HTLV-I was isolated in 1980 from a T cell lymphoma cell line from a patient originally thought to have cutaneous T cell lymphoma. Later it became clear that the patient had a distinct form of lymphoma (originally reported in Japan) called adult T cell leukemia/lymphoma (ATL). Serologic data have determined that HTLV-I is the cause of at least two important diseases: ATL and tropical spastic paraparesis, also called HTLV-I-associated myelopathy (HAM). HTLV-I may also play a role in infective dermatitis, arthritis, uveitis, and Sjögren's syndrome.
Two years after the isolation of HTLV-I, HTLV-II was isolated from a patient with an unusual form of hairy cell leukemia that affected T cells. Epidemiologic studies of HTLV-II failed to reveal a consistent disease association.
Biology and Molecular Biology
Because the biology of HTLV-I and that of HTLV-II are similar, the following discussion will focus on HTLV-I.
Human glucose transporter protein 1 (GLUT-1) functions as a receptor for HTLV-1, probably acting together with neuropilin-1 (NRP1) and heparan sulfate proteoglycans. Generally, only T cells are productively infected, but infection of B cells and other cell types is occasionally detected. The most common outcome of HTLV-I infection is latent carriage of randomly integrated provirus in CD4+ T cells. HTLV-I does not contain an oncogene and does not insert into a unique site in the genome. Indeed, most infected cells express no viral gene products. The only viral gene product that is routinely expressed in tumor cells transformed by HTLV-I in vivo is hbz. The tax gene is thought to be critical to the transformation process but is not expressed in the tumor cells of many ATL patients, possibly because of the immunogenicity of tax-expressing cells. Cells transformed in vitro, by contrast, actively transcribe HTLV-I RNA and produce infectious virions. Most HTLV-I-transformed cell lines are the result of the infection of a normal host T cell in vitro. It is difficult to establish cell lines derived from authentic ATL cells.
Although tax does not itself bind to DNA, it does induce the expression of a wide range of host cell gene products, including transcription factors (especially c-rel/NF-κB, ets-1 and -2, and members of the fos/jun family), cytokines [e.g., interleukin (IL) 2, granulocyte-macrophage colony-stimulating factor, and tumor necrosis factor], and membrane proteins and receptors [major histocompatibility (MHC) molecules and IL-2 receptor α]. The genes activated by tax are generally controlled by transcription factors of the c-rel/NF-κB and cyclic AMP response element binding (CREB) protein families. It is unclear how this induction of host gene expression leads to neoplastic transformation; tax can interfere with G1 and mitotic cell-cycle checkpoints, block apoptosis, inhibit DNA repair, and promote antigen-independent T cell proliferation. Induction of a cytokine-autocrine loop has been proposed; however, IL-2 is not the crucial cytokine. The involvement of IL-4, IL-7, and IL-15 has been proposed.
In light of the irregular expression of tax in ATL cells, it has been suggested that tax is important in the early phases of transformation but is not essential for the maintenance of the transformed state. The maintenance role is thought to be due to hbz expression. As is clear from the epidemiology of HTLV-I infection, transformation of an infected cell is a rare event and may depend on heterogeneous second, third, or fourth genetic hits. No consistent chromosomal abnormalities have been described in ATL; however, aneuploidy is common and individual cases with p53 mutations and translocations involving the T cell receptor genes on chromosome 14 have been reported. Tax may repress certain DNA repair enzymes, permitting the accumulation of genetic damage that would normally be repaired. However, the molecular pathogenesis of HTLV-I-induced neoplasia is not fully understood.
Features of HTLV-I Infection
HTLV-I infection is transmitted in at least three ways: from mother to child, especially via breast milk; through sexual activity, more commonly from men to women; and through the blood—via contaminated transfusions or contaminated needles. The virus is most commonly transmitted perinatally. Compared with HIV, which can be transmitted in cell-free form, HTLV-I is less infectious, and its transmission usually requires cell-to-cell contact.
HTLV-I is endemic in southwestern Japan and Okinawa, where >1 million persons are infected. Antibodies to HTLV-I are present in the serum of up to 35% of Okinawans, 10% of residents of the Japanese island of Kyushu, and <1% of persons in nonendemic regions of Japan. Despite this high prevalence of infection, only ∼500 cases of ATL are diagnosed in this area each year. Clusters of infection have been noted in other areas of the Orient, such as Taiwan; in the Caribbean basin, including northeastern South America; in northwestern South America; in central and southern Africa; in Italy, Israel, Iran, and Papua New Guinea; in the Arctic; and in the southeastern part of the United States (Fig. 188-4). An estimated 15–20 million persons have HTLV-I infection worldwide.
Global distribution of HTLV-I infection. Countries with a prevalence of HTLV-I infection of 1–5% are shaded darkly. Note that the distribution of infected patients is not uniform in endemic countries. For example, the people of southwestern Japan and northeastern Brazil are more commonly affected than those in other regions of those countries.
Progressive spastic or ataxic myelopathy developing in an individual who is HTLV-I positive (i.e., who has serum antibodies to HTLV-I) may be due to direct infection of the nervous system with the virus, but destruction of the pyramidal tracts appears to involve HTLV-I-infected CD4+ T cells; a similar disorder may result from infection with HIV or HTLV-II. In rare instances, patients with HAM are seronegative but have detectable antibody to HTLV-I in the cerebrospinal fluid (CSF).
The cumulative lifetime risk of developing ATL is 3% among HTLV-I-infected patients, with a threefold greater risk among men than among women; a similar cumulative risk is projected for HAM (4%), but with women more commonly affected than men. The distribution of the two diseases overlaps the distribution of HTLV-I, with >95% of affected patients showing serologic evidence of HTLV-I infection. The latency period between infection and the emergence of disease is 20–30 years for ATL. For HAM, the median latency period is ∼3.3 years (range, 4 months to 30 years). The development of ATL is rare among persons infected by blood products; however, ∼20% of patients with HAM acquire HTLV-I from contaminated blood. ATL is more common among perinatally infected individuals, whereas HAM is more common among persons infected via sexual transmission.
Four clinical types of HTLV-I-induced neoplasia have been described: acute, lymphomatous, chronic, and smoldering. All of these tumors are monoclonal proliferations of CD4+ postthymic T cells with clonal proviral integrations and clonal T cell receptor gene rearrangements.
About 60% of patients who develop malignancy have classic acute ATL, which is characterized by a short clinical prodrome (∼2 weeks between the first symptoms and the diagnosis) and an aggressive natural history (median survival period, 6 months). The clinical picture is dominated by rapidly progressive skin lesions, pulmonary involvement, hypercalcemia, and lymphocytosis with cells containing lobulated or “flower-shaped” nuclei (see Fig. 110-10). The malignant cells have monoclonal proviral integrations and express CD4, CD3, and CD25 (low-affinity IL-2 receptors) on their surface. Serum levels of CD25 can be used as a tumor marker. Anemia and thrombocytopenia are rare. The skin lesions may be difficult to distinguish from those in mycosis fungoides. Lytic bone lesions, which are common, do not contain tumor cells but rather are composed of osteolytic cells, usually without osteoblastic activity. Despite the leukemic picture, bone marrow involvement is patchy in most cases.
The hypercalcemia of ATL is multifactorial; the tumor cells produce osteoclast-activating factors (tumor necrosis factor α, IL-1, lymphotoxin) and can also produce a parathyroid hormone–like molecule. Affected patients have an underlying immunodeficiency that makes them susceptible to opportunistic infections similar to those seen in patients with AIDS (Chap. 189). The pathogenesis of the immunodeficiency is unclear. Pulmonary infiltrates in ATL patients reflect leukemic infiltration half the time and opportunistic infections with organisms such as Pneumocystis and other fungi the other half. Gastrointestinal symptoms are nearly always related to opportunistic infection. Strongyloides stercoralis is a gastrointestinal parasite that has a pattern of endemic distribution similar to that of HTLV-I. HTLV-I-infected persons also infected with this parasite may develop ATL more often or more rapidly than those without Strongyloides infections. Serum concentrations of lactate dehydrogenase and alkaline phosphatase are often elevated in ATL. About 10% of patients have leptomeningeal involvement leading to weakness, altered mental status, paresthesia, and/or headache. Unlike other forms of central nervous system (CNS) lymphoma, ATL may be accompanied by normal CSF protein levels. The diagnosis depends on finding ATL cells in the CSF (Chap. 110).
The lymphomatous type of ATL occurs in ∼20% of patients and is similar to the acute form in its natural history and clinical course, except that circulating abnormal cells are rare and lymphadenopathy is evident. The histology of the lymphoma is variable but does not influence the natural history. In general, the diagnosis is suspected on the basis of the patient's birthplace (see “Epidemiology,” above) and the presence of skin lesions and hypercalcemia. The diagnosis is confirmed by the detection of antibodies to HTLV-I in serum.
Patients with the chronic form of ATL generally have normal serum levels of calcium and lactate dehydrogenase and no involvement of the CNS, bone, or gastrointestinal tract. The median duration of survival for these patients is 2 years. In some cases, chronic ATL progresses to the acute form of the disease.
Fewer than 5% of patients have the smoldering form of ATL. In this form, the malignant cells have monoclonal proviral integration; <5% of peripheral blood cells exhibit typical morphologic abnormalities; hypercalcemia, adenopathy, and hepatosplenomegaly do not develop; the CNS, the bones, and the gastrointestinal tract are not involved; and skin lesions and pulmonary lesions may be present. The median survival period for this small subset of patients appears to be ≥5 years.
HAM (Tropical Spastic Paraparesis)
In contrast to ATL, in which there is a slight predominance of male patients, HAM affects female patients disproportionately. HAM resembles multiple sclerosis in certain ways (Chap. 380). The onset is insidious. Symptoms include weakness or stiffness in one or both legs, back pain, and urinary incontinence. Sensory changes are usually mild, but peripheral neuropathy may develop. The disease generally takes the form of slowly progressive and unremitting thoracic myelopathy; one-third of patients are bedridden within 10 years of diagnosis, and one-half are unable to walk unassisted by this point. Patients display spastic paraparesis or paraplegia with hyperreflexia, ankle clonus, and extensor plantar responses. Cognitive function is usually spared; cranial nerve abnormalities are unusual.
MRI reveals lesions in both the white matter and the paraventricular regions of the brain as well as in the spinal cord. Pathologic examination of the spinal cord shows symmetric degeneration of the lateral columns, including the corticospinal tracts; some cases involve the posterior columns as well. The spinal meninges and cord parenchyma contain an inflammatory infiltrate with myelin destruction.
HTLV-I is not usually found in cells of the CNS but may be detected in a small population of lymphocytes present in the CSF. In general, HTLV-I replication is greater in HAM than in ATL, and patients with HAM have a stronger immune response to the virus. Antibodies to HTLV-I are present in the serum and appear to be produced in the CSF of HAM patients, where titers are often higher than in the serum. The pathophysiology of HAM may involve the induction of autoimmune destruction of neural cells by T cells with specificity for viral components such as Tax or Env proteins. One theory is that susceptibility to HAM may be related to the presence of human leukocyte antigen (HLA)alleles capable of presenting viral antigens in a fashion that leads to autoimmunity. Insufficient data are available to confirm an HLA association. However, antibodies in the sera of HAM patients have been shown to bind a neuron-specific antigen [heteronuclear ribonuclear protein A1 (hnRNP A1)] and to interfere with neurotransmission in vitro.
It is unclear what factors influence whether HTLV-I infection will cause disease and, if it does, whether it will induce a neoplasm (ATL) or an autoimmune disorder (HAM). Differences in viral strains, the susceptibility of particular MHChaplotypes, the route of HTLV-I infection, the viral load, and the nature of the HTLV-I-related immune response are putative factors, but few definitive data are available.
Other Putative HTLV-I-Related Diseases
In areas where HTLV-I is endemic, diverse inflammatory and autoimmune diseases have been attributed to the virus, including uveitis, dermatitis, pneumonitis, rheumatoid arthritis, and polymyositis. However, a causal relationship between HTLV-I and these illnesses has not been established.
Women in endemic areas should not breast-feed their children, and blood donors should be screened for serum antibodies to HTLV-I. As in the prevention of HIV infection, the practice of safe sex and the avoidance of needle sharing are important.
Treatment: HTLV-I Infection
For the small number of patients who develop HTLV-I-related disease, therapies are not curative. In patients with the acute and lymphomatous types of ATL, the disease progresses rapidly. Hypercalcemia is generally controlled by glucocorticoid administration and cytotoxic therapy directed against the neoplasm. The tumor is highly responsive to combination chemotherapy that is employed against other forms of lymphoma; however, patients are susceptible to overwhelming bacterial and opportunistic infections, and ATL relapses within 4–10 months after remission in most cases. The combination of interferon α and zidovudine may extend survival. Because viral replication is not clearly associated with ATL progression, zidovudine is probably effective through its cytotoxic effects (as a chain-terminating thymidine analogue) rather than its antiviral effects. An experimental approach using an yttrium 90–labeled or toxin-conjugated antibody to the IL-2 receptor appears promising but is not widely available. Patients with the chronic or smoldering form of ATL may be managed with an expectant approach: treat any infections, and watch and wait for signs of progression to acute disease.
Patients with HAM may obtain some benefit from the use of glucocorticoids to reduce inflammation. Antiretroviral regimens have not been effective. In one study, danazol (200 mg three times daily) produced significant neurologic improvement in five of six treated patients, with resolution of urinary incontinence in two cases, decreased spasticity in three, and restoration of the ability to walk after confinement to a wheelchair in two. Physical therapy and rehabilitation are important components of management.
Features of HTLV-II Infection
HTLV-II is endemic in certain Native American tribes and in Africa. It is generally considered to be a New World virus that was brought from Asia to the Americas 10,000–40,000 years ago during the migration of infected populations across the Bering land bridge. The mode of transmission of HTLV-II is probably the same as that of HTLV-I (see above). HTLV-II may be less readily transmitted sexually than HTLV-I.
Studies of large cohorts of injection drug users with serologic assays that reliably distinguish HTLV-I from HTLV-II indicated that the vast majority of HTLV-positive cohort members were infected with HTLV-II. The seroprevalence of HTLV in a cohort of 7841 injection drug users from drug treatment centers in Baltimore, Chicago, Los Angeles, New Jersey (Asbury Park and Trenton), New York City (Brooklyn and Harlem), Philadelphia, and San Antonio was 20.9%, with >97% of cases due to HTLV-II. The seroprevalence of HTLV-II was higher in the Southwest and the Midwest than in the Northeast. In contrast, the seroprevalence of HIV-1 was higher in the Northeast than in the Southwest or the Midwest. Approximately 3% of the cohort members were infected with both HTLV-II and HIV-1. The seroprevalence of HTLV-II increased linearly with age. Women were significantly more likely to be infected with HTLV-II than were men; the virus is thought to be more efficiently transmitted from male to female than from female to male.
Although HTLV-II was isolated from a patient with a T cell variant of hairy cell leukemia, this virus has not been consistently associated with a particular disease and in fact has been thought of as “a virus searching for a disease.” However, evidence is accumulating that HTLV-II may play a role in certain neurologic, hematologic, and dermatologic diseases. These data require confirmation, particularly in light of the previous confusion regarding the relative prevalences of HTLV-I and HTLV-II among injection drug users.
Avoidance of needle sharing, adherence to safe-sex practices, screening of blood (by assays for HTLV-I, which also detect HTLV-II), and avoidance of breast-feeding by infected women are important principles in the prevention of spread of HTLV-II.