This study utilizes the comprehensive clinical and biomarker PPMI dataset to demonstrate that αSyn-SAA informs PD genetic and clinical heterogeneity and identifies at-risk individuals, possibly at a very early stage of degeneration. We confirmed that αSyn-SAA is highly accurate in differentiating PD from HCs, but observed variability among genetic sub-groups, in particular among LRRK2 PD patients. The proportion of cases with positive αSyn-SAA results was highest among patients with sporadic PD with typical olfactory deficit. By contrast, the proportion of αSyn-SAA positive results was lower in those with normal olfaction and in LRRK2 variant carriers. We found an αSyn-SAA positive gradient from almost certain in GBA PD to very likely in sporadic PD to only modestly more likely than not in LRRK2 PD. Preserved olfaction was associated with negative αSyn-SAA results across genetic sub-groups, but participants who were both normosmic and LRRK2 variant carriers were even more likely to be αSyn-SAA negative. This was particularly true among females, where only 12.5% of female normosmic LRRK2 carriers had a positive αSyn-SAA result.

In cases that came to post-mortem examination, all those with positive αSyn-SAA results had typical Lewy pathology, while the one case with negative αSyn-SAA, a LRRK2 carrier with preserved olfaction, had no Lewy pathology. These clinical and pathological data suggest that there may be different pathologies in αSyn-SAA positive and negative LRRK2 PD despite a similar clinical phenotype. Our results showing a lower frequency of αSyn-SAA positivity in PD patients carrying a LRRK2 variant is consistent with a prior study of 15 LRRK2 G2019S PD patients and 16 NMC.²⁴ Interestingly, the proportion of αSyn-SAA negative cases that we found (approximately one third) closely mirrors the frequency of LRRK2 PD cases without Lewy pathology at autopsy reported in the literature.^25,26 These consistent results indicate that αSyn-SAA is a marker of pathology characterized by α-Syn aggregates rather than a clinical PD phenotype, and enables ante-mortem differentiation of cases with atypical pathology.

Along with preserved olfaction, PD patients with a LRRK2 variant and a negative αSyn-SAA result were older, more likely to be female and had slightly less motor impairment. These results are consistent with a modest clinical-pathological correlation within LRRK2 PD, that can be stratified by αSyn-SAA. In addition, the excess of females among LRRK2 carriers who are αSyn-SAA negative may partially account for the observation that the expected male-to-female predominance seen in sporadic PD is not present in LRRK2 carriers.²⁷

Regarding prodromal and at-risk individuals, the results of this study confirm that a majority of RBD patients with abnormal PSG and DAT deficit are αSyn-SAA positive. We have extended our assessment to individuals at-risk for PD on the basis of impaired olfaction who also had DAT deficit, and they are equally likely to be αSyn-SAA positive. Of the five (out of 33) RBD cases that were αSyn-SAA negative, three had abnormal DAT imaging, four also had normal olfactory function, possibly explained by the observation that some RBD patients progress to MSA.²⁸ This disorder is associated with a less olfactory impairment than PD.²⁹ Prior studies evaluating the utility of αSyn-SAA in MSA have produced varying results, depending on assay condition, and MSA has been shown to have distinct αSyn-SAA aggregation properties relative to PD.^10,30 For this study, we used an assay that targeted detection of PD aggregates, and which has a lower sensitivity to detect glial cytoplasmic inclusions (GCIs) in MSA.

We observed several prodromal RBD and hyposmic cases with positive αSyn-SAA results but DAT imaging in the normal range. By contrast, no prodromal hyposmic participants had negative αSyn-SAA results and positive DAT imaging. Only those RBD participants potentially explained by MSA had negative αSyn-SAA and DAT binding consistent with PD. Approximately 7-10% of NMCs of either GBA or LRRK2 PD-related variants had positive αSyn-SAA results. This figure is relatively close to the lifetime penetrance of PD among GBA N370S carriers,³¹ but substantially lower than estimates of PD penetrance among LRRK2 G2019S carriers.^32,33 Taken together, these findings are consistent with a temporal pattern of biomarker abnormalities in which there may be a prolonged period in which abnormalities in αSyn-SAA are present before changes in physiological markers such as DAT-SPECT appear, and that this pattern may be more variable in LRRK2 carriers than GBA carriers. Interestingly, among the GBA NMCs, those few individuals who are αSyn-SAA positive are also more likely to show olfactory dysfunction even in the absence of a DAT deficit. Our results provide biomarker support for a prolonged prodromal period. This concept has been proposed in the literature, lacked supporting evidence from nigral neuropathological findings or dopamine imaging studies.³⁴ Subsequent studies in larger prodromal cohorts are needed to confirm this hypothesis.

The results of our study have immediate implications for clinical trial design. For LRRK2 targeted therapies, stratification based on the αSyn-SAA results may be necessary to assess therapeutic benefits. For GBA-targeted therapies it would be reasonable to exclude αSyn-SAA negative participants since GBA PD is overwhelmingly αSyn-SAA positive³⁵ and, thus, a negative result would raise questions about a PD diagnosis. αSyn-SAA could also be combined with other biomarkers, such as markers of Alzheimer’s pathology to classify individuals with mixed pathology.³⁰ Similarly, for αSyn targeted therapies, the possibility that αSyn-SAA negative participants will respond differently to treatment should be considered in patient selection and sample size estimates. For planned clinical trials targeting at-risk populations, αSyn-SAA results should be considered for identifying the earliest stages of synucleinopathy and those individuals likely to progress to a clinical disorder with typical Lewy pathology.

Our study has several methodological strengths. First, it uses samples from a large and well-characterized cohort consisting of several clinically relevant sub-groups. Participants were comprehensively evaluated with a battery of clinical evaluations of motor and non-motor features and biomarkers, including DAT imaging and fluid biomarkers, permitting interrogation of αSyn-SAA results against these measures. The number of PD participants and concurrent, matched controls and the inclusion of prodromal cohorts in PPMI allowed for sub-group analysis and extension of prior results to new populations. Second, αSyn-SAA analysis was performed in a blinded fashion using a robust, validated assay platform. Third, the participants were diagnosed by expert neurologists based on standardized criteria within the PPMI study.

A number of limitations of this study should be acknowledged, First, there was no a priori hypothesis or sample size estimation. Rather, we took advantage of the large and readily available PPMI cohort. Based on the diagnostic performance of the assay in prior studies,⁷ we expect a high power. We feel that this is a reasonable expectation given that this is the largest cohort that has ever been investigated in this manner. Second, the non-parametric Wilcoxon rank sum test was used to compare αSyn-SAA status in PD participants for some outcomes due to the low number of samples in some groups and skewed nature of the data, but additional samples would enable the use of more powerful methods. Third, the analyses presented in this report are all cross-sectional. Since PPMI has clinical and biological samples collected over time, a report on longitudinal data, including motor and non-motor features, will be an important topic for future analysis and may identify other clinical correlates. Fourth, the number of prodromal (hyposmia and RBD) cases with normal DAT imaging is relatively small, which prevents us from drawing definitive conclusions about the temporal ordering of αSyn-SAA abnormalities and those on DAT imaging. Fifth, several unanswered questions remain regarding genetic forms of PD. This cohort consisted of only G2019S LRRK2 carriers and N370S GBA carriers. αSyn-SAA results in patients with other LRRK2 or GBA variants, as well as variants in PD-associated genes such as PRKN and PINK1 could not be assessed.

Again, longitudinal studies and those in additional PPMI participants focusing on at-risk individuals, who are currently being recruited, may address these issues.

In summary, this study extends our understanding of the usefulness of αSyn-SAA for in vivo molecular assessment of PD. We show in a large, deeply phenotyped cohort that αSyn-SAA is highly accurate in typical PD, but results vary depending the presence of the LRRK2 G2019S variant as well as clinical features, particularly hyposmia. Within LRRK2 variant carriers, there are also differences in age, gender and motor performance that associate with αSyn-SAA status. Another key finding is that prodromal and NMCs, especially carriers of the GBA N370S variant, have evidence of abnormal α-synuclein aggregation prior to any other detectable clinical or biomarker change, including alterations in DAT imaging. One implication of this result is that the prodromal period in PD may be longer than had been projected previously and may start before there is loss of dopaminergic integrity, at least in some individuals. Taken together, these findings have immediate implications for clinical trial design, both to identify pathologically defined subgroups of PD patients and to establish biomarker-defined at-risk cohorts. Longitudinal research is needed to investigate the prognostic value of αSyn-SAA and whether changes in quantitative measures of αSyn aggregation indicate progressive pathology over time.