| Part Ⅰ Blood oxygenation level-dependent MR and 1H MR spectroscopy in human prostate cancer model:a preliminary studyPurpose:To establish a human prostate cancer model in nude rat, investigate the feasibility of BOLD MR and the characteristics of proton MR spectroscopy in the xenograft tumor model, explore possible connections between the two functional imaging methods.Materials and Methods:Human prostate cancer cell PC-3 were implanted subcutaneously in 24 male nude rats(4-5 weeks old). BOLD MR and MR spectroscopy were performed on a 3.0 Tesla MR scanner (Signa Excite HDX, GE) when tumor diameter reached l-3cm. A knee coil was used for BOLD scan, data were collected respectively for tumor-bearing rats inhaling air and 5 minutes after inhaling carbogen. All images were post-processed on GE ADW 4.3 workstation to measure tumor as well as contralateral muscle tissue baseline R2*(air), and calculate dynamic AR2 (R2 value changed before and after carbogen inhaling). A 3 inch animal coil was used for MR spectroscopy, applying single voxel proton MR spectroscopy in regions of interest, then measured signal-noise ratio of tCho peak by SAGE. Two R2* values were compared with paired t test, Pearson correlation between tCho SNR and R2* were assessed, P<0.05 was considered as statistically significant.Results:4-5 weeks after implanting,20 rats formed tumor with an average diameter 21.6mm. One rat died accidently during the imaging process, and 2 moved significantly, thus 17 rats completed BOLD scanning. R2* decreased after Carbogen inhaling, from(21.97±6.25)/s to(20.46±5.54)/s. R2* for muscle tissue was (38.94± 4.21)/s, which differed from tumor.15 qualified spectrum lines with low background noise and acceptable distortion were obtained, and tCho SNR was 7.98+2.33. No correlation was observed between baseline R2* and tCho.Conclusion:Hypoxia related parameters R2 and AR2* derived from BOLD MR can indirectly reflect tumor oxygenation status. BOLD parameters and tCho lacked distinct direct correlation, suggesting that choline metabolism was a complicated prosess except from hypoxia.Part Ⅱ Blood oxygenation level-dependent MR in human prostate cancer model comparing with hypoxia marker and micro PET/CT hypoxia imagingPurpose:To investigate the relationship between BOLD parameters in a human prostate cancer model and tumor hypoxia markers HIF-1α、CA IX and comparing to micro PET/CT hypoxia imaging.Materials and Methods:20 tumor-bearing rats received BOLD examination on a 3.0 Tesla MR scanner (Signa Excite HDX, GE). Data were collected respectively for rats inhaling air and 5 minutes after inhaling carbogen. All images were post-processed on GE ADW 4.3 workstation to measure tumor baseline R2 (air), and calculate dynamic △R2*(R2* value changed before and after carbogen inhaling). Micro PET/CT imaging were conducted within 1～2 days after MR scan. Each rat was injected about 2 mCi 18F-FMISO by caudal vein, static images were acquired 2 hours after injection in a three-dimensional pattern. Distributions of 18F-FMISO in tumor-bearing rats were obtained by reconstruction on Inveon Workstation through attenuation correction. Measuring SUVmax of tumor and contralateral muscle tissue, calculated tumor hypoxia volume. Rats were decapitated when completing all imaging examination, tumors were separated and fixed immediately. Hematoxylin-eosin staining and immunohistochemistry of HIF-1α、CA IX were performed on pathological specimen. Correlation between baseline R2* and HIF-1α、CA IX、SUVmax、HV were assessed, P<0.05 was considered as statistically significant.Results:One rat died accidently during the imaging process, and 2 moved significantly, thus 17 rats completed BOLD scanning, R2* decreased after carbogen inhaling, from(21.97±6.25)/s to(20.46±5.54)/s. Varying degrees of radioactivity concentration on tumor site were observed, tumor SUVmax was 1.46±0.28, and hypoxia volume was 2.87 ± 0.34, baseline R2* was correlated with hypoxia volume(r=0.46, P=0.04). Pathological examination showed significant HIF-1α expression but rare CA IX expression, neither had correlation with R2*.Conclusion:BOLD MR is a potential noninvasive method for evaluating prostate cancer, CA IX maybe not an appropriate hypoxia marker for prostate cancer. |
PDF Full Text Request