Dual hyperpolarized [1-13C]pyruvate and [13C]urea magnetic resonance imaging of prostate cancer

Abstract

Background

Although multiparametric (mp) ¹H magnetic resonance imaging (MRI) is increasingly used to detect and localize prostate cancer (PC), its correlation with tumor grade is limited. Hyperpolarized (HP) carbon-13 (¹³C) MR is an emerging imaging technique, which can be used to interrogate key biologic processes through in vivo detection of various HP probes. A distinct attribute of HP ¹³C MRI is the ability to detect multiple HP probes within a single acquisition. Here we report on the first simultaneous dual HP [1-¹³C]pyruvate and [¹³C]urea MRI with correlations to histopathologic findings in a patient with localized PC scheduled for radical prostatectomy.

Material and methods

Paired HP ¹³C and standard mp ¹H MRI were performed in a patient with biopsy-proven Gleason score 4 + 3 = 7 adenocarcinoma of the prostate scheduled for radical prostatectomy through a first-in-human pilot study of dual-agent HP MRI (NCT02526368). HP ¹³C MRI was performed using a clinical 3T scanner with ¹³C transmit-and-receive capabilities. Dynamic series of HP ¹³C pyruvate, lactate and urea imaging were acquired following intravenous (IV) injection of co-hyperpolarized [¹³C]urea (25 mM) and [1–¹³C]pyruvate (125 mM). The [1-¹³C]pyruvate-to-[1-¹³C]lactate conversion rate (k_PL) was calculated using an inputless two-site exchange model; AUC_urea was the [¹³C]urea signal summed over time. Following radical prostatectomy, whole-mount prostate histopathological slides were prepared and reviewed by an experienced genitourinary pathologist.

Results

Following informed consent, the patient underwent paired mp ¹H MRI and dual-agent HP MRI. mp ¹H MRI revealed a 1.2 cm lesion in the left apical posterior zone. Dual-agent HP MRI identified a focus of increased [1-¹³C]pyruvate-to-[1-¹³C]lactate conversion rate (k_PL) extending from the left apical posterior peripheral zone to the right gland. A corresponding area of abnormal tissue perfusion (AUC_urea) was seen in the left gland. Metabolism-perfusion mismatch (with several foci of increased k_PL/AUC_urea) was observed throughout the tumor. Tumor extension to the right midgland was confirmed at the time of radical prostatectomy and staining for lactate dehydrogenase-A was increased throughout the tumor relative to surrounding benign prostatic tissue.

Conclusion

This first-in-human radiopathologic study demonstrates the feasibility of dual-agent HP MRI in PC patients. Simultaneous assessment of tumor metabolism and perfusion was able to detect occult disease as well as to show a significant mismatch between intra-tumoral metabolism and tissue perfusion in high-grade PC. Prospective validation of these findings is warranted.