Ultrashort time-to-echo T2* and T2* relaxometry for evaluation of lumbar disc degeneration: a comparative study

Low back pain (LBP) is a leading cause of disability worldwide, placing a great burden on the global health care system [1, 2]. Intervertebral disc (IVD) degeneration (IVDD) is a significant contributor to nonspecific LBP, with a lifetime prevalence of over 80% [3, 4].

Early stages of IVDD are mainly in the form of biochemical changes, including proteoglycan (PG) reduction, dehydration, and collagen degeneration. It will lead to a decrease in hydrostatic pressure, resulting in nucleus pulposus (NP) dehydration and loss of the structural and mechanical properties of the IVDs. In advanced stages of IVDD, along with loss of hydration and the subsequent drop in disc pressure, IVD height decreases under load [5‐9]. These degenerative changes are accompanied by structural lesions, such as disc herniation, causing LBP, neurogenic claudication, and even cauda equina syndrome. At this stage, treatment strategy is limited to conservative treatment alone or surgical excision [10]. Early detection of alterations in IVDD is important for developing preventative strategies or reestablishing degenerated IVDs, such as gene therapy, stem cell therapy, and growth factor therapy [11‐13].

Conventional magnetic resonance imaging (MRI) is widely used for morphologic, qualitative assessment of IVDD in the clinical workup. Lumbar IVDD is commonly scored using the Pfirrmann grading system, which is based on the assessment of structure and loss of the signal intensity on T2-weighted imaging (T2WI). This grading system provides a standardized and reliable assessment of MRI disc morphology, but cannot detect early degeneration of IVDs characterized by a loss of PG quantitatively [14, 15].

Several quantitative MRI techniques to evaluate IVD degeneration objectively have been reported, such as diffusion-weighted imaging, diffusion tensor imaging, glycosaminoglycan chemical exchange saturation transfer, sodium, delayed gadolinium-enhanced MRI, T2/T2*, and T1rho mapping. Previous studies have demonstrated that T2* mapping could be quantitative imaging biomarkers for evaluating the biochemical state of the discs and correlating that with histology, water content, and degeneration [16].

Ultrashort time-to-echo (UTE) imaging as a novel MRI technique has the capacity to catch very short T2* signals (0.008 ~ 0. 50 ms) [17‐20], It has been confirmed to be sensitive to changes in the deep tissue matrix and to subtle and even preclinical degeneration [21]. To date, UTE-T2* imaging has been reported to be a reliable tool for quantitative assessment of the biochemical changes of short T2 tissues, including tendon, cartilage, and ligament, etc. [20‐25]. However, to the best of our knowledge, studies on UTE-T2* quantitative technique in evaluating IVDD are scarce. We hypothesized that quantitative UTE-T2* mapping is capable of revealing degenerative changes in the discs.

The present study aimed to assess whether lumbar IVDD can be evaluated using UTE-T2* mapping and to compare the potential of UTE-T2*and T2* values in the diagnosis of early IVDD.

This study is the first to investigate and compare the diagnostic efficacies of UTE-T2* and T2* mapping in detecting IVDD in humans. The results may help to confirm the feasibility and specificity of UTE-T2* as an objective and quantitative tool to identify early degenerative changes of the disc and show promise for clinicians to modify the diagnostics and therapeutic management strategies more accurately.

Conventional MRI, such as the Pfirrmann scale with T2WI, was limited in detecting ultrastructural alterations of early IVDD. Early stages of disc degeneration include biochemical changes, such as a loss or reduction of PG content, which can ultimately lead to dehydration. T2 relaxation reflected the integrated environment of the IVD, including water, protein, collagen, and other solutes [31], and was sensitive to water content and the composition of the collagen network structure. Researchers have reported that T2* relaxation time showed a good correlation with PG and collagen contents in 18 humans cadaveric IVDs [32]. Our findings confirmed an increase in the quantitative T2* values between the AAF and NP and a decrease between the NP and the PAF. In line with an earlier report [33, 34],T2 and T2* mapping provided roughly similar results. The inverse correlation of the T2 relaxation time in the disc with Pfirrmann grade has been reported by Welsch et al. [35] and Noebauer et al. [36], the early study reported a low-to-moderate correlation between Pfirrmann grades and T2 relaxation times, which were consistent with the Spearman's correlation coefficient between Pfirrmann grades and T2*value in our results. Both T2 and T2* mapping differ in the biochemical sensitivity of disc tissue, with T2 mapping being sensitive to tissue hydration, while T2* mapping being more sensitive to changes in tissue integrity [35].T2* mapping provide more valuable biochemical information on the IVDs ultrastructure, together with three-dimensional acquisition capability and higher spatial resolution in a short scan time [37].

UTE-T2* mapping was acquired using different echo times in the short (1–10 ms) and ultrashort echo time range [19, 22]. Because UTE-T2* mapping can catch the short T2* relaxations from tissues, it is more sensitive to biochemical collagen matrix changes compared with conventional MRI techniques, based on histologic standards [19].

Multiple large general-population–based studies have proved that UTE-T2* mapping can detect cartilage subsurface matrix changes, which can be indicative of reduced cartilage health from injury or early degeneration noninvasively [20, 23, 25]. Similar to the results reported by Detiger et al. [16], we observed a trend of decreasing UTE-T2* value with increasing degree of degeneration. The previous study revealed a significant correlation between T2* relaxation time and glycosaminoglycans (GAG) content in the nucleus pulposus, as well as histologic scoring with varying grades of degeneration [16]. During the aging process, the quantity and quality of PG and collagen contents diminish, along with a decrease in short T2* signal accordingly [21, 25, 34, 38]. T2* relaxometry appeared to be sensitive to water and PG contents. This may be the initial step in the degenerative cycle [25, 34, 38], which could be the underlying reasons for the decreased UTE-T2* and T2* value.

In NP and PAF, UTE-T2* mapping showed significantly higher diagnostic accuracy in differentiating early disc degeneration from normal than did T2*. Theoretically, both UTE-T2* and T2* mapping measured the T2* value of the tissue. UTE MRI technique mitigates the rapid signal loss from short T2* by reducing the TE to the scale of 0–200 microseconds to sample the free induction decay as early as possible. With considerably shorter TEs (0.032 ms in this study) than T2*, UTE-T2*mapping allows signals from very short T2 components to be detected [23]. Thus, UTE-T2*mapping is less sensitive to the magic angle effect and more sensitive to water protons and their local environment, making it a satisfactory method for evaluating disc generation. Because T2* relaxation time has been reported to reflect both the water content and PG content reduction [16], it is not hard to understand why the UTE-T2* value has better diagnostic accuracy than T2* for differentiating early disc degeneration.

Previous studies have reported that T2 relaxation time of Pfirrmann grades IV is significantly shorter than Pfirrmann grades III, and no significant differences were found between Pfirrmann grades IV and V, both of which show extremely low signal intensity [39]. The results of our study showed that compared to T2* values, UTE-T2* values conveyed significantly higher diagnostic performance in distinguishing the early disc degeneration from the advanced in PAF. Takashima et al. [40] reported that short T2* relaxation times with UTE are promising for assessing progressive IVD degeneration with poor water content, such as fibrosis change of IVDD with short T2 relaxation time. Our results are consistent with those findings quantitatively. Takashima et al. did not further discuss the quantitative evaluation of the early disc degeneration because their study population did not include grade I IVDs. Our results would appear to complement and refine their research.

A previous report on healthy ovine IVDs demonstrated that the T2 values show regional variation in discs and reported that high T2 values were observed in NP and low T2 values in the AAF and PAF when histologically evaluated [41]. Disc degeneration is believed to originates in NP with depletion of GAG, followed by a reduction in water content [5, 42]. Similar to previous reports, our study showed that correlations with Pfirrmann grade and UTE-T2* and T2* values were highest in NP and lowest in AAF. Our results also showed that NP had the highest diagnostic accuracy in predicting the early degeneration IVDs, while AFP and PFP were similar in predicting the early degeneration IVDs. These results suggest that the destruction of hydrophilic GAGs within NP was the main cause of the accumulation of cleaved extracellular matrix fragments with disc aging [16].

There were some limitations in this study. First, our study had no detailed histologic confirmation associated with IVDD changes. This is hard to achieve in humans. In addition, the relationship between T2* values and biochemical changes in IVDD has been previously established in human cadaveric lumbar discs [32]. Second, we were unable to compare the related clinical symptoms with the degree of degeneration in MRI quantitative parameters. Future research is warranted to explore these ideas. Third, no patient in this study showed grade V IVDs because grade V IVDs tend to have a collapsed disc space or a vacuum phenomenon, which is unable to measure the quantitative values due to susceptibility artifacts. If the complete grade V IVDs data set were available, quantitative evaluation of advanced IVD degeneration could be closer to reality. However, as our principal purpose was to detect early disc degeneration, the impact of incomplete grade V IVDs dataset on our results is within acceptable limits.

We demonstrated that UTE-T2*mapping was more accurate than T2* mapping in quantitatively diagnosing early intervertebral disc degeneration. In particular, UTE-T2* mapping allowed for precisely distinguishing disc degeneration, potentially providing a promising imaging biomarker with potential applications in intervertebral disc degeneration for the emerging cell-based therapies.