An objective fluctuation score for Parkinson's disease

Abstract

Introduction: Establishing the presence and severity of fluctuations is important in managing Parkinson's Disease yet there is no reliable, objective means of doing this. In this study we have evaluated a Fluctuation Score derived from variations in dyskinesia and bradykinesia scores produced by an accelerometry based system.

Methods: The Fluctuation Score was produced by summing the interquartile range of bradykinesia scores and dyskinesia scores produced every 2 minutes between 0900-1800 for at least 6 days by the accelerometry based system and expressing it as an algorithm.

Results: This Score could distinguish between fluctuating and non-fluctuating patients with high sensitivity and selectivity and was significant lower following activation of deep brain stimulators. The scores following deep brain stimulation lay in a band just above the score separating fluctuators from non-fluctuators, suggesting a range representing adequate motor control. When compared with control subjects the score of newly diagnosed patients show a loss of fluctuation with onset of PD. The score was calculated in subjects whose duration of disease was known and this showed that newly diagnosed patients soon develop higher scores which either fall under or within the range representing adequate motor control or instead go on to develop more severe fluctuations.

Conclusion: The Fluctuation Score described here promises to be a useful tool for identifying patients whose fluctuations are progressing and may require therapeutic changes. It also shows promise as a useful research tool. Further studies are required to more accurately identify therapeutic targets and ranges.

Conflict of interest statement

Competing Interests: MH has a financial interest in Global Kinetics Corporationl, the company that manufactures and supplies the PKG. SM's salary is in part funded by an open grant from GKC, the company that manufactures and supplies the PKG. This does not alter the authors' adherence to PLOS ONE policies on sharing data and materials.

Figures

Fig 1. The interquartile range of bradykinesia and dyskinesia scores.

A. The BKSIQR (green dots) and DKSIQR (green dots) for each patient was plotted against the IQRC. The error bars adjacent to the Y axis represent the median and interquartile range of the BKSIQR (green lines) and DKSIQR (red lines) shown in this figure. B. The BKSIQR (green dots) and DKSIQR (green dots) for each patient was expressed as a percentage of the IQRC and plotted against the IQRC. Note that there no IQRC values are less than 10. Note that the vertical line at IQRC = 22.5, where the BKSIQR contributes to about 60% to the IQRC, corresponds to the point that separates fluctuators and non fluctuators in Fig 2, showing that both BKSIQR and DKSIQR contribute significantly to the IQRC at this point.

Fig 2. The interquartile range of bradykinesia and dyskinesia scores in fluctuators and non fluctuators.

A. The BKSIQR, DKSIQR and IQRC of fluctuators (F and olive colour) and non fluctuating (NF and olive colour) patients were plotted. While the two populations can be separated using either DKSIQR or IQRC, the separation is more distinct for the IQRC. In each plot, the horizontal red dotted lines show values from receiver operator curves that resulted in the highest selectivity and sensitivity for separating the two populations. In the case of the IQRC plots the red dotted line corresponds to an IQRC of 31.4 which separates the two groups with greater sensitivity and specificity than either the BKSIQR or DKSIQR. Note that all IQRC >100 are shown as = 100. Error bars show median and interquartile range. B. This is a plot of the BKSIQR, DKSIQR and IQRC of patients with PD for less than 3 years (<3 and olive colour) and of patients on the waiting list for DBS (WL and olive colour). In each plot, the horizontal red dotted lines show values from receiver operator curves that resulted in the highest selectivity and sensitivity for separating the two populations. In the case of the IQRC plots the dotted line corresponds to an IQRC of 22.5 which separates the two groups with greater sensitivity and specificity than either the BKSIQR or DKSIQR. This is marked by an asterisk and is the value that becomes the FT and corresponds to an FS of 7.7 when the fluctuation formula is applied). Note that all IQRC >100 are shown as = 100. Error bars show median and interquartile range.

Fig 3. The Fluctuation Score before and after Insertion of Deep Brain Stimulators.

Graphs showing the median and interquartile range of rating scores and PKG measures before and after DBS in 15 PD patients. The Y axis is the relevant value of the scale or PKG measure. The FS threshold for transition from non fluctuators to fluctuators (see asterisk Fig 2B) is indicated by a dotted line marked 7.7. The shaded orange region is the area between the median and 75th percentile of FS (= 12.8) after DBS (See text for discussion). Note that the median FS after DBS (8.2) is very close to the FT: the value that separates fluctuators from non fluctuators in Fig 2 (FS = 7.7) Error bars show median and interquartile range. P values are obtained using Mann Whitney.

Fig 4. The change in Fluctuation Score with disease progression.

A: The Fluctuation Score of subjects described by the following captions were plotted. C: Control subjects (grey dots). ND: people with PD who were newly diagnosed (green dots). 3Y: non fluctuators with disease duration more than 3 years (>3Y) (green dots), EF: “early” fluctuators (EF) meaning modest non troublesome fluctuations (brown dots), NS: fluctuators but not suitable (NS) for DBS (dark red dots), PRE-S: on the waiting list for DBS (crimson red dots). POST-S: Post DBS (black dots). The dotted line with a black circle is the FS scores that separates fluctuators and non-fluctuators (the FT) and the grey shaded region is the region between the FS scores for the median and 75th percentile of Post DBS subjects (as in Fig 3). Note that this coincides with the interquartile rage of controls. Note the reduction of the FS score which is described in the results section and in the Discussion. B): The FS score of 177 subjects plotted against duration since first symptoms. Subjects whose FS is blow the FT are shown as green dots. Subjects with fluctuations within the RCMS are shown as tan dots and subjects with FS above the RCMS as red dots. The FS of Control subjects (from Fig 4A) are included (grey dots). Note that the black horizontal dotted line is a continuation of the median of controls. The FS scores within or above the RCMS became frequent after 3 years. Some subjects who had disease for many years had FS scores below the median of controls (dotted line) or even below the FT (see Discussion). C: This is a cartoon depicting the possible changes in FS over the course of PD and is based on Fig A and B above. The orange horizontal band represents the range in FS produced by an intervention such as DBS with the upper limit being the upper level of an acceptable response and the lower limit being FT (shown by the dotted line). Our proposal is that as non-PD subjects have an FS in this band and newly diagnosed PD are at a point below the cross-over point, then PD must initially produce a decline in FS until a diagnosis is made. With treatment and time, many subjects will have a progressive increase in their FS, eventually crossing the cross-over point and finally becoming a frank fluctuator and suitable for an intervention.