Detection of HTLV-1 proviral DNA in cell-free DNA: Potential for non-invasive monitoring of Adult T cell leukaemia/lymphoma using liquid biopsy?

Introduction: Fragmented genomic DNA is constitutively released from dying cells into interstitial fluid in healthy tissue. In cancer, this so-called 'cell-free' DNA (cfDNA) released from dying malignant cells encodes cancer-associated mutations. Thus, minimally invasive sampling of cfDNA in blood plasma can be used to diagnose, characterise and longitudinally monitor solid tumours at remote sites in the body. ~5% of carriers of Human T cell leukaemia virus type 1 (HTLV-1) develop Adult T cell leukaemia/lymphoma (ATL), and a similar percentage develop an inflammatory CNS disease, HTLV-1 associated myelopathy (HAM). In both ATL and HAM, high frequencies of HTLV-1 infected cells are present in the affected tissue: each carrying an integrated DNA copy of the provirus. We hypothesised that turnover of infected cells results in the release of HTLV-1 proviruses in cfDNA, and that analysis of cfDNA from infected cells in HTLV-1 carriers might contain clinically useful information pertaining to inaccessible sites in the body- e.g. for early detection of primary or relapsing localised lymphoma type ATL. To evaluate the feasibility of this approach, we tested for HTLV-1 proviruses in blood plasma cfDNA. Methods: CfDNA (from blood plasma) and genomic DNA (gDNA, from peripheral blood mononuclear cells, PBMC) was isolated from blood from 6 uninfected controls, 24 asymptomatic carriers (AC), 21 patients with HAM and 25 patients with ATL. Proviral (HTLV-1 Tax) and human genomic DNA (the beta globin gene, HBB) targets were quantified by qPCR using primer pairs optimised for fragmented DNA. Results: Pure, high quality cfDNA was successfully extracted from blood plasma of all study participants. When compared with uninfected controls, HTLV-1 carriers had higher concentrations of cfDNA circulating in their blood plasma. Patients with ATL who were not in remission had the highest levels of blood plasma cfDNA in any group studied. HTLV-1 proviral DNA was detected in 60/70 samples obtained from HTLV-1 carriers. The proviral load (percentage of cells carrying proviruses) was approximately tenfold lower in plasma cfDNA than in PBMC genomic DNA, and there was a strong correlation between the proviral load in cfDNA and PBMC genomic DNA in HTLV-1 carriers that did not have ATL. cfDNA samples in which proviruses were undetectable also had very low proviral load in PBMC genomic DNA. Finally, detection of proviruses in cfDNA of patients with ATL was predictive of clinical status: patients with evolving disease had higher than expected total amount of proviruses detectable in plasma cfDNA. Discussion: We demonstrated that (1) HTLV-1 infection is associated with increased levels of blood plasma cfDNA, (2) proviral DNA is released into blood plasma cfDNA in HTLV-1 carriers and (3) proviral burden in cfDNA correlates with clinical status, raising the possibility of developing assays of cfDNA for clinical use in HTLV-1 carriers.

CAR-iNKT cells targeting clonal TCRVß chains as a precise strategy to treat T cell lymphoma.

Introduction: Most T cell receptor (TCR)Vß chain-expressing T cell lymphomas (TCL) including those caused by Human T cell leukaemia virus type-1 (HTLV-1) have poor prognosis. We hypothesised that chimeric antigen receptor (CAR)-mediated targeting of the clonal, lymphoma-associated TCRß chains would comprise an effective cell therapy for TCL that would minimally impact the physiological TCR repertoire. Methods: As proof of concept, we generated CAR constructs to target four TCRVß subunits. Efficacy of the CAR constructs was tested using conventional T cells as effectors (CAR-T). Since invariant NKT (iNKT) cell do not incite acute graft-versus-host disease and are suitable for 'off-the-shelf' immunotherapy, we generated anti-TCRVß CAR-iNKT cells. Results: We show that anti-TCRVß CAR-T cells selectively kill their cognate tumour targets while leaving >90% of the physiological TCR repertoire intact. CAR-iNKT cells inhibited the growth of TCL in vivo, and were also selectively active against malignant cells from Adult T cell leukaemia/lymphoma patients without activating expression of HTLV-1. Discussion: Thus we provide proof-of-concept for effective and selective anti-TCRVß CAR-T and -iNKT cell-based therapy of TCL with the latter providing the option for 'off-the-shelf' immunotherapy.

Clonality of HIV-1- and HTLV-1-Infected Cells in Naturally Coinfected Individuals.

BACKGROUND: Coinfection with human immunodeficiency virus type 1 (HIV-1) and human T-cell leukemia virus type 1 (HTLV-1) diminishes the value of the CD4+ T-cell count in diagnosing AIDS, and increases the rate of HTLV-1-associated myelopathy. It remains elusive how HIV-1/HTLV-1 coinfection is related to such characteristics. We investigated the mutual effect of HIV-1/HTLV-1 coinfection on their integration sites (ISs) and clonal expansion. METHODS: We extracted DNA from longitudinal peripheral blood samples from 7 HIV-1/HTLV-1 coinfected, and 12 HIV-1 and 13 HTLV-1 monoinfected individuals. Proviral loads (PVL) were quantified using real-time polymerase chain reaction (PCR). Viral ISs and clonality were quantified by ligation-mediated PCR followed by high-throughput sequencing. RESULTS: PVL of both HIV-1 and HTLV-1 in coinfected individuals was significantly higher than that of the respective virus in monoinfected individuals. The degree of oligoclonality of both HIV-1- and HTLV-1-infected cells in coinfected individuals was also greater than in monoinfected subjects. ISs of HIV-1 in cases of coinfection were more frequently located in intergenic regions and transcriptionally silent regions, compared with HIV-1 monoinfected individuals. CONCLUSIONS: HIV-1/HTLV-1 coinfection makes an impact on the distribution of viral ISs and clonality of virus-infected cells and thus may alter the risks of both HTLV-1- and HIV-1-associated disease.

Quantification of T cell clonality in human T cell leukaemia virus type-1 carriers can detect the development of adult T cell leukaemia early.

Adult T cell leukaemia/lymphoma (ATL) arises from clonally expanded T cells that are infected with human T cell leukaemia virus type-1 (HTLV-1). Here, we show that ATL can be detected early in HTLV-1-carriers through quantification of T-cell receptor (TCR)Vß subunit diversity on T-cells infected with HTLV-1 (CD3+ CCR4+ CD26- T-cells) using an 'oligoclonality index' (OCI-flow). We established a reference range for OCI-flow by analysing peripheral blood mononuclear cells (PBMCs) from HTLV-1-carriers who had not developed ATL in a median of 10.5 years follow up (n = 38) and patients with ATL (n = 30). In the third cohort of HTLV-1-carriers with no history or clinical evidence of ATL (n = 106), 19% of high proviral load (PVL, ≥4 copies of HTLV-1/100 PBMCs) carriers had an OCI-flow in the ATL range, >0.770. Carriers with an OCI-flow >0.770 (n = 14) had higher lymphocyte counts and PVLs and were more likely to have a family history of ATL than carriers with OCI-flow ≤0.770. ATL subsequently developed in two of these 14 carriers but no carriers with OCI-flow ≤0.770 (p = 0.03, cumulative follow-up 129 person-years). This method can be used to identify a subset of high-PVL HTLV-1-carriers at increased risk of developing ATL who may benefit from intervention therapy, prior to the detection of disease.

Evolution of retrovirus-infected premalignant T-cell clones prior to adult T-cell leukemia/lymphoma diagnosis.

Adult T-cell leukemia/lymphoma (ATL) is an aggressive hematological malignancy caused by human T-cell leukemia virus type-1 (HTLV-1). ATL is preceded by decades of chronic HTLV-1 infection, and the tumors carry both somatic mutations and proviral DNA integrated into the tumor genome. In order to gain insight into the oncogenic process, we used targeted sequencing to track the evolution of the malignant clone in 6 individuals, 2 to 10 years before the diagnosis of ATL. Clones of premalignant HTLV-1-infected cells bearing known driver mutations were detected in the blood up to 10 years before individuals developed acute and lymphoma subtype ATL. Six months before diagnosis, the total number and variant allele fraction of mutations increased in the blood. Peripheral blood mononuclear cells from premalignant cases (1 year prediagnosis) had significantly higher mutational burden in genes frequently mutated in ATL than did high-risk, age-matched HTLV-1 carriers who remained ATL-free after a median of 10 years of follow-up. These data show that HTLV-1-infected T-cell clones carrying key oncogenic driver mutations can be detected in cases of ATL years before the onset of symptoms. Early detection of such mutations may enable earlier and more effective intervention to prevent the development of ATL.

Phosphatidylinositol 3-kinase-δ (PI3K-δ) is a potential therapeutic target in adult T-cell leukemia-lymphoma.

The prognosis of adult T-cell leukemia-lymphoma (ATL) remains very poor, and there is an urgent clinical need to investigate novel therapies for ATL. The expression of phosphatidylinositol 3-kinase-δ (PI3k-δ) is normally restricted to hematopoietic cells and is known as a key determinant of cell survival in certain cancers. The inhibitor of PI3k-δ, idelalisib, has been shown to be effective in the treatment of chronic lymphocytic leukemia. Here, we report the expression of PI3k-δ and the ability of idelalisib to promote apoptosis in ex vivo ATL samples. The activity of PI3K was quantified by a PI3-Kinase Activity ELISA kit. Although there was no significant difference in mean PI3K activity between healthy donors and patients with ATL, certain cases of ATL showed extremely high PI3K activities. The expression of PI3k-δ protein was detectable in most ATL cases. The freshly isolated cells from ATL patients were cultured with or without idelalisib for 0-10 days, and cell survival was then quantified. Idelalisib induced apoptosis in ATL cells in a time-dependent manner, and significantly reduced the frequency of viable ATL cells at 10 days. No time-dependent effects of idelalisib were observed in non-malignant T cells from the same patients. CCL22 has been reported to promote survival of ATL cells in part through the PI3K-AKT pathway. Idelalisib blocked this CCL22-induced phosphorylation of AKT and significantly inhibited the proliferation of ATL cells. These results validate the PI3K-AKT pathway as a potential therapeutic target in ATL.

Long-term clinical remission maintained after cessation of zidovudine and interferon-α therapy in chronic adult T-cell leukemia/lymphoma.

Globally, > 5-10 million people are estimated to be infected with Human T-lymphotropic virus type 1 (HTLV-1), of whom ~ 5% develop adult T-cell leukemia/lymphoma (ATL). Despite advances in chemotherapy, overall survival (OS) has not improved in the 35 years since HTLV-1 was first described. In Europe/USA, combination treatment with zidovudine and interferon-α (ZDV/IFN-α) has substantially changed the management of patients with the leukemic subtypes of ATL (acute or unfavorable chronic ATL) and is under clinical trial evaluation in Japan. However, there is only a single published report of long-term clinical remission on discontinuing ZDV/IFN-α therapy and the optimal duration of treatment is unknown. Anecdotal cases where therapy is discontinued due to side effects or compliance have been associated with rapid disease relapse, and it has been widely accepted that the majority of patients will require life-long therapy. The development of molecular methods to quantify minimal residual disease is essential to potentially guide therapy for individual patients. Here, for the first time, we report molecular evidence that supports long-term clinical remission in a patient who was previously treated with ZDV/IFN-α for 5 years, and who has now been off all therapy for over 6 years.

T Cell Receptor Vß Staining Identifies the Malignant Clone in Adult T cell Leukemia and Reveals Killing of Leukemia Cells by Autologous CD8+ T cells.

There is growing evidence that CD8+ cytotoxic T lymphocyte (CTL) responses can contribute to long-term remission of many malignancies. The etiological agent of adult T-cell leukemia/lymphoma (ATL), human T lymphotropic virus type-1 (HTLV-1), contains highly immunogenic CTL epitopes, but ATL patients typically have low frequencies of cytokine-producing HTLV-1-specific CD8+ cells in the circulation. It remains unclear whether patients with ATL possess CTLs that can kill the malignant HTLV-1 infected clone. Here we used flow cytometric staining of TCRVß and cell adhesion molecule-1 (CADM1) to identify monoclonal populations of HTLV-1-infected T cells in the peripheral blood of patients with ATL. Thus, we quantified the rate of CD8+-mediated killing of the putative malignant clone in ex vivo blood samples. We observed that CD8+ cells from ATL patients were unable to lyse autologous ATL clones when tested directly ex vivo. However, short in vitro culture restored the ability of CD8+ cells to kill ex vivo ATL clones in some donors. The capacity of CD8+ cells to lyse HTLV-1 infected cells which expressed the viral sense strand gene products was significantly enhanced after in vitro culture, and donors with an ATL clone that expressed the HTLV-1 Tax gene were most likely to make a detectable lytic CD8+ response to the ATL cells. We conclude that some patients with ATL possess functional tumour-specific CTLs which could be exploited to contribute to control of the disease.

CADM1/TSLC1 Identifies HTLV-1-Infected Cells and Determines Their Susceptibility to CTL-Mediated Lysis.

Human T cell lymphotropic virus-1 (HTLV-1) primarily infects CD4+ T cells, causing inflammatory disorders or a T cell malignancy in 5% to 10% of carriers. The cytotoxic T lymphocyte (CTL) response is a key factor that controls the viral load and thus the risk of disease. The ability to detect the viral protein Tax in primary cells has made it possible to estimate the rate at which Tax-expressing infected cells are eliminated by CTLs in persistently infected people. However, most HTLV-1-infected cells are Tax-at a given time, and their immunophenotype is poorly defined. Here, we aimed to identify a cell-surface molecule expressed by both Tax+ and Tax-HTLV-1-infected cells and use it to analyse the CTL response in fresh peripheral blood mononuclear cells. Cell adhesion molecule 1 (CADM1/TSLC1) was the best single marker of HTLV-1 infection, identifying HTLV-1-infected cells with greater sensitivity and specificity than CD25, CCR4 or ICAM-1. CADM1+CD4+ T cells carried a median of 65% of proviral copies in peripheral blood. In a cohort of 23 individuals, we quantified the rate of CTL-mediated killing of Tax+ and Tax-CADM1+ cells. We show that CADM1 expression is associated with enhanced susceptibility of infected cells to CTL lysis: despite the immunodominance of Tax in the CTL response, Tax+CADM1- cells were inefficiently lysed by CTLs. Upregulation of the CADM1 ligand CRTAM on CD8+ T cells correlated with efficient lysis of infected cells. Tax-CADM1+ cells were lysed at a very low rate by autologous CTLs, however, were efficiently killed when loaded with exogenous peptide antigen. High expression of CADM1 on most HTLV-1-infected cells in the face of enhanced CTL counterselection implies that CADM1 confers a strong benefit on the virus.

The retrovirus HTLV-1 inserts an ectopic CTCF-binding site into the human genome.

Human T-lymphotropic virus type 1 (HTLV-1) is a retrovirus that causes malignant and inflammatory diseases in ∼10% of infected people. A typical host has between 10(4) and 10(5) clones of HTLV-1-infected T lymphocytes, each clone distinguished by the genomic integration site of the single-copy HTLV-1 provirus. The HTLV-1 bZIP (HBZ) factor gene is constitutively expressed from the minus strand of the provirus, whereas plus-strand expression, required for viral propagation to uninfected cells, is suppressed or intermittent in vivo, allowing escape from host immune surveillance. It remains unknown what regulates this pattern of proviral transcription and latency. Here, we show that CTCF, a key regulator of chromatin structure and function, binds to the provirus at a sharp border in epigenetic modifications in the pX region of the HTLV-1 provirus in T cells naturally infected with HTLV-1. CTCF is a zinc-finger protein that binds to an insulator region in genomic DNA and plays a fundamental role in controlling higher order chromatin structure and gene expression in vertebrate cells. We show that CTCF bound to HTLV-1 acts as an enhancer blocker, regulates HTLV-1 mRNA splicing, and forms long-distance interactions with flanking host chromatin. CTCF-binding sites (CTCF-BSs) have been propagated throughout the genome by transposons in certain primate lineages, but CTCF binding has not previously been described in present-day exogenous retroviruses. The presence of an ectopic CTCF-BS introduced by the retrovirus in tens of thousands of genomic locations has the potential to cause widespread abnormalities in host cell chromatin structure and gene expression.

HTLV-1 drives vigorous clonal expansion of infected CD8(+) T cells in natural infection.

BACKGROUND: Human T-lymphotropic Virus Type I (HTLV-1) is a retrovirus that persistently infects 5-10 million individuals worldwide and causes disabling or fatal inflammatory and malignant diseases. The majority of the HTLV-1 proviral load is found in CD4(+) T cells, and the phenotype of adult T cell leukemia (ATL) is typically CD4(+). HTLV-1 also infects CD8(+) cells in vivo, but the relative abundance and clonal composition of the two infected subpopulations have not been studied. We used a high-throughput DNA sequencing protocol to map and quantify HTLV-1 proviral integration sites in separated populations of CD4(+) cells, CD8(+) cells and unsorted peripheral blood mononuclear cells from 12 HTLV-1-infected individuals. RESULTS: We show that the infected CD8(+) cells constitute a median of 5% of the HTLV-1 proviral load. However, HTLV-1-infected CD8(+) clones undergo much greater oligoclonal proliferation than the infected CD4(+) clones in infected individuals, regardless of disease manifestation. The CD8(+) clones are over-represented among the most abundant clones in the blood and are redetected even after several years. CONCLUSIONS: We conclude that although they make up only 5% of the proviral load, the HTLV-1-infected CD8(+) T-cells make a major impact on the clonal composition of HTLV-1-infected cells in the blood. The greater degree of oligoclonal expansion observed in the infected CD8(+) T cells, contrasts with the CD4(+) phenotype of ATL; cases of CD8(+) adult T-cell leukaemia/lymphoma are rare. This work is consistent with growing evidence that oligoclonal expansion of HTLV-1-infected cells is not sufficient for malignant transformation.

Cytotoxic T lymphocyte lysis of HTLV-1 infected cells is limited by weak HBZ protein expression, but non-specifically enhanced on induction of Tax expression.

BACKGROUND: Immunogenetic evidence indicates that cytotoxic T lymphocytes (CTLs) specific for the weak CTL antigen HBZ limit HTLV-1 proviral load in vivo, whereas there is no clear relationship between the proviral load and the frequency of CTLs specific for the immunodominant antigen Tax. In vivo, circulating HTLV-1-infected cells express HBZ mRNA in contrast, Tax expression is typically low or undetectable. To elucidate the virus-suppressing potential of CTLs targeting HBZ, we compared the ability of HBZ- and Tax-specific CTLs to lyse naturally-infected cells, by co-incubating HBZ- and Tax-specific CTL clones with primary CD4(+) T cells from HLA-matched HTLV-1-infected donors. We quantified lysis of infected cells, and tested whether specific virus-induced host cell surface molecules determine the susceptibility of infected cells to CTL-mediated lysis. RESULTS: Primary infected cells upregulated HLA-A*02, ICAM-1, Fas and TRAIL-R1/2 in concert with Tax expression, forming efficient targets for both HTLV-1-specific CTLs and CTLs specific for an unrelated virus. We detected expression of HBZ mRNA (spliced isoform) in both Tax-expressing and non-expressing infected cells, and the HBZ26-34 epitope was processed and presented by cells transfected with an HBZ expression plasmid. However, when coincubated with primary cells, a high-avidity HBZ-specific CTL clone killed significantly fewer infected cells than were killed by a Tax-specific CTL clone. Finally, incubation with Tax- or HBZ-specific CTLs resulted in a significant decrease in the frequency of cells expressing high levels of HLA-A*02. CONCLUSIONS: HTLV-1 gene expression in primary CD4(+) T cells non-specifically increases susceptibility to CTL lysis. Despite the presence of HBZ spliced-isoform mRNA, HBZ epitope presentation by primary cells is significantly less efficient than that of Tax.

Clonality of HTLV-2 in natural infection.

Human T-lymphotropic virus type 1 (HTLV-1) and type 2 (HTLV-2) both cause lifelong persistent infections, but differ in their clinical outcomes. HTLV-1 infection causes a chronic or acute T-lymphocytic malignancy in up to 5% of infected individuals whereas HTLV-2 has not been unequivocally linked to a T-cell malignancy. Virus-driven clonal proliferation of infected cells both in vitro and in vivo has been demonstrated in HTLV-1 infection. However, T-cell clonality in HTLV-2 infection has not been rigorously characterized. In this study we used a high-throughput approach in conjunction with flow cytometric sorting to identify and quantify HTLV-2-infected T-cell clones in 28 individuals with natural infection. We show that while genome-wide integration site preferences in vivo were similar to those found in HTLV-1 infection, expansion of HTLV-2-infected clones did not demonstrate the same significant association with the genomic environment of the integrated provirus. The proviral load in HTLV-2 is almost confined to CD8+ T-cells and is composed of a small number of often highly expanded clones. The HTLV-2 load correlated significantly with the degree of dispersion of the clone frequency distribution, which was highly stable over ∼8 years. These results suggest that there are significant differences in the selection forces that control the clonal expansion of virus-infected cells in HTLV-1 and HTLV-2 infection. In addition, our data demonstrate that strong virus-driven proliferation per se does not predispose to malignant transformation in oncoretroviral infections.

Is There a Role for HTLV-1-Specific CTL in Adult T-Cell Leukemia/Lymphoma?

ATLL is an aggressive malignancy of T cells that affects about 5% of individuals infected with HTLV-1. The precise mechanism of oncogenesis is not known, but there is evidence that two regulatory viral proteins, Tax and HBZ, are involved. A high set point proviral load is associated with development of ATLL or a chronic inflammatory condition, HAM/TSP. Several lines of evidence, including HLA class 1 association studies and in vitro killing assays, indicate that cytotoxic T lymphocytes are instrumental in determining this proviral load set point. Prior studies have focused chiefly on the CTL response to the immunodominant Tax protein: efficient lysis of Tax-expressing cells inversely correlates with proviral load in nonmalignant infection. However, a recent study showed that strong binding of peptides from HBZ, but not Tax, to HLA class 1 molecules was associated with a low proviral load and a reduced risk of developing HAM/TSP, indicating an important role for HBZ-specific CTL in determining infection outcome. In comparison with nonmalignant infection, HTLV-1-specific CTLs in ATLL patients are reduced in frequency and functionally deficient. Here we discuss the nature of protective CTL responses in nonmalignant HTLV-1 infection and explore the potential of CTLs to protect against ATLL.

HTLV-1-infected T cells contain a single integrated provirus in natural infection.

Human T lymphotropic virus type 1 (HTLV-1) appears to persist in the chronic phase of infection by driving oligoclonal proliferation of infected T cells. Our recent high-throughput sequencing study revealed a large number (often > 10(4)) of distinct proviral integration sites of HTLV-1 in each host that is greatly in excess of previous estimates. Here we use the highly sensitive, quantitative high-throughput sequencing protocol to show that circulating HTLV-1(+) clones in natural infection each contain a single integrated proviral copy. We conclude that a typical host possesses a large number of distinct HTLV-1-infected T-cell clones.

The tumor marker Fascin is strongly induced by the Tax oncoprotein of HTLV-1 through NF-kappaB signals.

Oncogenic transformation of CD4(+) T cells by human T-cell lymphotropic virus type 1 (HTLV-1) is understood as the initial step to adult T-cell leukemia/lymphoma, a process that is mainly initiated by perturbation of cellular signaling by the viral Tax oncoprotein, a potent transcriptional regulator. In search of novel biomarkers with relevance to oncogenesis, we identified the tumor marker and actin-bundling protein Fascin (FSCN1) to be specifically and strongly up-regulated in both HTLV-1-transformed and adult T-cell leukemia/lymphoma patient-derived CD4(+) T cells. Fascin is important for migration and metastasis in various types of cancer. Here we report that a direct link can exist between a single viral oncoprotein and Fascin expression, as the viral oncoprotein Tax was sufficient to induce high levels of Fascin. Nuclear factor-κB signals were important for Tax-mediated transcriptional regulation of Fascin in T cells. This suggests that Fascin up-regulation by Tax contributes to the development of HTLV-1-associated pathogenesis.

The avidity and lytic efficiency of the CTL response to HTLV-1.

In human T-lymphotropic virus type 1 (HTLV-1) infection, a high frequency of HTLV-1-specific CTLs can co-exist stably with a high proviral load and the proviral load is strongly correlated with the risk of HTLV-1-associated inflammatory diseases. These observations led to the hypothesis that HTLV-1 specific CTLs are ineffective in controlling HTLV-1 replication but contribute to the pathogenesis of the inflammatory diseases. But evidence from host and viral immunogenetics and gene expression microarrays suggests that a strong CTL response is associated with a low proviral load and a low risk of HAM/TSP. Here, we quantified the frequency, lytic activity and functional avidity of HTLV-1-specific CD8(+) cells in fresh, unstimulated PBMCs from individuals with natural HTLV-1 infection. The lytic efficiency of the CD8(+) T cell response-the fraction of autologous HTLV-1-expressing cells eliminated per CD8(+) cell per day-was inversely correlated with both the proviral load and the rate of spontaneous proviral expression. The functional avidity of HTLV-1-specific CD8(+) cells was strongly correlated with their lytic efficiency. We conclude that efficient control of HTLV-1 in vivo depends on the CTL lytic efficiency, which depends in turn on CTL avidity of Ag recognition. CTL quality determines the position of virus-host equilibrium in persistent HTLV-1 infection.