The Classification of Benign and Malignant Human Prostate Tissue by Multivariate Analysis of †̃HMagnetic Resonance Spectra
The Classification of Benign and Malignant Human Prostate Tissue by Multivariate Analysis of †̃HMagnetic Resonance Spectra
复制标题
DOI：
--
复制
发表时间：
--
期刊：
影响因子：
--
通讯作者：
Per Hahn;Ian C. P. Smith;Leonard;Leboldus;Charles;Littman;R. Somorjai;Tedros;Bezabeh
中科院分区：
文献类型：
--
作者：
Per Hahn;Ian C. P. Smith;Leonard;Leboldus;Charles;Littman;R. Somorjai;Tedros;Bezabeh
恶性肿瘤的发病（6<e:2> - 11）。我们在这里报告我们最近的进展，使用<s:2> ~ H MRS来表征和分类人类前列腺组织。有几项核磁共振研究报道了非常小样本量的前列腺组织提取物（12<e:2> - 17）。我们认为，全组织活检标本的离体MRS研究提供了更能代表体内情况的信息，因为一些组织生物化学可能依赖于形态和结构。迄今为止报道的前列腺MRS结果表明，代谢物如柠檬酸盐、胆碱、肌酸和肌醇的共振可能具有一定的诊断价值（12<e:2>€17）。然而，由于这些研究大多只针对少数病例进行，因此无法得出明确的结论。在这些共振中，柠檬酸盐的共振受到了最大的关注，因为发现前列腺腺癌瘤中的柠檬酸水平明显低于正常和前列腺增生标本中的浓度。然而，这只适用于腺组织起源的前列腺增生。已经发现基质前列腺增生中柠檬酸盐的含量与癌症中的含量一样低，因此仅根据柠檬酸盐浓度来区分两者的区别就不那么明显了。因此，不像传统的分析方法那样依赖于一个特定的共振或一组任意选择的共振，而是应该考虑光谱中的每一条信息。多变量分析方法，凭借其使用光谱中所有信息的能力，已经非常成功地以稳健，非主观和可靠的方式分析MRS数据（11,18 - 20）。在这里，我们提出了一个初步分析良性和恶性前列腺组织使用这些方法。材料和方法。本文收集了50例经尿道前列腺切除术和根治性前列腺切除术患者的前列腺组织标本。每个经尿道切除标本随机抽取2个芯片（平均40 <e:2> - 60个芯片）。在手术时，从切除的前列腺中获得根治性tatectomy标本的穿刺活检。这些活检是随机从前列腺后部获得的。所有活检仅在一个时间点进行，即没有患者再次活检。三名患者正在接受激素治疗。前列腺组织标本（n = 87）置于D20-PBS溶液中，在切除30 mm内快速冷冻于液氮中，保存…
onset of malignancy (6†" 1 1). We report here our recent progress using ‘H MRS to characterize and classify human prostatic tissue. There have been several nuclear magnetic resonance studies reported on prostate tissue extracts with very small sample sizes (12†" 17).We believe that an ex vivo MRS study of whole-tissue biopsy specimens provides information that is more representative of the in vivo situa tion because some of the tissue biochemistry may be morphology-and architecture-dependent. The prostate MRS results reported thus far suggest that resonances from metabolites such as citrate, choline, creatine, and inositol may have some diagnostic value (12†" 1 7). However, because most of these studies were performed on only a few cases, definitive conclusions cannot be drawn. Of these resonances, that of citrate has been pursued with the greatest interest because citrate levels were found to be present at significantly lower concentrations in prostate adenocarci noma than in both normal and BPH specimens. However, this is true only for BPH of glandular tissue origin. Stromal BPH has been found to have citrate levels as low as levels found in cancer and hence the distinction between the two, based solely on the citrate intensity, becomes less marked. Thus, instead of relying on a particular reso nance or a set of arbitrarily selected resonances, as is done in the conventional methods of analysis, every piece of information in the spectra should be considered. Multivariate analysis methods, by virtue of their ability to use all of the information in the spectra, have been very successful for analyzing MRS data in a robust, nonsubjective, and reliable manner (11, 18†" 20).Here, we present a preliminary analysis of benign and malignant prostatic tissue using these methods. MATERIALS AND METHODS MRS Specimens. Prostate tissue specimens were collected from transure thral resection of the prostate and radical prostatectomy from 50 patients. Two chips were selected randomly from each transurethral resection specimen (40†" 60 chips on average).A needle biopsy was obtainedfrom radicalpros tatectomy specimens at the time of surgery from the removed prostate. These biopsies were randomly obtained from the posterior aspect of the prostate. All biopsies were obtained at one time point only, i.e., there were no patients who were rebiopsied. Three patients were receiving hormonal therapy. Prostate tissue specimens (n = 87) were placed in D20-PBS solution, snap frozen in liquid nitrogen within 30 mm of excision, and stored …