Evaluation of Normal Changes in Pons Metabolites due to Aging Using Turbo Spectroscopic Imaging

Discussion

To develop a proton MR spectroscopic data base of in vivo changes in pons metabolite levels during aging, we used the TSI technique to analyze 78 healthy subjects. These data on normal pontine metabolite levels may be applied in future studies of pontine lesions. The results showed that a short TSI scan time could be used to produce good reconstructed spectra for clinical analysis. Our results indicate that the NAA/Cr in the pons rose rapidly in the 0- to 5-year-old group, reached a peak in the 10- to 20-year-old group, and then gradually decreased. The Cho/Cr decreased rapidly in the 0- to 5-year-old group, then continued to decrease, and remained stable after 20 years of age. The Cho/NAA decreased rapidly in the 0- to 5-year-old group, followed by a decline in the 6- to 20-year-old group, then the ratio slowly rose after 20 years of age.

Currently, MR spectroscopy analysis of normal brain tissue has focused on the frontal, temporal, and occipital lobes, and there are few relevant data for the pons.⁶ This lack of data is due to obstacles to using MR spectroscopy on the pons. The posterior fossa is surrounded by skull base bones, and the pons itself is a relatively small organ. These anatomic features are disadvantages for MR spectroscopy because it requires a uniform magnetic field. The TSI technique used in this study provides a solution to the above problems. Unlike ordinary MR imaging techniques that use frequency and 1 phase direction for spatial encoding, chemical shift imaging usually uses 2 phase directions for spatial coding, which leads to a relatively longer acquisition time; faster acquisitions typically require the sacrifice of spatial resolution.¹⁰

Compared with a standard 2D point-resolved spectroscopic sequence, TSI uses an additional 180° pulse (like fast spin-echo) to obtain an additional phase-encoding data-acquisition window, thus reducing the acquisition time. Also like the FSE sequence, with its multiecho technique, TSI is less sensitive to B0 inhomogeneity. Moreover, TSI can be combined with sensitivity encoding to further shorten the scan time and be more insensitive to susceptibility effects.^10,17 The major disadvantage of TSI is that it sacrifices spectral resolution because there is insufficient time between the 2 successive refocusing pulses to read the signal. However, by using TSI, 4.4-Hz spectral resolution can be obtained on a 1.5T system, which is sufficient to resolve NAA, Cr, Cho, and lactate spectra. However, for short TE metabolites such as glutamate and glutamine and myo-inositol, the signal rapidly decays with increasing TE, rendering the long echo spacing (at least 100 ms) of TSI impractical for their detection.^9,10 In this study, the TSI scan time was reduced to approximately 1 minute 20 seconds, the baseline of the reconstructed spectrum was flat, the metabolite peaks were sharp, and the NAA, Cho, Cr spectra were well-resolved.

The levels of various metabolites in the pons vary with age as the human brain undergoes a maturation process. In a neonate's brain, the Cho level is at its highest, and the NAA level is at its minimum. NAA mainly exists in mature neuronal cells, and its levels rise with the development of the brain. Cho is the precursor to phosphatidylcholine, which is the main component of cell membranes. During brain development, the maturation of myelin and the synthesis of cell membranes cause rapid consumption of Cho, and its level declines.^11,18,19 The decline of the Cho/Cr may also be related to the increase of Cr after birth, but Cr is mainly stable after the first year of life and can be used as an internal reference.¹⁸ In this study, the NAA levels increased rapidly in the 0- to 5-year-old group and the Cho levels declined rapidly, in support of the above viewpoint. The Cho/Cr value in the pons is relatively high (>1). The pons is characterized by a high attenuation of white matter because it contains many astrocytes and oligodendrocytes, which have a higher Cho content than neurons. This condition is similar to the higher Cho levels observed in the centrum semiovale compared with the cortex.^12,20,21

A previous MR spectroscopy study on a healthy population of subjects 4–88 years of age showed that the NAA/Cho of the white matter in the centrum semiovale increased rapidly in the first decade of life, reached a maximum by the second-to-third decades of life, and then declined steadily.¹¹ Although most maturation events in the human brain occur within the first few years of life, the brain is still establishing functional neuronal interconnections until puberty, and this maturation process is associated with significant alterations in the biochemical composition of the brain. As an example, the increase in the number of neuronal cells in the centrum semiovale reaches its peak by approximately 6 months after birth, but maturational changes continue to occur after infancy, with significant increases in the number and size of axonal and dendritic fluctuations.^8,11 In this study, the NAA and Cho levels in the pons reached a relatively stable phase by 10–20 years of age, which may correspond to the full maturation of the pons. NAA/Cr continued to decline in the 21- to 50- and the older than 50-year-old groups. This change implies that NAA decreases gradually after puberty. There were few changes in the Cho/Cr among the groups of 6–20, 21–50, and older than 50 years, implying that Cho is relatively stable after puberty. This result is similar to that reported previously by Moreno-Torres et al.¹³ The reduction in the level of NAA may be due to diminished neuronal attenuation and/or neuronal loss in the aged brain and also a diminished concentration of NAA in neurons.²¹ With the exception of neurons, Cho is widely present in various types of cells, particularly glial cells, which undergo less dramatic changes with aging than neurons. The minimal changes in Cho levels after puberty may also be explained by the release of Cho compounds related to membrane breakdown during the aging process.^13,21

There are some limitations of this study. First, the population sample enrolled in this study was relatively small, and more data should be collected in future studies to correct possible bias. However, the age range examined in this study was wide, with subjects from 0 to 78 years of age. In particular, we examined subjects during the first 5 years of life, which is a critical period for the developing brain, despite the difficulty in collecting data from children without sedation. Second, the regions of interest in TSI do not cover the whole pons. We attempted using 3D MR spectroscopy but the time required was not acceptable and the larger coverage to include the lower parts of the pons caused more difficulty in shimming. Therefore, we chose the region of interest covering as much of the pons as possible.