Молекулярные биомаркеры, связанные с ATLL и HAM HTLV-1 Инфицирование: систематический обзор.

INTRODUTION

Human T-lymphotropic virus type 1 (HTLV-1) was the first human retrovirus to be isolated and identified 1 , and it is estimated that there are between 5 and 10 million people living with HTLV-1 (PLHTLV-1) worldwide 2 , 3 . However, only about 3–5% of carriers develop severe associated complications, such as adult T-cell leukemia/lymphoma (ATLL) or HTLV-1-associated myelopathy (HAM), which poses challenges for early diagnosis and clinical management 4 , 5 .

HTLV-1 is primarily transmitted via sexual contact, blood transfusion, and breastfeeding, with higher prevalence in endemic regions such as southwestern Japan, sub-Saharan Africa, South America, the Caribbean, Austro-Melanesia, and specific foci in the Middle East 6 . HTLV-1 predominantly infects CD4+ T lymphocytes, promoting cell-to-cell transmission and clonal expansion of infected cells 7 , 8 . Diagnosis of infection is generally performed using serological assays such as enzyme-linked immunosorbent assay (ELISA), followed by western blot (WB) or real-time quantitative polymerase chain reaction (qPCR) to confirm and quantify the proviral load (PVL) 7 , 9 , which is widely used to infer the risk of clinical progression. However, PVL alone cannot predict disease development. Moreover, its utility is limited due to significant overlap between symptomatic and asymptomatic patients 10 , 11 .

Considering that currently available biomarkers still lack sufficient specificity and sensitivity to reliably identify individuals at high risk, particularly during the preclinical phase, there is growing interest in identifying additional markers that may improve diagnostic accuracy 12 . Omics sciences have emerged as approaches that enable the exploration of molecular mechanisms underlying HTLV-1 pathogenesis and its associated diseases. Potential biomarkers identified using these approaches may contribute to a better understanding of inflammatory, immunological, and metabolic processes associated with infection progression 13 . Thus, this systematic review aims to identify, categorize, and analyze the biomarkers reported in recent literature related to progression to ATLL and HAM, as well as to highlight existing gaps and potential strategies that may be applied in the future to address them.

MATERIALS AND METHODS

This systematic review formulated its research question based on the PICO framework. The population (P) comprised individuals living with HTLV-1 infection across different clinical conditions, including ATLL, HAM, and asymptomatic carriers. The intervention/exposure (I) involved the assessment of genomic, proteomic, and metabolic biomarkers. The comparison (C) considered the use of PVL alone or conventional single-biomarker approaches. The outcomes (O) focused on improved risk stratification for disease progression and potential applications in disease monitoring.

Based on this framework, the guiding research question was: “Which biomarkers have been described in recent literature as being associated with progression from HTLV-1 infection to ATLL or HAM, compared with the use of PVL alone, and how do these biomarkers perform in risk stratification and disease monitoring?”

Search strategies

The literature search was conducted up to May 2025 in the PubMed, Web of Science, Virtual Health Library (BVS), and CAPES Journals databases. The search period was restricted to the past five years to ensure that the review focused on the most recent and relevant evidence regarding biomarkers of interest. MeSH-based descriptors and related terms were used, combined with the Boolean operators ‘AND’ and ‘OR,’ and including synonyms to broaden coverage. The search strategy was as follows: ((“HTLV-1” OR “Human T-lymphotropic virus 1”) AND (“biomarkers” OR “markers” OR “genomics” OR “proteomics” OR “metabolomics” OR “omics”) AND (“TSP/HAM” OR “tropical spastic paraparesis” OR “ATLL” OR “adult T-cell leukemia” OR “ATL”)). No specific search for grey literature (e.g., theses or conference proceedings) was conducted to prioritize peer-reviewed studies, with preference given to open-access articles and full-text publications to facilitate analysis.

Eligibility criteria

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines 14 were applied to assist in article selection. Original studies published in English with full-text availability were included if they addressed biomarkers associated with progression to ATLL or HAM in PLHTLV-1. Studies were excluded if they used animal models, employed exclusively in silico approaches, involved pediatric populations, relied exclusively on publicly available datasets (e.g., KEGG, GEO, TCGA), or did not present original data. Additional exclusion criteria included studies addressing only serological or population genetics analyses, coinfections (such as HIV or hepatitis viruses), or those that did not compare at least two clinical groups (e.g., asymptomatic, HAM, ATLL, or infected symptomatic). Publications without original data (e.g., case reports, letters to the editor, comments, and reviews), duplicates, and studies that did not focus on molecular biomarkers or lacked sufficient information for qualitative synthesis were also excluded.

Methodological quality assessment

Data analysis was conducted narratively, with qualitative synthesis of the findings, due to the heterogeneity of study designs and methodological approaches. Biomarkers were categorized according to their nature (genomic, proteomic, and metabolic) to facilitate interpretation. Study quality was assessed using tools from the National Heart, Lung, and Blood Institute (NHLBI, National Institutes of Health, USA) 15 . For case-control studies, the Quality Assessment of Case-Control Studies tool was used, classifying studies as good (> 8 criteria met, without critical “no” responses in items 1, 2, 4, 5, 6, 9, 10), fair (7–8 criteria), or “poor” (< 7 criteria). For cohort and longitudinal studies, the Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies was applied, with studies classified as good (> 9 criteria, without critical “no” responses in items 1, 2, 4, 9, 11), fair (7–9 criteria), or poor (< 7 criteria). For cross-sectional studies, an adaptation of the latter tool was used, excluding items 8, 10, and 13. Studies were classified as good (> 7 criteria, without critical “no” responses in items 1, 2, 4, 9, 11), fair (6–7 criteria), or poor (< 6 criteria).

Data extraction

Following inclusion and exclusion criteria, two researchers (TL and LAR) independently selected the studies and applied the quality assessment tools. Any discrepancies were assessed by a third reviewer (SOGM) and resolved by consensus. Data were extracted on authorship, year of publication, study design, country of origin, analyzed biomarker, laboratory techniques employed, and associated clinical outcomes, which were organized into tables to facilitate qualitative and comparative analysis.

RESULTS

Record selection

Initially, 351 studies were identified across the databases ( Figure 1 ). After removing duplicates, 109 potentially relevant studies remained. During screening, articles were excluded based on title and abstract, leaving 57 studies for full-text review and qualitative analysis, ultimately resulting in 35 articles that met the criteria and were included in the review.

Study characteristics

The included studies comprised cross-sectional (n = 20), cohort (n = 9), case-control (n = 3), and longitudinal (n = 3) studies, conducted mainly in Brazil (n = 11), Japan (n = 11), Iran (n = 10), the United Kingdom (n = 3), and Argentina (n = 1). Population sizes ranged from 18 to 1,652 participants. Different samples—plasma, serum, peripheral blood mononuclear cells (PBMCs), cerebrospinal fluid (CSF), and biopsy tissue—were analyzed using molecular techniques. Genomic biomarkers were the most frequently investigated, followed by proteomic and metabolic markers. Recurring methodologies included ELISA, qPCR, enzyme-linked immunospot (ELISPot), flow cytometry, cytometric bead array (CBA), single-molecule array (SIMOA), and next-generation sequencing/RNA sequencing (NGS/RNA-seq), with additional complementary techniques. No biomarker demonstrated sufficiently high sensitivity and specificity for use in isolation. In total, 67 potential biomarkers were identified, including 37 genomic, 27 proteomic, and three metabolic, as shown in Table 1 . More studies focused on HAM cases (n = 23) than on ATLL (n = 15), with only three studies including both conditions. For HAM, the most frequently used samples were plasma (n = 8), serum (n = 7), CSF (n = 6), and PBMCs (n = 12). For ATLL, plasma (n = 6), serum (n = 2), PBMCs (n = 5), peripheral blood (n = 3), and biopsy (n = 1) predominated, with some studies using more than one matrix. Seven studies declared conflicts of interest, involving funding from pharmaceutical companies, participation in advisory boards, affiliations with biotechnology firms, and patent registrations related to the biomarkers, while the remaining 28 studies reported no conflicts.

Methodological quality assessment

Regarding methodological quality, 20 were rated as good and 15 as fair, no low-quality studies were included ( Table 2 ). All studies demonstrated high quality in the title, abstract, introduction, and description of participant characteristics, providing clear objectives and appropriate context. Studies rated as good were notable for well-defined populations, consistent analytical methods, and clear eligibility criteria. In contrast, studies rated as fair showed limitations, such as lack of sample size justification or absence of reporting on participation rates, partially compromising robustness. Despite methodological heterogeneity of study designs and variability in laboratory techniques, the exclusion of low-quality studies and adherence to most critical criteria ensure reliability of the findings.

DISCUSSION

This review demonstrates that HAM is predominantly associated with chronic inflammation and neurodegeneration, whereas ATLL is mainly linked to tumor-related and immunomodulatory markers. Although many available studies remain exploratory and require validation in larger cohorts, several biomarkers hold strong diagnostic and prognostic potential. However, only three studies adopted a longitudinal design, while the majority were cross-sectional, limiting the assessment of slow and gradual disease progression.

Our analysis identified genomic, proteomic, and metabolic biomarkers associated with progression to ATLL and HAM in PLHTLV-1, highlighting the potential of omics sciences for identifying novel and specific biomarkers. Among markers already used in clinical practice, PVL, measured by qPCR, was frequently reported across several studies 5 , 10 , 16 - 20 . Consistently elevated in patients with clinical outcomes—particularly those with HAM and acute forms of ATLL—PVL supports its role as a risk indicator, with its predictive value further enhanced when combined with other biomarkers 5 , 10 , 16 , 19 , 21 , 22 .

In the context of ATLL, key biomarkers identified included sTNFR2 and sCADM1. Two studies reported plasma concentrations of sTNFR2 to be approximately tenfold higher in patients with acute disease compared to asymptomatic individuals. Elevated levels were also frequently detected in ATLL cells, increasing progressively with disease advancement. Both biomarkers showed significant decreases following treatment initiation, supporting their potential use as monitoring tools.

Furthermore, sTNFR2 showed strong correlations with sCD25, sOX40, IL-10, and PVL, and sCADM1 correlated with PVL and other clinical parameters. These findings suggest that simultaneous increases may signal progression to acute ATLL 21 , 23 - 25 . Similarly, alterations in the humoral immune response against HTLV-1 proteins have been identified in ATLL and HAM patients. Decreased levels of anti-Tax and anti-Env antibodies were observed in ATLL patients compared to HTLV-1 carriers and HAM patients, while anti-Gag p19 and p24 levels were significantly elevated in ATLL. Conversely, HAM patients exhibited elevated anti-Gag and anti-Tax antibodies 26 . Plasma levels of cell-free DNA (cfDNA) were also found to be elevated in HTLV-1-infected individuals, including asymptomatic carriers and patients with HAM, with the highest levels observed in patients with aggressive ATLL. Notably, the lymphoma subtype exhibited higher cfDNA levels even when PVL was low, suggesting that cfDNA may serve as a more sensitive marker of tumor burden than PBMC-derived PVL in this subgroup 19 , 27 . Finally, the expression of the signal regulatory protein alpha (SIRPα), identified with biopsy analysis, revealed a potential prognostic marker: patients negative for SIRPα had a median survival of 16 months after diagnosis, whereas SIRPα-positive patients reached a median survival of 28 months 28 .

Several genomic biomarkers have been proposed for ATLL. In this context, qPCR validation confirmed significant upregulation of ADCY1 and ADCY3 in patients with ATLL. These genes encode adenylyl cyclases, key enzymes of the cyclic adenosine monophosphate (cAMP) production pathway, which catalyze the conversion of ATP into cAMP following G protein-coupled receptor activation. cAMP is responsible for cell proliferation, regulation of apoptosis, and response to DNA damage 29 . Additionally, experimental validation demonstrated increased expression of MAPK8 (JNK) together with its negative regulators EVI1, MKP-1, and PTPRR, indicating dysregulated MAPK/JNK signaling pathway in ATLL and further supporting the presence of altered intracellular signaling networks contributing to leukemogenesis 30 .

Moreover, qPCR analysis demonstrated significant downregulation of KMT2D, suggesting that epigenetic deregulation may also contribute to ATLL pathogenesis. In the same study, although UBB and RPS15A were experimentally validated, the authors concluded that their reduced expression levels do not support a pathogenic or biomarker role in ATLL 31 . RhoA and PRKACB were also found to be downregulated in ATLL patients compared with asymptomatic individuals, suggesting loss of their tumor-suppressive activity in infected cells, which may contribute to increased tumorigenicity, cancer progression, and reduced apoptosis 32 . However, all these genomic findings are limited by small sample sizes, and further studies in larger cohorts are required for validation.

For HAM, the combination of CXCL10 and Neopterin stood out as consistent indicators of disease progression, being detected in CSF and associated with central nervous system (CNS) inflammatory activity, as reported in five studies 22 , 33 - 36 . HAM patients exhibited CXCL10 levels up to sevenfold higher than asymptomatic individuals or controls, particularly in cases of accelerated clinical progression 22 , 35 . Together with Neopterin, CXCL10 proved superior to neuronal injury markers 34 , such as phosphorylated heavy neurofilament (pNf-H) 22 and light neurofilament (Nf-L) 35 , showing proportional increases with disease progression rate and serving as an indicator of clinical stage and degree of neurological impairment 22 . These findings reflect IFN-γ-mediated immune activation and persistent neuroinflammation in the CNS 34 , 37 . Supporting this scenario, Espíndola et al . 38 demonstrated a higher frequency and persistence of IFN-γ and IL-10 expression in infected clones, complementing recent data showing that lower plasma levels of IL-10 predict increased PVL and clinical progression, suggesting a protective role for this interleukin 17 .

In this context, IFN-γ has been a central focus of investigation. As one of the most abundant pro-inflammatory cytokines in patients with HAM, it has been hypothesized that its excessive production contributes to chronic CNS inflammation, leading to neuronal and myelin sheath damage 39 . Three studies included in this review investigated IFN-γ 16 , 40 , 41 , and, as previously described in the literature, it was consistently found at significantly higher levels in untreated HAM patients, including during intermediate disease stages, compared to asymptomatic individuals 40 , 42 . Using the ELISPot technique, one study demonstrated that spontaneous IFN-γ production by PBMCs was three- to fourfold higher in HAM patients compared to asymptomatic individuals 40 , reinforcing the findings of other studies 16 , 41 . This IFN-γ hyperactivity not only confirms its involvement in the neuroinflammatory process but also revealed a positive correlation with PVL, making IFN-γ a promising biomarker for risk stratification in these patients 43 . Moreover, it was observed that the combination of IFN-γ and IL-17A—both elevated in untreated HAM patients—can achieve a predictive accuracy of up to 62.1% for clinical progression when combined with PVL 16 .

Interestingly, although IL-17A was initially associated with HAM, its gene expression levels (mRNA) were also elevated in ATLL patients, indicating that its immunopathological role is not limited to a single disease. Furthermore, significant differences in expression were observed between analyzed groups (p < 0.05) 44 . As IFN-γ can be quantified or its activity measured in both plasma and PBMCs using minimally invasive methods 16 , 40 , 41 , it stands out as a promising core marker for future studies. These studies could investigate additional markers, such as IL-17A itself 16 , which could be combined with this cytokine to enhance predictive capacity regarding infection progression.

Other proteomic biomarkers have also been identified as indicators of HAM activity and progression. Elevated expression of V-CAM and CHIT-1 proteins has been reported, with CHIT-1 notably associated with pro-inflammatory responses in other neurodegenerative disorders. CHIT-1, in particular, has been proposed as a candidate marker for clinical stage and progression type, particularly in rapidly evolving cases with poor prognosis 45 . Similarly, pentraxin 3 (PTX3) protein was found at significantly elevated levels in the serum and plasma of patients with HAM compared to asymptomatic carriers and healthy controls, with no significant differences observed between the other groups 46 . Another relevant biomarker was the chemokine XCL1, whose plasma levels were significantly elevated in patients with HAM. This finding may be associated with facilitation of autoreactive T cell migration into the CNS. Notably, XCL1 levels were markedly higher in HAM patients than in individuals with multiple sclerosis, suggesting greater specificity in the context of HTLV-1 infection 47 .

In addition to these findings, other biomarkers were evaluated in combination with PVL. One example is the viral protein Tax, whose expression was found to be approximately 130-fold higher in symptomatic patients. When combined with PVL, Tax improved differentiation between symptomatic and asymptomatic individuals and was also associated with increased IRF-1 expression 5 . Another promising biomarker, AnxA1, was found at elevated levels in asymptomatic PLHTLV-1. Combined analysis of AnxA1 with PVL showed high sensitivity and specificity, suggesting potential for early diagnosis and estimation of the risk of clinical progression to HAM 20 .

From the genomic perspective, relevant biomarkers have also been identified. A notable finding is the association between the HLA-DRB1-GB-7 allele and increased disease susceptibility, as demonstrated in a genome-wide association study (GWAS). Individuals homozygous for this allele had up to a ninefold higher risk of disease development 48 . Additionally, HLA class I alleles have been implicated in modulating clinical outcomes. The HLA-B*35 allele has been associated with increased susceptibility to HAM, whereas HLA-A*02 has demonstrated a protective effect against ATLL. Independent studies have also reported a protective role of HLA-A*02 in HAM. Furthermore, HLA-C*07 has been linked to ATLL progression 49 . The HLA-A*24 allele was also presented as a relevant susceptibility marker, being more frequent among HAM patients despite its association with lower HTLV-1 proviral load 50 .

Beyond HLA variants, microRNA profiling revealed seven miRNAs (hsa-miR-29b-3p, hsa-let-7f-5p hsa-miR-32-5p, hsa-miR-192-5p, hsa-miR-141-3p, hsa-miR-342-3p, and hsa-miR-140-5p) that were significantly downregulated in patients with HAM. These miRNAs were correlated with PVL and involved in the regulation of cellular senescence and modulation of inflammatory pathways, such as NF-κB and TGF-β 18 . These findings are consistent with another study that reported reduced TGF-β1 and increased BDNF levels in carriers and patients with HAM, reinforcing the hypothesis that dysfunction of anti-inflammatory pathways may contribute to disease progression 34 . Another study identified five dysregulated small RNAs (miR-451a, miR-155, miR-146a, miR-150, and miR-223) associated with key signaling pathways relevant to both HAM and ATLL, including TGF-β, p53, MAPK, PI3K, and Wnt pathways, which are critical regulators of proliferation, apoptosis, and inflammatory responses 51 . Further experimental validation in ATLL demonstrated significant upregulation of miR-451a and downregulation of miR-142-3p in ATLL patients compared to healthy controls. These miRNAs were shown to regulate genes involved in immune modulation, cell migration, angiogenesis, and oncogenic signaling pathways, reinforcing the role of disrupted miRNA-mediated regulatory networks as a shared molecular feature underlying HTLV-1-associated disease progression, including both HAM and ATLL 52 .

Biomarkers derived from metabolic profiling have also gained attention due to their clinical applicability, as they reflect systemic aspects of the inflammatory response. In addition to these, GlycA—a nuclear magnetic resonance-derived marker that reflects circulating glycosylated acute-phase proteins—has recently emerged as a serum indicator of chronic systemic inflammation 53 . In asymptomatic individuals, a positive correlation between IL-6 and GlycA levels has been observed, consistent with previous findings in uninfected populations. However, this correlation was not detected in patients with HAM, possibly due to the predominantly neuroinflammatory nature of the disease. The lack of association between GlycA and IL-6 may indicate a distinct inflammatory pattern in this group 16 .

Oxidative stress has also been associated with HTLV-1 infection, particularly in cases progressing to HAM. Antioxidant enzymes, including superoxide dismutase (SOD), catalase (CAT), and glutathione peroxidase (GPX), exhibit reduced activity in PLHTLV-1, whereas serum levels of major oxidative damage markers—such as malondialdehyde and nitric oxide—are elevated, especially in symptomatic individuals. This imbalance suggests a progressive increase in oxidative stress across different infection stages. These findings highlight a potential therapeutic window based on modulation of the redox system to prevent or delay clinical progression of the infection 7 . Moreover, these results align with previous reports of reduced total antioxidant capacity in PLHTLV-1 54 , as well as with similar patterns observed in other chronic viral infections, such as human immunodeficiency virus (HIV) 55 . HIV-1 studies have demonstrated that, during the viral latency phase, specific metabolic pathways exhibit reduced activity, raising the hypothesis that similar mechanisms may also occur in HTLV-1 infection 56 .

The consistent identification of multiple biomarkers in this systematic review emphasizes the possibility of developing an integrated panel combining proteomic, genomic, and metabolic markers. Such an approach would provide a strategic framework for mapping research gaps and priorities, thereby facilitating the identification of underexplored areas. Among the omics sciences analyzed in this study, metabolomics was the least investigated, with only three metabolic markers identified, all associated with HAM, and only two obtained using metabolomic methods (nuclear magnetic resonance and mass spectrometry) 16 , 36 .

Given its potential to unravel complex molecular mechanisms, metabolomics represents a valuable approach to elucidate how alterations in metabolic pathways directly influence HTLV-1 pathogenesis and clinical outcomes. Exploratory strategies, such as untargeted metabolomics, enable the simultaneous profiling of multiple metabolites within a single sample 57 , facilitating the identification of novel patterns and correlations among metabolic pathways, previously reported consistent biomarkers, and clinical data. Hence, metabolomics may serve as a valuable tool to address current gaps in understanding pathophysiological mechanisms underlying HTLV-1 infection.

Limitations of this review include the heterogeneity of study designs, which limits direct comparisons between findings, as well as the predominance of small sample sizes. Methodological quality, assessed using NHLBI tools, indicated that 15 studies were rated as fair, mainly due to the lack of sample size justification and insufficient information regarding assessor blinding, control of confounding variables, and loss to follow-up. Furthermore, compared to HAM, considerably fewer studies focused on ATLL. This discrepancy may be explained by the short survival of patients who progress to aggressive clinical forms of ATLL—approximately eight to 10 months for the acute, lymphoma, and unfavorable chronic subtypes, and two to four years for indolent forms, such as favorable chronic and smoldering subtypes. This limited time frame poses challenges for cohort establishment and biological sample collection 23 , 58 . Despite these limitations, the overall quality and relevance of this systematic review remain robust. The inclusion of high and moderate methodological quality, supported by rigorous assessment, combined with a detailed narrative analysis and adherence to PRISMA guidelines 14 , ensures reliability of the data. Gray literature was excluded and language was restricted to minimize bias and avoid the inclusion of unvalidated data. While gray sources may provide additional insights, their limited quality and reproducibility justified exclusion, with relevant references still used to mitigate this limitation.

Finally, considering the observed gaps, particularly regarding metabolic markers, future studies are recommended to prioritize longitudinal cohorts of asymptomatic individuals to identify predictive markers during the preclinical phase 59 . The integration of machine learning-based approaches for omics data analysis could enhance the identification of predictive patterns, as has already been demonstrated in hematologic malignancies 60 . International collaborations, particularly in endemic regions such as Brazil and Japan, as exemplified by Silva et al . 33 , are important strategies to overcome sample size limitations and reduce geographic bias. Ultimately, clinical validation of identified markers will require well-designed multicenter prospective trials.

CONCLUSION

This systematic review, based on 35 studies of high and moderate methodological quality, compiled genomic, proteomic, and metabolic biomarkers with potential applications in risk stratification, prognosis, and therapeutic monitoring of progression to ATLL and HAM in PLHTLV-1. Although PVL remains the primary clinical marker, its limited specificity underscores the need for combination with other biomarkers such as IFN-γ, CXCL10, Neopterin, and sTNFR2. This review also reinforces that no single marker is sufficient; instead, multiparametric panels of correlated biomarkers are required. Metabolomics, although still underexplored, appears to be a promising approach to fill gaps in future investigations, enabling the discovery of novel markers with the understanding of metabolic pathways and infection pathogenesis, as well as their correlation with previously identified markers. The consistent findings across omics categories support the feasibility of developing integrated biomarker panels to enhance clinical decision-making. Additionally, longitudinal studies integrating omics approaches and artificial intelligence are needed to improve predictive accuracy, enable early intervention, and ultimately guide personalized clinical management in HTLV-1 infection, particularly during the asymptomatic phase.

Funding Statement

FUNDING: This study was supported by the Royal Society and Bolsa de Produtividade em Pesquisa of the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico – CNPq, category 1D, process Nº 301624/2022-4.

Footnotes

FUNDING: This study was supported by the Royal Society and Bolsa de Produtividade em Pesquisa of the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico – CNPq, category 1D, process Nº 301624/2022-4.

DATA AVAILABILITY :

The complete anonymized dataset supporting the findings of this study is included within the article.

REFERENCES

1. Coffin JM. The discovery of HTLV-1, the first pathogenic human retrovirus. Proc Natl Acad Sci U S A. 2015;112:15525–15529. doi: 10.1073/pnas.1521629112.

2. Legrand N, McGregor S, Bull R, Bajis S, Valencia BM, Ronnachit A, et al. Clinical and public health implications of human T-lymphotropic virus type 1 infection. Clin Microbiol Rev. 2022;35:e0007821. doi: 10.1128/cmr.00078-21.

3. Gessain A, Cassar O. Epidemiological aspects and world distribution of HTLV-1 infection. 388Front Microbiol. 2012;3 doi: 10.3389/fmicb.2012.00388.

4. Tu JJ, Maksimova V, Ratner L, Panfil AR. The past, present, and future of a human T-cell leukemia virus type 1 vaccine. 897346Front Microbiol. 2022;13 doi: 10.3389/fmicb.2022.897346.

5. Saffari M, Rahimzada M, Mirhosseini A, Ghezaldasht SA, Valizadeh N, Moshfegh M, et al. Coevolution of HTLV-1-HBZ, tax, and proviral load with host IRF-1 and CCNA-2 in HAM/TSP patients. 105337Infect Genet Evol. 2022;103 doi: 10.1016/j.meegid.2022.105337.

6. Eusebio-Ponce E, Anguita E, Paulino-Ramirez R, Candel FJ. HTLV-1 infection: an emerging risk: pathogenesis, epidemiology, diagnosis and associated diseases. Rev Esp Quimioter. 2019;32:485–496.

7. Yaghoubi N, Youssefi M, Rafat Panah H, Jarahi L, Zahedi Avval F. Evaluation of antioxidant status and oxidative stress markers in HTLV-1 infected individuals: correlation with the severity of virus-induced complications. 001408J Med Microbiol. 2022;71 doi: 10.1099/jmm.0.001408.

8. Bangham CR, Cook LB, Melamed A. HTLV-1 clonality in adult T-cell leukaemia and non-malignant HTLV-1 infection. Semin Cancer Biol. 2014;26:89–98. doi: 10.1016/j.semcancer.2013.11.003.

9. Rodrigues ES, Salustiano S, Santos EV, Slavov SN, Picanço-Castro V, Maçonetto JM, et al. Monitoring of HTLV-1-associated diseases by proviral load quantification using multiplex real-time PCR. J Neurovirol. 2022;28:27–34. doi: 10.1007/s13365-020-00924-2.

10. Harding D, Rosadas C, Tsoti SM, Heslegrave A, Stewart M, Kelleher P, et al. Refining the risk of HTLV-1-associated myelopathy in people living with HTLV-1: identification of a HAM-like phenotype in a proportion of asymptomatic carriers. J Neurovirol. 2022;28:473–482. doi: 10.1007/s13365-022-01088-x.

11. Furtado MS, Andrade RG, Romanelli LC, Ribeiro MA, Ribas JG, Torres EB, et al. Monitoring the HTLV-1 proviral load in the peripheral blood of asymptomatic carriers and patients with HTLV-associated myelopathy/tropical spastic paraparesis from a Brazilian cohort: ROC curve analysis to establish the threshold for risk disease. J Med Virol. 2012;84:664–671. doi: 10.1002/jmv.23227.

12. Asigbee TW, Nakamura-Hoshi M, Kuse N, Ishii H, Ishikawa K, Kawana-Tachikawa A, et al. Virus-host immune interaction in asymptomatic HTLV-1 carriers. Microbiol Spectr. 2025;13:e0250724. doi: 10.1128/spectrum.02507-24.

13. Babu M, Snyder M. Multi-omics profiling for health. 100561Mol Cell Proteomics. 2023;22 doi: 10.1016/j.mcpro.2023.100561.

14. Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, et al. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. n71BMJ. 2021;372 doi: 10.1136/bmj.n71.

15. National Heart, Lung, and Blood Institute Study quality assessment tools. [cited 2026 Feb 3]. https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools

16. Assone T, Menezes SM, Gonçalves FT, Folgosi VA, Prates GS, Dierckx T, et al. Systemic cytokines and GlycA discriminate disease status and predict corticosteroid response in HTLV-1-associated neuroinflammation. 293J Neuroinflammation. 2022;19 doi: 10.1186/s12974-022-02658-w.

17. Assone T, Menezes SM, Gonçalves FD, Folgosi VA, Braz M, Smid J, et al. IL-10 predicts incident neuroinflammatory disease and proviral load dynamics in a large Brazilian cohort of people living with human T-lymphotropic virus type 1. 1416476Front Immunol. 2024;15 doi: 10.3389/fimmu.2024.1416476.

18. Souza DR, Pessôa R, Nukui Y, Pereira J, Marcusso RN, Oliveira AC, et al. Identification of miRnas with possible prognostic roles for HAM/TSP. 2230015Virulence. 2023;14 doi: 10.1080/21505594.2023.2230015.

19. Katsuya H, Nakamura H, Maeda A, Ishii K, Nagaie T, Sano H, et al. HTLV-1 cell-free DNA in plasma as a potential biomarker in HTLV-1 carriers and adult T-cell leukemia-lymphoma. EJHaem. 2023;4:733–737. doi: 10.1002/jha2.725.

20. Santana BB, Queiroz MA, Cerveira RA, Rodrigues CM, Amoras ES, Costa CA, et al. Low Annexin A1 level in HTLV-1 infected patients is a potential biomarker for the clinical progression and diagnosis of HAM/TSP. 219BMC Infect Dis. 2021;21 doi: 10.1186/s12879-021-05917-y.

21. Guerrero CL, Yamashita Y, Miyara M, Imaizumi N, Kato M, Sakihama S, et al. Proteomic profiling of HTLV-1 carriers and ATL patients reveals sTNFR2 as a novel diagnostic biomarker for acute ATL. Blood Adv. 2020;4:1062–1071. doi: 10.1182/bloodadvances.2019001429.

22. Souza FS, Freitas NL, Gomes YC, Torres RC, Echevarria-Lima J, Silva-Filho IL, et al. Following the clues: usefulness of biomarkers of neuroinflammation and neurodegeneration in the investigation of HTLV-1-associated myelopathy progression. 737941Front Immunol. 2021;12 doi: 10.3389/fimmu.2021.737941.

23. Kato M, Imaizumi N, Tanaka R, Mizuguchi M, Hayashi M, Miyagi T, et al. Elevation of the plasma levels of TNF receptor 2 in association with those of CD25, OX40, and IL-10 and HTLV-1 proviral load in acute adult T-cell leukemia. 751Viruses. 2022;14 doi: 10.3390/v14040751.

24. Nakahata S, Syahrul C, Nakatake A, Sakamoto K, Yoshihama M, Nishikata I, et al. Clinical significance of soluble CADM1 as a novel marker for adult T-cell leukemia/lymphoma. Haematologica. 2021;106:532–542. doi: 10.3324/haematol.2019.234096.

25. Takenouchi N, Tanaka M, Sato T, Yao J, Fujisawa JI, Izumo S, et al. Expression of TSLC1 in patients with HAM/TSP. J Neurovirol. 2020;26:404–414. doi: 10.1007/s13365-020-00838-z.

26. Yamada A, Yasunaga JI, Liang L, Zhang W, Sunagawa J, Nakaoka S, et al. Anti-HTLV-1 immunity combined with proviral load as predictive biomarkers for adult T-cell leukemia-lymphoma. Cancer Sci. 2024;115:310–320. doi: 10.1111/cas.15997.

27. Joris T, Haddow J, Taylor GP, Cook LB, Rowan AG. Detection of HTLV-1 proviral DNA in cell-free DNA: potential for non-invasive monitoring of Adult T cell leukaemia/lymphoma using liquid biopsy? 1150285Front Immunol. 2023;14 doi: 10.3389/fimmu.2023.1150285.

28. Yanagida E, Miyoshi H, Takeuchi M, Yoshida N, Nakashima K, Yamada K, et al. Clinicopathological analysis of immunohistochemical expression of CD47 and SIRPa in adult T-cell leukemia/lymphoma. Hematol Oncol. 2020;38:680–688. doi: 10.1002/hon.2768.

29. Forghani-Ramandi MM, Mostafavi B, Bahavar A, Dehghankar M, Siami Z, Mozhgani SH. Illuminating (HTLV-1)-induced adult T-cell leukemia/lymphoma transcriptomic signature: a systems virology approach. 199237Virus Res. 2023;338 doi: 10.1016/j.virusres.2023.199237.

30. Shadabi S, Delrish N, Norouzi M, Ehteshami M, Habibian-Sezavar F, Pourrezaei S, et al. Comprehensive high-throughput meta-analysis of differentially expressed microRNAs in transcriptomic datasets reveals significant disruption of MAPK/JNK signal transduction pathway in Adult T-cell leukemia/lymphoma. 49Infect Agent Cancer. 2021;16 doi: 10.1186/s13027-021-00390-3.

31. Shayeghpour A, Forghani-Ramandi M-M, Solouki S, Hosseini A, Hosseini P, Khodayar S, et al. Identification of novel miRNAs potentially involved in the pathogenesis of adult T-cell leukemia/lymphoma using WGCNA followed by RT-qPCR test of hub genes. 12Infect Agent Cancer. 2023;18 doi: 10.1186/s13027-023-00492-0.

32. Mozhgani SH, Zarei Ghobadi M, Norouzi M, Rahimi H, Valizadeh N, Teymoori-Rad M, et al. Signaling factors potentially associated to the pathogenesis of adult T-cell leukemia /lymphoma: a network-analysis and novel findings assessment. 198875Virus Res. 2022;319 doi: 10.1016/j.virusres.2022.198875.

33. Silva SJ, Cabral-Castro MJ, Faria LC, Rosadas C, Araújo MF, Dutra AC, et al. CXCL-10 in cerebrospinal fluid detects neuroinflammation in HTLV-1-associated myelopathy with high accuracy. 89Viruses. 2025;17 doi: 10.3390/v17010089.

34. Freitas NL, Gomes YC, Souza FD, Torres RC, Echevarria-Lima J, Leite AC, et al. Lessons from the cerebrospinal fluid analysis of HTLV-1-infected individuals: biomarkers of inflammation for HAM/TSP development. 2146viruses. 2022;14 doi: 10.3390/v14102146.

35. Rosadas C, Zetterberg H, Heslegrave A, Haddow J, Borisova M, Taylor GP. Neurofilament light in CSF and plasma is a marker of neuronal damage in HTLV-1-associated myelopathy and correlates with neuroinflammation. Neurol Neuroimmunol Neuroinflamm. 2021;8:e1090. doi: 10.1212/NXI.0000000000001090.

36. Yamauchi J, Sato T, Yagishita N, Araya N, Hasegawa D, Tsutsumi S, et al. Use of cerebrospinal fluid CXCL10 and neopterin as biomarkers in HTLV-1-associated myelopathy/tropical spastic paraparesis treated with steroids. J Neurol Neurosurg Psychiatry. 2020;91:321–323. doi: 10.1136/jnnp-2019-321955.

37. Sato T, Coler-Reilly A, Utsunomiya A, Araya N, Yagishita N, Ando H, et al. CSF CXCL10, CXCL9, and neopterin as candidate prognostic biomarkers for HTLV-1-associated myelopathy/tropical spastic paraparesis. PLoS Negl Trop Dis. 2013;7:e2479. doi: 10.1371/journal.pntd.0002479.

38. Espíndola OM, Oliveira LC, Ferreira PM, Leite AC, Lima MA, Andrada-Serpa MJ. High IFN-?/IL-10 expression ratio and increased frequency of persistent human T-cell lymphotropic virus type 1-infected clones are associated with human T-cell lymphotropic virus type 1-associated myelopathy/tropical spastic paraparesis development. Intervirology. 2015;58:106–114. doi: 10.1159/000371766.

39. Yamano Y, Sato T. Clinical pathophysiology of human T-lymphotropic virus-type 1-associated myelopathy/tropical spastic paraparesis. 389Front Microbiol. 2012;3 doi: 10.3389/fmicb.2012.00389.

40. Apoliano CF, Assone T, Silva BC, Corral MA, Oliveira AC, Fonseca LA, et al. Interferon-? secretion enzyme-linked immunospot assay determined among human T cell lymphotropic virus type 1-infected subjects: a potential laboratory marker for early HTLV-1-associated myelopathy/tropical spastic paraparesis diagnosis. AIDS Res Hum Retroviruses. 2020;36:6–7. doi: 10.1089/AID.2018.0290.

41. Prates G, Assone T, Corral M, Baldassin MP, Mitiko T, Sales FC, et al. Prognosis markers for monitoring HTLV-1 neurologic disease. Neurol Clin Pract. 2021;11:134–140. doi: 10.1212/CPJ.0000000000000866.

42. Carvalho EM, Bacellar O, Porto AF, Braga S, Galvão-Castro B, Neva F. Cytokine profile and immunomodulation in asymptomatic human T-lymphotropic virus type 1-infected blood donors: J Acquir Immune Defic Syndr. 2001;27:1–6. doi: 10.1097/00126334-200105010-00001.

43. Mitre E, Thompson RW, Carvalho EM, Nutman TB, Neva FA. Majority of interferon-?-producing CD4 + cells in patients infected with human T cell lymphotrophic virus do not express tax protein. J Infect Dis. 2003;188:428–432. doi: 10.1086/376529.

44. Soltani S, Mozhgani SH, Siri G, Emadi MS, Foroushani AR, Jazayeri SM, et al. High expression of inflammatory cytokines and chemokines in human T-lymphotropic virus 1-associated adult T-cell leukemia/lymphoma. Jundishapur J Microbiol. 2022;15:e132348

45. Gomes YC, Freitas NL, Souza FS, Sandim V, Pereira DA, Nogueira FC, et al. Chitotriosidase 1 in the cerebrospinal fluid as a putative biomarker for HTLV-1-associated myelopathy/tropical spastic paraparesis (HAM/TSP) progression. 949516Front Immunol. 2022;13 doi: 10.3389/fimmu.2022.949516.

46. Manzarinejad M, Vahidi Z, Boostani R, Khadem-Rezaiyan M, Rafatpanah H, Zemorshidi F. Pentraxin 3, a serum biomarker in human T-cell lymphotropic virus type-1-associated myelopathy patients and asymptomatic carriers. Med Microbiol Immunol. 2023;212:271–278. doi: 10.1007/s00430-023-00770-z.

47. Saeidi M, Vahidi Z, Nahayati MA, Rezaiyan MK, Zemorshidi F, Mahdifar M, et al. XCL1, a serum biomarker in neurological diseases; HTLV-1-associated myelopathy and multiple sclerosis. 105962Microb Pathog. 2023;174 doi: 10.1016/j.micpath.2022.105962.

48. Penova M, Kawaguchi S, Yasunaga JI, Kawaguchi T, Sato T, Takahashi M, et al. Genome wide association study of HTLV-1-associated myelopathy/tropical spastic paraparesis in the Japanese population. Proc Natl Acad Sci U S A. 2021;118:e2004199118. doi: 10.1073/pnas.2004199118.

49. Benencio P, Fraile Gonzalez SA, Ducasa N, Page K, Berini CA, Biglione MM. HLA-B*35 as a new marker for susceptibility to human T-cell lymphotropic virus type 1 (HTLV-1) associated myelopathy/tropical spastic paraparesis (HAM/TSP) in patients living in Argentina. 29Retrovirology. 2020;17 doi: 10.1186/s12977-020-00536-y.

50. Tanaka M, Takenouchi N, Arishima S, Matsuzaki T, Nozuma S, Matsuura E, et al. HLA-A*24 increases the risk of HTLV-1-associated myelopathy despite reducing HTLV-1 proviral load. 6858Int J Mol Sci. 2024;25 doi: 10.3390/ijms25136858.

51. Nascimento A, Souza DR, Pessôa R, Pietrobon AJ, Nukui Y, Pereira J, et al. Global expression of noncoding RNome reveals dysregulation of small RNAs in patients with HTLV-1-associated adult T-cell leukemia: a pilot study. 4Infect Agent Cancer. 2021;16 doi: 10.1186/s13027-020-00343-2.

52. Ghobadi MZ, Emamzadeh R, Mozhgani SH. Deciphering microRNA-mRNA regulatory network in adult T-cell leukemia/lymphoma; the battle between oncogenes and anti-oncogenes. PloS One. 2021;16:e0247713. doi: 10.1371/journal.pone.0247713.

53. Guo Y, Liu Q, Zheng Z, Qing M, Yao T, Wang B, et al. Genetic association of inflammatory marker GlycA with lung function and respiratory diseases. 3751Nat Commun. 2024;15 doi: 10.1038/s41467-024-47845-w.

54. Shomali S, Avval FZ, Boostani R, Jarahi L, Youssefi M. Serum total antioxidant capacity status of HTLV-1 infected patients. Acta Virol. 2015;59:199–203. doi: 10.4149/av_2015_02_199.

55. Ivanov AV, Valuev-Elliston VT, Ivanova ON, Kochetkov SN, Starodubova ES, Bartosch B, et al. Oxidative stress during HIV infection: mechanisms and consequences. 8910396Oxid Med Cell Longev. 2016;2016 doi: 10.1155/2016/8910396.

56. Shytaj IL, Procopio FA, Tarek M, Carlon-Andres I, Tang H, Goldman AR, et al. Glycolysis downregulation is a hallmark of HIV-1 latency and sensitizes infected cells to oxidative stress. EMBO Mol Med. 2021;13:e13901. doi: 10.15252/emmm.202013901.

57. Canuto GA, Costa JL, Cruz PL, Souza AR, Faccio AT, Klassen A, et al. Metabolômica: definições, estado-da-arte e aplicações representativas. Quim Nova. 2018;41:75–91.

58. Katsuya H, Ishitsuka K, Utsunomiya A, Hanada S, Eto T, Moriuchi Y, et al. Treatment and survival among 1594 patients with ATL. Blood. 2015;126:2570–2577. doi: 10.1182/blood-2015-03-632489.

59. Odom JD, Sutton VR. Metabolomics in clinical practice: improving diagnosis and informing management. Clin Chem. 2021;67:1606–1617. doi: 10.1093/clinchem/hvab184.

60. Halicki P, Ciesluk K, Piszcz J. Metabolic profile analysis in hematological malignancies. Acta Haematol Pol. 2024;55:22–33.

Associated Data

Data Availability Statement

The complete anonymized dataset supporting the findings of this study is included within the article.