Secondary Logo

Journal Logo


HIV-1 infection transmitted by serum droplets into the eye: a case report

Eberle, Josefa; Habermann, Jürgenb; Gürtler, Lutz G.c

Author Information
  • Free

HIV is transmitted mainly by blood and sexual contact. Nosocomial infections occur most frequently by needle sticks or by contamination of eczematous skin lesions. The conjunctiva of the eye had been suggested to be a potential site of entry of HIV. This report presents evidence on the method of HIV transmission through the spillage of small amounts of contaminated serum into one eye. This mode of HIV transmission can be easily prevented by protecting the eyes with glasses and goggles.

A blood sample from a 26-year-old male laboratory technician, when tested for antibodies to HIV, showed a moderately high extinction in the screening enzyme-linked immunosorbent assay (ELISA; Abbott Laboratories, North Chicago, IL, USA; sample/cut-off ratio 2.8) and p24 and gp160 bands in the HIV-1 immunoblot, a pattern suggesting early seroconversion. A sample taken 1 week later showed high reactivity in two ELISA (Abbott and Dade–Behring, Marburg, Germany; sample/cut-off ratio > 8) and p24, p55 and gp160 immunoblot bands. A follow-up sample taken 1 month after the first bleed showed the same high ELISA reactivity, but all bands in the immunoblot, indicating unambiguous seroconversion.

The technician reported an accident approximately 5 weeks before the first specimen was drawn, when after centrifugation he was opening the stopper of a vacutainer tube by hand, and not using the tool supplied for that purpose. He was wearing gloves during this manipulation. He felt a droplet spill into his left eye, to which he reacted by blinking rapidly and not by washing the eye. At the time of the accident he was wearing contact glass lenses, which had induced a moderate conjunctivitis. He did not report the accident, nor was the number of the specimen identified.

It was possible to trace the source of the infection in the small hospital where he was working because at that time only one HIV-infected patient was present. This was a Thai women, who had been hospitalized after a severe traffic accident and had been treated in the intensive care unit for more than 1 month. At the time of hospitalization the CD4 cell count of the patient was 318/μl and viral load was 40 copies/ml, determined using a test that was probably not suitable to quantitate the HIV-1E virus accurately [1]. She was not receiving antiretroviral therapy because her HIV infection was not known before the traffic accident. The technician had no access to the intensive care unit and denied having had any sexual contact with the patient.

Nucleic acid sequence analysis of the C2V3 region of the HIV env gene from the patient and the technician by nested polymerase chain reaction (PCR), using the primers 5tat14 and 3env 155 in the first round and 5env72 and 3ed33 in the second round, revealed a sequence that clustered in the HIV-1E subtypes from Thailand when performed by computer analysis with the Treecon program as shown in Fig. 1[2]. Within the 350 nucleotides of the sequences of the patient (mvp7049–98) and technician two base pairs were different, leading to two amino acid mutations: glycine (GGT) to serine (AGT) and histidine (CAT) to proline (CCT) in the V3 loop of the technician's HIV (mvp7050–98). Re-analysis of the HIV nucleic acid sequence of the technician in a blood sample drawn after 3 weeks gave an identical sequence (mvp7581–98). Molecular analysis thus showed that both viruses were closely related, indicating that the technician harboured the HIV-1E of the Thai patient.

Fig. 1.:
Evolutionary tree of randomly selected HIV-1E viruses including the sequences of the index patient and the technician. It can easily be seen that the HIV of the Thai women (mvp7049–98) and the HIV of the technician (mvp7050 and 7581–98) cluster closely together.

The amino acid sequence of the V3 loop was as follows:





When the technician was interviewed in order to exclude other routes of HIV infection, he reported only one homosexual contact one year before the accident. From data available to us, HIV-1E is not present in the homosexual community in Munich, and acquisition of this HIV-1E virus by the technician through sexual contact is highly improbable. The C2V3 region sequence analysis and position in the evolutionary tree of HIV-1E documents the similarity between the two viruses, and the time course of the technician's infection indicates that the HIV-1E of the Thai woman was the source of his infection. Because no needle stick or other laboratory accident was reported by the technician, the only plausible manner of HIV transmission was via the reported serum droplet into his eye. During the past 10 years our laboratory has investigated seven spillages of blood from AIDS patients into the eye, mainly connected with venepuncture. All healthcare workers involved immediately rinsed the eye with tap water and one started antiretroviral therapy; none acquired an HIV infection. Why the technician did not rinse the eye after the accident is unclear, because he had been instructed twice within the previous 6 months about immediate action after eye contamination.

The possible transmission of HIV by contamination of the conjunctiva has been suspected previously as the mode of infection of a nurse in Italy [3]. Reports have also indicated that HIV can be cultured from cornea cells [4] and isolated from tears [4,5]. Our report indicates that after prolonged exposure with a sufficient quantity of virus, the mucous membrane of the eye seems to be susceptible to HIV entry. In this case viral entry might have been facilitated by the inflamed conjunctival cells. Protection of the eye can easily stop this rare mode of HIV transmission.

Josef Eberlea

Jürgen Habermannb

Lutz G. Gürtlerc


1. Oelrichs RB, Dyer WB, Downie JC, Harvey K, Sullivan JS, Workman C. Inaccurate HIV-1 viral load quantification by three major commercially available methods. AIDS 1999, 13: 727 –740.
2. Van de Peer Y, De Wachter R. Treecon for Windows: a software package for the construction and drawing of evolutionary trees for the microsoft windows environment. Comput Appl Biosci 1994, 10: 569 –570.
3. Gioannini P, Sinicco A, Cariti G, Luchini A, Paggi G, Giachino O. HIV infection acquired by a nurse. Eur J Epidemiol 1988, 4: 119 –120.
4. Salahuddin SZ, Palestine AG, Heck E, Ablashi D, Luckenbach M, McCulley JP, Nussenblatt RB. Isolation of the human T-cell leukemia/lymphotropic virus type III from the cornea. Am J Ophthalmol 1986, 101: 149 –152.
5. Fujikawa LS, Palestine AG, Nussenblatt RB, Salhuddin SZ, Masur H, Gallo RC. Isolation of a T-lymphotropic virus type III from the tears of a patient with the acquired immunodeficiency syndrome. Lancet 1985, ii: 529 –530.
© 2000 Lippincott Williams & Wilkins, Inc.