Intro & Objectives:
Diagnosis of prostate cancer relies on transrectal ultrasound (TRUS)-guided biopsy of
the prostate however given poor specificity of screening and insufficient resolution of
conventional ultrasound, many men undergo this procedure unnecessarily.
A novel 29 MHz high resolution micro-ultrasound (ExactVu™ micro-ultrasound,
Exact Imaging, Toronto, Canada) has been developed to significantly increase
the spatial resolution and tissue differentiation of TRUS. Combining patient
screening data with real-time visualization of the prostate (quantified using
ltrasound1 risk scores) may permit
leaving lower risk sections of the prostate, or entire patients with uniformly low risk glands,
un-biopsied without significantly increasing the procedure’s false negative rate.
Material & Methods:
Cine loops of 300 micro-ultrasound TRUS biopsies were examined from an ongoing
2,000 patient multi-center clinical trial (clinicaltrials.gov NCT02079025).
Patients were undergoing TRUS biopsy for suspicion of cancer due to elevated
PSA and/or abnormal DRE. Two investigators marked PRI-MUS scores on the
micro-ultrasound images to differentiate suspicious tissue from characteristically
benign tissue for all 300 loops, while blinded to pathology.
200 of these loops were a training set, and histograms of patient age, DRE result,
and PSA were examined for each PRI-MUS risk level to determine pre-screening
viability for potentially benign cases. Conditions were identified for lower-risk
individuals and/or areas of the prostate and applied to the 100-loop test set.
In the 100 sample test set (45 with clinically-significant cancer), each investigator
would have eliminated 11 biopsies while maintaining a per-core sensitivity of 96% by
applying age, DRE, and PSA to the PRI-MUS protocol. 3 of these 22 samples were positive
for cancer via pathology. One was a low-risk GS 6 lesion. Another was a GS 9 lesion
which would have been identified by surrounding high-PRI-MUS score samples.
The third was a small GS 7 lesion at 35% core length which would have been missed.
In total, diagnosis of cancer and grade of the index lesion would have been changed
in 1 of the 100 subjects for investigator 1, and in none of the 100 subjects for
investigator 2. This suggests an overall per-subject sensitivity of 98.7% (specificity 19.4%)
and NPV of 95.5% (PPV 47.0%).
This mini study suggests that the combination of clinical variables and micro-ultrasound may
allow better targeting of biopsies and avoidance of sampling certain low-risk areas of the
prostate in low-risk individuals. This technique could be done live during the biopsy procedure
and requires no additional equipment or personnel, assuming that a micro-ultrasound system and
the PRI-MUS protocol is employed for biopsy guidance.
- Ghai, S. et al., “Assessing Cancer Risk in Novel 29 MHz Micro-Ultrasound Images of the
Prostate”, Journal of Urology, Paper In Press: June 2016