Neonatal Inflammatory Profiles Associated With Perinatally Acquired Hepatitis C Viral Infection

Abstract

In this nested case–control study of infants with in utero hepatitis C exposure, we assessed 21 plasma cytokines/chemokines at age 2 months in 11 infants with perinatal transmission and 11 negative controls. Of 6 detectable biomarkers, cases exhibited higher plasma programmed death-ligand 1, a proposed marker of T-cell exhaustion that requires further evaluation, specifically in infants with perinatal hepatitis C infection.

Perinatal transmission of hepatitis C viral infection (HCV) occurs in approximately 6–8% of children born to pregnant people with HCV infection with viremia.¹ Approximately 20–40%, and in some reports up to 66%, of children with perinatally acquired HCV infection demonstrate spontaneous clearance of infection by age 5 years, compared with 15–45% of adults within 6 months of exposure.² Previously deferred until 18 months of age, but with high rates of loss to follow-up for children with perinatal HCV exposure, HCV testing of perinatally exposed infants at 2–6 months of life is now recommended in the United States.²

Despite the transition to earlier testing, currently available screening methods do not distinguish those neonates at the highest risk of perinatal acquisition. Studies evaluating the modulation of infant immune profiles involved in HCV exposure and perinatal transmission are limited.³ Compared with adults, infant immune function is immature, with less robust adaptive immune responses and greater reliance on the innate immune response.³ Cytokines and chemokines released from innate immune cells play key roles in infant immune regulation.³ Therefore, we sought to evaluate plasma inflammatory cytokine and chemokine profiles associated with perinatal HCV infection among infants with in utero HCV exposure.

METHODS

We performed a nested case–control study of infants born to pregnant people with HCV viremia (ie, detectable virus in maternal plasma) who presented to prenatal care and were screened for HCV before 23 weeks 6 days’ gestation in the Eunice Kennedy Shriver National Institute of Child Health and Human Development Maternal-Fetal Medicine Units (MFMU) Network’s prospective observational study of HCV in pregnancy (NCT01959321, 2012-2021). All participating hospitals received local Institutional Review Board approval before study initiation, and participants provided written informed consent.

The primary study’s eligibility criteria and methods were previously described.¹ Briefly, the parent study included pregnant people with a viable singleton gestation and an HCV-positive antibody screen who enrolled before 27 weeks 6 days’ gestation, and excluded those who planned termination of pregnancy, had known major fetal anomalies, planned to give birth at a non-MFMU network medical center, or were unwilling or unable to commit to 18 months of follow-up if perinatal transmission occurred. For this analysis, we further excluded infants without an available plasma sample at 2 months of life or who were exposed to HCV–HIV co-infection in the birthing parent. The cases were infants with perinatally acquired HCV infection, defined as having a positive HCV RNA polymerase chain reaction (PCR) measured in infant plasma at 2 months of life. Negative controls, defined as having a negative plasma HCV RNA PCR at 2 months of life, were randomly selected and matched on study site, term (≥37 weeks 0 days) versus preterm birth, and peak maternal viral load (HCV RNA >10⁶ IU/mL vs. HCV RNA ≤10⁶ IU/mL) during pregnancy. All plasma samples were centrifuged, aliquoted and stored at −80°C within 6 hours of collection, with aliquots shipped to the central laboratory at the University of Washington Molecular Virology Laboratory in Seattle, WA, for measurement of neonatal HCV RNA.

For measurement of cytokine and chemokine profiles, stored infant plasma samples were shipped to the Northwestern University Comprehensive Metabolic Core where they were analyzed using a custom multiplex assay (Human Luminex Discovery Assay, R&D Systems, Minneapolis, MN) per the manufacturer’s instructions for 18 pro-inflammatory cytokines and chemokines (interferon [IFN]-γ, IFN-α, tumor necrosis factor-α, granzyme B, interleukin [IL]-1α, IL-1β, IL-1RA, IL-2, IL-4, IL-6, IL-12, IL-15, IL-17A, IL-18, basic fibroblast growth factor [FGF-basic], programmed death-ligand 1 [PD-L1], C-C motif chemokine ligand 5 [CCL-5], C-X-C motif chemokine ligand-10/interferon-γ induced protein-10 [CXCL10/IP-10]) and 3 anti-inflammatory cytokines (IL-8, IL-10 and IL-13). All samples were plated in duplicate on a Luminex 100 System instrument (Luminex, Minneapolis, MN), with analyte concentrations determined by interpolation from a standard curve. In this exploratory analysis, all 21 inflammatory markers were screened without a prespecified hierarchy, and those that met the limits of detection were included in the analysis.

We performed descriptive statistics and compared baseline demographic and clinical characteristics of the birthing parent using the independent samples t test, Wilcoxon rank-sum test, χ2 test or Fisher’s exact test, as appropriate. Quantile regression was used to compare the distributions of cytokine and chemokine concentrations between infant cases and controls, and report medians and the difference in medians with 95% confidence intervals. Box plots were created to illustrate concentrations for cases and controls. Given the exploratory nature of this analysis, correction for multiple comparisons testing was not performed.

RESULTS

Of 432 pregnant people with HCV in the primary study, perinatal transmission occurred in 26 infants (6%). Two infants were excluded for exposure to HIV–HCV co-infection. Of the remaining 24 HCV RNA PCR-positive infants, 11 infants had postnatal follow-up with provision of a plasma sample at 2 months of age, and were matched to 11 HCV RNA PCR-negative controls (Figure, Supplementary Digital Content 2, https://links.lww.com/INF/G613). Baseline demographics and clinical characteristics for the birthing parents of the 22 infants in this analytic cohort were similar to those previously described for the overall cohort (Table, Supplementary Digital Content 3, https://links.lww.com/INF/G614).¹ The mean (±standard deviation) gestational age at birth was 38 weeks 2 days gestation overall, and median peak HCV viral load in the birthing parent did not differ in the 11 cases (15 log10 IU/ mL, interquartile range [IQR] 14–15) and 11 controls (14 log10 IU/mL, IQR 13–15, P = 0.47; Table, Supplementary Digital Content 3, https://links.lww.com/INF/G614).

We analyzed 6 cytokines and chemokines that met limits of detection in the assay of infant plasma at 2 months of life: tumor necrosis factor-α, IL-1RA, IL-18, FGF-basic, PD-L1 and CXCL10/IP-10. Median PD-L1 concentration was higher in cases than controls (79 vs. 58 pg/mL; difference in medians [95% confidence interval] 22 pg/mL [3–41]; P = 0.03, Table 1). Median CXCL10/IP-10 concentration was higher in cases than controls, but not statistically significantly (100 vs. 57 pg/mL, P = 0.05, Table 1). The distributions of other detectable cytokine concentrations did not differ between groups (Fig. 1). Concentrations of the remaining cytokines (IFN-γ, IFN- α, granzyme B, IL-1α, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12, IL-13, IL-15, IL-17A and CCL-5) assessed in infant plasma did not meet levels of detection (Table, Supplementary Digital Content 4, https://links.lww.com/INF/G615).

TABLE 1.

Distribution of Cytokine and Chemokine Concentrations in HCV RNA PCR-Positive Versus PCR-Negative Infants at 2 Months of Life

Analyte	2 Months HCV RNA PCR+ Median (IQR) N = 11	2 Months HCV RNA PCR− Median (IQR) N = 11	Difference in Medians (95% CI)	P Value
TNF-α	12 (11, 20)	11 (8, 14)	1 (−4, 5)	0.74
IL-1RA	533 (422, 706)	491 (422, 629)	42 (−164, 248)	0.67
IL-18	252 (164, 414)	267 (159, 295)	−14 (−153, 125)	0.83
bFGF	20 (13, 27)	11 (3, 44)	9 (−10, 27)	0.34
PD-L1	79 (66, 86)	58 (47,73)	22 (3, 41)	0.03
CXCL10/IP-10	100 (50, 115)	57 (30, 63)	44 (−1, 88)	0.05

All analyte concentrations, assessed at 2 months of life, are reported as pg/mL.

bFGF, basic fibroblast growth factor; CI, confidence interval; CXCL10/IP-10, C-X-C motif chemokine ligand-10, also known as interferon-γ-induced protein-10; IL-1RA, interleukin-1 receptor antagonist; TNF-α, tumor necrosis factor-alpha.

FIGURE 1. Distribution of cytokine and chemokine concentrations in HCV RNA PCR-positive versus PCR-negative infants at 2 months of life.^a.

^aQuantile regression was used to compare the distributions of cytokine and chemokine concentrations between infant cases and controls.

DISCUSSION

Compared with HCV-exposed infants without perinatal infection at 2 months of life, infants with perinatal HCV infection exhibited significantly higher PD-L1 plasma concentrations. Additionally, we identified a signal, albeit nonsignificant, for upregulation of CXCL10/IP-10 in infants with perinatal HCV infection at 2 months of life compared with negative controls.

Prior evaluation by Lutckii et al³ of 31 children ages 0.3–5.3 years (10 with HCV infection, 21 HCV-exposed uninfected) similarly identified children with perinatal HCV infection, but not exposure alone, had strongly altered immune profiles with upregulation of PD-L1 and CXCL10/IP-10. Although further study of these pathophysiologic mechanisms specific to perinatal HCV infection is limited, studies in adults with chronic HCV infection suggest upregulation of PD-L1 and CXCL10/IP-10 is a direct effect of the viral infection within hepatocytes.⁴

In adults, innate sensing of HCV in infected hepatocytes by pattern recognition receptors triggers transcriptional induction and upregulation of CXCL10.⁵ Considered to be an interferon-stimulated gene, CXCL10 upregulation is also mediated by IFN-γ.^4,5 In response to HCV infection, natural killer (NK) cells secrete IFN-γ that stimulates hepatocytes to produce CXCL10/IP-10, which attracts additional NK cells, type 1 helper T cells, and cytotoxic T cells that produce more IFN-γ, perpetuating the immune response in a positive feedback loop.^4,5

IFN-γ also drives upregulation of PD-L1, a marker of functional cytoxotic T-cell exhaustion that promotes HCV persistence.^6,7 Exposure to high levels of HCV viral antigen and chronic HCV infection are associated with greater PD-L1 expression on immune cells.⁶ In response to viral antigens, PD-L1 binds the PD-1 receptor on effector cytotoxic T cells and suppresses T cell responses via several mechanisms (eg, impaired proliferation, inactivation, anergy and/or apoptosis of cytotoxic T cells).^6,7 These dysfunctional T cells have decreased capacity to eliminate infected cells.^6,7 PD-1 is further expressed on innate immune cells (eg, NK cells, dendritic cells). Binding of PD-L1 to the PD-1 receptor on the surface of innate immune cells inhibits their cytotoxic activity and induces apoptosis, thereby promoting innate immune dysfunction.^6,7

Like that in adults, PD-L1 is a proposed marker of functional T cell exhaustion in children with other types of chronic viral infections (eg, human immunodeficiency virus).⁸ It is plausible that in infants with perinatal HCV exposure, high plasma PD-L1 may also represent an early sign of immune dysfunction marked by exhaustive T cell and/or innate immune responses, and a propensity for persistent infection. In the subset of HCV RNA PCR-positive infants, PD-L1 upregulation in response to high levels of replicating HCV antigen in utero may have led to suppressed antiviral immune responses and thereby viral persistence at 2 months of life. However, we only evaluated soluble, circulating PD-L1 as a surrogate marker of immune cell function. We have not directly evaluated PD-L1 and associated T cell effector function in HCV-specific T cell or NK cell assays, nor assessed how cellular sources of PD-L1 (eg, hepatocytes, myeloid cells, or lymphoid cells) in infants with perinatal HCV infection may differ from that in adults to be able to determine whether high plasma PD-L1 in infants develops secondary to high HCV antigen load, generalized IFN-γ activation, and/or immune checkpoint-mediated dysfunction. Understanding the role of the PD-1/PD-L1 pathway in regulating cytotoxic T cell and innate immune cell function requires further investigation in infants exposed to HCV viremia perinatally, including correlation with specific immune cell phenotypes.

Study strengths include a geographically diverse, prospective cohort of pregnant people with postnatal infant follow-up, with cases and controls matched based on peak HCV viral load in the birthing parent to control for the potential influence of viremia on cytokine and chemokine concentrations. However, only six of the twenty-one biomarkers screened met the limits of detection for analysis, and we did not correct for multiple comparisons given the exploratory nature of this analysis, raising the possibility of type 1 error with respect to the signals identified with PD-L1 and CXCL10. Additionally, requirement of prior participant consent for use of their samples in this ancillary study and loss to postnatal follow-up of cases in which perinatal transmission occurred limited our sample size of infants with available plasma samples at 2 months of life and our ability to perform adjusted regression models or subgroup analyses (e.g. by infant sex^9,10). Our findings thus require validation in larger cohorts.

Although the primary study followed children up to 18-24 months postnatally, even more significant loss to follow-up occurred by 2 years of life among the HCV RNA PCR-positive cases identified 2 months postnatally in this high-risk subset. The present analysis thus does not distinguish between those infants with persistence versus clearance of infection, and cannot determine the prognostic value of PD-L1 and/or CXCL10 for predicting HCV clearance in early childhood. Future studies should evaluate whether PD-L1 and other soluble inflammatory markers are associated with higher risk of chronic perinatal HCV infection and their ability to differentiate those infants who may benefit from enhanced surveillance and potentially earlier targeted direct-acting antiviral therapy to reduce loss to follow-up, missed treatment opportunities, and long-term complications of HCV infection. However, at present, our hypothesis-generating findings related to PD-L1 and CXCL10 are insufficient to change current screening, surveillance, or treatment recommendations. Large-scale longitudinal cohorts with paired immunologic and virologic outcome data are required before implementing these biomarkers into clinical management algorithms.

Supplementary Material

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website (www.pidj.com).