Neuroimaging in Aicardi-Goutières syndrome: Biomarkers for a progressive encephalopathy.

In 1984, 2 pediatric neurologists, Jean Aicardi and Françoise Goutières, published their seminal case report of 8 patients (from 5 families) with a devastating neonatal encephalopathy characterized by striking cerebral calcifications, white matter hypodensities, visualized on CT, accompanied by a persistent CSF lymphocytosis. Notably, the neuroradiologic findings suggested a perinatal toxoplasmosis, other (syphilis, varicella-zoster, parvovirus b19), rubella, cytomegalovirus, and herpes (TORCH) infection, and these patients often have an elevation of interferon-a in the CSF. Three decades of highly productive clinical and scientific investigation of Aicardi-Goutières syndrome (AGS) has led to the discovery of 7 causative genes and the realization that mutation in any of these leads to a genetically mediated autoimmune response to nucleic acid metabolism, analogous to systemic lupus erythematosus, in the developing brain. In this issue of Neurology®, La Piana and colleagues comprehensively analyze neuroimaging from 121 patients with genetically and clinically diagnosed AGS. The authors describe alterations detected by CT and MRI that correlate tightly with AGS and with specific genetic etiologies. For example, brain calcifications were present in 110/121 patients, with severe calcifications (defined as involving multiple locations beyond lentiform nuclei, deep white matter, and thalami, and having multiple and variable patterns) being much more prevalent in patients with TREX1 mutations. In contrast, the presence of RNASEH2B mutations was inversely associated with severe calcifications. Another notable finding was that a frontotemporal pattern of leukoencephalopathy had a significant association with clinical severity. These findings help the clinician if a mutation in TREX1 or RNASEH2B is identified, but with 7 genes implicated in AGS, including RNASEH2A, RNASEH2C, SAMHD1, ADAR, and IFIH1, these findings only point to the need for expanding the available cohort. Thus, while 121 patients is a large group, the capacity for clinically meaningful phenotype–genotype correlations is limited. Other important but not easy to acquire data that can enhance the ability to extract meaningful correlations from the data include a more complete timeline of disease progression, following patients from the perinatal through the early toddler years, as both clinical severity and imaging findings may progress. What can the clinician take away from these imaging findings that will help in the differential diagnosis and patient management? We are reminded again that imaging findings in AGS clearly mimic those of TORCH infections, and in all cases where a TORCH infection is considered, testing for AGS should commence promptly if no clear sign of infection is identified. Because most cases of AGS are mediated by autosomal recessive genetics, rapid identification of the etiology can immediately help with family planning. However, as the genetic community is discovering, there are also examples where de novo mutations function in a dominant fashion to cause disease. For example, autosomal recessive mutations in the gene KIF1A can lead to peripheral neuropathy, while dominant de novo mutations can lead to a severe and progressive encephalopathy with a level of clinical impairment analogous to AGS. By comparison, there are dominant heterozygous mutations in both IFIH1 and TREX1 that lead to AGS; others are likely to be identified. As prenatal genetic testing becomes more commonplace, the ability to gauge possible disease outcomes accurately from the genetic state will become increasingly important, particularly for diseases with considerable morbidity, such as AGS. Thus, following a cohort of participants who have mutations in the AGS genes, but do not immediately present with symptoms, would be an important next step toward a truly complete understanding of the phenotype–genotype correlation. Moreover, in addition to understanding the range of clinical outcomes of AGS through gene discovery, these advances may importantly point to a possible route for treatment. Thus, not only can autosomal dominant mutations in TREX1 lead to AGS, but these mutations can also lead to familial chilblain lupus, further strengthening this linkage to a progressive autoimmune disorder that could potentially be treated prior to symptoms. Apropos of this linkage, a recent article showed that inactivation of the TREX1