Inflammation is an overarching theme of many urological diseases including bladder-centric phenotype of Interstitial cystitis/painful bladder pain syndrome (IC/BPS) but bladder biopsy is the only reliable tool available to index bladder inflammation. However, biopsy is limited by invasiveness, complications and the site-selection bias can variability ranging from 25-75% in the detection of inflammation . The detection of inflammation can not only aid in the phenotyping of IC/BPS patients but also a specific evidence for the absence of inflammation can save patients from unnecessary exposure to toxicity of cyclosporine. On the contrary, a reliable surrogate of inflammation can not only predict the treatment success but could also provide an objective inclusion criterion for enrolling patients into clinical trial of new anti-inflammatory drugs.
One such surrogate is potassium sensitivity test (PST) before it fell out of favor because of pain and bladder spasm evoked in 75% of IC/BPS patients with the instillation of 0.3-0.4M KCl. Positive PST was linked to urothelial denudation of IC/BPS but drugs (e.g. oxybutynin) larger than K+ ion are also absorbed by healthy subjects.  Consequently, on the basis of histology of mucosal vasculature and laser Doppler blood flow, several groups (ref.1,2) have incriminated sluggish mucosal blood flow in positive PST of IC/BPS patients whereas the acceleration of mucosal blood flow dilutes diffused K+ for negative PST in healthy adults.  Though widely assumed to be integrated, perfusion and permeability are yet to be assessed simultaneously without using radiolabelled probe. Here, we examined the potential of intravesical contrast enhanced magnetic resonance imaging (ICE-MRI) to simultaneously measure mucosal perfusion and permeability of Gadobutrol as a minimally invasive surrogate for bladder inflammation.

We analyzed ICE-MRI data of three male subjects (age range 60-75 years) enrolled in an ongoing NCI funded clinical trial for suspicion of bladder cancer. After signing informed consent, ICE-MRI was performed with Siemens Biograph 3T before and after transurethral bladder instillation of 50mL contrast mixture (Gadobutrol and Ferumoxytol) with an uncoated 14Fr urethral straight tip catheter.  We acquired pre-and post-instillation  T1 weighted  Fast‐low‐angle‐shot‐3D volume interpolated breath hold examination (VIBE)and 1600 free breathing time-resolved angiography with interleaved stochastic trajectories (TWIST) scans over 450s in voxel volume of  0.5*0.5*1 mm3, repetition time 7.6ms, echo time 3.4ms, field of view 199*199 mm2, acquisition matrix size 192*154. Scans were  post processed by inhouse  MATLAB scripts to assess dynamic T1 signal intensity changes in the regions of interest (ROI) selected for transurethral resection and normal adjacent region.

The histopathology of the transected lesion confirmed the absence of cancer together with the existence of inflammation secondary to cystitis cystica, epithelial denudation or granuloma (Fig.1). The differences in the shape and brightness of normal and inflamed ROI on T1 VIBE and the corresponding dynamic changes in signal intensity over 450s in TWIST are linked to the inflammation driven dilatation of urothelial tight junctions facilitating entry of Gadobutrol into the expansive extracellular space. Histogram of inflamed pixels intensity revealed higher-than-normal distribution (lower kurtosis) as opposed to lower-than- normal distribution (higher kurtosis, fat tails) for healthy ROI.  
Owing to its diamagnetic nature, oxyhemoglobin lacks any signal and therefore pulsatile arrival of oxyhemoglobin in inflamed and healthy ROI causes the recurring dips in pixel intensity. The signal dips are sharper for normal ROI because while venous congestion slows the arrival of oxyhemoglobin to inflamed ROI, dilated urothelial tight junctions amplifies the paracellular entry of Gadobutrol into enlarged extracellular space of inflamed ROI.  
Since diffused gadobutrol from mucosa only enhances the signal of venules and veins but not of arteries, TWIST scans can visualize the dynamic contrast enhancement of veins from normal and inflamed areas of bladder wall for a radiation-free angiography.  As a result, we could quantitatively determine the differences in perfusion and permeability of normal and inflamed ROI after stitching 20 frames of the same slice, with each frame acquired 22s apart over 450s interval. Distinct contrast enhancement at the inflamed ROI in T1 weighted VIBE scan indexes Gadobutrol accumulation from faster diffusion but slower venous clearance whereas the absence of contrast enhancement at normal ROI reflects a rapid equilibrium between the rate of urothelial diffusion and its rate of venous clearance, which is corrobrated by the rapid temporal fluctuations in signal intensity of TWIST scans.  Moreover, mucosal blood flow in human bladder also prevents any contrast enhancement of the detrusor muscle which is noted with the ICE-MRI of excised pig bladder ex vivo with impaired mucosal blood flow( ref.3).

While umbrella cells covering 70% of the luminal surface of bladder wall are impermeable to instilled K+ in PST, drugs and dyes: Gadobutrol, Fluorescein and Evans blue dye, which can however diffuse through apico-lateral tight junctions of umbrella cells, covering 30% of the luminal surface. Since ICE-MRI relies on paracellular diffusion of smaller sized, paramagnetic, Gadobutrol (604.71 Daltons)  and luminal retention of large sized, superparamagnetic, Ferumoxytol (731 KiloDaltons), ICE-MRI provides clinical validation Stokesian diffusion into bladder mucosa of live humans with the inverse dependence of intravesical absorption on molecular weight.  In additon, ICE-MRI visualizes the inflammation dependent dilatation of tight junctions accelerating the Gadobutrol diffusion as predicted by the augmented absorption of instilled lidocaine and thiotepa in IC/BPS patients. In addition, ICE-MRI also visualizes any potential supplementation of paracellular diffusion by intravascular entry of Gadobutrol into Hunner lesion for phenotyping IC/BPS patients.  Dynamic contrast enhancement of mucosa and different branches of superior and inferior vesical veins in ICE-MRI visualizes the inflammation mediated dilatation of tight junctions between umbrella cells of urothelium and capillaries as the dilatation of former increases the paracellular diffusion of instilled Gadobutrol while latter causes venous congestion to slow the clearance of diffused Gadobutrol. The resilient vascular architecture erected by sub urothelial capillary plexus overlaid in a palisade pattern over mucosal plexus enables the acceleration of mucosal blood flow for the dilution of instilled irritants (K+) in PST and the mucosal blood flow entraps instilled drugs and probes (Gadobutrol) diffusing into mucosa for dilution and preclude their direct entry into detrusor muscle of living mammal, evident from the absence of Gadobutrol mediated contrast enhancment of detrusor muscle noted ex vivo (ref.3). 
The susceptibility induced signal loss in lumen from luminal retention of Ferumoxytol improves the image contrast of mucosa by  mirroring the signal loss from oxyhemoglobin (diamagnetic) which darkens arteries and arterioles and throws brightened veins draining diffused Gadobutrol into sharp relief. While TWIST measures both arterial and venous phase of contrast enhancement in an ROI after intravenous injection of Gadobutrol regularly in clinic, ICE-MRI after instillation of Gadobutrol exclusively measures the dynamic signal changes in venous phase as the diffused Gadobutrol only enters veins and not arteries.

Thus, ICE-MRI is not only capable of non-invasively probing the inflammation evoked urothelial hyperpermeability but the delay in the washout of diffused Gadobutrol due to venous stasis can be leveraged for radiation-free angiography of bladder wall. The gradient of signal intensity in ICE-MRI recapitulates the exponential decline of fluorescence signal from the decline in dye and drug concentration across bladder wall generated by the washout from mucosa by mucosal blood flow. Findings warrant future investigations to determine the potential of ICE-MRI as a minimally invasive, alternative to PST to confirm evidence of inflammation without biopsy and guarantee treatment success with cyclosporine.

Pontari MA, Hanno PM and Ruggieri MR. Comparison of bladder blood flow in patients with and without interstitial cystitis. J Urol 1999; 162: 330-334.
Rosamilia A, Cann L, Dwyer P, Scurry J and Rogers P. Bladder microvasculature in women with interstitial cystitis. J Urol 1999; 161: 1865-1870.
Tyagi P, Moon C, Singh N, Connell M, Maranchie J, Chermansky C, Yoshimura N and Kaufman J. High Resolution 3D T1-Mapping of Pig Bladder Wall by Intravesical Contrast Enhanced MRI At 3T. J Urol 2021; 206: e390.

Continence 7S1 (2023) 100856
DOI: 10.1016/j.cont.2023.100856