Secondary Logo

Journal Logo


Detection of HIV-1 nucleic acid and HIV-1 antibodies in needles and syringes used for non-intravenous injection

Rich, Josiah D.1,4; Dickinson, Brian P.1; Carney, John M.2; Fisher, Alvan3; Heimer, Robert2

Author Information



In the past decade, the percentage of AIDS cases by exposure category, has decreased amongst homosexual men and has increased amongst injecting drug users (IDU) [1]. The primary factor associated with HIV infection amongst IDU is the practice of sharing injection equipment, specifically needles and syringes (here-after referred to as syringes) [2]. Residual HIV-infected blood in shared syringes is a source of HIV infection.

This residual blood is introduced into the lumen of the syringe during intravenous administration through two practices known as ‘registering’ and ‘booting’. Registering means that once a needle is inserted, the injector will draw back on the plunger of the syringe and examine for the presence of blood in the syringe barrel to ensure that the needle has been properly placed in a vein [2]. Booting is performed after the drug has been administered: while the needle is still in the vein, the injector aspirates blood in the syringe to fill the barrel of the syringe and then reinjects the blood [2]. These processes contaminate the needle, barrel, and plunger of the syringe [2].

In the United States, laws regulating the sale, distribution, and possession of syringes prevent access to clean sterile syringes for those who inject illicit compounds [3]. When injection equipment is scarce, it is shared. Shared injection equipment has resulted in the transmission of HIV amongst intravenous IDU and has been implicated in anecdotal reports of HIV transmission amongst intramuscular IDU [4–6].

Subcutaneous and intramuscular injection of illicit drugs is a frequent occurrence. IDU whose veins have collapsed from long-term injecting drug use may resort to subcutaneous administration, ‘skin-popping’, of their drugs. Skin popping typically involves the insertion of a thin 27 or 28-gauge 0.5 inch insulin needle subcutaneously, followed by injection of the syringe contents. Anabolic steroid users often administer their drugs intramuscularly [7]. Intramuscular administration of steroids requires a longer needle (1–1.5 inch) to penetrate the muscle and a larger 22-gauge needle to accommodate the viscous oil-based compounds. Typically, during intramuscular injection the needle is inserted into the upper-outer quadrant of the gluteus maximus. Once inserted, the plunger of the syringe is drawn back and the syringe barrel is examined for the absence of blood to ensure that the needle has not been placed in a vein, a process known as ‘flagging’ [8]. Therefore, ideal subcutaneous and intramuscular administration of illicit drugs occurs without introduction of blood into the needle and barrel, or contamination of the plunger.

The volumes of blood transferred during intravenous drug injection and the presence of HIV DNA and HIV antibodies in syringes have been documented [9–11]. No studies have obtained similar information on syringes used for intramuscular or subcutaneous injection. We report an analysis of HIV antibodies, HIV proviral DNA, HIV RNA, and human DNA on syringes that have been used for intramuscular or sub-cutaneous injection in persons known to have HIV infection.

Materials and methods

Data collection

Disposable syringes that had been used for medically indicated intramuscular or subcutaneous injections in persons with known HIV infection [documented positive HIV, enzyme immunoassay (EIA) and confirmatory Western blot] were collected. Antibodies to HIV, HIV proviral or human DNA, and HIV RNA from these syringes were assayed. The study protocol was approved by the Institutional Review Board of The Miriam Hospital (Providence, Rhode Island).

Laboratory methods

To facilitate safe handling, a procedure approved by the Yale Office on Environmental Health and Safety was followed. Syringes were uncapped and placed directly into racked 1.5 ml microcentrifuge tubes. Residual material was removed from syringes by repeated flushing with 50 µl of a buffer containing non-ionic detergents (50 mM KCl, 10 mM Tris-HCl pH 8.3, 2.5 mM MgCl2, 0.5% vol/vol Tween-20, 0.5% vol/vol Nonidet P-40) that inactivate HIV. Extracts from 260 syringes were stored at −20°C until EIA and for the preparation of DNA for PCR. An additional 80 syringes were stored at −80°C until RNA was prepared for reverse transcriptase (RT)-PCR.

EIA were run using the Abbott viral based system modified as previously described [12]. The EIA assay uses a viral lysate containing all of the HIV viral proteins coated onto beads. The optical density cut-off ratios for a positive control were four times the background. In effect, the only change from the standard procedure was the omission of the initial dilution step. Instead, an aliquot of 10 µl of the syringe extract combined with 200 µl of the diluent supplied with the test kits was used for testing for anti-HIV antibodies. The limits of detection for this assay had been previously determined using duplicate serial dilution of 20 samples. The mean limit was estimated to be 0.84 nl blood (SEM range, 0.013–5.33 nl) [13].

DNA for PCR was recovered from the remaining volume of each syringe extract using protease digestion, organic extraction, and ethanol precipitation as previously described [13]. Precipitated DNA was dissolved in 20 µl Tris–EDTA (TE-80) buffer (10 mM Tris HCl pH 8.0, 0.1 mM EDTA) and amplified using a two-step procedure. First, HIV and β-globin sequences were co-amplified for 15 cycles in a 50 µl reaction containing 40 mM KCl, 10 mM Tris-HCl pH 8.3, 2.5 mM MgCl2, 0.01% weight/vol gelatin, 0.2 mM of each deoxynucleoside triphosphate, 1 U Taq DNA polymerase, 20 pmol of each globin primer pair and 30 pmol of each HIV-1 gag primer pair. An aliquot of 5 µl from the first round of PCR was then added to a 50 µl reaction containing 25 pmol of each nested HIV- 1 primer pair and amplified for 35 cycles. The HIV-1- specific nested product (191 base pairs) and the β-globin-specific product (395 base pairs) were separated by electrophoresis on 2.5% agarose gels and visualized with ethidium bromide.

RNA for RT-PCR was recovered from a second set of syringe extracts using a commercially available affinity chromatographic method (QIAamp blood kit, Qiagen Inc., Chatsworth, California, USA). RNA was eluted from the affinity column with 50 µl TE-80 buffer and amplified using a three-step procedure. First, cDNA of HIV and β-globin sequences were synthesized using specific primers and 5 U avian myeloblastosis virus RT in a 100 µl reaction containing the PCR buffer used above supplemented with 10 U placental ribonuclease inhibitor. Second, cDNA were directly amplified for 15 cycles. Finally, using nested PCR, aliquots of 5 µl from the first round of PCR were re-amplified and products were identified by electrophoresis as described above.

For the purposes of extrapolating the experimental results to the amounts of human blood and DNA, proviral DNA, and viral RNA detected in syringes, the limits of detection of each of the techniques were calculated using serial dilution. The limits of detection for HIV-1 proviral DNA and β-globin were determined using DNA from white blood cells of two HIV-infected individuals. The DNA was quantified spectrophotometrically and serially diluted from 20 000 cell equivalents down to 8 cell equivalents for use in PCR. DNA was amplified through two rounds of PCR and products were analyzed by gel electrophoresis. The limits of our ability to detect HIV RNA were measured using RNA extracted from stocks of two clinical isolates of HIV-1 that had been titrated in the laboratory using the protocol of the AIDS Clinical Trials Group [14].

Statistical analysis

Chi-squared statistical analyses were conducted to compare differences between antibodies recovered from intramuscular and subcutaneous syringes. The 95% confidence intervals were calculated by multiplying the SE by the standard normal deviation for α = 0.05. χ2 test for trend was performed using EpiInfo [Version 6.03, Centers for Disease Control and Prevention (CDC), Atlanta, Georgia, USA] to compare recovery of HIV RNA from syringes with HIV plasma viral load in source patients.


Of the 260 syringes initially analyzed for antibodies to HIV, only 16 syringes (6.2%) were positive by the EIA. Analyzing the syringes as a function of the gauge and length of their cannulae, it was determined that three (3.9%) out of 77 of the 25-gauge (0.5 inch) syringes, six (12%) out of 50 of the 29-gauge (0.5 inch) insulin syringes, one (2.1%) out of 48 of the 22-gauge (1 inch) syringes, and six (28.6%) out of 21 of the 25-gauge (0.625 inch) syringes had detectable HIV antibodies (Table 1). Prevalence of HIV antibodies in syringes used to inject medications subcutaneously was 7.9% (six out of 76) and was not significantly different from those used to inject medications intramuscularly (5.4%, 10 out of 184; P = 0.48; Table 1).

Table 1
Table 1:
. Detection of HIV antibodies among syringes used for medically indicated intramuscular and subcutaneous injection in HIV-infected individuals.

In contrast to the EIA findings, none of the syringes used for intramuscular or subcutaneous injection contained detectable β-globin or HIV proviral DNA. Serial dilution revealed that β-globin-specific products were detected in dilutions containing as few as 32 cell equivalents; HIV-1 specific products were detected only if many more cell equivalents of DNA were included in the nested PCR (Fig. 1). For one individual with advanced HIV disease, the 191 base-pair HIV-specific PCR product was detected when the DNA input was 4000 or 20 000 cell equivalents but not when the input was 800 cell equivalents or less. For the second individual, with less advanced disease, the 191 base-pair HIV-specific PCR product was detected only with 20 000 cell equivalents.

Fig. 1
Fig. 1:
. Limit of detection of human and HIV proviral DNA using nested PCR. DNA from the peripheral blood mononuclear cells of two HIV-1-infected patients was isolated, quantified spectrophotometrically, serially diluted, amplified by nested PCR using primers for β-globin in the first round, and primers for the HIV gag gene in both the first and second rounds. Lanes 1 and 7 show the PCR products obtained from the equivalent of 20 000 genomes; lanes 2 and 8 the equivalent of 4000 genomes; lanes 3 and 9 the equivalent of 800 genomes; lanes 4 and 10 the equivalent of 160 genomes; lanes 5 and 11 the equivalent of 32 genomes; and lanes 6 and 12 the equivalent of eight genomes. Lane M contains 100 base-pair ladder DNA.

In a second sample of 80 syringes, none had amplifiable β-globin sequences, but HIV RNA was detected in three (3.8%) used for intramuscular or subcutaneous injection. Syringes were divided into three categories: those used for injection of individuals with more than 105 HIV RNA copies/ml plasma, those with between 400 and 105 HIV RNA copies/ml, and those with less than 400 HIV RNA copies/ml. One of the three positive syringes had been used to inject individuals with titers over 105 copies/ml; two came from syringes that had been used to inject individuals with titers between 400 and 105 copies/ml, and no syringes were positive that had been used to inject individuals with less than 400 copies/ml of HIV RNA. A χ2 analysis comparing HIV RNA detection with plasma viral load revealed a non-significant trend (χ2 = 2.2, P = 0.14). Serial dilution of the RNA samples revealed that the nested RT-PCR assay achieved single copy sensitivity (Fig. 2).

Fig. 2
Fig. 2:
. Nested RT-PCR detection of HIV-1 RNA. Viral RNA was isolated from a clinical isolate of HIV-1 with a titer of 5.68 log10 IUPM/ml and serially diluted so that 197, 39.4, 7.9, 1.58, 0.32, 0.063, and 0.013 IUPM (infectious units per ml) were used for nested RT-PCR. Products from these seven reactions are shown in lanes 1–7. Cloned HIV-1 DNA diluted to 5 copies (lane A) served as a positive control. A 100 base-pair ladder was used as DNA size marker (lane M). The figure was cropped to include the bands of 200–600 base pairs in lane M.


We have examined the likelihood of HIV transmission associated with percutaneous exposure from a syringe used to administer medications intramuscularly or subcutaneously by testing for the presence of HIV or HIV antibodies in syringes that had been used for injection into known HIV-infected individuals. This is an important issue for situations in which needles may be shared, such as subcutaneous skin-popping or with intramuscular anabolic/androgenic steroid injection. In addition, health-care workers suffer occupational percutaneous exposure from syringes used for intramuscular or subcutaneous injection of HIV-infected patients. Data on the risks of such exposure are an important factor in decisions involving post-exposure prophylaxis.

We found that a small percentage of the syringes used for intramuscular or subcutaneous injections of HIV-infected individuals contained antibodies to HIV and HIV-1 virion RNA. Based on previous estimates for the mean limit of sensitivity for the antibody detection EIA [13], it appears that 6.2% of the syringes contained at least 0.1 nl blood. The absence of amplified human genomic or HIV-1-specific DNA suggests that even these syringes contained very little blood. Determination of the limits of the sensitivity of nested PCR demonstrated that we could detect β-globin-specific DNA if there were more than 32 cell equivalents. Thus, no syringe was found to have more than this number of cells. We conducted a calculation on the upper limit of blood detection. The mean number of nucleated cells in a microliter of blood is approximately 7000 in a healthy individual and less in individuals infected with HIV-1. Dividing 32 cell equivalents by 7000 cells/µl yields 4.6 nl as the mean upper limit for the volume of blood within these syringes. This volume is much less than the residual blood that remains inside a syringe after intravenous drug injection [10].

Our laboratory data suggests that infections can occur as a result of percutaneous exposure from needles that have been used to inject medications intramuscularly or subcutaneously, although they are less likely than those that were used to inject substances intravenously.

We detected HIV-1 viral RNA in syringes at approximately half the frequency with which we detected antibodies to HIV-1. It is unlikely that the viral RNA came from blood, since two of the positive syringes were used to inject individuals with titers of less than 105 copies/ml blood. For these syringes, at least 10 µl blood would be needed to yield a single virion, but the results from the PCR test for human β-globin revealed that no syringe contained more than 4.6 nl blood. Previous work has shown that infected individuals harbor between 1 and 36 copies of HIV proviral DNA per 100 follicular dendritic cells [15]. These cells form a minor component of the skin's immune system [16] and are unlikely to be aspirated into syringes during the flagging that precedes intramuscular or subcutaneous injection. This hypothesis is supported by the lack of detection of β-globin or HIV proviral DNA in any of the syringes tested. Therefore, we believe the syringes that tested positive contained viral RNA that came from virions in the interstitial fluid of the skin and were probably released by infected follicular dendritic cells of the skin.

The overall risk of HIV transmission from a percutaneous needlestick injury in a health-care setting is 0.3% [17]. This estimated risk includes exposures from needles that have been used for many types of injection, including intravenous injection. Epidemiologic data regarding accidental needlestick HIV transmission from needles used only to administer medications intramuscularly or subcutaneously are scarce. Through December 1996, the CDC documented 45 occupational acquisitions of HIV infection, as a result of percutaneous exposure [18]. Two of these HIV infections may have been from needles that were used for an intramuscular or subcutaneous injection (R. Metler, personal communication, May 1997). In addition, there are few case reports of HIV transmission associated with anabolic/androgenic steroid injection [4–6]. In either case, these data are limited by the absence of information on the number of percutaneous exposures. For accidental needlesticks, there are no estimates for the total number of individuals exposed to non-bleeding needlesticks. For steroid injections, we have neither reliable estimates for the number of injections nor data on the frequency of injections with the potential to transmit HIV.

In a case–control study of 31 health-care workers infected following percutaneous exposure, four risk factors were identified that were strongly associated with transmission: deep injury, the presence of visible blood, a procedure involving venous or arterial puncture, and terminal illness of the source patient [19]. Two of these risk factors, visible blood with an adjusted odds ratio of 5.2 and placement of the needle directly in a vein or an artery of the patient with an adjusted odds ratio of 5.1, suggest a lower likelihood of HIV transmission from a intramuscular or subcutaneous syringe.

The small fraction of syringes containing viral RNA and the fact that needlesticks are unlikely to transfer the entire contents of the syringe support the conclusion that a needlestick with a syringe used for intramuscular or subcutaneous injection is much less likely to result in the transmission of HIV-1 than syringes used for intravenous injection. On the other hand, sharing of syringes among steroid injectors or skin-poppers carries a greater risk, primarily because the entire contents of the contaminated syringe may be transferred.

We found HIV antibodies and HIV RNA in syringes used for intramuscular or subcutaneous injection. We estimated the volume of blood in these syringes. This volume is at least three orders of magnitude less than that found in syringes used during intravenous injection. The biological analysis of syringes used for intramuscular or subcutaneous injection is consistent with the scarce epidemiological data. The combination of the results presented in this report and the case findings leads to the conclusion that the risk of transmitting HIV via shared syringes used for intramuscular or sub-cutaneous injection may be low but is not zero.


1. Karon JM, Rosenberg PS, McQuillan G, Khare M, Guinn M, Petersen LR: Prevalence of HIV infection in the United States 1984–1992. JAMA 1996, 276:126–131.
2. Normand J, Vlahov D, Moses L: Preventing HIV Transmission: The Role of Sterile Needles and Bleach. Washington, DC: National Academy Press; 1995.
3. Gostin LO, Lazzarini Z, Jones TS, Flaherty K: Prevention of HIV/AIDS and other blood-borne diseases among injection drug users: a national survey on the regulation of syringes and needles. JAMA 1997, 277:53–62.
4. Sklarek HM, Mantovani RP, Erens E, Heisler D, Niederman MS, Fein AM: AIDS in a bodybuilder using anabolic steroids [letter]. N Engl J Med 1984, 311:1701.
5. Scott MJ, Scott MJ Jr: HIV infection associated with injections of anabolic steroids. JAMA 1989, 262:207–208.
6. Henrion R, Mandelbrot L, Delfieu D: Contamination par le VIH a la suite d'injections d'anabolisants [letter]. Presse Med 1992, 21:218.
7. Nemechek PM: Anabolic steroid users: another potential risk group for HIV infection [letter]. N Engl J Med 1991, 325:357.
8. Philips NW: Anabolic Reference Guide (Sixth Issue). Golden (CO): Mile High Publishing; 1991.
9. Hoffman PN, Larkin DP, Samuel D: Needlestick and needle–share: the difference. J Infect Dis 1989, 160:545–546.
10. Gaughwin MD, Gowans E, Ali R, Burrell C: Bloody needles: the volume of blood transferred in simulation of needlestick injuries and shared use of syringes for injection of intravenous drugs. AIDS 1991, 5:1025–1027.
11. Napoli V, McGowan J: How much blood is in a needlestick? [letter]. J Infect Dis 1987, 155:828.
12. Myers SS, Heimer R, Liu D, Henrard D: HIV DNA and antibodies in syringes from injecting drug users: a comparison of detection techniques. AIDS 1993, 7:925–931.
13. Heimer R, Myers SS, Cadman EC, Kaplan EH: Detection by polymerase chain reaction of HIV-1 proviral DNA sequences in needles of injecting drug users. J Infect Dis 1992, 165:781–782.
14. Hollinger FB: ACTG Virology Manual for HIV Laboratories: Bethesda: Division of AIDS, NIAID, NIH; 1993.
15. Cimarelli A, Zambruno G, Marconi A, Girolomoni G, Bertazzoni U, Giannetti A: Quantification by competitive PCR of HIV-1 proviral DNA in epidermal Langerhans cells of HIV-1 infected patients. J Acquir Immune Defic Syndr 1994, 7:230–235.
16. Henry M, Tschachler E: The skin immune system in the course of HIV infection. Photochem Photobiol 1996, 64:275–279.
17. Henderson DK, Fahey BJ, Willy M, et al.: Risk for occupational transmission of human immunodeficiency virus type 1 (HIV-1) associated with clinical exposures. A prospective evaluation. Ann Intern Med 1990, 113:740–746.
18. Centers for Disease Control and Prevention: HIV/AIDS Surveillance Report. Year-end Edition 1996. Atlanta: United States Department of Health and Human Services; 1996.
19. Centers for Disease Control Cooperative Needlestick Surveillance Group: Case–control study of HIV seroconversion in health-care workers after percutaneous exposure to HIV-infected blood – France, United Kingdom, and United States, January 1988–August 1994. MMWR 1995, 44:929–932.

HIV-1; injection; intramuscular; subcutaneous; intravenous; needlestick; anabolic steroids

© 1998 Lippincott Williams & Wilkins, Inc.