Association of T2w-images with Gleason Score?

/
Downes MR, Gibson E, Sykes J, Haider M, van der Kwast TH, Ward A. Acad Radiol. 2016 Nov;23(11):1412-1421. doi: 10.1016/j.acra.2016.07.013. PubMed PMID: 27639626.

Downes MR, Gibson E, Sykes J, Haider M, van der Kwast TH, Ward A. Acad Radiol. 2016 Nov;23(11):1412-1421. doi: 10.1016/j.acra.2016.07.013. PubMed PMID: 27639626.

Short summary 

This recent paper investigates whether it is possible to determinate Gleason grade sub-patterns based on T2 signal intensities. 

Conclusion

Due to the different histological tissue components, it is shown that the MRI T2 signal is dependent on the histological sub-patterns within Gleason grades 3 and 4 cancers.

Patient cohort

25 pre-radical prostatectomy MRI scans were compared with whole mount specimens, tumour and peripheral zone digital annotated by 2 pathologists, separated into the following 6 sub patterns: Gleason grade 3 (sparse, intermediate, packed); Gleason grade 4 (large cribriform, small cribriform, intra-ductal).

The log MRI signal for each contoured region (n=809) was measured and compared with histology.

The New and the Good

  • Attempt to analyse T2W signal to discern Gleason subgrades. This could perhaps support findings of the cribriform patterns.

Limitations

  • Small patient cohort.
  • No correlation with Diffusion/ADC.
  • Material samples from 2010-2011.
  • Large and complicated setup to ensure correct annotation of in-vivo and ex-vivo prostate MRI T2W with annotated focal sub patterns histology of Gleason grads. 

Possible consequences for clinical practice

If; as shown in this study; the T2 weighted signal is dependent on these histological sub-patterns within Gleason grade 3 and 4 cancers, then this can have implications on how we would direct biopsy sampling and thereby patient management

Possible implications for PIRADS v3

If this is shown in a larger study and re-producible, the signal intensities on T2 should be noted i.e. a colour map with thresholds for the most aggressive sub-pattern, or colour overlay of ADC and T2 and in some way integrated in the scoring.

Future study ideas derived from this paper

Further analyses using dedicated tissue analysis with this more precise co-registration technique. 

Personal comment

I think they have done a remarkable job in their effort to ensure a correct co-registration, considering, the tissue changes that arise in the preserving process of the prostate specimen.

BUT, I do miss the use of diffusion sequences although this was not part of this co-registration validation study, as they argue the need of high resolution T2 for accurate registration which excluded the use of DCE and DWI/ADC