Grimaldi Stephan, Boucekine Mohamed, Witjas Tatiana, Fluchere Frederique, Azulay Jean-Philippe, Guedj Eric, Eusebio Alexandre
Department of Neurology and Movement Disorders, University Hospital La Timone, Marseille, France
EA3279 Self-perceived Health Assessment Research Unit, Aix-Marseille University, Marseille, France.
J Neurol Neurosurg Psychiatry. 2021 Apr 23. doi: 10.1136/jnnp-2020-324823.
We aim to search for predictors of survival among clinical and brain F-FDG positron emission tomography (PET) metabolic features in our cohort of patients with multiple system atrophy (MSA).
We included patients with a 'probable' MSA diagnosis for whom a clinical evaluation and a brain PET were performed early in the course of the disease (median 3 years, IQR 2-5). A retrospective analysis was conducted using standardised data collection. Brain PET metabolism was characterised using the Automated Anatomical Labelling Atlas. A Cox model was applied to look for factors influencing survival. Kaplan-Meier method estimated the survival rate. We proposed to develop a predictive 'risk score', categorised into low-risk and high-risk groups, using significant variables entered in multivariate Cox regression analysis.
Eighty-five patients were included. The overall median survival was 8 years (CI 6.64 to 9.36). Poor prognostic factors were orthostatic hypotension (HR=6.04 (CI 1.58 to 23.12), p=0.009), stridor (HR=3.41 (CI 1.31 to 8.87), p=0.012) and glucose PET hypometabolism in the left insula (HR=0.78 (CI 0.66 to 0.92), p=0.004). Good prognostic factors were time to diagnosis (HR=0.68 (CI 0.54 to 0.86), p=0.001) and use of selective serotonin reuptake inhibitor (SSRI) (HR=0.17 (CI 0.06 to 0.46), p<0.001). The risk score revealed a 5-year gap separating the median survival of the two groups obtained (5 years vs 10 years; HR=5.82 (CI 2.94 to 11.49), p<0.001).
The clinical prognosis factors we have described support published studies. Here, we also suggest that brain PET is of interest for prognosis assessment and in particular in the search for left insula hypometabolism. Moreover, SSRIs are a potential drug candidate to slow the progression of the disease.
我们旨在探寻多系统萎缩（MSA）患者队列中临床及脑部F-FDG正电子发射断层扫描（PET）代谢特征的生存预测因素。
我们纳入了“可能”诊断为MSA的患者，这些患者在疾病病程早期（中位时间3年，四分位间距2 - 5年）接受了临床评估和脑部PET检查。使用标准化数据收集进行回顾性分析。使用自动解剖标记图谱对脑部PET代谢进行特征描述。应用Cox模型寻找影响生存的因素。采用Kaplan-Meier方法估计生存率。我们提议使用多变量Cox回归分析中纳入的显著变量，制定一个预测性的“风险评分”，分为低风险和高风险组。
纳入85例患者。总体中位生存期为8年（置信区间6.64至9.36）。不良预后因素包括体位性低血压（风险比[HR]=6.04（置信区间1.58至23.12），p = 0.009）、喘鸣（HR = 3.41（置信区间1.31至8.87），p = 0.012）以及左侧岛叶葡萄糖PET代谢减低（HR = 0.78（置信区间0.66至0.92），p = 0.004）。良好预后因素包括诊断时间（HR = 0.68（置信区间0.54至0.86），p = 0.001）和使用选择性5-羟色胺再摄取抑制剂（SSRI）（HR = 0.17（置信区间0.06至0.46），p < 0.001）。风险评分显示，所获两组的中位生存期相差5年（5年对10年；HR = 5.82（置信区间2.94至11.49），p < 0.001）。
我们所描述的临床预后因素支持已发表的研究。在此，我们还表明脑部PET对预后评估有意义，尤其是在寻找左侧岛叶代谢减低方面。此外，SSRI是延缓疾病进展的潜在候选药物。
