Background

Deficiency in oxidative metabolism is associated with the pathophysiology of many conditions ranging from neurologic diseases^1,2 to cardiovascular diseases³ in addition to liver⁴ and genetic dysfunction⁵. ³¹P MRS provides a noninvasive method of measuring metabolic markers such as Phosphocreatine (PCr), Inorganic Phosphate (Pi), Adenosine Triphosphate (ATP), etc. At 7T, additional metabolites such as Nicotinamide Adenine Dinucleotide it is oxidized and reduced form: NAD+/NADH can also be measured. Figure 1 shows an example of a ³¹P MR spectrum at 7T. This allows ³¹P MRS to reveal insights about cellular energy metabolism at the biochemical level in human subjects non-invasively⁶. In exercising skeletal muscle, dynamic ³¹P MRS can be used to obtain nearly real time measurements of phosphorous containing metabolites. One of the primary advantages of ³¹P MRS is that it is non-invasive and can therefore be used for treatment monitoring and to determine the effects of medical interventions. In this study we have quantified the aforementioned muscle metabolites and examined the consistency of these measures before and after an intervention.

Methods

Ergometer Apparatus: A supine leg extension ergometer was designed and constructed specifically for performing exercise targeted at the quadriceps femoris muscle group within the bore of the magnet. The apparatus was constructed with plastic, composite material, and ceramic. The resistance is provided by Kevlar cables which connect the leg extension bar to lead weights outside the 200 Gauss line of the magnet (Figure 2).
Acquisition and Exercise Protocol: All scans were acquired on a Siemens 7 T Terra MRI system with multinuclear capability and dual tuned ¹H and ³¹P surface coil placed at the quadriceps femoris muscle. Proton images were acquired to ensure accurate placement. The ³¹P MRS protocol consisted of a localizer followed by non-localized pulse and acquire sequences (FID). After optimization of the 90 degree flip angle, a TR=12 s FID sequence was run while the subject remained at rest. Then, a TR=2 s FID was run that lasted the full 600 s duration of the exercise challenge protocol (120 s of rest, 180 s of exercise, and 300 s of rest). The subjects consisted of 35 subjects with a wide range of heights and weights. An intervention in the form of exercise or medication was given In all but seven subjects, a 31P MRS scan was acquired after the intervention such that a total of 63 scans were performed.
Data Processing and Analysis: The data was processed with a software package written in Python using standard libraries, numpy, and lmfit (for fitting resonances). The NAD+/NADH (NADx) combined resonance overlaps with the alpha-ATP resonance. This region of the spectrum requires a multipeak model to fit and separate the NADx signal, whereas a single peak fit is sufficient for the other resonances of interest. The PCr measurement from the short TR FID acquisition (TR=2 s) provides dynamic estimates through the exercise challenge. During the rest phase following exercise, the exponential PCr recovery is fit and the time constant, τ, of this fit is calculated. The τ value provides a singular value in units of seconds that can be used as a marker for the state of metabolic function in the skeletal muscle. The dynamic intracellular pH in exercising skeletal muscle cells is calculated from the ppm shift between PCr and Pi.

Results and Discussion

Figure 3 shows an example of the dynamic ³¹P MR Spectra acquired during exercise and Figure 4 shows the changing PCr signal over time. Estimate of PCr and PCr recovery with exponential fitting showed consistent results between the pre-intervention scan measurements and post-intervention scan measurements (pre: 62.4+24.9, post: 61.0+27.5). The high variability is due to intrasubject changes due to intervention or placebo to which the analysis was blinded as an ongoing study however there were no major outliers demonstrating reliable measures. More importantly, the average Root Mean Square Error (RMSE) of the exponential fit for estimating τ values was 0.0230 pre-intervention and 0.0216 post-intervention showing no significant difference.
The NADx resonances were fit and separated from the adjacent alpha-ATP resonances as shown in Figure 5. The NADx peak area was normalized with the beta-ATP peak to obtain the NADx/beta-ATP ratio. The mean and standard deviation of the NADx/beta-ATP ratio was 0.217+ 0.187, before the intervention and 0.157+0.060, after.
Finally, the pH drop in skeletal muscle during the exercise was measured. Before the intervention the average minimum pH during exercise was 6.28 and after it was 6.22.

Conclusion

The results from the quantified values obtained the ³¹P MR Spectra show consistency before and after the intervention. This suggests that the ³¹P MRS acquisition protocol, exercise apparatus, and software pipeline a functioning as intended. Further data analysis from this study will utilize unblinded groups. Group differences between control subjects and subjects who received the intervention will be investigated. The preliminary results suggest ³¹P MRS can be a consistent and reproducible tool for measuring skeletal muscle energy metabolism.

Acknowledgements

No acknowledgement found.