2D JPRESS of Human Prostates Using an Endorectal Receiver Coil

Transcription

1 2D JPRESS of Human Prostates Using an Endorectal Receiver Coil Kenneth Yue, Alan Marumoto, Nader Binesh, and M. Albert Thomas* Magnetic Resonance in Medicine 47: (2002) A localized 2D J-resolved (JPRESS) MR spectroscopic sequence was evaluated in human prostates in vivo. Voxels of typically 2 ml were placed in the peripheral zone of the prostate. Eight healthy volunteers, three subjects with benign prostatic hyperplasia, and three patients with prostatic cancer were scanned on a 1.5T MR scanner, using a body coil for RF transmission and a pelvic phased-array coil combined with a disposable endorectal coil for signal reception. The total acquisition time for a 2D JPRESS spectrum was approximately 17 min. A major advantage of the endorectal 2D JPRESS was the ability to resolve the peaks of choline-containing compounds and those of spermine unequivocally. Spectral results clearly showed the biochemical changes in cancer and benign compared to healthy prostates, in conformity with ex vivo biochemical findings. The preliminary results suggest that the endorectal 2D JPRESS could be successfully implemented for the diagnostic examination of human prostates. Magn Reson Med 47: , Wiley-Liss, Inc. Key words: 2D JPRESS; prostate; spermine; endorectal coil The prostate is one of the most diseased human organs in men beyond the age of 60 (1). The current diagnostic routine usually begins with a digital rectal examination and is often followed by a prostate-specific antigen (PSA) blood test. Unfortunately, these tests suffer from the lack of specificity for differentiating prostate cancers (PCa) from benign conditions such as benign prostatic hyperplasia (BPH) or prostatitis. Furthermore, these tests are not particularly sensitive and PCa may only become evident when malignant cells have metastasized beyond the capsular region of the prostate (1,2). More sensitive and specific tests such as ultrasound-guided biopsy are available but are considerably more invasive and costly (2). A decade ago, Thomas et al. (3) used an endorectal coil for both excitation and reception to record proton ( 1 H) MR spectra in healthy human volunteers and PCa patients. Water-suppressed 1 H MR spectra were recorded using a binomial (1331, 2662) spin-echo sequence for a combined water suppression and signal excitation (3). Moreover, the earlier work of Schnall et al. (4) and the subsequent work of others have revolutionized prostate MR imaging and spectroscopy (MRI/MRS) by using an expandable endorectal receiver coil, which can be combined with an external pelvic phased-array (PPA) coil, providing MRI coverage of the prostate and the pelvis (4 11). These methodologies are now routinely used at many institutions. MR spectroscopic data can be threshold-adjusted and overlaid on MR images. Scheidler et al. (8) attempted to assess the efficacy of the combined MRI and 3D 1 HMR spectroscopic imaging (MRSI) in detection and localization of PCa. Using the (choline creatine)/citrate ratio, Kurhanewicz et al. (9,10) assessed whether MRSI in combination with MRI could improve PCa localization in postprostate biopsy cases. Parivar et al. (11) applied prostate MRI/MRSI for follow-up of PCa in patients who underwent cryosurgery. The polyamines putrescine, spermidine, and spermine are polycationic amines that are present in most living organisms and may be important physiological markers (12 16). Previous in vivo 1 H MR prostate spectra were one-dimensional and could not resolve these polyamines due to the overlapping resonance of these metabolites at 1.5T. In particular, the overlaps of creatine, polyamines, and choline-containing compounds severely limited an unambiguous differentiation of polyamines from trimethyl amines, namely, the choline residues. The purpose of our study is to investigate whether 2D JPRESS (17 19) can be applied to the human prostate in vivo using an endorectal probe. Furthermore, we attempt to demonstrate that 2D MRS can resolve the peaks due to choline-containing compounds (Cho), citrate (Cit), creatine (Cr), and polyamines such as spermine (Spm) for the detection of the underlying metabolic changes in PCa and in BPH vs. healthy prostates. MATERIALS AND METHODS Subject Selection A group of eight healthy volunteers, three BPH subjects, and three PCa patients were recruited. For the patient group, the MRS studies were done after fine-needle biopsy. All aspects of the study were explained to the subjects prior to the commencement of the study and informed consent was obtained in keeping with the institutional review board (IRB) guidelines. Department of Radiological Sciences, School of Medicine, University of California, Los Angeles, California. Data Acquisition Grant sponsor: Cancer Research Fund; Grant number: Interagency agreement (University of California, Contract V) with the De- 1.5T Signa Horizon MRI/MRS scanner (GE Medical Sys- A localized 2D JPRESS sequence was implemented on a partment of Health Services, Cancer Research Program. tems, Waukesha, WI) with a gradient strength of 2.2 G/cm Presented at both the 8th ISMRM Meeting, Denver, 2000, and the 86th RSNA Meeting, Chicago, and a rise time of 248 s for spectroscopy. Single voxel 3D *Correspondence to: M. Albert Thomas, Ph.D., Radiological Sciences, UCLA localization was achieved by the 2D analog of PRESS sequence, consisting of three slice-selective RF pulses (90, School of Medicine, Le Conte Avenue, Los Angeles, CA athomas@mednet.ucla.edu 180, 180 ), optimized using the Shinnar-Le Roux (SLR) Received 6 June 2001; revised 6 February 2002; accepted 7 February DOI /mrm algorithm. Water suppression was achieved by a CHESS Published online in Wiley InterScience ( sequence prior to voxel localization and 1024 t 2 complex 2002 Wiley-Liss, Inc. 1059

2 1060 Yue et al. FIG. 1. a: 2D JPRESS spectrum of a composite prostate phantom containing 25 mm citrate, 10 mm spermine, 5 mm creatine, and 2 mm free choline at ph 7.2 (27 ml, TR 2 sec, TE 30 ms). b: Simulated 2D JPRESS spectrum of citrate using J 15.6 Hz and 9.6 Hz at 1.5T (TR 2s,TE 30 ms, 256 F 1 points, and 4096 F 2 points). [Color figure can be viewed in the online issue, which is available at points were acquired for each FID. The additional J-resolved spectral dimension was obtained by inserting two symmetric t 1 increments of 5 ms each first between the two refocusing 180 RF pulses and, second, immediately before the acquisition filter. Two-step phase cycling schemes were used on all three pulses, resulting in a total of eight cycling steps in conjunction with signal averaging. Further details of the 2D JPRESS sequence can be found elsewhere (17 19). For the phantom study, a head coil (GE Medical Systems, Milwaukee, WI) was used for both RF transmission and reception and the total acquisition time was 34 min (TR 2 sec, TE 30 ms, 64 t 1 points, and 16 NEX). For the in vivo studies, the RF pulses were transmitted through a body coil and the signal was received by an endorectal surface coil configured with a PPA coil (MEDRAD, Pittsburgh, PA). The placement of the endorectal probe was done by an experienced radiologist (AM). The bulb was inflated with c.c. of air so that the plane of the surface coil, indicated by a marker along the handle, was in the closest proximity to the prostate tissue. MR spectroscopic voxel was prescribed on an axial fast spin-echo MRI optimized for prostate tissue contrast: 4-mm slice, TR 2.5 sec, TE 84 ms, FOV cm, acquisition matrix , and 4 NEX, resulting in an acquisition time of 4 min. The spectroscopic acquisition time per voxel location was min (TR 2 sec, TE 30 ms, t 1 points, and 8 NEX). Combined with an approximately 7-min voxel prescription, shimming, and prescan optimization, the total measurement time was within 30 min for one spectroscopic location and less than an hour for two locations. When the full width at half maximum (FWHM) of water peak was greater than 10 Hz, the voxel shimming procedure was repeated. Typically, a 7-Hz FWHM of unsuppressed water signal was observed prior to 2D JPRESS acquisition. Data Processing The 2D MRS raw files were transferred to an SGI O2 workstation (Silicon Graphics, Sunnyvale, CA) for offline data processing using Felix98 (Molecular Simulations, San Diego, CA). The raw matrix was apodized with phaseshifted sinebell squared functions along t 1 and t 2 and zero-filled to prior to Fast Fourier Transformation. All 2D spectra were presented as contour plots and the 2D spectral matrices were not skewed by 45 about J 0 Hz (18,20). Hence, the 2D J-resolved peaks due to weakly coupled protons were expected to be aligned on straight lines 45 with respect to the F 2 axis and more complex patterns were expected for strongly coupled spin systems (20). RESULTS Figure 1a shows a 2D JPRESS spectrum of a composite prostate phantom containing free Cho, Cit, Cr, and Spm at ph 7.2. A 27-ml voxel was localized. Shown in Fig. 1b is a simulated hard pulse version of 2D JPRESS (90, 180, 180 ) spectrum of Cit using the GAMMA simulation li-

3 2D JPRESS of Prostate Using Endorectal Coil 1061 FIG. 2. 2D JPRESS spectrum (a) of a 28-year-old healthy subject and (b) its corresponding 2 ml voxel placement. [Color figure can be viewed in the online issue, which is available at brary (21). Cit has two equivalent methylene groups. Each proton pair forms a strongly J-coupled AB spin system (20,22), resulting in eight J-resolved 2D peaks antisymmetric about F ppm, as evident in the experimental and simulated 2D spectra. The 2D peaks located along F1 1.6 Hz, 7.8 Hz, and 16.4 Hz were in agreement with a previous report (20). In addition, the projected 1D spectra onto F1 and F2 axes are also shown in Fig. 1b. Regarding other metabolites, the region between 3 4 ppm had several overlapping peaks along F1 0 Hz: Cr (3.0, 3.9 ppm), Cho tri-methyl singlet (3.2 ppm), Cho methylene multiplets (3.5, 4.0 ppm), and polyamines such as Spm (multiplets at 2.1 and 3.1 ppm). The J-resolved peaks of Cho at F2 3.5 ppm and 4.0 ppm were not readily observable here at 2 mm, although their presence has been confirmed at a higher concentration of 10 mm (18). On the other hand, the J-resolved peaks of Spm methylene protons were well isolated, as shown around the regions of F2 2.1 ppm and 3.1 ppm. Since the methylene protons of Spm follow weak J-coupling (Chemical shift difference Ⰷ J), the triplet centered at F2 3.1 ppm (F1 0 Hz and 7.3 Hz) was clearly visible. 2D J-resolved peaks due to Cr were not resolvable in Fig. 1a due to a limited spectral dispersion along F1. Figure 2 shows a 2D JPRESS spectrum of a 28-year-old healthy prostate using a 2-ml voxel. The presence of strong Cit peaks was consistent with the well-known fact of its high abundance in healthy prostate (3,10). Averaged over the eight healthy controls, the 2D peaks due to Cit were located along F1 1.6 Hz, 7.9 Hz, and 17.5 Hz. In addition, the triplet nicely resolved about F2 3.1 ppm along F1 0 Hz and 7.8 Hz was identified as Spm. Also, there were 2D peaks along F1 0 Hz due to Cr and Cho at F2 3.0 ppm and 3.2 ppm, respectively. Moreover, the J-resolved peaks due to Cho methylene protons were unobservable due to the small voxel size and reduced number of averages. Figure 3 shows a 2D JPRESS spectrum from a 2.25-ml voxel of a 60-year-old subject with BPH. More intense 2D cross peaks were indicative of an increase of Cit in BPH, as confirmed by previous 1D MRS reports (10). The locations of Cit peaks were comparable to that of healthy prostate. Although no considerable change of Cho was observed, a slight elevation of Spm was indicated. Figure 4 shows a 2D JPRESS spectrum from a 2-ml voxel of a 79-year-old patient with PCa. The three 2D peaks due to the methylene protons of Spm centered at F2 3.1 ppm were reduced remarkably in the two patients, along with a moderate increase of Cho. The 2D peak changes of Spm were minimal in the third PCa patient with a decrease of Cho. Compared to the spectra of healthy controls and BPH, a strong depletion of Cit was also evident in the entire group of PCa patients. Selected 2D peaks were quantified in the following regions (F1, F2): Cit ( 7.9 Hz, ppm), Spm (7.8 Hz, 3.05 ppm), and Cho Spm Cr (0 Hz, ppm). Their ratios (mean SD) in healthy controls, BPH, and PCa were: 1) Cit/Cho Spm Cr (n 8), (n 3), and (n 3); 2) Spm/ Cho Spm Cr (n 8), (n 3), and (n 3), respectively.

4 1062 Yue et al. FIG. 3. 2D JPRESS spectrum (a) of a 60-year-old subject with BPH and (b) its corresponding 2.25 ml voxel placement. [Color figure can be viewed in the online issue, which is available at FIG. 4. 2D JPRESS spectrum (a) of a 79-year-old patient with PCa and (b) its corresponding 2 ml voxel placement. [Color figure can be viewed in the online issue, which is available at

5 2D JPRESS of Prostate Using Endorectal Coil 1063 DISCUSSION A major problem with the 1D MRS or MRSI prostate data is the significant overlap of Cr, Spm, and Cho in the spectral range of ppm. 2D JPRESS clearly resolved the cross peaks due to Spm, separated by 8 Hz about F 1 0 ppm. Strong coupling effects in the 2D JPRESS spectra of human brain were reported earlier (18). The endorectal 2D JPRESS spectra of prostate showed similar effects for Cit and Cho. The T 2 value of Cho is ms (6), while that of polyamines is still unknown. Therefore, a short TE of 30 ms was chosen to optimize the initial signals from every metabolite, including the signals from the short T 2 metabolites. There is a definite need to quantify polyamines in vivo, since it has been reported that high concentration of Spm (10 20 mm) may be responsible for the slow growth of cancer cells in prostatic tissue. However, previous biochemical analyses of polyamines in PCa resulted in conflicting findings (12 15). Urinary excretion of spermidine was significantly elevated in prostatic carcinoma as compared to a control group of patients (13). Although there was no difference between the Spm levels of BPH and controls, Spm of PCa was elevated in expressed prostatic secretion (14). Plasma Spm levels were only occasionally elevated in PCa compared to normal prostate (15). In this study, the 2D cross peaks due to the methylene protons of Spm centered at F ppm were reduced remarkably in PCa patients (Fig. 4a), but slightly elevated in BPH patients (Fig. 3a). Our results were in good agreement with a recent ex vivo MRS and high performance liquid chromatography (HPLC) analysis of prostatic tissue (12). Although the polyamines include putrescine, spermine, and spermidine, the 2D JPRESS cross peaks of these different polyamines were not resolved. However, the MRS peaks due to different polyamines could be detected unequivocally using heteronuclear single quantum coherence spectroscopy (16). The 2D contour plots used in this study may not permit accurate measurements of peak positions and multiplet separations. Depending on the postprocessing parameters, the centroid of a 2D peak can easily shift along the F 1 axis. In strongly coupled spin systems (20), this dependency becomes even more pronounced, which accounts for the discrepancies between the in vitro and the in vivo F 1 results of Cit and Spm, as shown above. As an alternative, 2D cross sectional slices were used in the oversampled 2D J-resolved spectra of human prostate without presaturation of water (23). A drawback is that the sensitivity of 2D cross-sections is often affected by t 1 -noise or ridge (24). This causes noise bands at the 2D peak locations running parallel to the F 1 axis. The t 1 -noise results from either random fluctuation of the scanner or due to subject s motion. In our study, the strong t 1 -ridges recorded in two healthy controls were possibly due to the subject s movement. The single-voxel 2D JPRESS suffers from some drawbacks. The voxel size and its location must be selected at the time of acquisition and the time constraints limit the number of voxels to one or two per session. In this study, we selected one single voxel located predominantly in the peripheral zone of the prostate, where cancer lesions are commonly found. Another major drawback of a singlevoxel-based technique is that one may even miss the target lesion in PCa. Partial volume effect will be more severe while the voxel size is large ( 1 ml). Hence, it is worthwhile to investigate a 2D JPRESS analog of MRSI while keeping a reasonable acquisition time in prostate studies (25). In comparison with the recently proposed 2D L-COSY (26), 2D JPRESS has a narrower spectral window along F 1. Hence, the number of t 1 increments ( t 1 ) can be reduced in 2D JPRESS, resulting in a shorter total acquisition time. In addition, 2D JPRESS has superior sensitivity compared to 2D L-COSY, allowing a voxel size of as small as 1 2 ml. However, the strong coupling effect leads to more complex 2D J-resolved peaks at 1.5T than the weakly coupled protons, resulting in a difficult task of quantitation. In 2D L-COSY, there are no additional cross peaks due to the strong coupling; however, the cross peak intensities are weighted accordingly (20,24). Our concurrent effort focuses on the evaluation of endorectal 2D L-COSY in human prostates, where an improved spectral dispersion is expected to improve the quantitation of prostate metabolites. In conclusion, our observations suggest that spatially localized endorectal 2D JPRESS spectra can be successfully recorded using a clinical 1.5T MR scanner. The pilot results show that 2D JPRESS peaks due to Cit and Spm can be resolved and the changes in these metabolite levels can be successfully quantified in BPH and PCa patients compared to healthy controls. The results are consistent with the previous 1D MRS and biochemical findings. In particular, our results with a limited number of patients demonstrate that the Spm peaks are reduced in PCa and slightly increased in BPH, in agreement with a recent study using prostatic tissue (12). ACKNOWLEDGMENTS The authors thank Dr. Zoran Barbaric and Dr. Shantanu Sinha for scientific assistance and Dr. Dirk Mayer and Dr. Wolfgang Dreher for sample GAMMA programs. The authors also thank Mrs. and Mr. Raman for assistance in recording the earlier 2D JPRESS spectra. REFERENCES 1. Wilt TJ. Prostate cancer screening: practice what the evidence preaches. Am J Med 1998;104: Yarbro CH, Ferrans CE. Quality of life of patients with prostate cancer treated with surgery or radiation therapy. Oncol Nurs Forum 1998;25: Thomas MA, Narayan P, Kurhanewicz J, Jajodia P, Weiner MW. 1 HMR spectroscopy of normal and malignant human prostates in vivo. J Magn Reson 1990;87: Schnall MD, Lenkinski RE, Pollack HM, Imai Y, Kressel HY. Prostate: MR imaging with an endorectal surface coil. Radiology 1989;172: Heerschap A, Jager GJ, van der Graaf M, Barentsz JO, de la Rosette JJ, Oosterhof GO, Ruijter ET, Ruijs SH. In vivo proton MR spectroscopy reveals altered metabolite content in malignant prostate tissue. Anticancer Res 1997;17:

6 1064 Yue et al. 6. Heerschap A, Jager GJ, van der Graaf M, Barentsz JO, Ruijs SH. Proton MR spectroscopy of the normal human prostate with an endorectal coil and a double spin-echo pulse sequence. Magn Reson Med 1997;37: Lowry M, Liney GP, Turnbull LW, Manton DJ, Blackband SJ, Horsman A. Quantification of citrate concentration in the prostate by proton magnetic resonance spectroscopy: zonal and age-related differences. Magn Reson Med 1996;36: Scheidler J, Hricak H, Vigneron DB, Yu KK, Sokolov DL, Huang LR, Zaloudek CJ, Nelson SJ, Carroll PR, Kurhanewicz J. Prostate cancer: localization with three-dimensional proton MR spectroscopic imagingclinicopathologic study. Radiology 1999;213: Kurhanewicz J, Vigneron DB, Hricak H, Parivar F, Nelson SJ, Shinohara K, Carroll PR. Prostate cancer: metabolic response to cryosurgery as detected with 3D H-1 MR spectroscopic imaging. Radiology 1996;200: Kurhanewicz J, Vigneron DB, Nelson SJ, Hricak H, MacDonald JM, Konety B, Narayan P. Citrate as an in vivo marker to discriminate prostate cancer from benign prostatic hyperplasia and normal prostate peripheral zone: detection via localized proton spectroscopy. Urology 1995;45: Parivar F, Hricak H, Shinohara K, Kurhanewicz J, Vigneron DB, Nelson SJ, Carroll PR. Detection of locally recurrent prostate cancer after cryosurgery: evaluation by endorectal ultrasound, magnetic resonance imaging, and three-dimensional proton magnetic resonance spectroscopy. Urology 1996;48: van der Graaf M, Schipper RG, Oosterhof GO, Schalken JA, Verhofstad AA, Heerschap A. Proton MR spectroscopy of prostatic tissue focused on the detection of spermine, a possible biomarker of malignant behavior in prostate cancer. Magma 2000;10: Fair WR, Wehner N, Brorsson U. Urinary polyamine levels in the diagnosis of carcinoma of the prostate. J Urol 1975;114: Anderson RU, Fair WR. Physical and chemical determinations of prostatic secretion in benign hyperplasia, prostatitis, and adenocarcinoma. Invest Urol 1976;14: Chaisiri P, Harper ME, Griffiths K. Plasma spermine concentrations of patients with benign and malignant tumours of the breast or prostate. Clin Chim Acta 1979;92: Willker W, Flogel U, Leibfritz D. A 1 H/ 13 C inverse 2D method for the analysis of the polyamines putrescine, spermidine and spermine in cell extracts and biofluids. NMR Biomed 1998;11: Ryner LN, Sorenson JA, Thomas MA. 3D localized 2D NMR spectroscopy on an MRI scanner. J Magn Reson 1995;107B: Ryner LN, Sorenson JA, Thomas MA. Localized 2D J-resolved 1 HMR spectroscopy: strong coupling effects in vitro and in vivo [erratum published in Magn Reson Imag 1995;13:1043]. Magn Reson Imaging 1995;13: Thomas MA, Ryner LN, Mehta MP, Turski PA, Sorenson JA. Localized 2D J-resolved 1 H MR spectroscopy of human brain tumors in vivo. J Magn Reson Imag 1996;6: Kumar A. Two-dimensional spin-echo NMR spectroscopy: a general method for calculation of spectra. J Magn Reson 1978;30: Smith SA, Levante TO, Meier BH, Ernst RR. Computer simulations in magnetic resonance. An object-oriented programming approach. J Magn Reson 1994;106A: Wilman AH, Allen PS. The response of the strongly coupled AB system of citrate to typical 1 H MRS localization sequences. J Magn Reson 1995;107: Swanson MG, Vigneron DB, Tran TKC, Sailasuta N, Hurd RE, Kurhanewicz J. Single-voxel oversampled J-resolved spectroscopy of in vivo human prostate tissue. Magn Reson Med 2001;45: Ernst RR, Bodenhausen G, Wokaun A. Principles of nuclear magnetic resonance in one and two dimensions. Oxford: Clarendon Press; New York: Oxford University Press; Mayer D, Dreher W, Leibfritz D. Fast 1 H spectroscopic imaging combined with 2D correlation spectroscopy uncoupled in both frequency domains. In: Proc 8th Annual Meeting ISMRM, Denver, p Thomas MA, Yue K, Binesh N, Davanzo P, Kumar A, Siegel B, Frye M, Curran J, Lufkin R, Martin P, Guze B. Localized two-dimensional shift correlated MR spectroscopy of human brain. Magn Reson Med 2001; 46:58 67.