Hi all,
I wanted to share the results of a ‘thought experiment’ I’ve been playing with and get feedback on some inevitable follow-on questions.
Just to be clear I'm not advocating anyone does what I describe below - I'm posting it for feedback / discussion purposes only (and for someone to show me where it's already been done - and better - (but, hey, I enjoyed the journey :)).
Scenario: A PWP experiencing wearing-off of LD/CD is attributing this to the somewhat arbitrary default of evenly spaced doses suggested by the neurologist. What is the optimal medication schedule - one that reduces the variability of the levodopa equivalent dose (LED) value during the working day (~9:30am-5:30pm)? Can you create a ‘virtual LD pump’ dosing profile?
Based on my understanding of LED from the very useful parkinsonsmeasurement.org/t... the LED for LD/CD IR at evenly spaced doses could swing around during the day and the downward slopes could cause various, significant, wearing-off events during the day.
Assumptions:
A PWP with enough residual self-produced dopamine to ~1030, then a need for boost >=0.6LED during the working day.
Baseline medication schedule: With three LD/IR (100/25) (one at 0900, 1200 and 1500) you get the following profile with these spikes (and IMHO three ‘sharp’ wearing-off periods):
[FIGURE 1 - attached at end]
Medication ‘solution space’ constraints: Permutations of LD/CD 100mg IR and LD/CD 100mg CR up to a combined total of four pills.
Using the reported average blood plasma level profiles from the above website for these two medications I wrote a bit of code to find the optimum dosage times that aims for a minimum of 0.6LED and to minimise the standard deviation during the working day time period. (I did this a few months ago so need to double-check the nuances - if I don’t post text in this state I never will!)
This suggested the graph below (where the red line is 100mg LD/CD IR (0907) and green, blue and yellow lines are now 100mg LD/CD CR (at 1050, 1256 and 1524)). (Timings to the minute are a bit academic but hey this is a thought experiment :)) There's certainly less erratic LED levels.
[FIGURE 2 - attached at end]
Now the questions I have are:
(a) what assumption can you make about lunch? Do you assume that for the lunch window (30 mins before and 60 mins after) there is no absorption? Or do you assume no lunch?
(b) does anyone have any absorption data for dispersible LD? Could this be a way to further smooth out the profile with sips of a LD drink? (Don’t want to go too much off-topic on that one though)
(c) this is obviously a hypothetical situation for a non-existent ‘average’ PWP at a point in their 'journey' who wants a steady 0.6LED for a given window on an ‘average’ day. It would be interesting to try adding some noise factors (e.g. lunch, timings, LD absorption rates & half-lives) – to create an uncertainty profile to overlay. Does anyone have any suggestions what reasonable uncertainties might be? (I still need to read the references in the above website – just thinking ahead really))
To reiterate I'm not advocating anyone try what I describe - I'm posting it for feedback / discussion purposes only.
Thanks in advance for any suggestions / feedback.