Prostate MRI in the early diagnosis of prostate cancer : bi- or multiparametric ?

66 y/o man. Raising PSA (4.4ng/ml). Hyposignal on T2W-MRI (A, arrows) with a low ADC value (B, arrows) and a mildly hyperintense signal intensity on DWI images at a b-value of 2000 s/mm² (arrows, C). Progressive suppression of the lesion signal intensity at higher computed b-values (D-E, arrows), and enhancement of the lesion on DCE (arrows, F). Category assessment is PI-RADS 3 / LIKERT 4 on bpMRI and PI-RAD 3+1 (i.e 4) / Likert 4 or 5 on mpMRI. However, the array of b-values with increasing diffusion weighting show a progressive suppression of the signal intensity of both the lesion and the normal prostate, suggesting benign tissue. Inflammation was found on targeted biopsies. PSA level returned to base line two months later.

Zawaieh et al (1), from the University of Cambridge reports a series of 264 patients who underwent prostate multiparametric MRI (mpMRI), with T2 and DWI (biparametric MRI, bpMRI), at 1.5 and 3T, interpreted first, prior to full mpMRI using a Likert 1-5 scale. Readers were asked to give a prospective opinion on likely benefit of contrast prior reviewing the contrast enhanced (DCE) sequence, and a retrospective opinion following full mpMRI review. Results showed that bp- and mpMRI were assigned the same diagnostic value in 86% of cases. In lesions assigned a Likert score >2, concordance increased to 92.8%. At Likert score >2, bpMRI detected 93.5% of significant Ca (Gleason score>3) and mpMRI  94.6%. MpMRI had fewer false positives than bpMRI (11.4% vs 18.9%) and a lower Likert 3 rate (8.3% vs 17%), conferring higher specificity (74% vs 67%), but similar sensitivity (95% versus 94%) and ROC-AUC (90% vs 89%). DCE was prospectively considered of potential benefit in 27.3%, but readers would only recall 11% of patients for DCE-MRI, mainly to assess score 3 peripheral zone lesions. The authors conclude that DCE remains useful for an optimal characterization of PZ lesions.

Comments :

This well written article may suffer from a lack of homogeneity of the cohort and from the use of the Likert score. The PI-RADS user may be left without being convinced of the incremental value of contrast injection in a detection setting in patients with an elevated PSA level.

  • The 1.5T protocol is limited by the absence of b-values >1400s/mm² to read DW images, which makes difficult to assign a score to the DW signal (mildly of markedly bright?). The use of higher b-values, as done at 3T in the article,  is very helpful. The consensus now is to use a computed b-value of at least 2000s/mm² (2, 3), or even more (4), to obtain a sufficient diffusion weighting which increases the visibility of PZ lesions. If the contrast of signal intensity between the lesion and the rest of the prostate increases, it can be upgraded. If it does not, the lesion is classified non-suspicious (figure). Using PI-RADS, differentiating score 3 from score 4 lesions can be done more accurately (3).
  • The use of the Likert score makes it unclear whether the same conclusions could apply with the PI-RADS score. If both systems assign a score 4 in case of a positive DCE in case of focal PI-RADS 3 lesions (3+1), many Likert score 3 lesions are PI-RADS score 2, because the Likert score gives more value to a diffuse PZ enhancement and therefore PI-RADS 2 score is often upgraded to Likert 3 after contrast injection. Whether diffuse or focal, contrast enhancement of intermediate lesions,  has a low detection rate of significant Ca in expert centers (5) because the majority of these lesions correspond to focal prostatitis (figure 1).  A positive DCE represents the main cause of false positives and it is thus unclear for PI-RADS users how the specificity increased in the author’s article.

The series includes patients after treatment and patients in whom injection is recommended because of a risk of metallic artefact (hip prosthesis brachytherapy seeds…). Only naïve biopsy patients with unquestionable good quality of DW images may have been evaluated in the author’s, DCE being only be 66 y/o man. Raising PSA (4.4ng/ml). Hyposignal on T2W-MRI (A, arrows) with a low ADC value (B, arrows) and a mildly hyperintense signal intensity on DWI images at a b-value of 2000 s/mm² (arrows, C). Progressive suppression of the lesion signal intensity at higher computed b-values (D-E, arrows), and enhancement of the lesion on DCE (arrows, F). Category assessment is PI-RADS 3 / LIKERT 4 on bpMRI and PI-RAD 3+1 (i.e 4) / Likert 4 or 5 on mpMRI. However, the array of b-values with increasing diffusion weighting show a progressive suppression of the signal intensity of both the lesion and the normal prostate, suggesting benign tissue. Inflammation was found on targeted biopsies. PSA level returned to base line two months later.

Zawaieh et al (1), from the University of Cambridge reports a series of 264 patients who underwent prostate multiparametric MRI (mpMRI), with T2 and DWI (biparametric MRI, bpMRI), at 1.5 and 3T, interpreted first, prior to full mpMRI using a Likert 1-5 scale. Readers were asked to give a prospective opinion on likely benefit of contrast prior reviewing the contrast enhanced (DCE) sequence, and a retrospective opinion following full mpMRI review. Results showed that bp- and mpMRI were assigned the same diagnostic value in 86% of cases. In lesions assigned a Likert score >2, concordance increased to 92.8%. At Likert score >2, bpMRI detected 93.5% of significant Ca (Gleason score>3) and mpMRI  94.6%. MpMRI had fewer false positives than bpMRI (11.4% vs 18.9%) and a lower Likert 3 rate (8.3% vs 17%), conferring higher specificity (74% vs 67%), but similar sensitivity (95% versus 94%) and ROC-AUC (90% vs 89%). DCE was prospectively considered of potential benefit in 27.3%, but readers would only recall 11% of patients for DCE-MRI, mainly to assess score 3 peripheral zone lesions. The authors conclude that DCE remains useful for an optimal characterization of PZ lesions.

Comments :

This well written article may suffer from a lack of homogeneity of the cohort and from the use of the Likert score. The PI-RADS user may be left without being convinced of the incremental value of contrast injection in a detection setting in patients with an elevated PSA level.

  • The 1.5T protocol is limited by the absence of b-values >1400s/mm² to read DW images, which makes difficult to assign a score to the DW signal (mildly of markedly bright?). The use of higher b-values, as done at 3T in the article,  is very helpful. The consensus now is to use a computed b-value of at least 2000s/mm² (2, 3), or even more (4), to obtain a sufficient diffusion weighting which increases the visibility of PZ lesions. If the contrast of signal intensity between the lesion and the rest of the prostate increases, it can be upgraded. If it does not, the lesion is classified non-suspicious (figure). Using PI-RADS, differentiating score 3 from score 4 lesions can be done more accurately (3).
  • The use of the Likert score makes it unclear whether the same conclusions could apply with the PI-RADS score. If both systems assign a score 4 in case of a positive DCE in case of focal PI-RADS 3 lesions (3+1), many Likert score 3 lesions are PI-RADS score 2, because the Likert score gives more value to a diffuse PZ enhancement and therefore PI-RADS 2 score is often upgraded to Likert 3 after contrast injection. Whether diffuse or focal, contrast enhancement of intermediate lesions,  has a low detection rate of significant Ca in expert centers (5) because the majority of these lesions correspond to focal prostatitis (figure 1).  A positive DCE represents the main cause of false positives and it is thus unclear for PI-RADS users how the specificity increased in the author’s article.
  • The series includes patients after treatment and patients in whom injection is recommended because of a risk of metallic artefact (hip prosthesis brachytherapy seeds…). Only naïve biopsy patients with unquestionable good quality of DW images may have been evaluated in the author’s, DCE being only useful in case of expected poor quality DW-MRI (6)

The discussion section of the article seems to reconcile supporters and opponents of bpMRI, because it is well said that the quality of DWI is a crucial element for reading a bpMRI. To do so, radiologists should pay a particular attention to the DWI sequence by avoiding as much as possible diffusion artifacts (evacuating rectal air and/or feces by an enema), by respecting the PI-RADS guidelines for DWI acquisition parameters and by using high computed b-values (>2000s/mm²) to avoid the need of a contrast injection, which remains mandatory as soon as good quality DW images cannot be obtained.

So, prostate bpMRI before biopsy : yes, but only if you deserve it (7).

References

  1. Zawaideh JP, Sala E, Shaida et al. Diagnostic accuracy of biparametric versus multiparametric prostate MRI: assessment of contrast benefit in clinical practice. Eur Radiol. 2020.
  2. Jendoubi S, Wagner M, Montagne S, et al. MRI for prostate cancer: can computed high b-value DWI replace native acquisitions? Eur Radiol. 2019;29(10):5197-204.
  3. Pierre T, Cornud F, Colleter L et al. Diffusion-weighted imaging of the prostate: should we use quantitative metrics to better characterize focal lesions originating in the peripheral zone? Eur Radiol. 2017;28(5):2236-45.
  4. Rosenkrantz AB, Parikh N, Kierans et al. Prostate Cancer Detection Using Computed Very High b-value Diffusion-weighted Imaging: How High Should We Go? Acad Radiol. 2016;23(6):704-11.
  5. Kasivisvanathan V, Rannikko AS, Borghi M et al. MRI-Targeted or Standard Biopsy for Prostate-Cancer Diagnosis. N Engl J Med. 2018;378(19):1767-77.
  6. Han C, Liu S, Qin XB, Ma S, Zhu LN, Wang XY. MRI combined with PSA density in detecting clinically significant prostate cancer in patients with PSA serum levels of 4 approximately 10ng/mL: Biparametric versus multiparametric MRI. Diagn Interv Imaging. 2020.
  7. Cornud F. Bi-parametric prostate MRI before biopsy: Yes, but only if you deserve it. Diagn Interv Imaging. 2020;101(4):191-2.