A retrospective clinicopathological autopsy study from the University of California, San Francisco Neurodegenerative Disease Brain Bank has concluded that frontotemporal lobar degeneration with TDP-43 (FTLD-TDP) and limbic-predominant age-related TDP-43 encephalopathy neuropathologic change (LATE-NC) are best kept as separate diagnostic entities. Published on 30 June 2026 in Alzheimer's & Dementia: Diagnosis, Assessment & Disease Monitoring by D. Luke Fischer, Salvatore Spina, Bruce L. Miller, William W. Seeley and Lea T. Grinberg, the study compared 148 FTLD-TDP cases, including 39 of the type A subtype and 21 that could not be subtyped at all, with 42 cases classified as LATE-NC. FTLD-TDP type A cases were younger at symptom onset and at death, had shorter disease duration and frequently carried GRN or C9ORF72 mutations; the LATE-NC cases were older and mostly sporadic. Taking all neuropathologic features together, the authors report the two could be separated with greater than 95 percent confidence.

The question matters because type A FTLD-TDP and LATE-NC involve overlapping limbic circuitry and can look alike down the microscope. The authors note that both often show compact neuronal cytoplasmic inclusions across hippocampal subfields and the dentate gyrus, and that the diagnoses diverge only once the pathology leaves the hippocampus for other neocortical areas and the ventral striatum. That is where routine practice runs into trouble. When three raters, blinded to case details, judged TDP-43 staining in the middle frontal gyrus alone, the region most brain banks sample, one case originally classified as FTLD-TDP type A was called LATE-NC by all three, and eight more were misread by at least one rater. The case all three missed had little frontal pathology and carried a GRN mutation.

The authors are explicit about the weakness of that experiment: only one of their 42 LATE-NC cases reached stage 3, the neocortical stage, which blocked the head-to-head comparison they had planned and which they attribute to referral bias at a centre that draws younger-onset and atypical cases. They describe the frontal-slide analysis as exploratory for that reason. A larger multi-centre study reached a partly different conclusion. Ryan Shahidehpour, Peter Nelson and colleagues, writing in Acta Neuropathologica on 28 April 2025 using brain banks at the University of Kentucky, The 90+ Study at UC Irvine and the Mayo Clinic, reported that quantifying TDP-43 in the middle frontal gyrus classified more than 90 percent of cases as either LATE-NC stage 3 or FTLD-TDP, and proposed a diagnostic rubric on that basis.

What follows for criteria is a sampling question rather than a naming one. The UCSF group argues that ruling out FTLD-TDP in a brain with frontal TDP-43 pathology requires looking at further neocortical regions, and points to the inferior frontal gyrus (the opercular region) and the orbitofrontal gyri as candidates, since both are often involved early in FTLD-TDP presenting as behavioural variant frontotemporal dementia and late in LATE-NC. Nothing in the study changes the care of any living patient: all of this work is done on donated brain tissue after death.