There is Currently No Clear Evidence That Linear or Macrocyclic Gadolinium-Based Contrast Agents Cause Parkinson’s Disease or Parkinson Syndrome

I read with interest the article by Shin et al. [1] on a retrospective cohort study examining the causal relationship between exposure to linear or macrocyclic gadolinium-based contrast agents (GBCAs) and the development of Parkinson’s disease (PD) or Parkinson’s syndrome (PS). It was found that exposure to linear GBCAs slightly increased the risk of parkinsonism of any cause, whereas exposure to macrocyclic GBCAs did not increase the risk of PD/PS [1]. It was concluded that caution is warranted when using linear GBCAs until further evidence becomes available [1]. The study is forward-looking, but some points require further discussion.

The first point concerns the retrospective study design [1]. Retrospective designs have several drawbacks [2]. These include poor data quality, missing data, the inability to establish a causal relationship (merely an association), susceptibility to recall bias and selection bias, difficulties in controlling for confounding variables, and a generally lower level of evidence compared to prospective studies [2].

The second point is that I disagree with the view that linear GBCAs increase the risk of developing PD/PS. Numerous studies have already shown that GBCAs do not increase the risk of developing PD/PS [3].

The third point is that only a limited number of alternative risk factors for the development of PD/PS were included in the analysis (diabetes, hyperlipidemia, hypertension, malignancies) [1]. Other known risk factors for the development of PD/PS include age, male sex, mutations in LRRK2, PARK7, PINK1, PRKN, SNCA, or GBA1, long-term exposure to pesticides, herbicides, solvents, or heavy metals such as manganese, lead, or iron, previous traumatic brain injuries, a high-fat diet, a diet high in processed meat, cardiovascular disease, hepatitis C, diabetes, tobacco use, and possibly caffeine [4]. As long as these additional risk factors have not been ruled out as possible contributing causes for the development of PD/PS, the presented results remain unreliable. Since the number of patients with hypertension, diabetes, and hyperlipidemia was higher in the linear GBCA group, this could simply explain the higher prevalence of PD/PS in the linear GBCA group. The higher proportion of male subjects in the linear GBCA group compared to the macrocyclic GBCA group could also explain the increased risk of PD/PS in the former group.

The fourth point is that there is no pathophysiological concept that impaired function of globus pallidus or the dentate nucleus is involved in the development of PD/PS. PD/PS is generally regarded due to loss of dopaminergic neurons originating in the substantia nigra, but not in the basal ganglia.

The fifth point is that the cumulative doses received by the patients were not specified and were not included in the analysis [1]. Since it is conceivable that the risk of PD/PS is dose-dependent and increases with the cumulative gadolinium dose, this parameter should be included in the analysis.

The sixth point is that it is unclear why patients with secondary parkinsonism were included in the study. Secondary parkinsonism is defined as a consequence of medications (e.g., metoclopramide, chlorpromazine), other neurological disorders (e.g., traumatic brain injury, Lewy body dementia, meningitis, multiple system atrophy, progressive supranuclear palsy, stroke, Wilson’s disease), or other conditions (e.g., human immunodeficiency virus, carbon monoxide poisoning, mercury poisoning, narcotics, MTPT). Patients with secondary parkinsonism should be excluded from the study to avoid skewing the results.

The seventh point is that patients were followed up for three years during the pandemic [1]. Since SARS-CoV-2 infections can be complicated by PD/PS [5], we should know how many of the included patients developed PD/PS after January 2020 and could be attributed to COVID-19. These patients should be excluded from the analysis.

The eigths point is that the assessment of renal function two years prior to the index MRI does not necessarily rule out the possibility that renal insufficiency - and thus gadolinium overload - was present at the time of the index MRI.

Another argument against the hypothesis that GBCAs cause PD/PS is that gadolinium is deposited not only in the substantia nigra but also in the putamen, globus pallidus, nucleus ruber, thalamus, colliculi, and superior cerebellar peduncle; however, no other movement disorders such as chorea, choreoathetosis, isolated athetosis, tremor, ataxia, dystonia, or RLS were reported. Such movement disorders would be expected based on the pathophysiological findings.

Finally, we should know on what basis patients who had undergone cerebral or spinal MRI were excluded.