INTRODUCTION — Human T-lymphotropic virus (HTLV-I), the first human retrovirus to be discovered [1], is estimated to infect 5 to 10 million people worldwide [2,3]. This virus is considered the most oncogenic human pathogen [4] and is the causative agent of two typically fatal diseases: adult T cell leukemia-lymphoma (ATL) [5-7] and HTLV-I-associated myelopathy [8], which is also known as tropical spastic paraparesis [9].
The epidemiology, virology, and pathogenic features of HTLV-I will be reviewed here. The diagnosis, disease associations, and treatment are discussed separately. (See "Human T-lymphotropic virus type I: Disease associations, diagnosis, and treatment".)
VIROLOGY — HTLV-I is an enveloped, single-stranded RNA virus of the Retroviridae family, the only human pathogen of the subfamily oncovirus, which includes HTLV-II, bovine leukemia virus, simian T cell leukemia virus, HTLV-III, and HTLV-IV [10]. It has a diploid genome, comprised of two identical strands of positive sense RNA, each 9032 base pairs long. The genomic organization is similar to other retroviruses, with two long terminal repeat (LTR) sequences flanking gag, pol, and env genes (figure 1).
However, HTLV-I also possesses a unique 1.6 kb region, termed pX, which is located between env and the 3'LTR [11]. This region encodes a number of regulatory proteins: p40tax (Tax), p27rex, p21rex, p12, p13 and p30. HTLV-I basic leucine zipper factor (HBZ) is encoded by the minus (complementary) strand of pX region [12]. Of these, Tax and HBZ have been most implicated in viral pathogenesis. (See 'Pathogenesis' below.)
Cell entry and replication — Viremia in HTLV-I infection is of an extremely low level, in contrast to HIV infection. New infection is the result of transmission of infected lymphocytes rather than cell-free viral particles. HTLV-I displays CD4 T cell tropism, but virion entry to CD4 T cells occurs most efficiently by direct cell-to-cell contact via a virological synapse [13] rather than from virions free in plasma. Cell-free HTLV-I can also infect dendritic cells (via heparin sulphate proteoglycans and neuropilin-1) [14] and can subsequently be transmitted from dendritic cells to CD4 T cells [15]. The glucose transporter, glut-1 [16,17], has also been identified as a receptor for the HTLV-I envelope glycoprotein (env). However, cells lacking GLUT-1 expression can be infected by HTLV-I [18]. HTLV-I-infected CD4 T cells can produce CCL22, a CCR4 ligand. CCL22 attracts CCR4-expressing CD4 cells, resulting in the preferential transmission of HTLV-I to CCR4-positive, CD4-positive T cells [19].
After entering the cell, the RNA genome of HTLV-I is reverse transcribed, generating a DNA product that integrates in a non-random manner into the host genome. Following integration, viral replication can occur in two ways. First, re-expression of the integrated provirus forms a new intracellular virion (also termed "infectious" replication) [20]. Second, the integrated provirus replicates with each mitotic cell division. This results in a low viral replication rate and, given that mitotic replication is driven by host rather than by viral DNA polymerase, a high fidelity of transcription. These factors combine to explain the high genetic stability of HTLV-I [21].
HTLV-I regulates its own transcription via Tax, which acts as a transcriptional activator [22,23], and Rex, which acts to suppress transcription by regulating viral mRNA splicing. The opposing actions of Tax and Rex result in transient expression of viral gene products, which may assist in evasion of the host immune response [24]. When integrated as provirus, persistent infection is established, and subsequent translation of viral products can induce numerous cellular changes favoring cell survival and proliferation. Infection does not induce any cytopathic effect.
Proviral load determination — There is rapid clonal expansion in the first few weeks of infection until an effective host immune response is established. After this point, most hosts carry between 104 and 105 different clones [25]. In asymptomatic infection, approximately 0.1 to 1 percent of peripheral blood mononuclear cells (PBMCs) carry provirus (viral DNA within the host cell genome), while higher proviral loads (up to 30 percent of PBMCs) are associated with HAM/TSP. Asymptomatic carriers with high proviral loads are more likely to progress to HAM/TSP [26].
Proviral load is independent of age or sex [27], but varies with the natural history of the infection. A prospective study of 10 patients who acquired transfusion-related HTLV-I showed a median proviral load of 212 copies per 105 lymphocytes within three months of infection, decreasing to 27 copies per 105 lymphocytes at 14 months of infection or longer [28]. Another study examining HTLV-I viral load in 30 individuals over a median of 10.4 years found a minor decrease over time (-0.017 log copies/106 PBMCs/year) [29]. In a study of 36 asymptomatic pregnancy HTLV-1 carriers, proviral load remained stable during pregnancy but rose three-fold at one month postpartum [30]. It is unknown how long this elevation lasts.
Differences in proviral load between individuals infected with identical viral sequences suggest that host factors contribute most to proviral loads [31]. Polymorphisms in certain host MHC class I molecules have been shown to influence the proviral set-point [32], implicating antigen presentation to CD8 T cells as an important step in viral control.
PATHOGENESIS — HTLV-I shares many similarities with human immunodeficiency virus type 1 (HIV-1) but differs significantly from HIV-1 in the spectrum of diseases caused and the manner by which it causes them. Two virologic differences between HIV-1 and HTLV-I help to account for this:
●The low replication rate and the high fidelity of replication of HTLV-I result in a relatively low viral burden and high genetic stability, thereby reducing the possibility of immune escape.
●HTLV-I does not induce death of T cells, but rather causes cell proliferation and transformation.
The two diseases predominantly associated with HTLV-I, ATL and HAM/TSP, rarely overlap. Further, each is associated with a different route of transmission (ATL with breast feeding, and HAM/TSP with blood transfusion), and each has different pathogenic and immunologic correlates. Overall, this suggests that HTLV-I infection can result in two fundamentally different disease processes. (See "Human T-lymphotropic virus type I: Disease associations, diagnosis, and treatment".)
Cellular transformation — Although HTLV-I is oncogenic, it neither induces transformation by encoding oncogenes, nor does it integrate into the host genome to disrupt host gene expression. Rather, oncogenesis is mediated through viral gene products interacting with host proteins (often transcription factors), thereby altering their function. This is typified by Tax, a 40 kDa nuclear phosphoprotein. It is the molecule most associated with HTLV-I pathogenesis and has been shown to be both necessary and sufficient to induce cellular transformation [33-36].
Tax has no human homologue [37], but it is able to affect several signaling pathways, including NF-kB, CREB, SRF and AP-1 [38], in three distinct ways:
●Tax can transactivate many genes favoring T cell proliferation (including interleukin [IL]-2, IL-2Ra and IL-15), and inhibiting apoptosis (Bcl-XL).
●Tax can trans-repress genes controlling cell cycle (p18 INK4c and lck) and DNA repair (DNA polymerase b) [39].
●Tax can also bind to and functionally inactivate proteins, such as the cell cycle regulator p16INK4a, the oncosuppressor TGF-beta and the mitotic checkpoint controller MAD-1 [40].
Tax can also impair the action of the cell cycle regulator p53, although the mechanism is not fully understood. The net outcome of these effects in the infected cell is to promote cell proliferation, induce genetic instability, and inhibit apoptosis, leading to the persistent clonal expansion of infected cells and the promotion of oncogenesis.
HBZ has also been implicated in the pathogenesis of ATL and, unlike Tax, is uniformly expressed in ATL cells [12,41]. Both HBZ RNA and the protein it encodes have been shown to induce changes with the RNA altering cell proliferation and the protein suppressing Tax-mediated viral transcription. The precise mechanisms by which HBZ participates in leukemogenesis have not yet been elucidated, although its selective inhibition of cellular senescence via the classical NF-kappaB pathway [42] and ability to induce T-cell proliferation [43] may be relevant.
Immunologic control — Both cellular and humoral responses are mounted against HTLV-I. Antibodies to gag proteins are produced within the first two months of infection, followed by antibodies specific for env proteins [44]. Anti-tax antibodies develop later in approximately half of those infected. Tax antibodies may play a role in pathogenesis of HAM/TSP, given that higher anti-Tax antibody levels are associated with higher proviral load [45] and development of HAM/TSP [46].
HTLV-specific cytotoxic T lymphocytes (CTLs) can readily be found in most HTLV-associated diseases, as well as in asymptomatic carriers, and are thought to be important in viral control. These cells primarily respond to epitopes from the Tax protein [47,48]. Mutations in Tax resulting in viral escape from CTL killing have been described [49]. Epitopes within other proteins, such as the gp46 envelope glycoprotein [50], have also been described. The significance of HTLV-specific T cells in HAM/TSP or ATL remains poorly defined. Impaired function of CTLs in both disease states has been described [51-53], although it is unclear what factors result in this functional impairment.
HTLV-I can induce a proinflammatory, type 1 T helper (Th1)-like state in CCR4-positive, CD4-positive T cells from both blood and central nervous system in patients with HAM/TSP that may play a role in pathogenesis [54,55]. Given its role in cellular transmission and immune activation, CCR4 is a promising therapeutic target in HTLV-I infection [56].
EPIDEMIOLOGY AND TRANSMISSION
Global prevalence — HTLV-I infects an estimated 5 to 10 million people worldwide [2,3]. The true prevalence in many populous areas, such as East Africa, China, and South Asia, is currently unknown due to a lack of systematic testing. Therefore, there is substantial uncertainty surrounding any global estimate of prevalence.
HTLV-I is primarily transmitted by breast feeding, although spread via blood transfusion, sharing of needles, and sexual intercourse also occurs. This predominance of vertical transmission results in clustering of cases in familial or geographically discrete groups. It is endemic in southern Japan, the Caribbean, South America, the Melanesian islands, Papua New Guinea, the Middle East, and in West, Central, and Southern Africa [57]. In these endemic areas, seroprevalences range from three to five percent in Trinidad [58] up to 30 percent in rural Miyazaki in southern Japan [59].
By contrast, non-endemic areas, such as the United States and Europe, have prevalence rates of less than 1 percent [60,61]. As an example, in a study of over two million first time blood donors in the United States between 2000 and 2009, only 104 individuals (0.005 percent) were confirmed antibody positive for HTLV-I [62]. These data may underestimate the true prevalence, as potential blood donors are screened, and those with high-risk characteristics are excluded.
Seroprevalence tends to increase with age, and women are nearly twice as likely to be infected as men [59,62,63], although the sex difference is only seen after 30 years of age and probably reflects the relative efficiency of sexual transmission from male to female [64]. In Southern Japan, the seroprevalence rate among pregnant women is approximately 5 percent. This can be contrasted with much lower seropositivity in non-endemic areas in Japan (0.1 to 1 percent), a result similar to serologic studies from Europe.
HTLV-I displays little genomic variability during the course of infection or between patients in the same geographic area. Mother/child and spouse pairs from Okinawa, Japan have been shown to be infected with highly conserved, and often identical, viruses based upon direct sequencing of the viral genome [21]. Studies in France [65], Papua New Guinea, the Solomon Islands [66], and Zaire [67] have shown similarly low genomic variability based upon the less accurate sequencing of PCR products. There is recognized strain variation between geographic areas, but this variation is small.
The risk of transmission appears to be independent of HLA class 1 type, but is modestly increased by greater HLA class 1 matching between mother and child (threefold increased risk with five or six shared HLA class 1 types) [68]. Susceptibility of infection in African children has been linked to chromosome 6q27, [69].
Mother to child transmission — HTLV can be transmitted vertically from mother to child, mostly through breastfeeding.
Breast-fed infants — HTLV-I antigen is found in the milk of infected mothers [70]; animal models have shown that breast milk alone can transmit HTLV-I [71]; and risk of infection is higher in breast-fed infants than in those who are bottle fed [72]. Over a one-year period, breast-fed infants have a fourfold increase risk of infection (15.7 versus 3.6 percent). These findings strongly support the hypothesis that virus can be spread though breast feeding, probably through lymphocytes in milk. Overall, the risk of transmission via breast feeding is estimated to be approximately 16 to 30 percent [73,74].
A longer duration of breast feeding increases transmission risk. In a prospective study of 780 Japanese infants, those who were breast fed for seven months or more had a rate of seroconversion to HTLV-I of 14.4 percent, while those breast fed for six months or less seroconverted at an equivalent rate as bottle-fed infants (4.4 and 5.7 percent, respectively) [75]. Similar findings have been reported in smaller but prospective studies in Japan [76] and Jamaica [77].
Another important risk factor is the provirus load in breast milk. This was illustrated in a report from Jamaica in which the risk of transmission ranged from 4.7 to 28.7 per 1000 person months at low and high provirus loads, respectively [78].
It is likely that the presence of maternal antibodies transferred in utero is protective for some months after birth. A prospective study of breast-fed infants born to seropositive mothers showed that infants seropositive at two years of age remained seropositive (for at least 10 years), but for infants who were seronegative at 2 years, no further transmission was seen [79]. These results suggested that after the period of breastfeeding, the risks of transmission were small prior to puberty.
Animal models have demonstrated the effectiveness of virus-specific immunoglobulin given prophylactically to prevent HTLV-I transmission [80,81]. Similar studies have not been performed in humans. Freeze-thawing of breast milk is an effective way to reduce transmission [79].
Formula-fed infants — Since 3 to 4 percent of bottle-fed infants born to HTLV-I infected mothers are HTLV-I seropositive [75], transmission to children by routes other than breast feeding is possible. The period of risk appears to be greatest within the first two years of life. A prospective study of 154 bottle-fed infants born to seropositive mothers showed that the risk of infection was 0.6 percent at one year and 4.6 percent at two years, with no new acquisition of infection for the 10 years following this two-year period [82].
The route of this transmission remains uncertain. HTLV-I provirus can be identified in cord blood and maternal saliva, but convincing evidence that either of these account for infection in the absence of breast feeding is lacking. In a prospective study, all seven children found to have HTLV-I proviral DNA in cord blood were HTLV-I seronegative at one to two years [83]. One reason for this may be that cord blood provirus is frequently defective and unable to establish infection [84]. Saliva and mouthwash specimens have also been shown to harbor proviral DNA [85-87]. However, lymphocytes are infrequent in saliva, and saliva itself frequently contains HTLV-I antibody [88] and can inhibit cell transmission in vitro [89].
Mode of delivery — Vaginal delivery may be a further route of transmission via placental microtransfusions or mucosal exposure to infected lymphocytes. A retrospective, nonrandomized study of 261 pregnant women in Brazil found a protective effect of Caesarian section (20 percent versus 9 percent transmission rate) [72]. It remains unknown whether Caesarian section provides any added benefit for reducing perinatal transmission when breastfeeding is also avoided. When Caesarian section is offered, breastfeeding should still be avoided to maximize the reduction in transmission risk.
Sexual transmission — HTLV-I can be transmitted via sexual intercourse, but the frequency of this means of transmission has been difficult to define. The prevalence of antibodies to HTLV-I in female sex workers differs by location, ranging from 3.2 to 21.8 percent in Kinshasa, Zaire and Callao, Peru, respectively [90,91].
As mentioned above, male to female transmission is more efficient than the reverse. As an example, the infection rate was considerably higher in the seronegative heterosexual partner when a man was the HTLV-I positive partner compared to a woman (4.9 versus 1.2 per 100 person-years) [92]. The presence of coexisting genital ulceration and cervicitis is associated with an increased risk of transmission [63,91].
Transmission via sexual intercourse also appears to be related to viral or proviral DNA load [31,93]. In one study, men who transmitted the virus to their heterosexual partners had significantly higher geometric mean viral loads than non-transmitters [93].
In one study of 2655 Peruvian men who have sex with men (MSM), HTLV-I infection was detected in 1.8 percent of subjects, while HIV infection was detected in 12.4 percent; 7 percent of HIV-infected men had HTLV coinfection [94].
Blood transfusion — Transfusion of infected cells, but not plasma [89], is an efficient means of HTLV-I transmission. The estimated probability of seroconversion by this route is 40 to 60 percent [95-97], with the time to seroconversion reported as 21 to 47 days [96] with a median of 51 days [97]. The risk of transmission decreases the longer blood is stored [96,97]. For countries with high HTLV-I seroprevalence rates, routine screening of blood donations for this virus is an effective way of reducing transfusion-related infection [98]. However, those countries that would most benefit from screening are often the least able to resource it. (See "Blood donor screening: Laboratory testing".)
The seroprevalence of HTLV-I in blood donors is generally much lower than in the general population (5.1 per 100,000 person years in the United States [62], 32 per 100,000 in Taiwan [99]), since patients at risk of bloodborne disease are usually excluded from donating. Where screening occurs, overall risks of transmission are low [100,101]. Countries that routinely screen all blood products for HTLV-I and -II include the United States, Canada, Brazil, Australia, New Zealand, France, and the United Kingdom [102]. In the United States from 2007 to 2008, the residual risk was determined to be 0.29 cases per 106 blood products [103].
Tissue donation — There are few data on the risk of transmission of HTLV-I from transplanted organs, as only rare cases have been reported [104-106]. A survey of 99 patients identified through the Japanese Renal Transplant Registry identified a high rate of transmission among 10 patients who were HTLV-1 seronegative at the time of transplant and received an organ from a seropositive donor; after a median follow-up of 4.5 years, HTLV-1-associated myelopathy/tropical spastic paraparesis was reported in 4 of 10 and seroconversion was reported in 7 of 8 (with those data missing in 2 patients) [107]. No HTLV-1-related disease was identified in 30 HTLV-1-positive recipients who received an HTLV-1-positive transplant.
A small transplant-related transmission cluster was identified in Germany; it included one liver and two kidney recipients who received their organs from a single donor who was retrospectively discovered to be seropositive for HTLV-I [105]. Two of the three recipients developed a cutaneous lymphoproliferative disease within three years following the transplant, an atypical presentation of HTLV-I-associated disease within an unusually rapid time frame; the lymphoproliferative lesions had high HTLV-I DNA concentrations. One of the recipients did not develop antibodies to HTLV-I for several years despite having high levels of HTLV-I DNA in the peripheral blood mononuclear cells. This report suggests the possibility that HTLV-I transmitted to transplant recipients may be associated with an unusual clinical presentation as well as delayed seroconversion and thus be difficult to detect.
In the United States, organ donations are evaluated for HTLV-I/II. The likelihood of tissue donors having viremia due to HTLV-I/II (as well as hepatitis B, hepatitis C, and HIV) was evaluated in 11,391 tissue donors to five tissue banks in the United States [108]. The estimated probability of HTLV-I/II viremia at the time of donation that would be undetected by screening with current serologic methods was 1 in 128,000 for HTLV-I/II.
Injection drug use — The impact of injection drug use (IDU) as a route of transmission varies widely among countries and is dependent upon the baseline seroprevalence in the region. Certain areas (eg, Italy [109], Eastern Europe [110], Thailand [111]) have low seroprevalences among drug users. Higher rates have been reported in areas with a high seroprevalence in the general population (eg, 20 to 40 percent in one Brazilian study [112]). Many studies fail to distinguish between HTLV-I and HTLV-II infection, although those that do suggest that HTLV-I is consistently less prevalent than HTLV-II among IDUs in Western Europe and the United States.
Occupational exposure — There is a potential risk of HTLV-I transmission following occupational exposure to infected blood through needlestick injury or mucosal splash. However, occupational acquisition has only very rarely been described, and the risk of transmission has never been accurately quantified. In a retrospective study from Australia, 53 health care workers were exposed to HTLV-I-infected blood and received serologic follow-up, and none demonstrated evidence of seroconversion [113]. Reassurance with serological follow-up alone appears reasonable for most occupational HTLV-I exposures, particularly in the absence of any proven, effective postexposure intervention.
Zoonotic transmission — HTLV-I is thought to have originated from a simian T lymphotrophic virus that was transmitted from a nonhuman primate to man. In Central Africa, where HTLV-I is endemic and close contact with nonhuman primates still occurs, particularly among hunters, zoonotic transmission may be another source of new infections [114,115]. As an example, in a cohort study of 269 individuals who had been bitten by nonhuman primates, the prevalence of HTLV-I was higher than that of controls matched for age, sex, and ethnicity [114].
SUMMARY AND RECOMMENDATIONS
●General – Human T-lymphotropic virus (HTLV-I) is estimated to infect 5 to 10 million people worldwide. This virus is the causative agent of two typically fatal diseases: adult T cell leukemia-lymphoma and HTLV-I-associated myelopathy, which is also known as tropical spastic paraparesis. (See 'Introduction' above.)
●Proviral load determination – Lower levels of viral DNA within peripheral blood mononuclear cells (PBMCs) correlate with asymptomatic infection while higher levels correlate with HTLV-associated diseases, such as HTLV-I-associated myelopathy. (See 'Proviral load determination' above.)
●Pathogenesis – Like human immunodeficiency virus type 1 (HIV-1), HTLV-I infects CD4 T-cells; however, unlike HIV-1, which causes T-cell death, HTLV-I leads to cellular proliferation and transformation and disease states such as adult T cell leukemia-lymphoma. These two viruses also differ in viral replication rates and genetic stability; HIV is associated with high levels of viremia and genetic heterogeneity, while HTLV-I is associated with a low replication rate and genetic stability of its viral progeny. (See 'Pathogenesis' above.)
●Epidemiology and transmission – HTLV-I is primarily transmitted by breast feeding, although spread via blood transfusion, sharing of needles, and sexual intercourse also occurs. Measures to reduce the risk of transmission exist for each of these routes. (See 'Epidemiology and transmission' above.)
●Tissue donation – The likelihood of tissue donors having viremia due to HTLV-I is low in the United States. (See 'Tissue donation' above.)
ACKNOWLEDGMENT — The UpToDate editorial staff acknowledges David T Scadden, MD, who contributed to an earlier version of this topic review.
Do you want to add Medilib to your home screen?