An Interferon-γ Release Assay for Evaluation of Cell-mediated Immunity in Infants With Congenital Cytomegalovirus Infection

(See the Major Article by Capretti et al on pages 367–73.)

Congenital cytomegalovirus (cCMV) infection is the most common infection acquired in utero, with a birth prevalence of nearly 0.7% in the United States [1]. Of all infected infants with or without CMV symptoms at birth, 75% are born to seropositive women, and almost 20% will have permanent neurodevelopmental effects such as sensorineural hearing loss (SNHL), motor disability, or cognitive delay [2]. While behavioral modifications can lower the risk of maternal CMV infection during pregnancy, no current intervention can prevent transmission from mother to fetus or reduce severity of disease in affected infants. More effective prevention and treatment approaches are needed, informed by a deeper understanding of the mechanisms that underly CMV transmission and of virus–host interactions during early congenital infection.

T cells can be detected in human samples by parameters such as gene expression, receptor clonotype, surface markers, or functional capacity. In particular, commercial interferon-γ (IFN-γ) release assays (IGRAs) have become a common approach in healthcare settings and are typically based on enzyme-linked immunospot (ELISpot) or enzyme-linked immunosorbent assay (ELISA) platforms. The ELISpot platform does include counting peripheral blood mononuclear cells (PBMCs) in the sample, while the ELISA platform does not include counting PBMCs in the sample. Despite requiring more time and operator skill, incubating PBMCs rather than whole blood in the ELISpot protocol allows direct measurement of the proportion secreting IFN-γ in response to antigen, thus identifying precursor frequencies and accounting for leukopenia or lymphopenia in the sample. In contrast, the ELISA protocol measures bulk IFN-γ secretion reported qualitatively or as indeterminant if the positive control does not reach a threshold level of IFN-γ secretion.

Based on the ELISA platform, the QuantiFERON-TB assay (QFN-TB; QIAGEN Sciences Inc, Germantown, MD) was approved by the US Food and Drug Administration (FDA) in 2001 as a test for tuberculosis infection. Similarly, the QuantiFERON-CMV assay (QFN-CMV; QIAGEN Sciences Inc, Germantown, MD; Cellestis Ltd, Victoria, Australia) was developed for CMV disease risk stratification and monitoring in solid organ transplant (SOT) or hematopoietic stem cell transplant (HSCT) recipients. The assay measures IFN-γ secretion in whole blood after stimulation with CMV peptides that represent CD8 T-cell epitopes. Several studies of SOT or HSCT patients [3] have used IGRAs to demonstrate an association between CMV-specific T cells and risk or severity of CMV disease, although some [4–6] have reported limited clinical utility of the IGRA approach. Use of QFN-CMV in a clinical setting was first reported by Walker et al in healthy individuals and SOT patients [7] and by others more recently in both SOT and HSCT recipients [3, 8, 9] and is part of some clinical trials (clinicaltrials.gov NCT03858907, NCT02784756, NCT03699254). The T-SPOT.CMV (Oxford Immunotec Ltd, Oxfordshire, United Kingdom) and T-Track CMV (Lophius, Regensburg, Germany) assays measure CMV-specific T cells using the ELISpot platform. However, QFN-CMV is not currently available in the United States, none of the CMV IGRAs are FDA approved, and their use in routine clinical practice has not been widely adopted in transplant centers.

In this issue of Clinical Infectious Diseases, Capretti and colleagues examined the relationship between neonatal T-cell responses measured by QFN-CMV at 2 time points (<14 and 2 –60 days of age) and several clinical parameters. Previous studies have described CD4 or CD8 T-cell responses in young children with cCMV infection, some of which have shown that CMV-specific T cells are detectable in the fetus and neonate, increase in frequency and breadth of antigen specificity over time, and have various deficits in functional capacity, including IFN-γ secretion [10–12]. However, none have clearly delineated the protective and/or pathologic role of these T cells in the breadth of clinical outcomes observed in cCMV infection.

In a cohort of 30 infants, Capretti et al reported a significant correlation between QFN-CMV reactivity and symptoms or viral load. Of note, their binary designation of “asymptomatic” or “symptomatic” cCMV infection included isolated SNHL in the latter category based on some [13] but not other [14] consensus guidelines. All 16 infants with a positive QFN-CMV were asymptomatic and had normal clinical outcomes, which translated to a negative predictive value (NPV) of 100% for symptomatic infection with implied relatively poor prognosis. In contrast, a minority of 14 infants with a negative or indeterminant QFN-CMV were asymptomatic, resulting in a positive predictive value of 71.4% at T0 and 75% at T1 for symptomatic infection (or lower if the 2 infants with indeterminate QFN-CMV are removed). In addition to lower viral load at T1 for infants with a positive test, these data suggest that neonatal CD8 T cells are protective for cCMV disease.

The QFN-CMV system that involves a few viral peptides, single T-cell type and function but lack of enumeration, and a qualitative result is not designed to characterize CMV-specific T cells in detail or to identify mechanisms of cCMV disease severity. Moreover, the threshold for a positive result has not been established for infants. IFN-γ or cytokine secretion in general may not be the best T-cell function to measure in this population, and infants with cCMV infection or who receive antiviral therapy may have low total white cell or lymphocyte counts. Caution should also be taken in assuming global immunosuppression for patients with an indeterminate result in any QFN or other ELISA platform, which is based on a technical feature of the assay. Even in transplant recipients, no single assay is widely used to measure global immunosuppression. However, QFN-CMV or similar IGRA may fill the need for a cCMV prognostic tool if validated in larger studies as the authors suggest.

The historical practice of designating infants with cCMV infection as “symptomatic” or “asymptomatic” hampers more than facilitates patient care and research in the field, primarily due to lack of accepted definitions for these terms. The approach to risk stratification and assessment of prognosis at birth and in clinical trials should be grounded in CMV biology and pathogenesis rather than subjective features on physical examination. Despite the limitations of the QFN-CMV assay, data reported by Capretti et al support the possibility that a test based on neonatal CMV-specific T-cell responses could be used in clinical practice. Given that healthcare providers currently have few objective criteria to assess risk or prognosis, a test with such a high NPV would be a welcome tool for offering parents anticipatory guidance. As the authors suggest, a major goal of future studies should be to inform development of a composite biomarker including viral, maternal, and fetal/neonatal measures that can accurately predict clinical outcomes for infants with cCMV infection. Their study reporting a correlation between T-cell IFN-γ secretion and severity of cCMV disease represents significant progress toward this goal.

Note

Potential conflicts of interest. L. G. reports personal fees from Moderna Therapeutics, Merck, Roche, RTI Health Solutions, and Aclimed outside the submitted work. The author has submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References