Reply to Letter to the Editor: “Disease Progression in Multiple System Atrophy: The Value of Clinical Cohorts with Long Follow-Up”

Abstract

We are pleased to read the letter by Saulnier and colleagues related to our article “Disease Progression in Multiple System Atrophy—Novel Modeling Framework and Predictive Factors,”¹ and we thank the authors for the impressive work they have done to replicate and expand on our research in a larger and more contemporary cohort.

As Saulnier and colleagues rightly point out, our study relied on a cohort of 121 patients with a relatively advanced disease stage and a limited follow-up period of 2 years. We acknowledge these limitations and fully agree on the importance of both larger sample size and longer individual follow-up duration for achieving high-quality estimates of the natural history disease trajectory in multiple system atrophy (MSA).

The authors’ replication of our disease progression modeling using the larger French MSA cohort of 663 patients with a follow-up period of up to 11 years estimated a longer duration of the MSA progression trajectory compared to our findings. Although differences between the cohorts were limited in terms of average baseline disease severity and symptom duration (5.4 years in the European MSA study cohort vs. 4.5 years in the French MSA cohort), some differences are to be expected due to other cohort differences (years in which the study was ongoing, study locations, etc.). Interestingly, the authors explored how restricting longitudinal follow-up data to 2.5 years affected the estimates. The authors found that restricting samples resulted in a compressed trajectory with smaller interpatient differences compared to using long-term follow-up data. Based on these findings, Saulnier and colleagues suggest that studies of restricted follow-up time could overestimate progression rate and underestimate interpatient differences, which we agree may be the case. Our disease progression model performs temporal recalibration on a latent timescale, which requires a trade-off between the (vertical) deviation from the estimated mean trajectory and the (horizontal) deviation on the timescale. The trade-off is determined in a fully data-driven approach based on the maximum likelihood principle. When long-term patterns are not sufficiently clear in individual patient-level trajectories, the model may prioritize minimization of vertical differences over estimating temporal patterns. One aspect of the estimation that was not addressed by Saulnier and colleagues is the role of multiple outcome measures. In our study, we aligned patient samples using six different outcome measures, whereas only two outcome measures were used for replication in the French MSA cohort. Including a greater number of disease-relevant measures increases the number of observed data points per patient used to predict their latent disease time, which could possibly alleviate some of the issues related to shorter-term follow-up.

This replication study based on the French MSA cohort underscores the importance of long-term follow-up data to accurately describe long-term disease progression in MSA. These findings are important not only for the design of observational studies but also for interventional studies of potentially disease-modifying therapies, where longer individual follow-up will enable a better understanding of how treatment changes the course of disease.

L.L.R. is currently an employee at Lilly and was employed at H. Lundbeck A/S when the work of the research paper was done. D.O.Å., A.-K.B., I.H.H., and J.L.M. are employees at H. Lundbeck A/S. L.K. is an employee at BEC Financial Technologies and was employed at H. Lundbeck A/S when the work of the research paper was done. F.K. received personal fees from Institut de Recherches Internationales Servier, Clarion Healthcare, Takeda Pharmaceuticals, Sanofi, Teva, Bial, and the Austrian Society of Neurology and received grant support from the MSA Coalition, Austrian Science Fund, and the National Institutes of Health outside of the submitted work. G.K.W. reports speaker fees, consulting fees, and honoraria from Biohaven, Biogen, Inhibikase, Lundbeck, Ono, Takeda, and Theravance and grants from the Austrian Science Funds, Parkinson funds, MSA Coalition, NIH ATB, the International Parkinson Disease and Movement Disorders Society, Austrian Parkinson Society, Austrian Autonomic Society, Austrian Autonomic Society, Eisai, Minoryx, Novartis, Vaxxinity; royalties from Springer Nature Publishing Group and Thieme Verlag. K.S. is a recipient of grants from the FWF Austrian Science Fund, The Michael J. Fox Foundation, International Parkinson and Movement Disorder Society, AOP Orphan Pharmaceuticals AG, and EU (all to the institution). K.S. has received payment for consultations from Ono Pharma UK, Lundbeck, and Ever Pharma. K.S. has received payment for lectures from Teva, UCB, AOP Orphan Pharmaceuticals AG, Roche, Grünenthal, Stada, Licher Pharma, Biogen, Bial, and AbbVie. K.S. has received payment for participation in advisory boards from Bial, Stada, and AbbVie, and honoraria from the International Parkinson and Movement Disorders Society. W.P. reports consultancy and lecture fees from AC Immune, Alterity, AbbVie, Affiris, Bial, Biogen, Britannia, Lilly, Lundbeck, Neuroderm, Neurocrine, Roche, Sunovion, Stada, Takeda, UCB, and Zambon in relation to clinical drug development programs for Parkinson's disease; personal fees from AstraZeneca, Boston Scientific, Intec, Ipsen, Denali Pharmaceuticals, Novartis, Orion Pharma, Prexton, Teva; as well as royalties from Thieme, Wiley Blackwell, Oxford University Press, and Cambridge University Press and grant support from the MJFF and EU FP7 and Horizon 2020 programs.

Roles of conception and design (C), drafting (D), critical review and revision (R) were:

L.R.: CDR

I.H.: CDR

L.K.: R

D.Å.: R

A.B.: R

F.K.: R

G.W.: R

K.S.: R

W.P.: R

J.M.: R