Prognosis Markers for Monitoring HTLV-1 Neurologic Disease

Abstract

Background

Human T-cell lymphotropic virus type 1 (HTLV-1) infection is associated not only with some severe manifestations, such as HTLV-1-associated myelopathy (HAM) and ATLL, but also with other, less severe conditions. Some studies have reported neurologic manifestations that did not meet all the criteria for the diagnosis of HAM in individuals infected with HTLV-1; these conditions may later progress to HAM or constitute an intermediate clinical form, between asymptomatic HTLV-1 carriers and those with full myelopathy. This study evaluated the prognostic value and looked for a possible association of those parameters with the intermediate syndrome (IS) status and HAM status.

Methods

Proviral load (PVL), spontaneous lymphoproliferation, interferon (IFN)-γ spontaneous production was quantified in samples of asymptomatic and HAM patients, as well as patients with IS.

Results

The critical age range was 50–60 years for IS outcome and more of 60 years for HAM outcome, with an increased risk of 2.5-fold for IS and 6.8-fold for HAM. IFN-γ was increased in patients with IS compared with asymptomatic carriers (ACs) (p = 0.007) and in patients with HAM compared with ACs (p = 0.03). Lymphoproliferation was increased in patients with HAM vs ACs (p = 0.0001) and patients with IS (p = 0.0001). PVL was similar between groups.

Conclusion

IFN-γ has high specificity of prediction of subject remain asymptomatic compared with PVL and lymphoproliferation assay tests. IFN-γ has been shown to be a biomarker of progression to intermediate stage and to HAM. The association of other markers with manifestations associated with HTLV-1 infection that does not meet the HAM criteria should be verified.

It is estimated that 5–10 million people worldwide carry the human T-cell lymphotropic virus type 1 (HTLV-1)¹ and may cause the HTLV-1–associated myelopathy (HAM). It is a chronic disease characterized by a progressive demyelination of the spinal cord and affects 1%–2% of patients with HTLV-1^2,3 in the fourth and fifth decades of life.^4–6

Despite the relatively low rates of HTLV-1 morbidity, many HTLV-1–infected patients may present complaints that do not fulfill the criteria for HAM.⁶ These conditions may later progress to HAM or constitute an intermediate clinical form, between asymptomatic HTLV-1 carriers and those with full myelopathy.^7,8 A study conducted in Rio de Janeiro reported neurologic abnormalities in 36% of HTLV-1–positive individuals among blood donors.⁸ Sensory, motor, urinary, or autonomic manifestations are the most common complaints, but they have not been well defined as new clinical outcomes related to early inflammation process.^9–11

In studies comparing HTLV-1–infected carriers with uninfected control subjects was found the presence of higher frequency of motor and bladder dysfunctions in patients with HTLV-1.^8–14 In some patients, such symptoms precede the diagnosis of HAM for years.⁸ These data suggest that individuals with HTLV-1 may exhibit a wide variety of neurologic manifestations distinct from the classic HAM. Thus, this study evaluated the prognostic value of proviral load (PVL), spontaneous lymphoproliferation, and interferon (IFN)-γ spontaneous production for the prediction of HTLV-1–related symptomatic manifestations.

Methods

The outpatient clinic of the Institute of Infectious Diseases “Emílio Ribas” (IIER) cares for 659 HTLV-1–infected individuals. From that cohort, 328 volunteers agreed to participate in the current study, which ran from January 2017 to September 2019. From the included patients, 130 were asymptomatic, 63 have the intermediate syndrome (IS), and 135 had HAM. Age, sex, and self-reported skin color data were collected from the hospital records. Participants were subdivided according to the degree of neurologic disability determined by neurologists. For this study, patients with HTLV-1 who presented more 3 signs, such as dermatologic, ophthalmologic, rheumatologic, urinary, dysautonomic, and oral changes, were classified as patients with IS.⁷

The clinical diagnosis of HAM was made using criteria described by Guide of clinical management of HTLV patient⁴ gradual onset of progressive spastic paraparesis; symmetrical, predominantly upper motor neuron disorders with mild sensory and bladder disorders; high antibody titers to HTLV-1 antigens in the serum and CSF.

All individuals included were older than 16 years, had a positive serology for HTLV-1, and were in regular follow-up at the IIER outpatient clinic treated or not with methylprednisolone. Volunteers with myelopathies non–HTLV-1–related, illiterate, and/or incapacitated by mental disorders to understand the informed consent form, pregnant women, and patients coinfected with other viral infections (HIV, HBV, and HCV) were excluded.

Standard Protocol Approvals, Registrations, and Patient Consent Forms

The study was approved by the Ethics Committees of the Institute of Tropical Medicine of São Paulo #2012/133, Institute of Infectious Diseases “Emilio Ribas” #13/2011, and University of São Paulo School of Medicine #356/12.

HTLV-1 PVL

HTLV-1 PVL was quantified in 118 AC, 61 IS, and 135 HAM samples (total = 314). The quantification was performed by real-time PCR, using primers and probes targeting the pol gene: SK110 5′-CCCTACAATCCAACCAGCTCAG-3′ and SK111 5′-GTGGTGAAGCTGCCATCGGGTTTT-3′ and the internal HTLV-1 TaqMan probe 5′-CTTTACTGACAAACCCGACCTACCCATGGA-3′, selected using Oligo (National Biosciences). All samples were run in duplicate, and results were expressed as HTLV-1 DNA copies/10⁴ peripheral blood mononuclear cells (PBMCs), following the protocol described elsewhere,¹⁵ including collection and preparation of samples.

Spontaneous T-cell Lymphoproliferation Aassay

The levels of spontaneous lymphoproliferation assay (LPA) were determined in 112 AC, 58 IS, and 99 HAM samples (total = 269). For the LPA, a suspension of 2 × 10⁶ PBMCs was cultured with tritiated thymidine. Phytohemagglutinin (PHA) was used as a positive control was performed as described previously,¹⁶ including collection and preparation of samples. The results were expressed in counts per minute using triplicate samples.

Detection of Spontaneous IFN-γ Production Using Spot-forming Cells by ELISpot Assay (ELISpot)

Quantification of IFN-γ was performed for 47 AC, 44 IS, and 51 HAM samples (total = 142) by ImmunoSpot with bound enzyme (ELISpot). PBMCs were cultured in triplicate at a concentration of 2 × 10⁶ cells to verify spontaneous IFN-γ production. PHA was used as a positive control, whereas PHA without cells served as a negative control. Details and specifications of the materials used, including collection and preparation of samples, are described in a base article for IFN-γ quantification.¹⁷

Statistical Analysis

Data of age, sex, and self-reported skin color were treated with mean, SD, absolute and relative frequencies, and odds ratio (OR). The ANOVA test was used for comparison between groups of PVL, LPA, and IFN data. The receiver operating characteristic (ROC) curve was used to determine the sensitivity and specificity of laboratory data and the Fisher test to determine the OR and confidence interval (CI). p Value <0.05 was set as the significance threshold. Analyzes were performed using GraphPad Software 5.0 (La Jolla, CA).

Data Availability

All the data are available for this research on request to the corresponding author (J.C.).

Results

There was a wide distribution of age frequencies in the asymptomatic carrier (AC) group; however, 50.8% of subjects with IS are aged between 50 and 60 years, and 45.1% of subjects with HAM are aged between 60 and 70 years. Mean age was 49 years (SD ± 13.4) for the AC group, 51 years (SD ± 14.0) for the IS group, and 63 years (SD ± 13.0) for the HAM group. The latter group was older than ACs and patients with IS (p < 0.001). Most of the participants were older than 45 year, and the risk is increased 2.5 times for IS status (95% CI 1.2–4.9) and 6.8 times for HAM development (95% CI 3.5–13.1) (table 1). Female sex was more frequent in all groups, from 65% in HAM, 77% in AC and IS. Self-reported white skin color as according to the criteria of the Instituto Brasileiro de Geografia e Estatística, are majority among the ACs (53.8%), patients with IS (45.7%), and patients with HAM (55.5%). Sex e skin color did not increase the risk of disease development (table 1).

Table 1.

Risk Association of Age, Sex, and Skin Color Self-reported With Outcome

There was no difference among groups as to HTLV-1 PVLs (p > 0.05) (figure 1A). LPA was higher for patients with HAM compared with asymptomatic subjects and patients with IS (p < 0.001). Comparing asymptomatic vs IS, there was no difference (figure 1B). The spontaneous production of IFN-γ was higher for both IS and HAM compared with asymptomatic patients (p = 0.007) and (p = 0.03), respectively (figure 1C).

Figure 1. Biomarkers Associated with HAM Development.

AC = asymptomatic carrier; HAM = HTLV-1–associated myelopathy; HTLV-1 = human T-cell lymphotropic virus type 1; IFN = interferon; IS = intermediate syndrome; PBMC = peripheral blood mononuclear cell; SFC = spot-forming cell.

ROC Analyses and Optimal Cutoffs of LPA and IFN-γ Levels

In the ROC curve analysis, LPA has 47.0% of sensitivity and 72.4% of specificity, with a cutoff of 1,108 counts per minute (p = 0.0001; AUC = 0.6524 (figure 2A). The positive predictive value (PPV) is 47.0%, the negative predictive value (NPV) is 72.3%, and accuracy is 49.0%. IFN-γ has 44.2% of sensitivity and 89.4% of specificity, with a cutoff of 1,659 SFC/2 × 10⁴ cells (p = 0.0001; AUC = 0.7288 (figure 2B). The PPV was 44.2%, the NPV was 89.3%, and accuracy was 59.1%.

Figure 2. ROC Curve Analysis of LPA and IFN-γ.

AUC = area under the curve; LPA = lymphoproliferation assay; IFN = interferon; ROC = receiver operating characteristic.

In OR analysis, subjects with LPA above cutoff have an increased risk of 2.5-fold for development of IS (95% CI 1.9–5.4; p value = 0.01) and an increased risk of 4.3-fold for development of HAM (CI 2.1–8.89 p = 0.0001). Subjects with IFN-γ above the cutoff have an increased risk of 7.67-fold for development of IS (CI 2.5–23.0; p = 0.0001) and an increased risk of 5.8-fold for development of HAM (CI 1.9–17.3; p = 0.0006) (table 2).

Table 2.

OR and 95% CI Calculated from Biomarker Cutoffs

Discussion

PVL has been widely used as the main prognostic marker related to the virus, indicating correlation with HAM and time of disease progression.^4,18–20 We found it difficult to use the PVL as a gold standard predictor of risk for development of neurologic damage. It was demonstrated that in the chronic phase of HTLV-1 infection, PVL becomes stable in most infected individuals and may vary among individuals,^21,22 where HTLV-1 PVL was assessed over 10 years.²¹

Spontaneous lymphoproliferation is one of the immunologic hallmarks of HAM and is considered to be an important factor related to the pathogenesis.^23–25 We found that a higher level of lymphoproliferation was associated with the development of HAM, although we could not demonstrate its association with the IS, which does not favor the use of this marker in monitoring asymptomatic patients.

IFN-γ is described as capable of discriminating patients with HAM from asymptomatic HTLV-1 carriers^26,27; however, few studies have investigated the relationship with intermediate state.¹⁷ The spontaneous production of IFN-γ was increased in patients with IS and HAM, demonstrating that if subjects with IS were considered as asymptomatic, they tend to polarize to HAM if classified by IFN-γ quantification, which agrees with clinical evaluation. In vitro production of IFN-γ cytokine may be a stronger marker candidate for monitoring asymptomatic and intermediate-stage patients in risk assessment for HAM progression. Previous studies have demonstrated that HTLV-1–associated symptoms occur in the fourth and fifth decades of life of patients.^20,28 We verified the relationship of age with disease stages. Intermediate-stage patients were also of intermediate age.

HTLV-1 infection causes disorders in the immune system, and some cytokines are involved in the pathogenesis of those disorders. The exacerbated production of IFN-γ,^19,20 tumor necrosis factor-α,^26,29 interleukin-2 (IL-2), and interleukin-1 (IL-1), which are produced by Th1 cells, plays a critical role in cellular immunity.^30,31 This immune activation could be the initial factor of damage on the spinal cord myelin.^32,33

Classically, HTLV-1 infection can induce spontaneous T-cell proliferation due to increased secretion of IL-2 and IL-2R expression,³¹ potentially host factors responsible for the myelin membrane damage. However, such cellular activation may alter the function and/or differentiation of CD4 T cells.³¹ This chronic proinflammatory process may be considered a precursor of CNS damage.^32,34,35 Therefore, spontaneous production of IFN-γ and other immunologic mediators should be investigated for their ability to predict the development of neurologic lesions, especially in the early stages.^17,30,36

Despite the complexity, the host vs HTLV-1 interaction is of fundamental importance in understanding the clinical prognosis of asymptomatic carriers. Current diagnostic and treatment options are not adequate to determine the risk of progression to HAM.^22,37 Few markers have been evaluated in clinical practice, usually by isolated groups of researchers, with contradictory results and high costs. In turn, lymphoproliferation^33,38 indicates the activation of the immune system, which may justify the use of corticosteroids in the clinical practice to slow progression of HAM.^39,40 However, LPA presents issues of complex elaboration and requires the use of radioactive materials, making it difficult to perform in routine laboratories in large scale. Because of the neurologic findings that may be associated with the development of HAM in individuals previously considered free of the disease, according to current criteria, there is a need to monitor individuals infected by HTLV-1 for the presence of early markers of clinical onset. Finally, these findings may be important for clinical practice. Gamma interferon levels can distinguish clinical groups with potential for predict neurologic damage in asymptomatic patients. The verification of IFN-γ in conjunction with clinical assessment, age between 40 and 60 years and female sex, may support the initiation of treatment with methylprednisolone in attempt for delay or interrupt the HAM progression.

Appendix. Authors

Study Funding

No targeted funding reported.

Disclosure

The authors report no disclosures relevant to the manuscript. Full disclosure form information provided by the authors is available with the full text of this article at Neurology.org/cp.