Neurochemical changes may play a role in the pathophysiology of IBS. We aimed to characterise neurochemical changes associated with Irritable Bowel Syndrome (IBS) using in-vivo dimensional localized correlation spectroscopy (2D L-COSY) at 3T with a 64 channel head and neck coil.IBS is a characterised intermittent chronic abdominal pain and altered bowel habit in the absence of an organic cause. It has been suggested that the central nervous system may influence the development or expression of IBS, that is IBS can be conceptualised as a ‘brain gut disorder’ (1).

In vivo 2D L-COSY was acquired from the posterior cingulate cortex (PCC), using a 3T MR scanner (Siemens Prisma) equipped with a 64 channel head and neck coil, on three patients diagnosed with IBS (Rome III diagnostic questionnaire) and three healthy controls. Localized shimming was performed prior to data acquisition using automatic adjustment of first and second order shim gradients, followed by manual shimming of the zero-order shim gradient to achieve a peak width of water at half-maximum that was 14Hz or better. Data was acquired from the PCC with a 27cm3 voxel using the following parameters: RF carrier frequency at 2.0 ppm; TR 1.5 s; spectral width of 2000 Hz; increments size of 0.8 ms in 96 t1 increments giving an indirect spectral width of 1250 Hz; 8 averages per increment; and 1024 data points. The WET water-suppression sequence was applied prior to data acquisition. Scan time for the 2D L-COSY was 20 minutes.

Raw COSY spectra were transferred to MATLAB (2) for signal combination followed by row concatenation into a 2D matrix. Felix 2007 was used for the processing and analysis of the 2D COSY data. The data was processed in Felix using the parameters as described in Ramadan et al (3). The creatine methyl resonance (F2=3.02-F1=3.02ppm) was used as the internal standard for chemical shift and for cross and diagonal peak volumes. Cross and diagonal peaks were assigned and peak volumes ratios to creatine measured using Felix NMR 2007. Average peak volumes were calculated for each assigned metabolite and statistical significance was calculated using a t-test in Stata (4).

The results of this pilot study are shown below in Table 1, which compares the percentage difference peak volume/creatine ratio between IBS participants and healthy controls. A typical 2D spectrum from healthy control and a spectrum obtained from an IBS participant are shown in Figure 1. The metabolites of interest are denoted in Figure 1. The differences are large and statistically significant (albeit from a small cohort). The two cases of IBS examined both showed a reduction in fucosylated glycans (Fuc-II and Fuc-III), as assigned previously (5).

For the small IBS cohort, we recorded a reduction in NAA, combined glutamine+glutamate peak (Glx), glycerophosphocholine (GPC), and phenylalanine signals. There is a decrease in GABA that trends towards significance and a statistically significant increase in isoleucine. Many of these findings are in keeping with previous L-COSY and 1D MRS studies of chronic pain (6,7) where a decrease in NAA was associated with chronic pain. Interestingly, we found a decrease in Glx crosspeak, in keeping with the findings of Niddam et al (8), but this is in contrast to other chronic pain conditions where Glx has reported to increase (9). Additionally, the fucoslyated glycans (Fuc II and Fuc III) (5) were both affected by the IBS but not
Fuc I+threonine as reported for chronic pelvic pain (10), nor was any free α-fucose detected. This study is now fully underway with a larger cohort recruited to increase the statistical power. The goal is three fold; to distinguish IBS from the healthy cohort and from other chronic pain conditions to identify those neurochemical pathways that are deregulated as a function of IBS; and to study the effect of therapy.

Our new data in IBS patients is consistent with the concept that neurochemical alterations occur in IBS. However, they are different to other types of chronic pain condition such as (Fibromyalgia, CPPS etc).