the disease and a poor prognosis (1, 2). It may i Department of Microbiology, and 2Infectious Diseases Unit, Puerto Real University Hospital, University of CAdiz, Plaza Fragela s/n, 11003 CAdiz,Spain.
Results and Discussion. In the HIV antibody positive sera from asymptomatic seropositive in-
23. Biaser M J, Smith PF, Kohler PF: Susceptibility of Campylobacter isolates to the bactericidal activity of human serum. Journal of Infectious Diseases 1985,151: 227-235. 24. Endtz HP, van der Meet JWM, Mouton RP: Campylobacter jejuni bacteraemia in hypogammaglobulinaemia: emergence of ciprofloxacin resistance and results of serum bactericidal assays. In: Kaijser B, Falsen E (ed): Proceedings of the Fourth International Workshop on Campylobacter Infections. University of G6teburg, Ktingalve, Sweden, 1988, p. 150--151. 25. Perlman DM, Ampel NM, Schifman RB, Cohn DL, Patton CM, Aguirre ML, Wang WL, Blaser M J: Persistent Campylobacterjejuni infections in patients infected with human immunodefieieney virus (HIV). Annals of Internal Medicine 1988, 108: 540-546. 26. Endtz HP, Mouton RP, van der Reyden T, Ruijs G J, Biever M, van Klingereu B: Fluoroquinolone resistance in Campylobacter spp. isolated from human stools and poultry products. Lancet 1990, 335: 787.
Prosthetic Joint Infection due to Candida parapsilosis in the UK: Case Report and Literature Review J. Paul 1, S.H. White 2, K.M. Nicholls 1, D.W. C r o o k 1.
Candida infection of joint replacements is a rare but increasingly reported phenomenon. A case of Candida parapsilasis prosthetic knee joint infection occurring in the UK is described. Cure followed removal of the prosthesis and treatment, first with a combination of amphotericin and 5fluorocytosine, then ketoconazole.
The overwhelming majority of cases of infection of total joint arthroplasties are ascribed to bacterial pathogens (1). However, there is a growing North American literature referring to cases of prosthetic joint candidosis. A favourable outcome of these infections is seen after removal of the prosthesis and administration of specific antifungal treatment. However, such successful treatment depends on isolating Candida and recognising it as the aetiologic agent. We describe a case in 1Public Health Laboratory, Levels 6/7, John Radcliffe Hospital, Oxford OX3 9DU, UK. 2Nuffield Orthopaedic Centre, Windmill Road, Headington, Oxford OX3 7LD, UK.
847
the UK of infection of a total knee arthroplasty with Candida parapsilosis where our culturebased diagnosis was supported by histological and serological evidence, and where serial serological tests for candida allowed monitoring of the therapeutic response. The literature on candida joint infection, its diagnosis and treatment is reviewed.
Case Report. A 63-year-old man underwent a bicompartmental Oxford knee replacement for severe osteoarthritis. Continued pain led to revision to an Insall Burstein total condylar replacement 20 months later. Loosening of the prosthesis but no evidence of infection was noted at the time of the revision. Three months after revision, the joint became hot and red suggesting infection. Exploration showed inflamed granulations, but routine microbiological cultures were negative. Four months later persistent pain resulted in removal of all foreign material and the knee was fused using compressive biplanar external fixation. Tissue was taken for histological and microbiological investigation. Culture of debrided cancellous bone, joint capsule and periarticular tissue revealed Candida parapsilosis in pure growth. Histological examination of debrided bone showed yeasts and chronic, non-specific inflammatory changes. Sera taken 8 and 105 days before the time of removal of the prosthesis were sent to the Central Public Health Laboratory, London, UK, to test for the presence of candidal precipitins and agglutinins. The results proved to be highly supportive of a diagnosis of infection with Candida parapsilosis. A culture of the organism sent to the Central Public Health Laboratory, London, was found to be sensitive to amphotericin, 5-fluorocytosine and ketaconazole and the identification was confirmed as Candida parapsilosis. One week after removal of the prosthesis, treatment with intravenous amphotericin, 50 mg daily, and 5fluorocytosine, 2500 mg 6-hourly, was started. This regimen was well tolerated and continued for 40 days, and was followed by treatment for 24 days with ketoconazole, 200 mg 12-hourly. The knee became painless and results of serial serological tests for candida indicated a therapeutic response. At follow-up two years after removal of the prosthesis, the patient's fused knee was trouble-free and serological tests for candida were negative.
Discussion. Fourteen cases of prosthetic joint candidosis including our own case have been reported in the world literature. Apart from the
848
case reported here, all were from the USA. As we are unaware of any major differences in the incidence of systemic candidal infections between the USA and other developed regions of the world, there is no apparent reason to explain this bias in cases reported. Recognition of this infectious complication in the UK should alert practitioners to the potential of further cases in European countries. The Candida species most commonly observed in clinical practise is Candida albicans, which occurs in all types of candidosis (2). Compared with candida infection as a whole, a disproportionate number of cases of prosthetic joint infection have been attributed to Candidaparapsilosis. Six of the 14 infections reviewed here including the present case involved Candida parapsilosis (3-7), four Candida albicans (7-10), one Candida (Torulopsis) glabrata (11), and three Candida tropicalis (11, 12). The clinical manifestations of infection is reported to vary for the different candidal species. Lambertus et al. (12) showed that there is a more indolent course in Candida parapsilosis infections of arthroplasties than in cases caused by the more invasi'~e species Candida albicans and Candida tropicalis, which present with unequivocal signs of infection much sooner after surgery. As Candida parapsilosis is a member of the human skin flora, direct inoculation into a joint or haematogenous spread are considered the likely modes of prosthetic joint infection. As an example of direct inoculation, MacGregor et al. (3) describe a case where arthrocentesis led to joint infection. Haematogenous seeding of the joint was the likely route of Candida albicans infection in two cases with long-term indwelling intravenous catheters (8, 9, 13) and also in an intravenous drug abuser with a Candida parapsilosis shoulder joint infection (4). Patients with rheumatoid arthritis are perceived to be more likely to develop fungal infections of their joints than patients with other types of joint disease, although the numbers are too small to establish a definitive association. Of the 14 cases reported here, four received a joint replacement because of rheumatoid disease (3, 9, 8, 10), three because of degenerative changes associated with trauma (12, 4, 7), one because of avascular necrosis (5), and six because of mechanical wearing of the joint (12, 7, 11, 6) including the present case. Diagnosis of candida joint infection depends almost entirely on culture of the organism from joint fluid or tissue and visualisation of it in his-
Eur. J. Clin. Microbiol. Infect. Dis.
tological sections. Unfortunately, the serodiagnosis of eandida infection is rendered difficult due to the confusion generated by false positive reactions seen in uninfected individuals whose immune systems have been exposed to the antigens of these ubiquitous commensals and false negative reactions seen in infected immunodeficient patients (2). However, Murray et al. (14) demonstrated the value of serial serological tests (immunodiffusion and latex agglutination) for confirming the diagnosis and for monitoring the therapeutic response in a case of candida arthritis and bursitis. Similarly, when serology was applied to the management of the present ease to detect candida, it was found to be of assistance for the same reasons. All the cases reviewed here were treated by removing the prosthesis and by giving antifungal therapy, mostly amphotericin supplemented by 5fiuorocytosine or ketoconazole, which was reported to result in eventual cure. As experience accumulates with the newer antifungal agents, such as fluconazole, these may prove to be excellent alternative agents for treatment of serious candidosis. References 1. Giilespie WJ: Infection in total joint replacement. Infectious Diseases Clinics of North America 1990, 4:
465-484, 2. Odds FC: Candida and candidosis. Balli6re Tindatl, London, 1988. 3. MacGregor RR, Schimmer BM, Steinberg ME:
Results of combined amphotericinB/5-fluorocytosine therapy for prosthetickneejoint infectedwith Candida parapsilosis. Journal of Rheumatology 1979, 6: 451455. 4. IJchtman EA: Candida infection of prosthetic shoulder joint. Skeletal Radiology1983, 10: 176--177. 5. Morley DC, Patlerson A: Candida parapsilosis infection of total hip replacement: a ease. Orthopaedic Reviews 1983, 12: 61--64. 6. Younkln S, Evarts CMcC, Ste|gbigel RT: Candida parapsilosis infection of a total hip-joint replacement: successful re-implantation after treatment with amphotericin B and 5-Iluorocytosine.Journal of Bone and Joint Surgery 1984, 66A: 142-143. 7. Lira EVA, Stern PJ: Candida infection after implant arthroplasty: a case report. Journal of Bone and Joint Surgery 1986, 68A: 143-145. 8. lskander MK, Khan M: Candida albicans infectionof a prosthetic knee replacement.Journal of Rheumatology 1988, 15: 1594-1595. 9. Koch AE: Candida albicans infection of a prosthetic knee replacement:a report and reviewof the literature. Journal of Rheumatology 1988, 15: 362-365. 10. Levine M, Rehm SJ, Wilde AH: Infection with Candick: albicans of a total knee arthroplasty.ClinicalOrthopaedics 1988, 226: 235-239.
Vol. 11, 1992
849
11. Goodman JS, Seeibert DG, Reahl GE, Geclder RW: Fungal infection of prosthetic joints: a report of two eases. Journal of Rheumatology 1983, 10: 494-495. 12. Lambertus M, Thordarson D, Goetz MB: Fungal prosthetic arthritis: presentation of two eases and review of the literature. Reviews of Infectious Diseases 1988, 11: 1038-1043. 13. Koch AE: Reply to the editor. Journal of Rheumatology 1988, 15: 1595. I4. Murray HW, Fialk MA, Roberts RB: Candidaarthritis: a manifestation of disseminated candidiasis. American Journal of Medicine 1976, 60:. 587-595.
furthermore be a useful tool for monitoring antiviral drug therapy (3-5). To improve the sensitivity of the test it may be useful to separate the immuno-complexes without damaging the binding capacity of the free antigen. Nishanian et al. (6) have developed a simple method based on low-pH treatment of the test samples before performing a standard enzyme immunoassay (EIA) to detect HIV p24 antigen, which has little or no effect on HIV antigen reactivity. We evaluated the use of acid treatment to improve p24 antigen detection in HIV-1 seropositive subjects, in particular asymptomatic individuals.
Improved Detection of HIV p24 Antigen in Serum after Acid Pretreatment
Materials and Methods. Five hundred and fiftyfour serum samples were used, 509 of which were from HIV seropositive individuals belonging to the following high risk groups: intravenous drug users (n = 478), homosexuals (n = 5), heterosexuals (n = 11), prostitutes (n = 9), and pediatric cases (n = 6). Clinically the status of the subjects was classified as asymptomatic (n = 390), A R C (n = 49) and AIDS (n = 70). Sera from a control group of 45 healthy blood donors not belonging to high risk groups were also tested. Sera were tested for the presence of antibodies to HIV-1 using a commercial E I A (Rapid Elavia Mixt, Pasteur, France) and results confirmed by Western blot analysis (Bioblot, Biokit, Spain). HIV-1 p24 antigen was detected using a commercial kit (HIV p24 Ag Assay, Coulter, USA) following the manufacturer's instructions. Although the sensitivity level of the technique is 5 pg/ml, we preferred to use a cut-off value of 10 pg/ml owing to the fact that the acid treated samples were diluted 1:2. The results were registered quantitatively. Positive samples were confirmed by a neutralization assay using the HIV p24 antigen (Coulter). The specimen was confirmed as positive for HIV-1 p24 antigen if the reduction of absorbance of the aliquot containing confirmatory reagent was at least 50 % compared with that containing the control reagent. The samples were treated according to the method proposed by Nishanian et ai. (6). In this technique two volumes of test serum were mixed with one volume of 0.5 N HCI (pH 2.5-3.0), the mixture incubated for 60 rain at 37 *C and then neutralized with one volume of 0.5 N N a O H to about pH 7. The treated samples were tested parallel to nontreated sera as described in detail above to detect p24 antigen. The controls and standards were not acid pretreated.
M.A. Rodriguez-Iglesias 1., J.R, Alvarez 1, A. Ver~ara 2, M.S. Garcia-Valdivia 1, I. Jesfis 1, J. Mira I
HIV-1 p24 antigen was detected in 554 sera (509 from HIV-1 seropositive individuals and 45 sera from seronegative controls) using a conventional method with acid pretreatment of the sample in order to separate the p24 antigen/anti-p24 antibody immune complexes. In asymptomatic individuals there was a substantial increase in antigen detection (48.2 % vs 8.4 %). Similar results were also observed in A R C (59.1% vs 12.2 %) and AIDS patients (85.7 % vs 37.1%). It can be concluded that the acid treatment improves the sensitivity of conventional techniques to detect HIV-1 p24 antigen.
Following seroconversion the HIV-1 p24 antigen diminishes to usually undetectable levels due to formation of immune complexes with the antibody and remains thus for lengthy periods of time in clinically asymptomatic seropositive individuals. The onset of AIDS related complex (ARC) or AIDS may be associated with detectable antigenemia. A steadily increasing antigen concentration and decrease in anti-p24 antibody titre correlates strongly with progression of the disease and a poor prognosis (1, 2). It may i Department of Microbiology, and 2Infectious Diseases Unit, Puerto Real University Hospital, University of CAdiz, Plaza Fragela s/n, 11003 CAdiz,Spain.
Results and Discussion. In the HIV antibody positive sera from asymptomatic seropositive in-
