Mehrdad Pourfathi1,2, Yi Xin1, Stephen J Kadlecek1, Maurizio Cereda3, Harrilla Profka1, Sarmad M Siddiqui1,4, Hooman Hamedani1,4, and Rahim R Rizi1
1Radiology, University of Pennsylvania, Philadelphia, PA, United States, 2Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, PA, United States, 3Anesthesiology and Critical Care, University of Pennsylvania, Perlman School of Medicine, Philadelphia, PA, United States, 4Bioengineering, University of Pennsylvania, Philadelphia, PA, United States
Synopsis
Ventilator
induced lung injury (VILI) results from mechanical trauma and secondary
inflammatory responses. Previous
imaging studies focused on the effects of ventilator settings on structural or
functional changes; however, no MRI studies have been conducted to elucidate
the metabolic changes during VILI. In this study, the potential of HP [1-13C]-pyruvate
and its conversion to [1-13C]-lactate as a marker for the lung inflammation
was investigated in a two-hit model of acid aspiration and VILI.Introduction
Ventilator
induced lung injury (VILI) results from mechanical trauma and secondary
inflammatory responses. Lung protective ventilation attenuates cytokine
concentrations and inflammatory infiltrates in the lungs [1, 2]. Previous
imaging studies have focused on the effects of ventilator settings on
structural or functional changes; however, no MRI studies have been conducted
to elucidate the metabolic changes that occur during VILI. In this study, the conversion
of HP [1-
13C]-pyruvate and to [1-
13C]-lactate was
investigated as an
in-vivo marker of lung inflammation was investigated in a
two-hit model of acid aspiration and VILI.
Materials and Methods
Five
Sprague-Dawley rats (402 ± 85g) were ventilated with tidal volume (TV) of 6mL/kg
and positive end-expiratory pressure (PEEP) of 10cmH
2O (protective
ventilation) for t = 60min, until tracheal hydrochloric acid (pH 1.5, 2.0
mL/kg) instillation at t = 70min. The ventilation scheme was maintained until t
= 150min. Then n = 2 rats were ventilated for 90 minutes with TV = 12mL/kg and
PEEP=3cmH
2O to induce VILI; the remaining 3 rats continued with protective
ventilation until the end of the experiment. A summary of the experimental
setup is shown in
Figure 3. Rats were imaged in
supine position using a
1H/
13C Quadrature birdcage coil
(m2m) at 4.7T (Varian Inc.). Proton images were acquired using a multi-slice gradient
echo (TR/TE = 120/4ms, α=20°) images (128x128 voxels). HP [1-
13C]-pyruvate
imaging was conducted at t = 60min (baseline), t = 150min (post-HCl), and t =
240min (post-second-hit). ~22 µL of [1-
13C]-pyruvate (Cambridge Isotopes
Inc.) was polarized using HyperSense DNP polarizer (Oxford Instruments) to over
15%. The sample was melted using 4mL of a dissolution buffer (40mM Trizma,
160mM NaOH, 50mM NaCl, 0.1 g/L EDTA) at 180°C to yield an isotonic solution of
80mM [1-
13C]-pyruvate with neutral pH at 37°C. 10 seconds after the
dissolution 3mL of the hyperpolarized agent was administered via the tail-vein
over 12 seconds. Three single-slice axial carbon-13 chemical shift images were acquired
using a 16x16 single-point CSI sequence with (TR/TE=110/0.35ms, α=12°), 25
seconds after the start of the injection (in-plane FOV = 50x50mm
2, 25mm slice thickness). This
yielded an in-plane resolution of 3.1x3.1mm
2. A 1.5M [1-
13C]-sodium propionate (Cambridge
Isotopes) sample (185ppm) was placed inside the coil to ensure proper overlay
of the proton and carbon images. Three 10s breath-holds were applied to minimize the
respiratory motion artifact during carbon imaging (28s). Blood oxygen
level was monitored using a pulse oximeter during the scan to ensure sufficient
oxygenation (SPO
2 > 90%). A custom outward spiral k-space phase-encoding order was used to acquire the center of the k-space every eleventh acquisition.
Multiple FIDs acquired at k
xy = 0 were used to partially compensate
the signal loss due to T
1 decay and mitigate the associated blurring
artifact to improve localization of the spectra. Metabolites
were quantified by integrating the peaks in the real spectra after zero and first
order phasing and 4
th-order polynomial baseline correction. Spectroscopic
data were processed using custom routines in MATLAB2014b (MathWorks Inc.).
Processed carbon images were converted to DICOM format and overlaid on the
proton image using OsiriX 7.0 (
Figure 1).
Results and Discussion
Figure 1 shows images of
the metabolites in a rat post HCl injury. The intensity of each image is
normalized to its own maximum intensity. The images suggest global injury in
the lungs. Although the signal intensity in the heart is dominant (and in the
muscle for alanine), the combination of breath-hold and de-blurring using the
custom k-space trajectory allows for proper localization of metabolites to the
lungs (
Figure 2. Lactate-to-pyruvate
ratio was measured from the averaged spectra in the voxels over the lungs
(highlighted voxels in
Figure 2). The average lactate-to-pyruvate ratio as a function of time is shown in
Figure 3. Our measurements show an increase of 20.8 ± 12.1% in the
lactate-to-pyruvate ratio after HCl instillation, compared to the baseline. The
observed effect size for this measurement was 1.79 (Cohen’s d). This initial
effect was likely due to the presence of inflammatory infiltrate [3]. Furthermore, our
results indicate a further increase in the lactate-to-pyruvate ratio after
exacerbation of the injury due to injurious ventilation leading to VILI (43.0 ± 21.1% greater than HCl cohort). This increase was
significantly smaller in rats that underwent protective ventilation (14.1 ± 10.8% greater than HCl
instillation).
Conclusions
This study demonstrates that HP [1-
13C]-pyruvate
imaging can be used as a marker for metabolic changes due to lung injury in
this two-hit model of acid aspiration and VILI,
providing critical
information for detection and management of
acute lung injury that is otherwise difficult to acquire from other conventional modalities.
Acknowledgements
No acknowledgement found.References
[1] Pinhu, et al.
Lancet. 361:332–40 (2003); [2] Tremblay et al. J. Clinical Investigations.
99(5):944-952 (1997); [3] Shaghaghi et. al., NMR in biomedicine, 27(8):939-947 (2014).