Table 1.

Utility of multivariate risk prediction tools, with and without MRI, in the prostate cancer diagnostic work-up

Risk stratification by	Stratifying men to indicate the need for:	Stratified risk (test result)		Action for further testing	Purpose of action	(Dis)advantages
Multivariate risk prediction tool not including MRI (Fig. 1d)	Biopsy and MRI	High	Yes Yes Yes	MRI Targeted biopsy Systematic biopsy	Maximizing diagnostic yield	Pros  1. Owing to increased pre-test probability, the post-test probability (PPV) of both MRI and biopsies will increase  2. When stratified to low-risk, both diagnostic tests (MRI and biopsy) can be omitted. Resulting in reductions of low-risk PCa detection Cons  1. Does not distinguish between the indication for biopsy or for MRI  2. No discrimination of added value for MRI or biopsy can result in too much testing  3. Missed clinically significant prostate cancers in non-biopsy patients
		Intermediate	Yes Yes Yes	MRI Targeted biopsy Systematic biopsy	Maximizing diagnostic yield
		Low	No No No	MRI Targeted biopsy Systematic biopsy	Reducing MRI and biopsies  Reducing targeted biopsies  Reducing systematic biopsies  Reducing low-risk PCa detection
Multivariate risk prediction tool including MRI (Fig. 1e)	Biopsy	High	Yes No	Targeted biopsy Systematic biopsy	Optimizing diagnostic yield  Reducing systematic biopsies	Pros  1. Optimal pre-biopsy information gathered for biopsy indication.  2. Large PI-RADS category 5 lesions will be categorized in the ‘high-risk’ group, and will probably be biopsied in targeted approach only, omitting systematic biopsies  3. The added value of combining MRI within a multivariate prediction tool is mostly focussed on reducing the high false-positivity of MRI (low specificity and low PPV). Probably PI-RADS category 3 and 4 lesions will fall in this ‘intermediate’ category. With the combination of elevated risk, based on clinical parameters, these lesions will be biopsied in a targeted approach but also systematic biopsies will be performed to gain maximal diagnostic yield  4. Negative MRI with a high-risk profile on clinical parameters could be biopsied in a systematic manner  5. Low-risk profile (negative MRI and low-risk on clinical parameters) may save further testing Cons  1. No reduction of prostate MRI
		Intermediate	Yes Yes	Targeted biopsy Systematic biopsy	Maximizing diagnostic yield
		Low	No No	Targeted biopsy Systematic biopsy	Reducing targeted biopsies  Reducing systematic biopsies  Reducing low-risk PCa detection
Multivariate risk prediction tool (step 1) not including MRI (Fig. 1f)	MRI	High	Yes	MRI	Maximizing diagnostic yield	Pros  1. Distinguish between indication of biopsy and MRI. The first multivariate prediction tool (step 1) stratifies who should get an MRI. This may lead to reduction of MRI (and subsequent reduction of biopsies in the low-risk group. Men in the high-risk group will get an MRI  2. The results from this MRI will be combined in a second multivariate prediction tool (step 2) after MRI, and may further lead to further risk stratification for the indication who to biopsy, and in which approaches to use  3. Men with a negative MRI and with a high-risk profile on clinical parameters could be categorized in the intermediate or low-risk group, depending on the thresholds that will be chosen and if biopsy is required, a systematic approach could be used  4. Thresholds in step 1 might differ from step 2 Cons  1. Missing significant cancers as the trade-off for the reduction of MRI scans done
		Intermediate	Yes	MRI	Maximizing diagnostic yield
		Low	No No No	MRI Targeted biopsy Systematic biopsy	Reducing MRI  Reducing targeted biopsies  Reducing systematic biopsies  Reducing low-risk PCa detection
Multivariate risk prediction tool including MRI (step 2) (Fig. 1f)	Biopsy	High	Yes No	Targeted biopsy Systematic biopsy	Optimizing diagnostic yield  Reducing systematic biopsies
		Intermediate	Yes Yes	Targeted biopsy Systematic biopsy	Maximizing diagnostic yield
		Low	No	Systematic biopsy if needed	Reducing targeted biopsies  Reducing systematic biopsies  Reducing low-risk PCa detection