COMPARISON OF ULTRASONOGRAPHIC AND RADIOGRAPHIC

MEASUREMENTS OF CANINE PROSTATE DIMENSIONS

ATALAN,
GULTEKIN DR. VET.MED.,FRANCES
J. BARR,MA, VETMB,PHD,MRCVS,
E. HOLT,BVMS, PHD,CBIOL,MIBIoL,FRCVS
PETER

The purposes of this study were to: 1) compare prostatic dimensions measured on radiographic and
ultrasonographic images and 2) compare a subjective radiographic assessment of prostate size with a
previously-described objective method. Thirty-four male dogs undergoing investigation of prostatic
disorders were used. Prostate length and depth were measured from ultrasonographic and radiographic
images. A subjective assessment of prostate size (‘small’, ‘normal’, or ‘enlarged’) was made in 29
animals by one of the authors who was unaware of radiographic or ultrasonographic measurements. In
addition, the distance from sacral promontory to the pubic brim was also measured. A prostate length
or depth of >70% of this distance was defined as ‘enlarged’ and ~ 7 0 %
as ‘normal’. After the effects of
magnification on radiographic measurements were eliminated, there were no significant differences
between prostatic length measured by the two methods. However, a significant difference was obtained
between prostatic depth measurements. The subjective assessment agreed with a previously described
objective assessment of prostatic size in 21/29 dogs for prostate length but in only 12/29 for depth.
Prostatic length varied from 46.6 to 116.4% (mean 75.7%) of the distance from the pubic brim to the
sacral promontory. Prostatic depth varied from 33.0% to 94.6% (mean 59.7%) of the same distance. It
is recommended that prostate length, rather than depth, be used when evaluating prostate size from
lateral abdominal radiographs. Veterinary Radiology & Ultrasound, Vol. 40, No. 4, 1999, p p 408412.

Key words: dog, prostate, measurements, radiography, ultrasound.

Introduction cilitates the differentiation of radiographically identifiable

P ROSTATE DISEASE IS a common problem in older, sexu-
ally-intact male dogs. 132 Diseases associated with the
most severe prostatomegaly are cysts, abscessation and neo-
p~asia.~
Radiographically, the prostate gland is of soft-tissue
opacity and its identification is influenced by the differential
subject opacity of surrounding tissue^.^ It is normally in-
trapelvic and may, therefore, be difficult to see radiographi-
cally.s However, generalized enlargement of the prostate
gland produces cranial and possibly dorsal displacement of
the urinary bladder.5
Ultrasonographically, the prostatic parenchyma is nor-
mally moderately echoic with a coarse but even texture
throughout.6 Linear echogenic streaks may be detected run-
ning longitudinally through the middle of the prostate, rep-
resenting peri-urethral fibrous tissue.‘ Ultrasonography fa-
From the Division of Companion Animals, Department of Clinical Vet-
erinary Science, University of Bristol, Langford House, Langford, Bristol
BS40 5DU, U.K.
Address correspondence and reprint requests to Dr. P. E. Halt at the
above address.
Received April 21, 1998; accepted for publication October 29, 1998.
Mr. Atalan was supported by a grant from Higher Educational Coun-
cellors in Ankara and Kafkas University, Kars, Turkey.
prostatomegaly attributable to abscessation or neoplasia
from other causes of pro~tatomegaly.~
Radiographic techniques are useful in determining the
size, shape and location of the prostate gland. However,
radiographically-visualized lesions are nonspecific; there is
no indication of whether the changes are due to hyperplasia,
infection or neoplasia. Abdominal radiography allows a
limited assessment of prostatic d i ~ e a s e . ~
Radiographic assessment of prostatic dimensions have
been reported previously. The prostate gland is said to be
large if the prostatic dimensions exceed 70% of the pubic
brim-sacral promontory dimension in survey lateral radiog-
raphy.* The size of the normal prostate gland has been
defined as not exceeding 50% of the width of the pelvic
inlet on a ventrodorsal radiograph.9 Furthermore, it has been
suggested that prostatic enlargement on lateral radiographs
of greater than 90% of the pubic brim-sacral promontory
distance is suggestive of neoplasia, abscessation or parapro-
static cysts.*
Estimation of canine prostate gland size may provide im-
portant information in the clinical evaluation of benign
prostatic disease and in monitoring the response to medical
therapy. Work presented elsewhere” suggests that ultraso-
nographic measurements can be used to calculate prostate

408
VOL. 40, No. 4 ULTRASONOGRAPHIC
VERSUSRADIOGRAPHIC
PROSTATE
MEASUREMENTS 409

sured using a ruler (Fig. 2). In addition, the distance from

the sacral promontory to the pubic brim was also measured
(Fig. 3). To eliminate the effect of magnification error on
measurements, the following distances were also measured:

a) Focal film distance (always 100 cm)

b) The distance from the cassette to the table top (always 11
cm)
c) The distance from the mid-perineum of the animal to the
table top in the lateral position.

To allow for magnification of the image during radiogra-

phy, the following formula was used to calculate the actual
prostatic dimensions:
FIG. 1. Sagittal sonographic image of an adult canine prostate, illustrat- 100 - (b i-
C)
ing where the length and depth measurements were obtained. The trans- (radographic
ducer was placed on the ventral abdominal wall, just cranial to the pubic Actual prostate dimension = 100 prostate dimension)
brim and to one side of the penis.
Length (L): + to +
Depth (D): x to x This formula was also used to calculate the actual distance
from the pubic brim to the sacral promontory.
size. However, ultrasonographic equipment may not readily Diagnostic Value of Radiographic Measurements
be available in general practice and an alternative might be
to use radiographic measurements. One of the authors (FJB) examined the radiographs and
Therefore, the aims of this study were to: 1) compare classified the prostate as small, normal or enlarged, without
prostatic dimensions measured on radiographic and ultraso- knowledge of the above measurements. This classification
nographic images and 2) assess the value of comparison of was subjective and was based on the position of the prostate
prostate dimensions with the distance between the sacral within the pelvis or abdomen, and it's relation to pelvic inlet
promontory and the pubic brim in the diagnosis of normal or size. The findings of this subjective assessment were then
enlarged prostate glands. compared with radiographic measurements. A prostate di-
mension (length or depth) of greater than 70% of the sacral
Materials and Methods
This study was carried out on 34 male dogs referred to the
University of Bristol, Department of Clinical Veterinary
Science for investigation of a prostatic disorder. There was
a wide variety of breeds, and body weight ranged from 11.3
to 86 kg (median 32.4 kg) and in age from 4.0 to 15.0 years
(mean 8.1 years). Five dogs were neutered and the remain-
der intact.
Ultrasonographic Measurements of Prostate Dimensions
The protocol for the ultrasonographic imaging and mea-
surement of prostate dimensions was described elsewhere.''
The prostate was always imaged through the caudoventral
abdominal skin, usually with the dog in dorsal recumbency
(26 dogs). Eight dogs which were difficult to maintain in
dorsal recumbency were imaged in lateral recumbency. A
standard sagittal image was obtained. From the image,
length and depth were measured (Fig. 1).
Radiographic Assessment of Prostatic Dimensions FIG. 2. Lateral caudal abdominal radiograph of an adult male dog illus-
trating where prostate measurements were made. This radiograph shows
A survey caudal abdominal radiograph was made in each prostatic enlargement, with some small areas of mineralisation. Cystic
calculi are also present.
dog in right lateral recumbency. Craniocaudal (length) and Length: - - - - - - -
ventrodorsal (depth) dimensions of the prostate were mea- Depth - - - - -
410 ATALANET AL. 1999

TABLE1. Summary of Ultrasonographic and

Radiographic Measurements"
Ultrasonographic Radiographic
Dimensions Measurements Measurements
Prostate length (cm) 5.4 k 1.30 5.2 k 1.64
(3.5-8.3) (2.8-8.6)
Prostate depth (cm) 4.3 k 1.11 4.5 k 0.99
(2.4-7.0) (2.5-7.1 )
*All are expressed as mean f SD (range).

Comparison of Ultrasonographic and

FIG. 3. Lateral caudal abdominal/pelvic radiograph of an adult male
dog, illustrating the position of the pubic brim-sacral promontory measure-
ment (arrow).
promontory-pubic brim distance was defined as enlarged.8
Below 70% was defined as normal.
Statistical Analyses
The data were first examined for normal distribution. All
data were distributed normally.
The differences between lengths and depths measured
using the two methods were analysed for statistical signifi-
cance by means of the paired Student's t-test using a Com-
puting system." Prostatic length and depth measured on the
survey radiographs were expressed as a percentage of the
distance between the sacral promontory and the pubic brim.
Spearman's rank correlation was used to examine the
relationship between the radiologist's assessment and the
radiographic measurements.
Results
Ultrasonographic Measurements of Prostate Dimensions
The ultrasonographic prostatic measurements varied from
3.5 to 8.3 cm (mean 5.4 cm) for length and from 2.4 to 7.0
cm (mean 4.3 cm) for depth (Table 1). There were signifi-
cant relationships between radiographic pubic brim-sacral
promontory distance and prostate length (r = 0.394; p <
0.05) and depth (r = 0.399; p < 0.01) measured ultrasono-
graphically.
Radiographic Assessment of Prostatic Dimensions
Radiographic prostate length varied from 2.8 to 8.6 cm
(mean 5.2 cm) and depth from 2.5 to 7.1 cm (mean 4.5 cm)
(Table I).
There were significant relationships between pubic brim-
sacral promontory distance and prostate length (r = 0.446;
p < 0.02) and prostate depth (r = 0.552; p < 0.005).
"Minitab, Pennsylvania State University, State College, Pennsylvania.
Radiographic Measurements
There was a significant correlation between prostatic
lengths (r = 0.847; p < 0.001) measured by the two meth-
ods and the t-test showed no significant difference between
radiographic and ultrasonographic lengths (p > 0.05). Al-
though ultrasonographic and radiographic prostate depths
were significantly correlated (r = 0.830; p < O.OOl), there
was a significant difference between the actual measure-
ments (p < 0.05) (Table 2).
Diagnostic Value of Radiographic Measurements
On the subjective radiographic assessment, in 21 dogs the
prostate was classified as enlarged, in 10 as normal, in one
as small and in two the prostate could not be visualised.
Radiographic measurements could not be obtained from the
prostate classified as small, in the two dogs where the pros-
tate was not seen, or in two other dogs (one normal, one
enlarged prostate) where the prostate margin was indistinct.
When results from these five dogs were excluded from the
analysis, there remained 29 animals with complete data. In
these dogs, prostatic length varied from 46.6 to 116.4%
(mean 75.7%) of the distance from the pubic brim to the
sacral promontory and prostatic depth varied from 33.0% to
94.6% (mean 59.7%) of the same distance.
In the 29 dogs, a previously published objective defini-
tion of prostatic enlargement* agreed with our subjective
assessment in 21/29, when prostate length was the objective
measurement used (r = 0.397; p < 0.05). However, there
was no relationship between subjective assessment and ob-
jective enlargement when prostate depth was used (r =
0.030; p > 0.05); there was agreement in only 12/29 dogs.
When length and depth measurements (whichever was the
2. Results of Paired Student's T-test between Ultrasonographic
TABLE
and Radiographic Prostate Measurements
Difference Between
Ultrasonographic
Measurements and Radiographic
(cm) n Measurements" P
Prostate length 24 0.09 c 0.88 0.61
(-1.5, 1.4)
Prostate depth 32 0.26 c 0.62 0.023
(-1.8,0.8)
"Values expressed as mean ? SD (range)
VOL. 40, No. 4 ULTRASONOGRAPHIC
VERSUSRADIOGRAPHIC
PROSTATE
MEASUREMENTS
greater % of the sacral promontory-pubic brim distance)
were used, there was agreement with the subjective assess-
ment in 21/29 dogs (r = 0.356; p < 0.05).
Discussion
The canine prostate gland has been measured previously
both grossly, radi~graphically'~~~" and ultrasonographi-
cally~7,10.12, I 3 Transabdominal ultrasound has been found to
be a simple and quick method for estimating prostate size in
the dog.7.10,Ix i 4
In this study we found that, if radiographic magnification
error is eliminated, there is no difference between prostate
length measured by radiographic and ultrasonographic
methods. However, there was a significant difference be-
tween measurement of prostatic depth using these two
methods. This may be due to imprecise organ boundaries,
particularly on radiographs, or the presence of adjacent vis-
cera, especially the colon, which may obscure the dorsal
radiographic border of the p r o ~ t a t eA
. ~ sonographically in-
distinct dorsal and ventral prostate margin has also been
reported, likely because these margins have similar acoustic
impedances to other tissues.I2 Radiographically, prostatic
outline may also be affected by disease processes. For ex-
ample, prostatic inflammation or neoplasia with extension
into surrounding tissues may lead to indistinct prostatic
margins on abdominal radiographs.
Radiographic assessment of prostatic dimensions has
been performed in normal male dogs.* Prostate depth and
length were measured from lateral radiographs and ex-
pressed as a percentage of the distance between the cranial
aspect of the pubic bone and the sacral promontory. The
mean percentage was 56% in 24 normal dogs. However,
prostatic dimensions (length and depth) exceeding 70%
were considered to indicate prostatomegaly in 30 dogs with
various prostatic diseases. When this objective assessment
of prostatic enlargement' was compared with the subjective
assessment of an experienced radiographer, there was good
agreement when length was used as the objective measure-
ment but not when depth was used. The combination of
length and depth did not significantly improve the level of
agreement. This suggests that prostatic enlargement does
not occur uniformly but results in greater increases in length
REFERENCES
1. Hornbuckle WE, Mackoy DM, Allan GS, Gunther R. Prostatic dis-
ease in the dog. Cornell Vet 1978;68:284-305,
2. Greiner TP, Johonson RG. Diseases of the prostate gland. In: Text-
book of Veterinary Internal Medicine (ed. 2) Ettinger, SJ. (ed) Philadel-
phia, Saunders. 1983; pp, 1459-1493.
3 . Stone EA, Thrall DE, Barber DL. Radiographic interpretation
of prostatic disease in the dog. J Am Anim Hosp Assoc 1978;14:113-
118.
4. Johnston GR, Feeney DA, Rivers B, Walter PA. Diagnostic imaging
of the male canine reproductive organs. Vet Clin North Am Small Anim
Pract 1991;21:553-589.
41 1
than depth. This is difficult to explain although there may be
less resistance to longitudinal enlargement than that pro-
vided in the dorsoventral plane by adjacent structures such
as the pubis and rectum. Conversely, longitudinal shrinkage
when prostate size decreases after castration may explain
why the bladder neck is more caudally positioned in such
animals. l 5
There was no 'gold standard' of definitive prostatome-
galy or normality in the dogs in this series. Physical mea-
surement is the only real gold standard but could not be
justified in these clinical cases since it would involve, at the
least, surgical if not post mortem intervention. However, the
results suggest that radiographic prostatic depth is an unre-
liable estimate of prostate size. This supports the findings of
other work from our laboratory.'"
In the evaluation of prostate size, radiographic methods
may be expensive and pose exposure hazards to both animal
and examiner. Ultrasonographic methods are non-invasive,
simple and quick procedures for the assessment of prostate
0.14 but the equipment required may not readily be
available in practice. Feeeney et al.' considered that pros-
tatic size can be evaluated better with ultrasonography than
with radiography. The prostate can be readily visualised
ultrasonographically since there are few other structures be-
tween the abdominal skin and it, and since the urinary blad-
der can be used as a landmark. Prostatic consistency can
also be evaluated better using ultrasonography than radiog-
r a p h ~ ~since
, ' radiography cannot differentiate between dif-
ferent soft tissues and fluid. However, this study reveals that
radiographic assessment of prostate length may be as useful
as ultrasonographic prostate length measurement since there
was good correlation between length measured by the two
methods.
In conclusion, there is a significant difference in prostatic
depth measurements between ultrasonography and radiog-
raphy and depth appears to be a poor indicator of prostate
size. If a previously published radiographic method of pros-
tate measurement8 is used to assess prostate size, it is rec-
ommended that prostate length, rather than depth, is com-
pared with the pubic brim-sacral promontory distance. This
measurement appears to be reliable, whether measured ra-
diographically or ultrasonographically.
5. Root CH. Prostate gland. Thrall DE (ed). Textbook of Veterinary
Diagnostic Radiology. W.B. Saunders Company, Philadelphia. 1986; pp,
375-378.
6 . Barr F. Diagnostic Ultrasound in the Dog and Cat. Blackwell Sci-
entific Publications, Oxford. 1990; pp. 65-77.
7. Feeney DA, Johnston GR, Klausner, Perman V, Leinenger JR, Tom-
linson MJ. Canine prostatic disease: Comparison of ultrasonographic ap-
pearance with morphologic and microbiologic findings. J Am Vet Med
Assoc 1987;190:1027-1034.
8. Feeney DA, Johnston GR, Klausner JS, Perman V, Leininger JR,
Tomlinson MJ. Canine prostatic disease: Comparison of radiographic ap-
412 ATALAN
ET
pearance with morphologic and microbiologic findings. J Am Vet Med
Assoc 1987;190: 1018-1026.
9. Lattimer JC. The prostate. In: Textbook of Veterinary Diagnostic
Radiology. Thrall, DE. (ed) Philadelphia. W.B. Saunders 1986; pp. 444-
457.
10. Atalan G, Barr FJ, Holt P, Brown PJ. Ultrasonographic estimation of
prostatic size in canine cadavers. Res Vet Sci (Accepted for publication).
11. O’Shea JP. Studies on the canine prostate gland. Factors influencing
its size and weight. J Comp Pathol 1962;72:321-331.
12. Cartee RE, Rowles T. Transabdominal sonographic evaluation of
the canine prostate. Vet Radio1 1983;24:156-164.
AL. 1999
13. Diquklou A, Razaire 0, Guelfi JF, Concordet D, Berthelot X. Vol-
ume measurement of the canine prostate using ultrasonography: determi-
nation and usual values. In: WSAVA, BSAVA & FECAVA WORLD
CONGRESS-SYNOPSES. 1997; p. 242.
14. Atalan G, Holt PE, Barr F. Ultrasonographic estimation of prostate
size in normal dogs and its relationship to body weight and age. J. Small
Anim. Pract. (accepted for publication).
15. Power SC, Eggleton KE, Aaron AJ, Holt PE, Cripps PJ. Urethral
sphincter mechanism incompetence in the male dog: importance of bladder
neck position, proximal urethral length and castration. J Small Anim Pract
1998;39:69-72.