Reply to ‘Neuropathological progression of clinical Parkinson disease subtypes’

In our recent News & Views assessment of the important work by De Pablo-Fernández and colleagues on 111 pathology-proven Parkinson disease (PD) cases from the Queen Square Brain Bank (Clinical Parkinson disease subtyping does not predict pathology. Nat. Rev. Neurol. 15, 189–190 (2019))^1,2, we stated that although the study served to confirm the predicted trajectory of the three data-driven severity subtypes of PD — mild, intermediate and diffuse malignant — it failed to suggest there were pathological correlates to each subtype.

In their Correspondence article (Neuropathological progression of clinical Parkinson disease subtypes. Nat. Rev. Neurol. https://doi.org/10.1038/s41582-019-0197-x (2019))³, De Pablo-Fernández and colleagues state that we failed to mention that these pathological changes were reached over a considerably shorter disease duration in the diffuse malignant subgroup than in the other subgroups. We did acknowledge that the mean rates of progression and survival mirrored the early subtypes: longest survival and slowest progression in the mild motor-predominant group, shortest survival and fastest progression in the diffuse malignant group, and intermediate progression and survival in the intermediate group. Inferences on the timing of pathological changes before death, however, are not possible from this cross-sectional post-mortem evaluation of pathology.

Importantly, any differences in the rate of neuropathological deterioration among subgroups cannot be used to conclude that the neuropathology features themselves were important determinants or correlates of clinical PD subtypes. In fact, the neuropathological features were similar across the three subtypes.

Protein aggregates that are identified at post-mortem and are relied upon for nosology might not be pathogenic but could potentially be universal compensatory responses to a wide range of biological stressors⁴. This idea was supported by the data from the study by De Pablo-Fernández and colleagues². Amyloid and tau pathology were associated with older age at death: individuals with more Alzheimer disease pathology lived longer². This finding is paradoxical to our current disease model but aligns with a future systems biology approach to diseases of brain ageing⁵. If ‘PD’ is to be accepted as the nosological entry point to many diseases, most of which have Lewy pathology in common, we will need to vigorously engage in an individualized search ‘upstream’ to identify molecular and biological subtypes and truly usher in the era of precision medicine⁶.