Added value of amide proton transfer-weighted magnetic resonance imaging to Prostate Imaging Reporting and Data System version 2.1 in differentiating clinically significant prostate cancer

Introduction

Prostate cancer (PCa) ranks second in the global incidence among male tumors and has a relatively high mortality rate, accounting for 358,989 deaths worldwide in 2018 (1,2). PCa has shown an increasing incidence in recent years, gradually affecting younger age groups (3). The introduction of prostate-specific antigen (PSA) screening has reduced the mortality rate of patients with PCa, but its low specificity often leads to unnecessary biopsy of the prostate (4). This emphasizes the need for more accurate and effective diagnostic methods for guiding treatment.

Multiparametric magnetic resonance imaging (MRI) is currently the most effective imaging modality for the diagnosis of PCa (5). Prostate Imaging Reporting and Data System (PI-RADS) was introduced in 2012 (6), providing a standardized and validated method of investigating PCa with MRI. According to the requirements of the PI-RADS, each lesion should be scored by T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), and dynamic contrast-enhanced MRI (DCE-MRI). Radiologists qualitatively score suspicious lesions based on the signal characteristics (hypointensity or hyperintensity) observed in these sequences (7-9). Numerous studies have confirmed the ability of PI-RADS to differentiate clinically significant PCa (csPCa) (10-12). However, the interpretation remains subjective and requires extensive experience, which limits the overall accuracy and interobserver consistency (13,14).

Recently, some studies have attempted to incorporate quantitative parameters from DCE-MRI as a complementary tool for PI-RADS, but the results have not been satisfactory (15-19). As a type of chemical-exchange saturation-transfer MRI, amide proton transfer-weighted (APTw) MRI allows for the indirect quantitation of endogenous proteins based on the exchange rate between amide proton presaturated by a chemical shift of 3.5 ppm and free water (20,21). This technique has been successfully applied in the diagnosis and pretreatment evaluation of different tumors, such as brain tumors, lung cancer, bladder cancer, and uterine and adnexal tumors (22-26). Additionally, Yang et al. reported a model based on APT and DWI values that had a satisfactory area under the curve (AUC) of 0.880 (5) in differentiating between benign and malignant lesions of the prostate. The study by Qin et al. also demonstrated the efficacy of APTw-MRI in in predicting the grade of PCa (27).

Although the value of APTw-MRI in the early diagnosis of PCa has been widely acknowledged, its potential ability in differentiating csPCa and added value to PI-RADS scores remains unclear. Therefore, this study aimed to validate the value of APTw-MRI in differentiating csPCa and to investigate the its added value to PI-RADS v. 2.1. We present this article in accordance with the TRIPOD reporting checklist (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1121/rc).

Methods

Patients

From April 2023 to April 2024, data from 544 participants who underwent multiparametric MRI at a single institution were retrieved for retrospective analysis. The MRI scans performed between January and April 2024 were selected as the validation cohort. The inclusion criteria were patients who (I) had complete pathological information; (II) received a prostate MRI scan, including T2WI, DWI, DCE, and APT sequences; and (III) underwent prostate biopsy and/or radical prostatectomy within 1 month after the MRI scan. The exclusion criteria were patients who had (I) hormone or radiation treatment before MRI examination; (II) insufficient image quality; (III) pathology of lesions acquired 1 month after MRI scan that might have progressed; or (IV) diffuse lesions with complex intratumoral conditions rendering an appropriate region of interest (ROI) placement difficult. The flowchart of participant selection is displayed in Figure 1. This study was registered at China Clinical Trials Registry (registration number: IIT-2024-060). The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study was approved by the institutional and governmental ethics committee of The Second Affiliated Hospital of Nanchang University. The requirement for individual consent was waived due to the retrospective nature of the analysis.

Figure 1 Patient selection flowchart. PSA, prostate-specific antigen.

Image acquisition

All examinations were performed using 3-T MRI device (Discovery MR750w; GE HealthCare, Chicago, IL, USA) with a 32-channel body coil. Before the examination, the participants were instructed to urinate to empty the bladder. The scan sequences are shown in Table S1. T2WI and DWI images were obtained at a matched position under the same parameter settings to allow for direct comparisons. The location of the APT images was also determined on T2WI. The apparent diffusion coefficient (ADC) map was automatically calculated using DWI sequences.

The APTw-MRI used in this study was a work-in-progress sequence provided by GE HealthCare, did not require additional hardware software, and was conducted based on the fast spin-echo sequence. A pseudocontinuous saturation pulse with a power of 2.0 mT and a duration of 2 seconds was used in the module of chemical-exchange saturation-transfer preparation, with the image being obtained at the following frequency offsets: ±7, ±4, ±3.5, and ±3.0 ppm. APT effects were evaluated using magnetization transfer ratio asymmetry (MTRasym) at a bias of 3.5 ppm, which was calculated according to the following formula: MTRasym (%) = [S-3.5 ppm − S+3.5 ppm]/S0, where S-3.5 ppm represents the signal strength when the frequency offset is –3.5 ppm, and S+3.5 ppm represents the signal strength when the frequency offset is at +3.5 ppm. MTRasym (3.5 ppm) was also defined as the APT value (28).

Image analysis

After uniform training in lesion identification and PI-RADS scoring, two radiologists with at least 2 years of experience in examining prostate magnetic resonance (MR) images who were blinded to the pathological results independently evaluated the MR images of each participant and assigned PI-RADS scores. When multiple lesions were found in a participant, the tumor with the highest PI-RADS score (for lesions with the same PI-RADS score, the one with the largest size was evaluated) was selected as the target lesion. Finally, the PI-RADS score was confirmed by a senior radiologist with 18 years of experience in prostate imaging.

The two above-mentioned junior radiologists (one of them resegmented 60 participants selected at random from the total patient population using a computer-generated random number generator) were blinded to the pathological results and independently assessed the conventional MR images of each participant. The ROI was outlined on the largest level of the selected lesion and on the section above and below this level on the T2W images. The ROIs were set as large as possible, with hemorrhage, necrosis, and edema being avoided. The ROIs in T2WI were then copied and pasted to the APT and ADC images to ensure the consistency of the sizes and slices of the ROIs. Subsequently, the parameters [minimum APT value (APTmin), maximum APT value (APTmax), mean APT value (APTmean), minimum ADC value (ADCmin), maximum ADC value (ADCmax), and mean ADC value (ADCmean)] were recorded. This process was repeated three times, and the mean value was used for further analysis (29). All measurements were performed on an Advantage Workstation 4.6 (GE HealthCare).

Histopathology

Prostate specimens were obtained through targeted standard transrectal ultrasonography (TRUS)-MRI fusion, cognitive fusion targeted biopsy, or prostate cystectomy. All pathological evaluations were performed according to the International Society of Urological Pathology (ISUP) 2014. In this study, csPCa was defined as lesions with an ISUP Gleason Grade Group (GGG) greater than or equal to 2. Lesions with GGG <2 or no PCa were defined as not clinically significant PCa (ncsPCa) (30).

Statistical analysis

In this study, P<0.05 was considered statistically significant. Statistical analyses were performed with SPSS 26.0.0 (IBM Corp., Armonk, NY, USA), and graphics were drawn using R version 3.6.0 (with packages “rmda”, “rms”, and “pROC”; The R Foundation for Statistical Computing), and Python version 3.9.13 (with packages “seaborn” and “pyplot”; Python Software Foundation, Wilmington, DE, USA).

Weighted k analysis was employed to evaluate the interobserver agreement of the PI-RADS scores, and the intraclass correlation coefficient (ICC) was used to evaluate the interobserver agreement of the parameters from APT or ADC. The Kolmogorov-Smirnov test or the Shapiro-Wilk test was used to assess the normality of the distribution. In addition, the independent samples t-test (for normally distributed data) or the Mann-Whitney test (for nonnormally distributed data) was employed to analyze statistical differences in the demographic and MRI parameters. Receiver operating characteristic (ROC) curve analysis was performed to evaluate the ability of six MR parameters (APTmin, APTmax, APTmean, ADCmin, ADCmax, and ADCmean) in identifying csPCa. Finally, the two parameters with the highest areas under the ROC curve (AUC) or the greatest comprehensive performance in the ADC and APT values were selected for subsequent analyses.

Multivariate logistic regression analysis was utilized to construct models based on selected parameters and PI-RADS, including (I) ADC + PI-RADS (ADC PI-RADS), (II) APT + PI-RADS (APT PI-RADS), and (III) ADC + APT + PI-RADS (combined model). The Delong test was applied to compare the AUCs between the models. The performance and clinical utility of the models were assessed with calibration curves, the Hosmer-Lemeshow goodness-of-fit test, and decision curve analysis (DCA). Parameter analysis and model construction were conducted based on the training cohort. The validation cohort was employed solely to evaluate the efficacy of the nomogram.

Results

Participant characteristics

A total of 161 patients (mean age 69.34±7.70 years), including 102 with ncsPCa and 59 with csPCa, were ultimately enrolled in the study. The median PSA was 11.80 (IQR, 8.04–25.00) ng/mL. Among the participants, 65 underwent systematic biopsy, and 96 underwent prostatectomy. The results revealed that lesions with a PI-RADS score greater than 3 points were more likely to be diagnosed as csPCa. Among the 75 participants with PI-RADS scores greater than 3, 50 (66.67%) were determined to have csPCa and 25 (33.33%) to have ncsPCa. Among the 86 patients with PI-RADS scores less than or equal to 3, 9 (10.47%) had csPCa and 77 (89.53%) had ncsPCa. The clinical and pathological characteristics of these participants are displayed in Table 1.

Interobserver agreement

The PI-RADS score showed good interobserver agreement (k=0.808; P<0.001). For the quantitative parameters, the ICCs of the two radiologists’ measurements were 0.845 (95% CI: 0.753–0.904), 0.884 (95% CI: 0.814–0.929), and 0.953 (95% CI: 0.923–0.972) for APTmin, APTmax, and APTmean, respectively. The ICCs for ADCmin, ADCmax, and ADCmean were 0.901 (95% CI: 0.839–0.939), 0.959 (95% CI: 0.279–0.665), and 0.977 (95% CI: 0.961–0.986), respectively.

Comparisons of APT and ADC values between csPCa and ncsPCa

The APT and ADC values in the csPCa and ncsPCa groups are shown in Table 2 and Figure 2. The APT values (APTmin, APTmax, APTmean) were noticeably higher in patients with csPCa than in those with ncsPCa, whereas all ADC values (ADCmin, ADCmax, ADCmean) were significantly lower in patients with csPCa than in those with ncsPCa. Figure 3 shows the MR images of two representative patients.

Figure 2 Box plots of the APT and ADC values. APT, amide proton transfer; ADC, apparent diffusion coefficient; APTmin, minimum APT value; APTmax, maximum APT value; APTmean, mean APT value; ADCmin, minimum ADC value; ADCmax, maximum ADC value; ADCmean, mean ADC value; csPCa, clinically significant prostate cancer; ncsPCa, not clinically significant prostate cancer.

Figure 3 Two representative cases from this study. Case 1: an 80-year-old man with dysuria whose physical examination showed an increase in PSA level (PSA: 12.8 0 ng/mL). Axial T2-weighted imaging showed a localized hypointense signal lesion in the left PZ (A) with a markedly hyperintense signal in DWI (focus, b=1,400) (B). (C) An early DCE image did not indicate marked enhancement within the lesion. (D) The ADC map showed a markedly hypointense signal within the lesion, (E) with an isointense signal in the APTw image. (F) Microscopic image of hematoxylin and eosin staining (original magnification 40×). PI-RADS assessment category =5. The APTmean and ADCmean of the above lesion were 0.652 and 0.807, respectively. Case 2: a 63-year-old man whose physical examination showed an increase in PSA level (PSA: 8.27 ng/mL). Axial T2-weighted imaging showed a localized hypointense signal lesion in the right PZ (G) with a markedly hyperintense signal in DWI (focus, b=1,400) (H). (I) Early DCE imaging showed marked enhancement within the lesion. The ADC map showed a markedly hypointense signal within the lesion (J) with a hyperintense signal in the APTw image (K). (L) Microscopic image of hematoxylin and eosin staining (original magnification 40×). PI-RADS assessment category =4. The APTmean and ADCmean of the above lesion were 0.750 and 9.931, respectively. The red circles on the images indicate the position of the target lesion. PSA, prostate-specific antigen; PZ, peripheral zone; DWI, diffusion-weighted imaging; ADC, apparent diffusion coefficient; DCE, dynamic contrast-enhanced; APTw, amide proton transfer-weighted image; PI-RADS, Prostate Imaging Reporting and Data System; APTmin, minimum APT value; APTmax, maximum APT value; APTmean, mean APT value; ADCmin, minimum ADC value; ADCmax, maximum ADC value; ADCmean, mean ADC value.

Univariable performance

Table 3 shows the univariable ROC analysis of the APT and ADC values. The AUCs of APTmax, APTmin, and APTmean were 0.647, 0.663, and 0.723, respectively. The AUCs of ADCmax, ADCmin, and ADCmean were 0.773, 0.703, and 0.759, respectively. The parameters with the greatest comprehensive performance in predicting csPCa were APTmean (AUC: 0.723; 95% CI: 0.634–0.812; accuracy: 0.724; sensitivity: 0.717; specificity: 0.727) and ADCmean (AUC: 0.759; 95% CI: 0.666–0.852; accuracy: 0.806, sensitivity: 0.565; specificity: 0.932), with cutoff values of >1.644 and <0.656, respectively. Given the simplicity and practicality of the models, APTmean and ADCmean were included in further analyses.

Prognostic performance of the models

Multivariate logistic regression analysis was used to construct models based on (I) ADCmean + PI-RADS, (II) APTmean + PI-RADS, and (III) ADCmean + APTmean + PI-RADS. Table 4 and Figure 4 display the ROC analysis results of the models. The AUCs of the PI-RADS, APT PI-RADS, ADC PI-RADS, and the combined model were 0.813 (95% CI: 0.743–0.884), 0.867 (95% CI: 0.806–0.928), 0.833 (95% CI: 0.759–0.908), and 0.875 (95% CI: 0.813–0.936), respectively. With the introduction of APTmean, the AUC of APT PI-RADS (AUC: 0.867) was significantly improved compared to that of PI-RADS alone (AUC: 0.813) (P=0.002); moreover, the positive predictive value (PPV: 0.757 vs. 0.667) and negative predictive value (NPV: 0.814 vs. 0.727) were markedly improved. With the introduction of the ADCmean, the combined model yielded a higher AUC (0.875), specificity (0.841), and NPV (0.839) as compared to the APT PI-RADS, but the difference in AUCs was not statistically significant (AUC: 0.875 vs. 0.867; P=0.69).

Figure 4 ROC analysis of models in differentiating csPCa. PI-RADS, Prostate Imaging Reporting and Data System; CI, confidence interval; ADC, apparent diffusion coefficient; APT, amide proton transfer; csPCa, clinically significant prostate cancer; ROC, receiver operating characteristic.

Nomogram construction and validation

Furthermore, a nomogram was constructed based on the combined model to facilitate the prediction of csPCa in clinical practice (Figure 5A). As noted previously, APTmean had a greater impact than did ADCmean in the combined model. The calibration curves (Figure 5B) demonstrated that the predictions of the combined model were consistent with the actual results (P=0.95). The DCA results (Figure 5C) indicated that the combined model could provide greater benefits for patients than could the PI-RADS.

Figure 5 Nomogram and diagnostic performance of the combined model. (A) Nomogram of the combined model based on PI-RADS, APTmean, and ADCmean. The total score is the sum of all point of variables as represented on the scale in the uppermost point. The total points projected on the bottom scale represent the probability of csPCa. (B) Calibration curves of the combined model. (C) Decision curve analysis of the combined model and PI-RADS model. The horizontal axis shows the threshold for high risk, and the vertical axis shows the net benefit to the patients. The colored lines represent the benefit of models. “All” indicates the benefit of the intervention for all patients. PI-RADS, Prostate Imaging Reporting and Data System; APT, amide proton transfer; ADC, apparent diffusion coefficient; csPCa, clinically significant prostate cancer; ADCmean, mean ADC value; APTmean, mean APT value.

We further verified the efficacy of the nomogram in the validation cohort. The result of ROC analysis (Figure S1) demonstrated that the nomogram constructed based on PI-RADS score, APTmean, and ADCmean retained high diagnostic efficacy (AUC: 0.835) in the validation cohort. The detailed results are provided in Appendix 1.

Discussion

Precise differentiation of csPCa is crucial for follow-up treatment and patient prognosis. This study evaluated the added value of APT and ADC values in differentiating csPCa. Patients with csPCa exhibited significantly higher APT values and lower ADC values compared to patients with ncsPCa. The APTmean and ADCmean were the most accurate parameters in predicting csPCa. The APTmean could enhance the ability of PI-RADS v. 2.1 in the differentiation of csPCa. However, the inclusion of the ADCmean only provided a marginal improvement to APT PI-RADS.

Numerous have demonstrated the good performance of PI-RADS v. 2.1 in differentiating csPCa (10-12). A meta-analysis also reported that PI-RADS v. 2.1 performed well in both the zonal location and peripheral zone (31). In this study, PI-RADS v. 2.1 yielded relatively high accuracy (0.784), sensitivity (0.848), and specificity (0.750), but the positive predictive value (PPV) of 0.667 was relatively low, similar to the results found in the study of Westphalen et al. (32). This may be attributed to the subjectivity of MRI interpretation, as the evaluation is based on qualitative parameters. In this study, both APT-w MRI and the ADC exhibited good performance in differentiating csPCa, which is consistent with prior studies (18,33-35). APTmean and ADCmean showed relatively high AUCs of 0.723 and 0.759, with a cutoff value of >1.644 and <0.656, respectively. However, there was a discrepancy in their efficacy for PI-RADS enhancement. The results showed that the introduction of APTmean significantly improved the efficacy of PI-RADS v. 2.1 in differentiating csPCa as compared to the PI-RADS alone (AUC: 0.867 vs. 0.813; SPE: 0.818 vs. 0.750; PPV: 0.757 vs. 0.667; NPV: 0.814 vs. 0.727). This supports the feasibility of using APTw-MRI as a supplementary sequence for PI-RADS v. 2.1. This can be attributed to the high sensitivity of APT to amide proton in the microenvironment as evidenced by the higher APT value of csPCa, which reflects the tumor cellularity, proliferative activity, invasiveness, and tumor burden of csPCa (36,37). The quantitative analyses of the intratumor environment enable APT to provide a more comprehensive representation of PI-RADS.

In contrast, the introduction of the ADCmean, as a supplement to either the PI-RADS or APT PI-RADS, did not significantly improve the performance of the original model (AUC: 0.875 vs. 0.867; P=0.69). This could be attributed to the multiple influence on the ADC values, such as tissue components, membrane permeability, blood perfusion, and cell density (38). Moreover, although the quantitative features of ADC values were not incorporated, PI-RADS v. 2.1 considers signal characteristics from ADC maps, which may also be a contributing factor. In contrast, APT values provide quantitative measurements of phenotype and are not sensitive to these factors.

As the combined model yielded the best comprehensive performance (exhibiting the highest AUC, accuracy, specificity, and NPV of 0.875, 0.821, 0.841, and 0.839, respectively), a nomogram was constructed based on the combined model. The DCA revealed that the combined model showed greater standardized net benefit than did the PI-RADS v. 2.1 alone. Furthermore, an additional validation cohort also confirmed the stability of the nomogram, supporting the clinical application of the model.

Limitations

The limitations of the present study should be acknowledged. First, the samples in this study were obtained from a single center, and the sample size was relatively small. Although the findings demonstrated the independent predictive value of APTw-MRI for csPCa and its complementary value to PI-RADS v. 2.1, further validation with a larger sample cohort and multicenter studies is still necessary before clinical application. Second, this study primarily focused on the differentiation of csPCa and did not consider the staging and grading of PCa. Finally, due to the small sample size, this study did not categorically discuss the lesions in different zones of the prostate, which requires further investigation with a larger sample size. In addition, biopsy pathology was used for results of lesions in some patients. Although a few studies have demonstrated this to be reliable for PCa detection (39), a larger-sample cohort with prostatectomy treatment will be invaluable in future studies. Finally, some technical issues and challenges should be addressed in subsequent research, including faster imaging methods for 3D APT-weighted imaging to acquire the full spectrum of lesion characteristics and purer APT values isolated from other influences (such as the pH in the microenvironment and the nuclear overhauser-mediated chemical-exchange saturation-transfer effect).

Conclusions

To the best of our knowledge, this is the first study to examine the application of APTw-MRI in the prediction of csPCa and its additional value to PI-RADS v. 2.1. The findings in this study indicated that APTw-MRI is beneficial for the differentiation of csPCa and can improve the comprehensive efficacy of PI-RADS v. 2.1. In future research, we will aim to validate the stability of the results and the effects in both zones of the prostate, with the goal of conducting further prospective studies.

Acknowledgments

Funding: This study was supported by the Natural Science Foundation of Jiangxi Province (No. 20212B A B206053).

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://qims.amegroups.com/article/view/10.21037/qims-24-1121/rc

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1121/coif). J.D. is an employee of GE HealthCare and has a collaborative relationship with this study. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki (as revised in 2013) and was approved by the institutional and governmental ethics committee of the Second Affiliated Hospital of Nanchang University. The requirement for individual consent was waived due to the retrospective nature of the analysis.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Mottet N, van den Bergh RCN, Briers E, Van den Broeck T, Cumberbatch MG, De Santis M, et al. EAU-EANM-ESTRO-ESUR-SIOG Guidelines on Prostate Cancer-2020 Update. Part 1: Screening, Diagnosis, and Local Treatment with Curative Intent. Eur Urol 2021;79:243-62. [Crossref] [PubMed]
2. Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2018;68:394-424. [Crossref] [PubMed]
3. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011;61:69-90. [Crossref] [PubMed]
4. Sciarra A, Cattarino S, Gentilucci A, Salciccia S, Alfarone A, Mariotti G, Innocenzi M, Gentile V. Update on screening in prostate cancer based on recent clinical trials. Rev Recent Clin Trials 2011;6:7-15. [Crossref] [PubMed]
5. Yang L, Wang L, Tan Y, Dan H, Xian P, Zhang Y, Tan Y, Lin M, Zhang J. Amide Proton Transfer-weighted MRI combined with serum prostate-specific antigen levels for differentiating malignant prostate lesions from benign prostate lesions: a retrospective cohort study. Cancer Imaging 2023;23:3. [Crossref] [PubMed]
6. Dickinson L, Ahmed HU, Allen C, Barentsz JO, Carey B, Futterer JJ, Heijmink SW, Hoskin P, Kirkham AP, Padhani AR, Persad R, Puech P, Punwani S, Sohaib A, Tombal B, Villers A, Emberton M. Scoring systems used for the interpretation and reporting of multiparametric MRI for prostate cancer detection, localization, and characterization: could standardization lead to improved utilization of imaging within the diagnostic pathway? J Magn Reson Imaging 2013;37:48-58. [Crossref] [PubMed]
7. Mehralivand S, Bednarova S, Shih JH, Mertan FV, Gaur S, Merino MJ, Wood BJ, Pinto PA, Choyke PL, Turkbey B. Prospective Evaluation of PI-RADS™ Version 2 Using the International Society of Urological Pathology Prostate Cancer Grade Group System. J Urol 2017;198:583-90. [Crossref] [PubMed]
8. Woo S, Suh CH, Kim SY, Cho JY, Kim SH. Diagnostic Performance of Prostate Imaging Reporting and Data System Version 2 for Detection of Prostate Cancer: A Systematic Review and Diagnostic Meta-analysis. Eur Urol 2017;72:177-88. [Crossref] [PubMed]
9. Seo JW, Shin SJ, Taik Oh Y, Jung DC, Cho NH, Choi YD, Park SY. PI-RADS Version 2: Detection of Clinically Significant Cancer in Patients With Biopsy Gleason Score 6 Prostate Cancer. AJR Am J Roentgenol 2017;209:W1-9. [Crossref] [PubMed]
10. Gaudiano C, Rustici A, Corcioni B, Ciccarese F, Bianchi L, Schiavina R, Giunchi F, Fiorentino M, Brunocilla E, Golfieri R. PI-RADS version 2.1 for the evaluation of transition zone lesions: a practical guide for radiologists. Br J Radiol 2022;95:20210916. [Crossref] [PubMed]
11. Byun J, Park KJ, Kim MH, Kim JK. Direct Comparison of PI-RADS Version 2 and 2.1 in Transition Zone Lesions for Detection of Prostate Cancer: Preliminary Experience. J Magn Reson Imaging 2020;52:577-86. [Crossref] [PubMed]
12. Xu L, Zhang G, Zhang D, Zhang X, Bai X, Yan W, Zhou Y, Zhou Z, Xiao Y, Jin Z, Sun H. Comparison of PI-RADS version 2.1 and PI-RADS version 2 regarding interreader variability and diagnostic accuracy for transition zone prostate cancer. Abdom Radiol (NY) 2020;45:4133-41. [Crossref] [PubMed]
13. Greer MD, Brown AM, Shih JH, Summers RM, Marko J, Law YM, Sankineni S, George AK, Merino MJ, Pinto PA, Choyke PL, Turkbey B. Accuracy and agreement of PIRADSv2 for prostate cancer mpMRI: A multireader study. J Magn Reson Imaging 2017;45:579-85. [Crossref] [PubMed]
14. Rosenkrantz AB, Ginocchio LA, Cornfeld D, Froemming AT, Gupta RT, Turkbey B, Westphalen AC, Babb JS, Margolis DJ. Interobserver Reproducibility of the PI-RADS Version 2 Lexicon: A Multicenter Study of Six Experienced Prostate Radiologists. Radiology 2016;280:793-804. [Crossref] [PubMed]
15. Wei C, Jin B, Szewczyk-Bieda M, Gandy S, Lang S, Zhang Y, Huang Z, Nabi G. Quantitative parameters in dynamic contrast-enhanced magnetic resonance imaging for the detection and characterization of prostate cancer. Oncotarget 2018;9:15997-6007. [Crossref] [PubMed]
16. Fusco R, Sansone M, Petrillo M, Setola SV, Granata V, Botti G, Perdonà S, Borzillo V, Muto P, Petrillo A. Multiparametric MRI for prostate cancer detection: Preliminary results on quantitative analysis of dynamic contrast enhanced imaging, diffusion-weighted imaging and spectroscopy imaging. Magn Reson Imaging 2016;34:839-45. [Crossref] [PubMed]
17. Peng Y, Jiang Y, Yang C, Brown JB, Antic T, Sethi I, Schmid-Tannwald C, Giger ML, Eggener SE, Oto A. Quantitative analysis of multiparametric prostate MR images: differentiation between prostate cancer and normal tissue and correlation with Gleason score--a computer-aided diagnosis development study. Radiology 2013;267:787-96. [Crossref] [PubMed]
18. Tavakoli AA, Hielscher T, Badura P, Görtz M, Kuder TA, Gnirs R, Schwab C, Hohenfellner M, Schlemmer HP, Bonekamp D. Contribution of Dynamic Contrast-enhanced and Diffusion MRI to PI-RADS for Detecting Clinically Significant Prostate Cancer. Radiology 2023;306:186-99. [Crossref] [PubMed]
19. Oto A, Kayhan A, Jiang Y, Tretiakova M, Yang C, Antic T, Dahi F, Shalhav AL, Karczmar G, Stadler WM. Prostate cancer: differentiation of central gland cancer from benign prostatic hyperplasia by using diffusion-weighted and dynamic contrast-enhanced MR imaging. Radiology 2010;257:715-23. [Crossref] [PubMed]
20. Zhou J, Payen JF, Wilson DA, Traystman RJ, van Zijl PC. Using the amide proton signals of intracellular proteins and peptides to detect pH effects in MRI. Nat Med 2003;9:1085-90. [Crossref] [PubMed]
21. Ward KM, Aletras AH, Balaban RS. A new class of contrast agents for MRI based on proton chemical exchange dependent saturation transfer (CEST). J Magn Reson 2000;143:79-87. [Crossref] [PubMed]
22. Park JE, Kim HS, Park KJ, Choi CG, Kim SJ. Histogram Analysis of Amide Proton Transfer Imaging to Identify Contrast-enhancing Low-Grade Brain Tumor That Mimics High-Grade Tumor: Increased Accuracy of MR Perfusion. Radiology 2015;277:151-61. [Crossref] [PubMed]
23. Meng N, Fu F, Feng P, Li Z, Gao H, Wu Y, Zhang J, Wei W, Yuan J, Yang Y, Liu H, Cheng J, Wang M. Evaluation of Amide Proton Transfer-Weighted Imaging for Lung Cancer Subtype and Epidermal Growth Factor Receptor: A Comparative Study With Diffusion and Metabolic Parameters. J Magn Reson Imaging 2022;56:1118-29. [Crossref] [PubMed]
24. Takayama Y, Nishie A, Togao O, Asayama Y, Ishigami K, Ushijima Y, Okamoto D, Fujita N, Sonoda K, Hida T, Ohishi Y, Keupp J, Honda H. Amide Proton Transfer MR Imaging of Endometrioid Endometrial Adenocarcinoma: Association with Histologic Grade. Radiology 2018;286:909-17. [Crossref] [PubMed]
25. Chen W, Liu G, Chen J, Wei Q, Ye Y, Du X, Feng J, Yan Z, Deng K, Liu X. Whole-tumor amide proton transfer-weighted imaging histogram analysis to predict pathological extramural venous invasion in rectal adenocarcinoma: a preliminary study. Eur Radiol 2023;33:5159-71. [Crossref] [PubMed]
26. Li S, Liu J, Zhang Z, Wang W, Lu H, Lin L, Zhang Y, Cheng J. Added-value of 3D amide proton transfer MRI in assessing prognostic factors of cervical cancer: a comparative study with multiple model diffusion-weighted imaging. Quant Imaging Med Surg 2023;13:8157-72. [Crossref] [PubMed]
27. Qin X, Mu R, Zheng W, Li X, Liu F, Zhuang Z, Yang P, Zhu X. Comparison and combination of amide proton transfer magnetic resonance imaging and the apparent diffusion coefficient in differentiating the grades of prostate cancer. Quant Imaging Med Surg 2023;13:812-24. [Crossref] [PubMed]
28. Zhou J, Blakeley JO, Hua J, Kim M, Laterra J, Pomper MG, van Zijl PC. Practical data acquisition method for human brain tumor amide proton transfer (APT) imaging. Magn Reson Med 2008;60:842-9. [Crossref] [PubMed]
29. Peng Y, Ye J, Liu C, Jia H, Sun J, Ling J, Prince M, Li C, Luo X. Simultaneous hepatic iron and fat quantification with dual-energy CT in a rabbit model of coexisting iron and fat. Quant Imaging Med Surg 2021;11:2001-12. [Crossref] [PubMed]
30. Epstein JI, Egevad L, Amin MB, Delahunt B, Srigley JR, Humphrey PA. The 2014 International Society of Urological Pathology (ISUP) Consensus Conference on Gleason Grading of Prostatic Carcinoma: Definition of Grading Patterns and Proposal for a New Grading System. Am J Surg Pathol 2016;40:244-52. [Crossref] [PubMed]
31. Park KJ, Choi SH, Kim MH, Kim JK, Jeong IG. Performance of Prostate Imaging Reporting and Data System Version 2.1 for Diagnosis of Prostate Cancer: A Systematic Review and Meta-Analysis. J Magn Reson Imaging 2021;54:103-12. [Crossref] [PubMed]
32. Westphalen AC, McCulloch CE, Anaokar JM, Arora S, Barashi NS, Barentsz JO, et al. Variability of the Positive Predictive Value of PI-RADS for Prostate MRI across 26 Centers: Experience of the Society of Abdominal Radiology Prostate Cancer Disease-focused Panel. Radiology 2020;296:76-84. [Crossref] [PubMed]
33. Wang HJ, Cai Q, Huang YP, Li MQ, Wen ZH, Lin YY, Ouyang LY, Qian L, Guo Y. Amide Proton Transfer-weighted MRI in Predicting Histologic Grade of Bladder Cancer. Radiology 2022;305:127-34. [Crossref] [PubMed]
34. Guo Z, Qin X, Mu R, Lv J, Meng Z, Zheng W, Zhuang Z, Zhu X. Amide Proton Transfer Could Provide More Accurate Lesion Characterization in the Transition Zone of the Prostate. J Magn Reson Imaging 2022;56:1311-9. [Crossref] [PubMed]
35. Huang H, Liu Z, Ma Y, Shao Y, Yang Z, Duan D, Zhao Y, Wen S, Tian J, Liu Y, Wang Z, Yue D, Wang Y. Based on PI-RADS v2.1 combining PHI and ADC values to guide prostate biopsy in patients with PSA 4-20 ng/mL. Prostate 2024;84:376-88. [Crossref] [PubMed]
36. Goffeney N, Bulte JW, Duyn J, Bryant LH Jr, van Zijl PC. Sensitive NMR detection of cationic-polymer-based gene delivery systems using saturation transfer via proton exchange. J Am Chem Soc 2001;123:8628-9. [Crossref] [PubMed]
37. Hambrock T, Somford DM, Huisman HJ, van Oort IM, Witjes JA, Hulsbergen-van de Kaa CA, Scheenen T, Barentsz JO. Relationship between apparent diffusion coefficients at 3.0-T MR imaging and Gleason grade in peripheral zone prostate cancer. Radiology 2011;259:453-61. [Crossref] [PubMed]
38. Razek AA, Gaballa G, Denewer A, Nada N. Invasive ductal carcinoma: correlation of apparent diffusion coefficient value with pathological prognostic factors. NMR Biomed 2010;23:619-23. [Crossref] [PubMed]
39. Wegelin O, Exterkate L, van der Leest M, Kummer JA, Vreuls W, de Bruin PC, Bosch JLHR, Barentsz JO, Somford DM, van Melick HHE. The FUTURE Trial: A Multicenter Randomised Controlled Trial on Target Biopsy Techniques Based on Magnetic Resonance Imaging in the Diagnosis of Prostate Cancer in Patients with Prior Negative Biopsies. Eur Urol 2019;75:582-90. [Crossref] [PubMed]