Serum uric acid and total bilirubin as putative biomarkers of resistance in Prodromal Parkinson's disease: Longitudinal data from the PPMI study.

Background: The role of blood uric acid and more recently bilirubin as biomarkers in symptomatic motor PD has been increasingly established in the literature. Objective: Our present study assessed the role of serum uric acid and total bilirubin as putative biomarkers in a prodromal PD cohort followed longitudinally. Methods: Longitudinal 5-year serum uric acid and total bilirubin measurement data of 65 Prodromal PD patients (including REM Sleep Behavior disorder (RBD), N=39 and Hyposmia, N=26) with an abnormal DATSCAN imaging were downloaded from the Parkinson's Progression Markers Initiative (PPMI) database. This cohort was compared with 423 de novo sporadic PD patients and 196 healthy controls enrolled in the same study. Results: After adjusting for age, sex and Body Mass Index (BMI), baseline and 5-year longitudinal serum uric acid levels were higher in the Prodromal cohort and RBD subgroup as compared to the motor PD cohort. This was also true for longitudinal measurements in the Hyposmic subgroup. In contrast, baseline and longitudinal serum total bilirubin did not differ between each prodromal group and the PD cohort. Conclusions: Our results are indicative of a role of serum uric acid (but probably not of total bilirubin) as a marker of neuroprotection, in a certain subgroup of premotor patients exhibiting exclusively non motor features (hyposmia or RBD). It is possible that an inherent antioxidant resistance of a subset of RBD or hyposmia patients with high serum uric acid level delayed or precluded the emergence of a motor PD phenotype as opposed to the PD cohort.

the Prodromal cohort and RBD subgroup as compared to the motor PD cohort. This was also true for longitudinal measurements in the Hyposmic subgroup. In contrast, baseline and longitudinal serum total bilirubin did not differ between each prodromal group and the PD cohort.
Conclusions: Our results are indicative of a role of serum uric acid (but probably not of total bilirubin) as a marker of neuroprotection, in a . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 4 certain subgroup of premotor patients exhibiting exclusively non motor features (hyposmia or RBD). It is possible that an inherent antioxidant resistance of a subset of RBD or hyposmia patients with high serum uric acid level delayed or precluded the emergence of a motor PD phenotype as opposed to the PD cohort.
Keywords: Prodromal Parkinson's disease; Uric acid; Total Bilirubin; REM sleep Behavior disorder; Hyposmia . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint Introduction Prodromal Parkinson's disease (PD) corresponds to the premotor phase of PD lasting many years. During this phase the neurodegeneration process is active and ongoing and typical prodromal non-motor symptoms like REM sleep behavior disorder (RBD), depression, hyposmia and constipation emerge. The progression to the nonmotor and motor manifestations of PD aligns well with the neuropathological staging system proposed by Braak [1]. Numerous risk factors have been thoroughly described in previous studies and they facilitate the assessment of the probability to develop motor PD [2]. Such markers have been incorporated in the recent MDS criteria for prodromal PD [3].
Oxidative stress is thought to play a vital role in the pathogenesis of Parkinson's disease (PD) [4]. Two endogenous antioxidant molecules, uric acid and bilirubin, are considered to be buffering agents and to likely protect dopaminergic neurons from oxidative stress in PD. Uric acid exerts its antioxidant function mainly by means of ferrum chelation, although other mechanisms may also contribute to this effect. In vitro studies have shown that uric acid could induce autophagy activation and ameliorate alpha-synuclein (SNCA) accumulation [5]. In an in vivo PD model, uric acid demonstrated neuroprotective properties for dopaminergic neurons by means of . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) High serum uric acid levels have been correlated with lower risk of PD, a milder disease course and with the more benign tremor dominant (TD) motor subtype [11,12]. As far as cognitive outcomes are concerned, lower levels of serum uric acid in the early disease stages were reported to be associated with the later occurrence of mild cognitive impairment (MCI) in an early PD cohort [13]. However, the causal relationship between high plasma urate and low risk of PD and its putative prognostic value have been questioned by other researchers [14,15].
Increased serum bilirubin levels might also be related to PD etiopathology as a possible serum biomarker of oxidative stress dysregulation. The heme oxygenase (HO) -bilirubin pathway is one of the major antioxidant defense mechanisms. Heme oxygenase-1 (HO-1), an enzyme that converts heme to free iron, carbon monoxide (CO) and biliverdin (bilirubin precursor) is expressed in response to various stressors including neurodegenerative disorders like Parkinson's disease [16]. Individuals with decreased serum bilirubin probably lack the endogenous defense system to prevent free radicals from damaging . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 7 dopaminergic cells in the substantia nigra. Under control of different transcription factors but with a prominent role played by Nrf2 (common mediator also implicated in uric acid antioxidant pathway), HO-1 induction is crucial in nervous system response to damage [17].
In previous studies, PD subjects showed higher levels of bilirubin compared with controls, possibly due to HO overexpression as a compensatory mechanism from the early stages of PD; higher bilirubin levels in PD were not likely to be related to dopaminergic replacement therapy [18,19]. Lee and co-authors showed using imaging correlates, that bilirubin was the most significant antioxidant marker and it was correlated to the degree of dopaminergic deficit [20]. The Prodromal cohort in the PPMI study included premotor subjects with prodromal PD symptoms (REM sleep behavior disorder-RBD or hyposmia/reduced olfaction) [21,22]. RBD was diagnosed based on clinical history and sleep related scales along with polysomnography, if available. In the hyposmia cohort, olfaction was measured using the University of Pennsylvania Smell Identification Test (UPSIT).
Hyposmia was defined as a score of <10th percentile for age and sex. Hyposmic individuals with a normal DATSCAN during the screening visit were excluded from follow up in PPMI (however serum uric acid . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 9 and total bilirubin measurements from samples of this cohort collected during the screening visit were also available in the database of the study).
Biochemical analyses (including measurements of uric acid and total bilirubin level in serum) have been carried out in Covance laboratories in a uniform fashion, as per the study protocol. Finally, in order to assess the putative impact of uric acid levels on clinical progression of patients in the prodromal cohort and determine . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 1 1 Regarding each prodromal subgroup, baseline and 5-year longitudinal serum uric acid measurements were higher in the RBD cohort as compared to the PD cohort (p=0.028 and p=0.017 respectively). There . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 1 3 Similarly, baseline and longitudinal serum total bilirubin measurements did not differ between RBD and PD cohorts (p=0.348 and 0.219) and between RBD subjects and healthy controls either (p=0.84 and 0.644).
Baseline and longitudinal serum total bilirubin measurements did not differ between Hyposmic cohort and PD cohort (p=0.264 and 0.872) and between hyposmic subjects and healthy controls (p=0.101 and 0.975 respectively). Moreover, there was no significant effect of Time or Time*Group interaction on total bilirubin in any of the subgroups.

Discussion
The role of blood uric acid and more recently bilirubin as biomarkers in symptomatic motor PD has been increasingly established in the literature. Our present report assessed the role of serum uric acid and total bilirubin as putative biomarkers in a prodromal PD cohort followed longitudinally. It appears that both the entire Prodromal PD cohort and the RBD subgroup participants exhibited increased serum uric acid level as compared to sporadic PD both at baseline and during a 5-year follow up (this was also true for the hyposmic subgroups regarding follow up). The level of uric acid decreased longitudinally in the prodromal group, the RBD group and the Hyposmic group almost reaching that of sporadic PD at year 5. Moreover, despite a trend for higher uric acid, all three prodromal cohorts did not statistically differ from healthy controls following age, sex and BMI adjustments. Uric . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 1 4 acid level increased in healthy controls as opposed to prodromal and PD subjects. On the other hand, total bilirubin did not differ between sporadic PD and either of the three prodromic cohorts (Prodromal, RBD and Hyposmic subgroups) and level of total bilirubin remained stable during the follow up period across the groups. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 1 5 into 2 groups according to the presence or absence of PD and it appeared that patients without PD and those who had more than 5 years of RBD exhibited higher levels of uric acid than patients with PD. The authors concluded that higher levels of plasma urate were associated with a longer duration of RBD without converting to PD, exhibiting resistance to the development of motor symptoms [23]. Moreover, in patients without PD, there was a positive correlation between years of evolution of RBD and the levels of uric acid. However, in our study, no such correlation between level of uric acid and duration of RBD symptomatology could be established in the RBD subgroup.
The notion of uric acid homeostasis contribution in the resistance of prodromal patients to PD could also apply to non manifesting carriers of pathogenic PD mutations. Bakshi and co authors evaluated the putative role of uric acid as a candidate biomarker of PD risk modulation in pathogenic LRRK2 mutation carriers using data from the

LRRK2 Cohort Consortium and the Parkinson's Progression Markers
Initiative. According to their observations, non manifesting LRRK2 mutation carriers had significantly higher levels of uric acid than those who developed PD and this applied to both sexes. The authors reached the conclusion that uric acid monitoring could be used as a biomarker of resistance to PD among LRRK2 mutation carriers [24].
Regarding the association of uric acid level in prodromal PD with cognitive outcomes, no firm correlations could be established in our report. A previous study assessed how levels of serum uric acid affect cognition in patients with RBD. Regarding the memory domain, the . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 1 6 low uric group had worse scores than the healthy controls, whereas the scores of high urate group were intermediate between the low uric acid group and the healthy controls. As far as other cognitive features like executive function and language domain are concerned, lower scores were found in the patients with low uric acid, whereas the scores of the patients with high urate were similar to those of the healthy controls.
According to the aforementioned study, uric acid levels affect cognitive function in patients with RBD, partly owing to its antioxidant and neuroprotective function [25]. In another study during the motor PD phase, uric acid and uric acid/creatinine levels in the early and medium stage PD patients were significantly higher than in the advanced stage ones [26]. Accordingly, if we consider prodromal and motor Parkinson's disease to be the two edges in a continuum, higher uric acid could possibly mark the initial phases of the disorder and a gradual decrease of urate level might simply reflect disease progression.
The absence of significant differences in total bilirubin level between prodromal and motor PD might implicate a limited significance of this biomarker for the prodromal phase of the disorder. All current data on bilirubin focus on symptomatic PD patients with motor deficits [18,19].
Alternatively, since increased total bilirubin is a hallmark of early motor PD, it is possible that an adequate oxidative stress buffering during the prodromal phase of PD might hamper a rise in total bilirubin via the heme oxygenase 1 pathway. Clearly, further research is required in order to decipher the role of bilirubin as putative marker . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint 1 7 also for prodromal PD. In a study assessing the level of uric acid, uric acid/creatinine ratio, total bilirubin and indirect bilirubin, only bilirubin was associated negatively with PD stages. However, the lack of association between uric acid in the serum and total or indirect bilirubin suggests that these two biomarkers may play a different role in the etiopathogenesis of PD [27].
An important merit of our report is that data we used from the PPMI database have been collected and processed uniformly across PPMI centers. On the other hand, limitations include the relatively small number of Prodromal patients of the cohort studied in spite of a thorough clinical and laboratory assessment. Given the fact that clinical studies focused on uric acid related neuroprotective pharmacological treatments, like inosine administration, are underway in motor PD [28], the prodromal PD phase seems to be a promising field for early interventions as well. Future longitudinal studies targeted to elucidate the role of uric acid or bilirubin as putative biomarkers also for premotor PD might pave the way for relevant therapeutic interventions.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted December 5, 2021.  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted December 5, 2021. . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted December 5, 2021. ; https://doi.org/10.1101/2021.12.04.21267290 doi: medRxiv preprint