Diagnostic Challenges and Pitfalls in Parkinson’s Disease
Despite significant advances in our knowledge into the causes and mechanisms of neurodegeneration leading to the development of Parkinson’s disease and the availability of sophisticated molecular determinations and imaging techniques, every time we are faced for the first time with a patient presenting with parkinsonian symptomatology, it still remains a diagnostic challenge. Even more so, if we pretend to prognosticate with a certain degree of accuracy the future development of Parkinson’s disease (PD) in the so called “at risk population”.
We know that the pathological process of PD starts decades before the typical clinical features, that allow us to make a diagnosis, appear, namely slowness of movement (bradykinesia), stiffness (rigidity), and abnormal movements (tremor). The disease, during its course, evolves through different stages; a pre-clinical stage in which no symptoms or signs are present, a pre-motor stage characterized by non-motor symptoms such as olfactory loss (hyposmia), acting-out dreams (Rem Behavior Disorder, RBD), depression, and constipation. Finally, motor signs appear, at first subtly, heralding the motor stage, and it is then that the patient seeks the opinion of his primary care physician and is referred to a neurologist to confirm the suspected diagnosis.
To better understand the diagnostic challenges and pitfalls in PD, there are a number of concepts that we need to understand. First, Parkinsonism is an umbrella designation that includes a wide variety of disorders including PD, therefore, not every Parkinsonism is PD, at the risk of being redundant. PD itself is a very heterogeneous disorder, as the mechanisms leading to the pathological process (neurodegeneration) that causes a loss of dopamine producing cells, in a region of the brain called Substantia Nigra, which are key regulators of motor function, are multifactorial and may not be the same in each individual patient (genetic, environmental, age-related, etc.). Moreover, neurodegeneration, also involves other cell types, such as acetylcholine, serotonin, and noradrenaline, subserving other important functions in the brain (cognition, motivation, regulation of autonomous processes, sleep, etc.).
The major challenge we are faced with is that the diagnosis of PD is, at present, done mostly through pattern recognition, and diagnostic criteria are often based on recognition of these typical distinctive patterns and supportive features (phenomenology based). One of our major limitations is the lack of specific biomarkers (laboratory determinations or imaging features) to differentiate PD from other atypical forms of Parkinsonism or controls (except in those cases in whom there is a specific causative genetic mutation, monogenic PD, which can be identified through specific laboratory techniques). Presently, research is focused towards developing different biomarkers through the analysis of certain molecular components of body fluids (blood, saliva, cerebrospinal fluid), abnormal protein deposits in tissues (skin biopsy, salivary gland biopsy), subtle structural changes detected by magnetic resonance imaging (MRI), biochemical changes identified by functional neuroimaging, and abnormal protein deposits in the brain using molecular markers by means of Positron Emission Tomography (PET imaging). Hopefully when these techniques are perfected and become part of the routine diagnostic tools available to the neurologist, we will be able to overcome the pitfalls that we will be discussing here.
We can subdivide them into:
A. Pitfalls at the pre-clinical stage
B. Pitfalls at the pre-motor stage
C. Pitfalls at the motor stage
Pitfalls at the pre-clinical stage
Even in individuals carrying some of the genetic mutations responsible for the “monogenic forms of PD” identified on the basis of a family history of the disease we cannot ascertain if and when they will develop the disease. LRRK2 and GBA-PD mutation carriers, which are the two most common, are not fully penetrant, meaning that having the mutation does not necessarily mean they will develop the disease, as there are genetic modifiers (type of gene mutation, the presence of other gene variants in the same individual), as well as non-genetic modifiers (age, sex, environment, lifestyle), that are critical determinants of its development and the age at which the disease will manifest.
More problematic yet, is the possibility of ascertaining the risk of disease development in the population at large (sporadic or non-genetic cases). In sporadic cases, no single gene mutations are responsible, however, analysis of genetic variations (variants or polymorphisms) in multiple genes has identified 90 “risk loci” (regions of the genes) that are associated with the risk of developing PD. As mentioned in the previous paragraphs the presence of these variations in individual cases does not mean they will develop PD as the so called “polygenic risk scores” are not necessarily informative; moreover, an individual may also carry genetic modifiers that are protective, and also non-genetic modifiers may influence the risk of developing the disease.
Pitfalls at the pre-motor stage
Hyposmia, RBD, constipation, depression, and anxiety (one or all) are usually present in the pre-motor phase of PD. The critical question is how robust are these prodromal non-motor symptoms in predicting “phenoconversion” (developing the typical motor symptoms of the disease) to the motor phase? The strongest predictor, by far, is the presence of RBD (acting out dreams), followed by olfactory loss. However, the presence of one or both of these symptoms does not inform us as to when or if the person will develop PD with a high degree of certainty. Longitudinal follow-up of individuals complaining of olfactory loss showed that only 5.37 % of hyposmics phenoconverted to PD up to 17 years after the initial presentation of the symptom (metanalysis of 3272 cases). The case for RBD is somewhat more complicated, even though the risk of phenoconversion to Parkinsonism is 28% up to 12 years, not all cases phenoconvert to PD, as some of them develop Multiple System Atrophy (MSA) or Lewy-body Dementia (LBD), (two distinct and different forms of Parkinsonism having similar pathology to PD).
Pitfalls at the motor stage
Let us assume that a patient seeking a definitive diagnosis of PD has either the classic presentation triad of asymmetric (on one side of the body) bradykinesia, rigidity and rest tremor or some of the “atypical” presentations of PD such as Isolated unilateral tremor, or Postural Instability Gait Disturbance (PIGD), or symmetric bradykinesia and rigidity without tremor, or pure bradykinesia, or symmetric tremor, without any additional atypical features that may suggest a different diagnosis.
As our diagnosis is based on phenomenology, first we have to differentiate PD from other frequent causes of slowness or tremor such as bradykinesia of the elderly, Depression associated slowness, Hypothyroidism, Essential Tremor, Dystonic tremor, or Drug-induced parkinsonism. After ruling out these diagnoses through careful history taking and neurologic examination, we have to apply the diagnostic criteria established by the International Parkinson and Movement Disorder Society in 2015. These criteria indicate that first we have to demonstrate the presence of parkinsonian signs (bradykinesia + rigidity, and/or tremor) in the absence of “exclusion criteria” (any neurological sign considered atypical or not expected to be present in PD) or “red flags” (additional elements provided by both the history and the clinical exam, such as rapid progression or lack of progression, early falls, early development of autonomic dysfunction like hypotension while standing, deep inspirations, laryngeal stridor, lack of response to levodopa, absence of non-motor symptoms, symmetric parkinsonism). The use of supportive criteria (response to levodopa, dyskinesia, rest tremor, olfactory loss, RBD) helps in increasing the accuracy of our diagnosis.
Despite taking all these precautions, evidence shows that in the early stages of the disease (< 5 years) our diagnostic accuracy is low, going up to 88% after 5 years, meaning that during the first 5 years we have to repeatedly examine and interrogate our patients looking for telltale signs indicating the possibility of diagnoses other than PD (e.g. “Atypical Parkinsonisms”).
An additional factor that may lead to misdiagnosis is using the presence of a strong levodopa response or lack of it as an unequivocal supportive or exclusionary criteria, as reports in the medical literature show that even pathologically proven PD may show an initial poor response to levodopa, and cases of Atypical Parkinsonism may initially show a significant response to the drug.
Another important question that is frequently asked is: Can we predict progression of the disease in an individual patient? Many attempts have been made to classify patients according to subtypes, such as “Tremor predominant” or “Akinetic-rigid predominant” in the belief that those cases in whom tremor is the predominant clinical feature are more benign and therefore carry a better prognosis. This is quite controversial as patients may evolve in the course of the disease, changing the pattern of their clinical features. There are however certain factors that may predict a worse outcome, as it is the case of concomitant Diabetes, Small Vessel Disease (vascular lesions detected with MRI), very late onset of the disease, or early cognitive changes.
The last pitfall I would like to discuss is related to the use of Neuroimaging in certifying the diagnosis. We have several Neuroimaging techniques that may aid in the diagnosis, including classic MRI, Midbrain Sonography, Iron Susceptibility MRI (to detect changes in the Nigrosome 1), DAT-Scan, FD-PET, Cardiac Scintigraphy (to detect sympathetic denervation), however, all of them have limitations. A positive DAT-Scan or FD-PET does not diagnose PD, what they do is provide evidence of a dopamine deficit, which can be seen in Parkinsonisms other than PD. Suggestive imaging findings, using the other techniques previously mentioned, may not be present at all, or if present, they may become evident late in the course of the disease and not help in early diagnosis. None of these other studies are always highly sensitive or specific as there are false positives and false negatives which render them ineffective in offering us an unequivocal diagnosis in individual patients.
In conclusion, for the time being, we still have to deposit our trust in the experience and clinical ability of the treating physician who will have to perform a comprehensive neurological and clinical examination, careful history taking, and resort when necessary to additional studies, and closely follow-up the patient.
Oscar S. Gershanik, MD has been involved in past World Parkinson Congresses for more than 16 years. He was on the steering committee for the Fourth World Parkinson Congress in Portland, Oregon, was the Program co-chair for the Third World Parkinson Congress in Montreal, Canada and was a speaker at the First World Parkinson Congress in Washington, DC. He is presently Professor of Neurology and Scientific Director, Institute of Neuroscience, Favaloro Foundation University Hospital in Argentina.
Ideas and opinions expressed in this post reflect that of the author solely. They do not necessarily reflect the opinions or positions of the World Parkinson Coalition®