THE JOURNAL OF INFECTIOUS DISEASES · VOL. 160. NO. 6 · DECEMBER 1989
© 1989 by The University of Chicago. All rights reserved. 0022-I899/89/6006-0020$01.00
Absence of Infectious HIV-1 in the Urine of Seropositive Viremic Subjects
Paul R. Skolnik, Barry R. Kosloff, Larry J. Bechtel, Kenneth R. Huskins, Theresa Flynn, Nancy Karthas, Kenneth Mcintosh, and Martin S. Hirsch
From the Division of Geographic Medicine and Infectious Diseases, Department of Medicine, Tufts University, New England Medical Center, and the Infectious Disease Units, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, and Department of Pediatrics, Children's Hospital and Harvard Medical School, Boston; and the Medical Products Department, E. I. du Font de Nemours, Wilmington, Delaware
Urine and peripheral blood samples from 48 human immunodeficiency virus type 1 (HIV-1) seropositive individuals (38 adults and 10 children) were evaluated for the presence of HIV-1 by cocultivation and for HIV-1 p24 antigen by ELISA. None of the urine samples contained replication-competent HIV-1; 41 (85%) of 48 simultaneously obtained peripheral blood mononuclear cell samples contained replication-competent HIV-1. None of 26 urine samples available for analysis contained HIV-1 p24 antigen as determined by ELISA; 12 (34%) of 35 simultaneously obtained peripheral blood samples had detectable serum HIV-1 p24 antigen. Two of the individuals studied had HIV nephropathy, three had pyuria, and five had microscopic hematuria. Culture sensitivity was maximal when mycostatin (and not amphotericin B) was used as an antifungal agent. Our findings indicate that urine from HIV-1-seropositive individuals is unlikely to contain infectious HIV-1. This would imply that the risk of transmission of HIV-1 by urine is low to nonexistent.
Human immunodeficiency virus (HIV), the etiologic agent of the acquired immunodeficiency syndrome (AIDS), has been isolated in tissue culture from several body fluids including blood, semen, tears, saliva, cerebrospinal fluid, cervical secretions, and breast milk. With the rapidly rising incidence of HIV infection in infants and young children, the question of whether HIV can be found in urine becomes increasingly important because it relates to the potential for transmission of this virus to health care workers. We examined urine from HIV-1-seropositive adults and children for the presence of replication-competent HIV-1 and the HIV-1 p24 antigen.
Study subjects and sample acquisition. Peripheral blood and urine samples were collected from individuals known to be HIV-1-seropositive by ELISA assays confirmed by Western blotting. Specimens were collected from 51 individuals. Three urine samples were contaminated by fungus and were excluded from the study, leaving samples from 48 subjects for evaluation. Ten subjects were <6 y of age (mean, 2.2 y; range, 0.33-5.8 y); 38 were greater than or equal to 18 y of age (mean, 38 y; range, 21-61 y).
All pediatric patients had Centers for Disease Control (CDC) class P-2 infection. Of the adult patients, 17 had AIDS (CDC group IVB-E); 5 had AIDS-related complex (CDC group IVA); 3 had generalized lymphadenopathy (CDC group III); and 13 were asymptomatic (CDC group II). Two of the adults with CDC group IVE disease had nephrotic syndrome consistent with AIDS nephropathy, although renal biopsy was not done. Ten adults were receiving antiretroviral therapy with either zidovudine or dideoxycytidine at the time of study, 23 were not receiving therapy, and 5 were enrolled in placebo-controlled trials ofzidovudine therapy (which remain blinded at this time). Three pediatric patients were receiving intravenous gammaglobulin; the other children were not. No child was receiving zidovudine or other antiretroviral agents.
Culture of peripheral blood for HIV-1. Mononuclear cells were isolated by ficoll-hypaque density centrifugation from 50 ml of heparinized peripheral blood (<10 units/ml) from adults and 3-5 ml from children. Cells (10 x 106) were cocultivated with 5 x 106 peripheral blood mononuclear cells (PBMC) from HIV-1-seronegative control donors in RPMI-1640 medium containing 2 µg/ml of polybrene (Sigma, St. Louis) and 10% interleukin-2 (IL-2; Electronucleonics, Silver Springs, Md), supplemented with penicillin (250 units/ml), streptomycin (250 µg/ml), hepes buffer (0.01 M), L-glutamine (2 mM), and 20% heat-inactivated fetal calf serum (GIBCO, Grand Island, NY). The PBMC from the HIV-1-seronegative control donors had been previously harvested by ficoll-hypaque centrifugation and stimulated for 3-4 d with 10% IL-2 and 10 µg/ml of phytohemagglutinin-P (PHA-P; Difco, Detroit) in RPMI-1640 medium supplemented as above.
Cultures were maintained with twice-weekly addition of fresh media and weekly addition of fresh prestimulated PBMC. Cultures were tested twice weekly for the HIV-1 p24 antigen by ELISA (NEN-DuPont, Boston). Cultures were considered positive if two consecutive determinations of optical density by ELISA were more than twice the mean of the negative control wells, with the second positive value greater than the first. The cultures were maintained for 4 w.
Analysis and culture of urine for HIV-1. An aliquot of urine was analyzed by dipstick for microscopic hematuria and leukocyte esterase activity. The urinary sediment was examined microscopically for cellular material and urinary casts. When sufficient urine was available, an additional aliquot was frozen at -70°C for subsequent p24 antigen determination. The remaining urine from each subject was processed in three ways for culture in an effort to maximize HIV-1 recovery.
1. Urine (1 ml) was treated with 0.1 ml of a penicillin and streptomycin solution in phosphate-buffered saline (10,000 units/ml and 10,000 µg/ml, respectively) and 0.1 ml of either mycostatin (1,000 units/ml) or amphotericin B (25 µg/ml) for 10 min at room temperature after adjusting the pH to 7.0 with Na2HC03. This was then cultured with 10 x 106 PBMC from an HIV-1-seronegative individual (prestimulated with PHA-P and IL-2 as above) in RPMI-1640 medium with 2 µg/ml of polybrene and l0% IL-2 and supplemented as in the peripheral blood culture protocol. The cultures were maintained in 25-cm2 flasks (Falcon, Oxnard, Calif) in a total volume of 10 ml.
2. Urine (36 ml) was spun in an ultracentrifuge (Beckman Instruments, Palo Alto, Calif; model L8-70M, rotor SW28) at 27,000 rpm for 1.5 h at 4°C. The pellet was resuspended in 2.5 ml of RPMI-1640 medium with 2 µg/ml of polybrene, l0% IL-2, 0.1 ml of the penicillin-streptomycin solution, and 0.1 ml of mycostatin or amphotericin B as in method 1. This mixture was incubated at room temperature for 10 min and then cultured with PHA-P-stimulated PBMC as above.
3. Urine (36 ml) was ultracentrifuged and processed as in method 2, but 5'-iodo-2'-deoxyuridine (IUDR) was added to the culture at a final concentration of 25 µg/ml in an attempt to activate potentially latent HIV-1.
Occasionally, pediatric specimens contained <75 ml of urine; in these cases, a minimum of 20 ml was processed by ultracentrifugation or the ultracentrifuged culture with IUDR was omitted. The cultures were incubated at 37°C with 5% CO2. Cultures were maintained for 4 w with twice-weekly addition of fresh media and weekly addition of PHA-P stimulated PBMC; they were assayed twice weekly for HIV-1 p24 antigen by ELISA with the criteria for positivity as described above.
Direct ELISA for HIV-1 p24 antigen in urine and serum. Urine samples were passed through a 0.45-µm filter (Millipore, Bedford, Mass) to reduce non-specific background signal (data not shown) and then processed according to the manufacturer's protocol (NEN-DuPont). Regression analysis of the optical densities was performed using antigen standards diluted in RPMI-1640 or in control urine from an HIV-1 seronegative donor. Twenty-six urine specimens (2 pediatric and 24 adult) were available for direct ELISA analysis. HIV-1 p24 antigen in serum was assayed according to the manufacturer's protocol (NEN-DuPont).
Determination of urine culture sensitivity (spiking experiments). Urine was obtained from an HIV-1-seronegative donor. The HTLV-IIIB strain of HIV-1, which had been previously titered by a virus yield assay, was added at various concentrations ranging from 2.78 x 10-2 tissue culture infectious dose (TCID)50/ml to 1.38 x 102 TCID50/ml of urine. The samples were then processed as above with ultracentrifugation and culture without IUDR. Initial treatment of the urine was either no antifungal agent or mycostatin or amphotericin B at the concentrations used in the patient urine cultures.
None of the 48 urine samples contained HIV-1 as determined by culture with PHA-stimulated PBMC. In addition, none of 26 urine samples (2 pediatric and 24 adult) assayed directly by ELISA for HIV-1 p24 antigen was positive. In contrast, HIV-1 was isolated by cocultivation of concurrently obtained peripheral blood specimens from 41 (85%) of the 48 seropositive subjects. This included peripheral blood from all 10 children and 31 (82%) of 38 adults. HIV-1 p24 antigen was detected in the serum of 12 (34%) of 35 adult subjects; pediatric serum samples were not tested for HIV-1 p24 antigen.
Urinalysis was performed on all 10 pediatric specimens. Three patients had pyuria (1-10 white blood cells [WBC]/high power field [HPF]) and two had hematuria (2-10 red blood cells [RBC1/HPF). One of the children with pyuria also had 2-4 WBC casts/HPF. Urinalysis was done on 30 of the 38 specimens from adults. Two had nephrotic syndrome with proteinuria (AIDS-associated nephropathy). One of these had slight microscopic hematuria as graded by dipstick analysis and 3-5 RBC/HPF in the urinary sediment. The other had moderate microscopic hematuria and 20-50 RBC/HPF in the urinary sediment. Neither had pyuria. Three other adults had 1-2 WBC/HPF with negative leukocyte esterase activity and one had 5-6 RBC/HPF with slight microscopic hematuria on dipstick analysis.
The sensitivity of the urine culture protocol for HIV-1 recovery was determined by adding known quantities of the HLTV-IIIB strain of HIV-1 to urine from HIV-1-seronegative donors. The sensitivity of this assay was dependent on whether mycostatin or amphotercin B was used for culture. We were able to detect 2.78 x 10-1 TCID50/ml when mycostatin was used as the antifungal agent and 2.48 x 101 TCID50/ml when amphotercin B was the antifungal agent (figure 1).
Overall, mycostatin was used in 3 pediatric urine cultures and 25 adult urine cultures, whereas amphotericin B was used in 7 pediatric cultures and 12 adult cultures. Urine from the two pediatric patients with hematuria was cultured with mycostatin, as was the pediatric sample with 5-10 WBC/HPF. Samples from the two other pediatric patients with pyuria, including the subject with WBC casts, were cultured with amphotericin B. The urine samples from the two patients with nephrotic syndrome were cultured with amphotericin B. One of three urine samples from the adult patients with 1-2 WBC/HPF was cultured with mycostatin; the other two were cultured with amphotericin B. The urine from the adult with 5-6 RBC/HPF was cultured with amphotericin B.
Transmission of HIV-1 infection is known to occur through sexual intercourse (both heterosexual and homosexual), via blood and blood products or needles contaminated with HIV-1, prenatally (presumably by transplacental passage of virus), and perinatally during birth or perhaps breast feeding. Transmission of infection is likely dependent on several factors including the titer of infectious virions within the body fluid, the degree of parenteral or mucous membrane exposure to the body fluid, and perhaps as-yet-undefined host resistance factors to infection. Thus, transfusion of contaminated blood products, which contain large amounts of infectious HIV-1, usually results in transmission of infection to the recipient.
In contrast, there are no well documented cases of transmission of HIV-1 infection by saliva or tears, although small amounts of HIV-1 have been isolated from these body fluids. We were unable to detect HIV-1 (or the HIV-1 p24 antigen) in urine, even though most of these patients were viremic with HIV-1 at the time of culture. We are aware of a single report of isolation of HIV-1 from urine. One of five urine specimens was reported to be positive for virus at "low titer," although details of the source of this specimen, the culture and assay methods used for viral detection, and whether it was possible to passage the viral isolate were not provided. Taken together, these data suggest that the likelihood of HIV-1 transmission by exposure to urine from HIV-1-seropositive individuals is low to nonexistent.
An interesting observation in this study was the apparent antiretroviral activity of amphotericin B. Other investigators have reported similar findings. These observations led us to substitute mycostatin for amphotericin B in our urine cultures. Our data indicate that mycostatin is the preferred antifungal agent when attempting to isolate HIV-1 in potentially contaminated cultures.
It should be noted that 40% of our urine cultures for HIV-1 included amphotericin B. This would be expected to decrease the sensitivity of detection of HIV-1 in these cultures as indicated by experiments using known quantities of exogenously added HIV-1. In addition, it is not known what effect, if any, the presence of antiretroviral agents or their metabolites would have on the culture recovery of HIV-1. Of our study subjects, 21% were known to be receiving antiretroviral therapy at the time of culture.
The lack of detectable HIV-1 p24 antigen in urine from HIV-1-seropositive individuals, as assayed by ELISA, may represent the absence of antigen or, alternatively, the presence of antigen-antibody complexes. Antibody to HIV-1, without detectable HIV-1 p24 antigen, has recently been demonstrated in urine. It would also be interesting to study additional patients with HIV nephropathy, especially those with detectable serum HIV-1 p24 antigen, for the presence of p24 antigen in urine. These individuals typically have proteinuria and might therefore be more likely to have antigen or antigen-antibody complexes in their urine.
Our findings have relevance to appropriate precautions for health care workers, particularly in pediatric and day care settings where urine contamination is frequent. Although the absence of demonstrated infectious HIV-1 in urine specimens is encouraging, we would still recommend that "universal precautions" be practiced for all body fluids, especially for any body fluid (including urine) that may be contaminated with blood.
We thank R. Byington for advice regarding culture techniques for HIV, D. Merrill for technical assistance, A. Francke and L. Micklay for help in obtaining urine samples, and J. Donoghue for secretarial assistance.
Received for publication 14 February 1989 and in revised form 28 June 1989.
This work was supported by grants from the Life and Health Insurance Medical Research Fund and the Carter-Wallace Foundation (to P. R. S.) and by grants CA-35020 and CA-12464 from the National Cancer Institute (to M. S. H.).
Please address requests for reprints to Dr. P. R. Skolnik, Division of Geographic Medicine and Infectious Diseases, New England Medical Center, Box 67, 750 Washington Street, Boston, MA 02111.
Figure 1. Recovery of HIV-1, as determined by ELISA for the HIV-1 p24 antigen, after addition of HIV-1 strain HTLV-IIIB to urine from an HIV-1 seronegative donor. Urine was processed for culture with: A, no antifungal agent; B, mycostatin (final concentration, 10 units/ml); or C, amphotericin B (final concentration, 0.25 µg/ml). HIV-1 was added at concentrations as indicated in key.
Home Page Events Links Health Site Index