Author's address (for correspondence): M Russo, Obstetric Unit, Department of Veterinary Clinical Sciences, Veterinary School, University of Naples, Napoli, Italy. E-mail: email@example.com
Ultrasound is a common imaging technique used for examination of the dog's prostate gland that has proven to be successful for diagnosis of many clinical conditions. Whilst B-mode ultrasound is useful for assessment of the prostate gland, there appears to be substantial variation in the appearance of some pathological conditions, although individual experienced ultrasonographers often have a high aptitude for making a correct diagnosis. Here, we report the normal volume and appearance of the canine prostate gland with B-mode ultrasound and use thematic analysis to identify the categories described by experienced ultrasonographers when reporting both normal and abnormal findings. Four thematic categories were identified: background echotexture, parenchymal stippling, generalized appearance and focal changes. We found that a relatively narrow variety of thematic variables (descriptors) were used for reporting of prostatic pathology within these thematic categories. There was also poor association between these descriptors and the underlying pathology, with the only unique descriptors being ‘cysts’ for benign prostatic hyperplasia and ‘parenchymal mineralization’ for adenocarcinoma. In comparison with the limitations of B-mode ultrasound, we document the value of measurement of vascular perfusion kinetics using contrast-enhanced ultrasound and demonstrate how perfusion kinetics differ between benign and malignant prostatic disease.
Disease of the prostate gland is frequently seen in the entire dog, and the conditions most commonly presented include benign prostatic hyperplasia, prostatitis, prostatic cysts and prostatic neoplasia (Smith 2008).
B-mode ultrasound is commonly used for examination of the shape, size, position, margination and internal architecture of parenchymatous organs and has been applied to examination of the prostate gland to diagnose and differentiate disease and monitor response to treatment (Cartee and Rowles 1983; Feeney et al. 1987; Kawakami et al. 1993). Whilst prostatic ultrasound is commonly used in clinical practice (Lacreta et al. 2012), much information relating to image interpretation originates from textbook chapters or a limited number of small case series. The consensus appears to be that several diseases have a similar ultrasonographic appearance and are difficult to differentiate (Burk and Ackerman 1996; Mattoon and Nyland 2002), and yet there has been little objective assessment of the appearance of different prostatic pathologies. Gathering such information is undoubtedly difficult and analysis is fraught with problems but is important especially for the diagnosis of malignant prostatic disease as benign and malignant lesions may, at least initially, have similar clinical and imaging appearances.
In this work, we review relevant literature demonstrating how different prostatic pathology might be differentiated using B-mode and contrast-enhanced ultrasound, as well as undertaking a qualitative thematic analysis of the descriptive terminology used within ultrasound reports from experienced ultrasonographers. Using this method, we aim to establish whether a structured analysis can identify the thematic categories (the aspects that are described) and the thematic variables (the descriptors used), and furthermore to investigate whether the thematic variables differ between benign and malignant prostatic disease.
The prostate is a bilobed gland that sits astride the prostatic urethra. Although it appears to be relatively simple to assess prostatic size from linear measurements of prostatic height, width or length, or to estimate prostatic volume (Juniewicz et al. 1989; Suzuki et al. 1998; Kamolpatana et al. 2000), there is some debate about the exact relationship between the size of the prostate gland and size of the dog (Ruel et al. 1998; Atalan et al. 1999). Most likely this relates to comparing a linear measurement of prostate dimensions (e.g. total prostatic width) with body weight; however, there remains some debate about the relation between prostatic volume and body weight (Atalan et al. 1999), possibly because of the potential complication of an age-related component to prostatic volume (Berry et al. 1986; Korodi et al. 2010). It is important, however, that some standard measurement of prostatic size is established as many pathological conditions have the potential to cause increased size of the gland.
We investigated the usefulness of ultrasound for estimation of prostatic volume in 14 dogs with histologically normal prostate glands (age 3.1 ± 1.1 SD years and weight 10–41 kg). Ultrasound examination using a 10 MHz transducer was performed within 12 h of euthanasia and transverse plane images were used to measure width and height, and longitudinal plane images were used to measure the length of each lobe. The volume of each lobe was calculated using the formula for volume of an ellipse V = (length × width × height) × 0.523 and total prostatic volume was estimated by summing measures for the left and right lobes. The prostate glands were removed at post-mortem examination and the actual prostatic volume measured by water displacement. There was no significant difference between the volume calculated for the left or right lobes and when combined there was a significant correlation between the actual total prostatic volume measured by water displacement and the calculated total volume estimated with ultrasound (coefficient of correlation = 0.96) (Fig. 1).
To establish the normal range of prostatic volumes found in healthy dogs, we examined a cohort of 49 young entire dogs (age 3.3 ± 1.3 SD years and weight 9–49 kg) with no clinical signs of prostatic disease. Dogs were examined using a 10 MHz ultrasound transducer and prostatic volume was calculated as previously described. There was no difference in the volume of the left and right lobes and when summed there was a relationship between body weight and total prostatic volume (coefficient of correlation = 0.31) (Fig. 2). There was a widespread of data around these values (demonstrated by the large confidence intervals) despite the similar ages of the dogs, and we propose that these values be used as ‘normals’ for dogs within this weight range. Furthermore, we suggest that older dogs are measured against these normal volumes enabling reporting of prostatic enlargement even if this is considered a normal ageing process.
Normal B-mode Appearance
The normal prostate gland has a characteristic appearance commonly defined as an hypoechoic parenchyma with moderately echogenic stippling present in a uniform pattern throughout the gland (Mattoon and Nyland 2002). The prostatic capsule is relatively echogenic and can easily be identified when positioned at 90° to the ultrasound beam. When the prostate gland is imaged in the transverse plain, the two lobes appear adjacent to each other, but they are imaged separately in the longitudinal plane. In the caudal part of the gland, the urethra can be identified as a circular structure that is present within the dorsal part of the gland (Garcia Real et al. 2007).
To undertake qualitative analysis of ultrasonographic descriptions to establish the thematic categories (the particular aspects that were described) as well as the thematic variables (the descriptions used within these categories), we undertook a detailed prostatic ultrasound examination of 25 dogs (age 2.9 ± 1.3 SD years and weight 11–41 kg) that had no clinical signs of prostatic disease. Each ultrasonographer was asked to provide a written report for each dog using profuse descriptive terminology that was examined using a structured strategic thematic analysis (Castro et al. 2011) for the words and terms that had been used. Three thematic categories were identified for these descriptions of the normal prostate gland: background echotexture, parenchymal stippling and generalized appearance. The common normal prostate thematic variables (descriptors) for background echotexture were ‘hypoechoic’ ‘moderately hypoechoic’, for parenchymal stippling were ‘regular’, ‘even’, ‘uniform’ and for generalized appearance was ‘homogenous’. There was infrequent descriptive reporting of the appearance of prostatic symmetry, appearance of the capsule or urethra.
Abnormal B-mode Appearance
Whilst some reports demonstrate specific changes in ultrasonographic appearance relating to prostatic pathology (Lacreta et al. 2012), commonly it is suggested that ultrasonographic examination has limited ability to differentiate between prostatic diseases (Smith 2008), and it appears that many pathologies result in similar changes often manifest by increased echogenicity (Paclikova et al. 2006). To evaluate whether any descriptive criteria could be ascribed to specific pathological processes, we undertook detailed prostatic ultrasound examinations of 18 dogs with prostatic pathology (age 5.9 ± 3.2 SD years and weight 19–37 kg) that had been killed for a variety of clinical reasons. The ultrasonographers provided detailed descriptive reports of the appearance of the prostate gland and measured prostatic volume as described previously. The dogs were retrospectively placed into four categories, namely benign prostatic hyperplasia (eight dogs), prostatitis (two dogs), adenocarcinoma (four dogs) and mixed benign pathology (four dogs) based upon histological examination. Thematic category evaluation was performed as previously described and it was found that as well as the previous three categories an additional category (focal changes) was always reported. Evaluation of the thematic variables within the thematic categories for the different diseases demonstrated that very few descriptors were frequently used, with the only unique descriptors being reported for benign prostatic hyperplasia (‘cysts’) and adenocarcinoma (‘mineralized opacities’) (Table 1). There were no thematic categories relating to symmetry, margination or peri-glandular changes; the authors do identify these changes in cases of neoplasia and prostatitis; however, they were not present in these cases.
Table 1. Thematic analysis of ultrasonographic descriptions from 25 normal dogs and 18 abnormal dogs reported by experienced ultrasonographers. Four thematic categories were identified and common thematic variables are shown for each category. The only unique descriptors were ‘cysts’ (reported for benign prostatic hyperplasia) and ‘mineralized opacities’ (reported for adenocarcinoma)
Normal prostate (n = 25)
Hypoechoic Moderately hypoechoic
Regular Even Uniform
Being prostatic hyperplasia (n = 8)
Increased stippling less prominent
Prostatitis (n = 2)
Increased stippling less prominent
Increased echogenicity hypoechoic regions
Focal echogenic regions
Adenocarcinoma (n = 4)
Increased echogenicity Hypoechoic regions Patchy hypoechoic
Focal echogenic regions Focal hypoechoic regions
Mixed benign pathology (n = 4)
Focal echogenic regions Focal hypoechoic regions
None of the dogs with prostatic pathology had a total prostatic volume that exceeded the normal values previously reported; for all dogs regardless of pathology, total prostatic volume was within the 95% confidence limits described in Fig. 2. Two of the dogs with prostatic neoplasia did, however, have asymmetry of volume measurements between the two prostatic lobes.
The development of prostatic pathology is in many cases associated with changes in the vascular supply to the prostate as well as changes to the prostatic parenchyma; indeed, Günzel-Apel et al. (2001) found that benign prostatic hyperplasia was characterized by a significant increase of prostatic artery systolic and diastolic peak velocity compared with normal dogs. Interestingly though, these relatively coarse measures of prostatic vascularization do not appear to be disturbed in all cases of pathology, and Newell et al. (1998) found no difference in blood flow or resistive index in dogs with prostatitis compared with normal dogs. Most recently, contrast-enhanced ultrasound (CEUS) has been developed as a technique and shown to be able to quantify vascular perfusion, essentially by measurement of peak flow and transit time within tissues using video-densitometric analysis of real-time images. This technology offers substantial advantages over simple measurement of flow within vessels supplying the organ and enables targeted measurement specific to suspected lesions. We evaluated the use of CEUS in normal dogs and those with a variety of prostatic diseases (Russo et al. 2009; Vignoli et al. 2011). Ten normal dogs (age 2.2 ± 0.9 SD years and weight 6–37 kg), and 25 dogs with prostatic disease (age 9.1 ± 2.1 SD years and weight 7.5–43.0 kg), were anaesthetized and injected intravenously with a second-generation contrast agent, sulphur hexafluoride microbubbles (SonoVue; Bracco Imaging, Milano, Italy), and examined with dedicated contrast-enhanced ultrasound analytical software (Contrast Tuned Imaging Technology, Esaote, Milano, Italy). Prostate vascular flow was measured including peak intensity expressed as a percentage (PPI), and the time to peak (TTP) expressed in seconds. Thereafter, a single trans-abdominal ultrasound-guided Tru-Cut biopsy of a random portion of the gland (normal dogs) or a targeted biopsy (dogs with prostatic disease) was taken. Various pathologies were found in the 25 abnormal prostates (11 benign prostatic hyperplasia, 1 prostatitis, 4 adenocarcinoma and 9 mixed benign pathology). For normal dogs, PPI was a mean of 16.8% (± 5.8 SD) with a median of 17.80%, and the mean TTP was 33.6 ± 6.4 s with a median of 34 s. For dogs with benign prostatic disease, there were no statistical differences to values for the normal dogs; for benign prostatic hyperplasia the PPI was a mean of 16.9 + 3.8% (median 15.8%) and the TTP was a mean of 26.2 ± 5.8 s (median 24.4 s), for mixed benign pathology the PPI was a mean of 14.8 ± 7.8% (median 12.2%) and the TTP was a mean of 31.9 ± 9.7 s (median 32.8 s), and for prostatitis the PPI was 14.2% and the TTP was 25.9 s. Most interestingly for adenocarcinomas, values were significantly higher than the normal dogs and the dogs with benign pathology; the PPI was a mean of 23.7 ± 1.9% (median 24.2%) and the TTP was 26.9 ± 4.8 s (median 21.1 s).
Prostatic ultrasonography is an important diagnostic technique in small animal practice. There has, however, often been confusion about the interpretation of ultrasonographic findings in prostatic disease. In this study, we demonstrate that estimation of total prostatic volume by summing separate measurements of the two prostatic lobes is a useful method that correlates well with actual prostatic volume. Furthermore, we document the wide range of normal prostatic volumes that occur in normal dogs. These data are particularly important as they originate from young dogs and will undoubtedly be useful for comparison with older dogs when prostate volume may increase (Brendler et al. 1983; Berry et al. 1986; Korodi et al. 2010) as well as for comparison with cases of prostatic pathology. Surprisingly though, in our study of prostatic disease, none of the 18 dogs with confirmed histological changes had a prostatic volume that exceeded the 95% confidence limits of the normal ranges described here. It is possible that prostatic volume is not as useful a measure of prostatic pathology as previously thought. Interestingly, in 2 of the 4 dogs with adenocarcinoma, the lobes of the prostate were asymmetrical, perhaps validating our proposition that lobe volume should be measured separately.
There are many reports of the ultrasonographic appearance of different prostatic diseases and yet to date no attempts to analyse whether ultrasonographic appearance can be useful to differentiate different pathological processes. In this study, we undertook a structured thematic analysis of the descriptive ultrasonographic reports from normal dogs. Clearly, whilst a qualitative thematic approach has the ability to generate rich descriptive data that can be evaluated in the context that they are reported, it can result in difficulty assessing links and associations that occur between observations. In the present study, we identified three thematic categories used for describing the normal prostate and four thematic categories used for describing prostatic pathology. Interestingly, there were no thematic categories relating to symmetry, margination or peri-glandular changes. This probably represents the fact that such changes were not present in the cases examined, but perhaps suggests that, methodologically, thematic analysis might be improved by providing the ultrasonographers with the thematic categories required, and then examining the data for the thematic variables (descriptors).
In the present study whilst four thematic categories were identified, it was interesting that there was poor relation between the thematic variables (descriptors) and the underlying pathology; only the term ‘cysts’ was associated with benign prostatic hyperplasia and ‘parenchymal mineralization’ with adenocarcinoma. These findings support previous observations that the ultrasonographic appearance (descriptions) relates relatively poorly to the underlying disease process (Mattoon and Nyland 2002; Vignoli et al. 2011) and confirms the diagnostic value of parenchyma mineralization for detecting prostatic neoplasia as has previously been reported (Bradbury et al. 2009).
Whilst measurement of prostatic volume and the ultrasonographic appearance did not appear to be particularly useful for differentiating prostatic pathologies, the measurement of vascular perfusion kinetics using contrast-enhanced ultrasound illustrates some exciting possibilities. There was a clear increase in both the percentage peak intensity and time to peak in dogs with malignant prostatic disease compared with dogs with benign disease and normal dogs. Further studies into vascular perfusion kinetics seem warranted.
The continued advances made in diagnostic imaging technology have led to substantial improvement in disease investigation, and it appears that particularly contrast-enhanced ultrasound imaging will significantly improve our ability to diagnose prostatic pathology in dogs.
Conflicts of interest
None of the authors have any financial or personal relationships with other people or organizations that could inappropriately influence or bias the content of this work. This study received no funding from external bodies.