Mother-to-Child Transmission of Human T-Lymphotropic Virus Type II (HTLV-II)

Methods

Blood specimens were obtained from women who were positive for HTLV-II (three prostitutes [donors Y06, Y01, Y03] and one woman with a history of cervical cancer [Y08]) and their family members from Yucatan, Mexico, who consented to free testing. Information regarding sexual behavior, drug use, and breast-feeding history was obtained for each participant. The serum samples were tested for antibodies to HTLV by a modified Western blot assay, incorporating purified recombinant transmembrane protein and HTLV type-specific external glycoproteins specific for HTLV-I (rgp46^I) or HTLV-II (rgp46^II) protein with a whole virus lysate [6].

The amplification and detection of HTLV sequences by the nested polymerase chain reaction were performed on DNA specimens from selected persons who had Western blot profiles suggestive of HTLV infection. Nested amplification was performed in three gene regions, pol, env, and tax.

Results

Western blot analyses of the serum samples from family members Table 1 showed isolated r21e reactivity, followed by some rgp46^II reactivity. Three of the four sexual partners had r21e reactivity. However, the 8-year-old son (donor Y17) of a prostitute (donor Y06) had antibodies to both gag (p24) and env (rgp46^II, r21e) gene products, indicating HTLV-II positivity. Human T-lymphotropic virus type II-specific genomic sequences were also detected from peripheral blood lymphocytes of donor Y17, further confirming HTLV-II infection in this 8-year-old boy (Table 1).

Evaluation of the various risk factors for HTLV-II infection in the 8-year-old child (donor Y17) showed no history of intravenous drug abuse, blood transfusion, or use of nondisposable needles for childhood vaccinations. He is not homeless and attends an elementary school. Repeated questioning related to sexual behavior ruled out the possibility of sexual abuse. This child's only recognizable risk factor for acquisition of HTLV-II infection appears to be breast-feeding during the first 4 years of his life. In contrast, his 3-year-old brother (donor Y20), who was seronegative, was breast-fed for only 2 months. The breast-feeding of this child for up to 4 years suggests rather unusual bonding between mother and child, thus raising the possibility of incestuous behavior.

Discussion

A recent study showed that the HTLV-II genome can be found in the breast milk of women infected with HTLV-II [4], thus suggesting the possible transmission of this virus through breast-feeding. That the breast-feeding might be the major route of mother-to-child transmission of HTLV-II is further supported by a recent study in which no non-breast-fed babies born to women infected with HTLV-II had the HTLV-II genome [3]. Although the exact mechanism of transmission is not known, some of the virus-infected cells could have penetrated the mucosal barriers during their passage from the oral cavity to the gastrointestinal tract. Furthermore, oral administration of HTLV-I-infected human breast milk lymphocytes to marmoset monkeys showed transmission of HTLV-I in this animal model [7].

Nested polymerase chain reaction analyses were performed using primers in the pol, env, and tax gene regions to identify the HTLV genome in those persons who had any band on Western blot analyses. Although all of the specimens with antibodies to both gag (p24) and env (rgp46 and r21e) contained the HTLV-II genome (donors Y08, Y06, Y01, and Y03), all of the specimens with isolated r21e reactivity (Y26, Y29, Y34, Y18, Y20, and Y36) and with rgp46^II and r21e reactivity (Y31) were negative by nested polymerase chain reaction analysis, suggesting that isolated gag and env reactivities do not represent true HTLV infection.

Serologic analyses of the spouse or partners (or both) of these women infected with HTLV-II revealed antibodies to r21e in three of the four partners; however, no HTLV-specific genomic sequences could be amplified, even by nested polymerase chain reaction analysis. Although presence of the r21e band previously represented an early marker of seroconversion [8], lack of the HTLV-I and HTLV-II genomes in all of these specimens supports the absence of true HTLV infection in the sexual partners of women infected with HTLV-II. Antibodies to r21e have also been documented in persons with no evidence of HTLV infection and presumably reflect antigenic mimicry to a closely related antigen [9]. Previous studies analyzing sexual transmission showed that HTLV-I transmission between spouses occurred more frequently from husband to wife and rarely from wife to husband [10]. A similar study of HTLV-II infection among spouses also showed predominantly male-to-female transmission [2]. However, large cohort studies are needed to confirm these findings.