Best Ways to Interpret Test Results in PPP

©KieferPix/Shutterstock

A recent study supports the use of the McManis method over the Fournier method, and has proposed cutoffs for abnormal test results for confirmation of primary periodic paralysis (PPP) using the long exercise test.¹ PPP is a rare genetic disorder in which affected individuals experience flaccid weakness, often accompanied by abnormal (high or low) potassium levels.

While clinical diagnosis may be relatively straightforward in patients who present during acute episodes, many patients present days or weeks after an attack when neurological and potassium levels have normalized. Additionally, symptoms may overlap with other neurological or medical conditions, further complicating diagnosis. Genetic testing is confirmatory in only 60% to 70% of cases diagnosed clinically.

The long exercise test (LET) is often used to confirm suspected diagnoses of PPP. Muscles in most individuals with PPP become less excitable over time, resulting in a decline in the amplitude and area of the action potential that triggers muscle contraction. The LET measures this decline for a sustained, isometrically exercised muscle.

The best cutoffs for abnormal LET test results have not been defined, however, and various methods are used to calculate the results:

• One technique is called the McManis method, also called the peak-to-nadir method, which defines the LET amplitude as a percentage of the post-exercise peak value with a cutoff of >40%.
• A second technique is the Fournier method, also called the baseline-to-nadir method, which defines the LET amplitude as a percentage of the pre-exercise baseline value with a cutoff of >20%. A third method calculates the decrease in area, which uses the peak-to-nadir method and has a cut-off of 47%.

Most studies define abnormal cut-offs as two standard deviations above the mean, which is calculated from healthy controls. These cutoffs can result in tests that are good at weeding out PPP in individuals who do not actually have the disorder but also suffer from high rates of false negatives.

To provide better guidance on how to calculate and interpret the LET, first author Daniel Simmons, MD, University of Rochester Medical Center, and colleagues conducted a retrospective study that included 55 individuals with PPP (32 of which were genetically confirmed), 82 controls with other neuromuscular conditions, and 43 healthy controls. 

Analyses showed that the peak-to-nadir (McManis) method was consistently more accurate for both the amplitude and area LET measurements compared to the baseline-to-nadir (Fournier) method. There was no consistent trend favoring amplitude over area.

Researchers then estimated optimal cutoff levels based on various pre-test probabilities, or the clinical impression that the patient actually had PPP. For a pre-test probability of ≤50%, they estimated that the optimal cutoffs using the peak-to-nadir method were >40% (amplitude) or >50% (area). Analyses suggested that these cutoffs would increase diagnostic certainty and yield a true negative rate of essentially 100%. When the pre-test probability is ≥90%, more liberal cutoffs of >25% (amplitude) and >36% (area) could be used.

Results also suggested that combining LET results on both amplitude and area is more useful than either one alone and can better identify true positives without affecting the true negative rate.

Although 67% of patients with PPP in the study were male, results showed no statistically significant differences based on sex, suggesting the cutoffs can be applied equally to both.

Quick take homes

• Retrospective study found the peak-to-nadir (McManis) method is the best way to calculate results for the long-exercise test in PPP.

• Optimal cutoffs using the peak-to-nadir method were found to be >40% (amplitude) or >50% (area), for a pre-test probability of ≤50%.

• More liberal cutoffs could be used for patients with higher pre-test probability.

References:

1. Simmons DB, Lanning J, Cleland JC, et al. The Long Exercise Test in periodic paralysis: A Bayesian analysis. Muscle Nerve. 2018 May [Epub ahead of print].