An article just published online in the UroToday International Journal appears to be the first prospective study to indicate, in a large group of patients, that 3D color Döppler ultrasound imaging (3D-CDI) can diagnose the presence of prostate cancer with an accuracy close to that of prostate biopsy.
Merkle reports that he has carried out 3D-CDI of the prostates of 418 patients and correlated diagnosis based on these imaging data to diagnoses based on subsequent biopsies. The entire paper is available on line. The critical information about this study can be summarized as follows:
- All patients tested (since 2005) had either a PSA ≥ 4 ng/ml or a positive family history for prostate cancer and/or breast cancer and a PSA ≥ 3 ng/ml.
- PSA levels in primary cancer patients ranged from 1.4 to 20 ng/ml (mean 8.2 ± 6.5 ng/ml).
- 211/418 men were sonographically classified as having prostate cancer and then histologically verified.
- 160/418 were classified as having benign lesions, of which 153 were also histologically verified (95.2 percent).
- 7/160 patients initially classified with benign lesions were found to have malignant disease on biopsy.
- 3D-CDI diagnosed these patients with a sensitivity of 0.82 and good specificity (0.91).
Now we should be clear that this is not a “perfect” paper. In the first place, Merkle did all of the ultrasounds and all of the biopsies himself. The paper would have been “purer” if a physician who was not aware of the results of 3D-CDI had also biopsied the patients independently, so that we could definitely differentiate between pure biopsy and pure 3D-CDI diagnostic results. However, this would have required all patients to undergo two sets of biopsies in close succession. In the second place, the protocol of the study appears to have changed somewhat over time. In particular, Merkle reports a protocol change after the first 300 patients.
What is important about this paper, however, is that it suggests the possibility that 3D-CRI really may be able to identify the presence of prostate cancer with a relatively high degree of accuracy. We are therefore forced to ask the next questions in the chain, as follows:
- Do the ultrasound data allow us to make better judgments about the need for biopsies in these patients? In other words, can they help us to differentiate between men with low, intermediate, and high risk prostate cancer? Merkle implies that they can.
- Is it possible to repeat this study quickly in a multi-center setting to see if the results reported by Merkle can be replicated? That would seem to depend on the number of 3D-CDI machines currently available, and at which centers. It is The “New” Prostate Cancer InfoLink’s understanding that the number of these machines in clinical use at this time is small, and that the number of physicians with the experience to “read” these images with accuracy is even smaller.
The “New” Prostate Cancer InfoLink definitely encourages the development of an appropriate clinical trial to repeat Merkle’s study at the earliest possible date. Ideally, such a study should be constructed in such a way as to “blind” the physician who takes the first set of biopsy data from knowing the resulst of the 3D-CDI scan. The second set of biopsy data should then be taken by a physician with full awareness of the 3D-CRI image data.