The Role of Transabdominal Ultrasound in the Prostatic Hypertrophy

0022-534 7/86/1356-1190$02.00/0 Vol. 135, June Printed in U.S.A.

THE JOURNAL OF UROLOGY

Copyright© 1986 by The Williams & Wilkins Co.

THE ROLE OF TRANSABDOMINAL ULTRASOUND IN THE PREOPERATIVE EVALUATION OF PATIENTS WITH BENIGN PROSTATIC HYPERTROPHY CLAUS G. ROEHRBORN, HERBERT K. W. CHINN, PAT F. FULGHAM, KENNY L. SIMPKINS PAUL C. PETERS

AND

From the University of Texas Health Science Center at Dallas and Veterans Administration Medical Center, Dallas, Texas

ABSTRACT

We evaluated 59 consecutive patients for obstructive voiding symptoms with physical examination, excretory urography, urethrocystoscopy, post-voiding residual volumes, uroflowmetry and transabdominal ultrasound of the prostate. Of the patients 53 underwent transurethral resection of the prostate and 6 underwent open prostatectomy. Followup uroflowmetry was done 4 weeks postoperatively. The best predictor of the actual prostatic weight was transabdominal ultrasound (r equals 0.975), and with digital rectal examination and urethrocystoscopy there was a tendency to overestimate small and underestimate large glands. Symptoms such as nocturia do not allow any conclusions about the size of the prostate. The correlation between post-void residual and specimen weight also is poor and only patients in retention (Foley catheter) have a significantly larger prostate. Preoperative uroflowmetry provides limited information about the prostatic size but, although the difference between the preoperative and postoperative flow rate index is highly significant (p less than 0.001, mean difference test), the correlation between specimen weight, and the difference between preoperative and postoperative flow rate index is not significant. We recommend transabdominal sonography of the prostate, a simple and noninvasive procedure, as a useful adjunct in the preoperative evaluation of patients with benign prostatic hypertrophy. In addition to the fact that only this method provides reliable and reproducible estimations of the weight of the prostate, further information, such as measurement of the post-void residual volume, can be obtained without additional costs or loss of time. An important step in the evaluation of patients with obstructive voiding symptoms is to determine whether the prostate gland is enlarged. Accurate assessment of the size of the prostate is important to avoid difficulties during suprapubic or retropubic prostatectomy, when the gland is smaller than expected and, on the other hand, to minimize the risk of complications during prolonged transurethral resection, when the gland is larger than expected. 1 Digital rectal examination and urethrocystoscopy can provide erroneous results in predicting the size of the prostate. 2 Measurement of the weight by means of retrograde urethrography has an error of about 20 per cent. 3 Since nocturia is one of the earliest symptoms of prostatism, one might expect that the evidence of this and other symptoms gives somes clues about the size of the prostate. Likewise, the amount of post-voiding residual urine and the result ofuroflowmetry could be expected to allow a rough estimation of the prostatic weight. A correlation between symptom and obstructive score, and the weight of the surgical specimens has been described recently.4 The interest of urologists in ultrasound of the prostate has been focused mainly on whether early cancer could be detected. A recent review article indicated that transrectal sonography might be helpful to detect subclinical stage A cancer but at the same time it concluded that this currently is not practical without a mass screening program,5 as proposed previously by others. 6 The use of transabdominal sonography of the prostate to estimate the prostatic size also has been described previously7-9 and was found to be reasonably accurate. Sonography also has the advantages of not requiring any preparation and being Accepted for publication January 20, 1986. Read at annual meeting of American Urological Association, Atlanta, Georgia, May 12-16, 1985.

noninvasive. We determined the accuracy of different modalities to predict the weight of the prostate gland in patients scheduled for an operation. MATERIALS AND METHODS

A total of 59 consecutive nonrandomized patients underwent a standard preoperative evaluation, including history and physical examination with digital rectal examination, urethrocystoscopy, excretory urography, uroflowmetry (23) and measurement of the post-voiding residual by catheterization, and transabdominal ultrasound of the prostate. The sbnographer was not given information about the cystoscopic or digital rectal estimation of the prostatic size. Uroflowmetry was obtained more than 4 weeks postoperatively. Sonography was done with a real-time sector scanner (Diasonics DS) and a 3.5 MHz. trarisducer. The patient was asked not to urinate for 3 hours before the study and then was placed in the supine position on the examination table. The transducer was placed in a suprapubic caudally angled position, which is necessary to visualize the prostate under the pubic bone. The surgical capsule of the gland served as the limit for measurement and 3 diameters (breadth, height and length) were measured with transverse and longitudinal scans (fig. 1). For the calculation of the weight the formula for the ellipsoid was applied (weight = specific gravity X % X 1r X rl x r2 X r3). The specific gravity of the prostate is 1.05 gm./cc9 and% X 1r = 4.4. Therefore, the formula can be simplified to weight = 0.55 X dl X d2 X d3, where dl to d3 equals diameters 1 to 3. It is well known that there is a weight loss of prostatic tissue during transurethral prostatic resection. 1°For our purposes the formula was simplified further to weight = (dl X d2 X d3)/2, thereby taking a 10 per cent weight loss into account. Thus, the calculated weight represents the expected weight of the specimen after a complete transurethral prostatic resection.

1190

1191

FIG. 1. Positioning of transducer, and sonographic images of bladder and prostate in transverse (A and C) and longitudinal (B and D) directions. Predicted and actual specimen breadths (B) were 6.8 and 6.7 cm., respectively, predicted and actual lengths (L) were 6.3 and 6.4 cm., respectively, and predicted and actual heights (H) were 3.9 and 3.7 cm., respectively. Reprinted with permission from Resnick, M. I. and Sanders, R. C.: Ultrasound in Urology. Baltimore: The Williams & Wilkins Co., p. 222, 1979.

100-

-

"

90 -

Y=63+Q,06 r =0,222 p=0,09

weight via the mean difference test and the least square method of correlation. The relationship among uroflowmetry, improvement of the flow rate and prostatic size was studied with the flow rate index 11 for this purpose. The flow rate index reduces peak flow rate, flow time and volume to a single figure, and correlates highly with peak and mean flow rates, thereby obviating the need for volume adjustment before different flow rates are compared. This method allows easier statistical evaluation with the mean difference tesL All clinical tests and the operation were done resident staff with different levels of training. All tests and correlations were done on a Commodore computer with the statistical program DIANA .

·X

"

,_

If)

0 Cl)

:::,,..

.

(l)

<(

.,

50

" " 10

" @

". " "

O>

....."'

RESULTS

50

100

140

Specimen Weight (gms) FIG. 2. Correlation between specimen weight and patient age

No patient had to be eliminated because of unsatisfactory results of the ultrasonic evaluation. Correlations between the weight of the surgical specimen and the estimated weight by rectal examination, urethrocystoscopy and sonography were done with the least square method. Furthermore, we investigated whether the frequency of nocturia, patient age or the amount of post-void residual correlated with the specimen

Patient age ranged from 50 to 92 years, with a mean age of 65.3 years. There was no significant correlation between patient age and the weight of the specimen (r = 0.222, p = 0.09) (fig. 2).

The correlation between the post-void residual and the specimen weight was negative and poor only (r = -0.138). When the mean weights of the specimens were compared in 7 groups of patients with less than 20 to more than 300 cc post-void residuals no statistical difference was found (mean difference test). However, the group with indwelling Foley catheters had a significantly higher specimen weight (p <0.05) (fig. 3, A). The correlation between preoperative flow rate index and specimen weight was negative (r = -0.46) and significant (p =

1192

ROEHRBORN AND ASSOCIATES

150

150

y=47.4-32,6·X r =-0.46

p=0.005

CJ)

E 125

*

Cl

-;;, E .:?: 100 •

E 100 Cl

a.>

:s:

75

I= Flowrote - Index before surgery F = Foley Catheter (pot.in retention)

C

a.>

E

50

(.)

a.>

a.

25

(/)

0

I <20

I

II

C
I

E

I

·u

I


i

0.

en

50-100 150-200 >300 20-50 100-150 200-300 Foley

A

PVR (cc)

50 •

: . ~ ... . ·

.. .. ... ..-~ .... . ·.

B

0 '-'--'--'--'--'--'--'--'--'--'---'--'---'---'-- I F

.1 .2 .3 .4 .5 ,6 ,7 ,8 .9 1,0 1,1 1,21.3

FIG. 3. A, means and standard deviations of specimen weight in 8 groups of patients with increasing post-void residuals. Group 8 represents patients in acute urinary retention. Only this group had significantly higher specimen weights (mean difference test). Asterisk indicates p <0.05. B, negative correlation between preoperative flow rate index (I) in 25 patients and specimen weight. Nine patients in acute urinary retention with Foley catheter (F) had significantly larger glands.

Mean, standard deviation and range of estimated weight by rectal examination, urethrocystoscopy and sonography, and weight of surgical specimens Mean Gm.± Standard Deviation (Range) Rectal examination Cystoscopy Sonography Specimen wt.

29.11 31.96 38.40 34.31

± 11.76 (10.0-50.0) ± 14.84 (15.0-100.0)

± 28.38 (10.0-130.0) ± 26.89 (1.0-130.0)

0.05) (fig. 3, B). Although the difference between the preoperative and postoperative flow rate indexes (0.53 versus 1.36) was highly significant (p <0.001) there was no significant correlation between the improvement of the flow rate (measured as the difference of postoperative minus preoperative flow rate index) and the specimen weight. Nevertheless, the correlation coefficient, r = 0.53, indicated that the improvement of the flow rate correlated with the amount of removed tissue. The mean difference test showed no difference in the weight of the surgical specimen in 8 groups of patients without nocturia or with 1 to more than 6 episodes of nocturia per night. Results of the weight estimation by all 3 methods were compared to the specimen weight (see table). Figure 4 shows the correlation among these parameters. Rectal examination and cystoscopy tend to overestimate small and underestimate large glands, as demonstrated by comparison with the ideal (dotted) regression line in figure 4, A and B. The sonographic estimation provided the best results (r = 0.975, p <0.001) and all values are within the residual variance lines (fig. 4, C). In 14 patients with a difference of more than 10 gm. between the sonographic prediction and the actual specimen weight sonography was repeated about 4 weeks postoperatively (in 1 patient transurethral resection was repeated after a similar interval). In all re-studied patients residual tissue was demonstrated sonographically. The sum of the weight of the first surgical specimen and repeated sonography (or the reoperative specimen) correlated well with the original prediction based on the preoperative sonography (r = 0.97). Three patients in whom the transurethral resection was believed to be incomplete by the attending staff or the surgeon were excluded from the statistical calculations. DISCUSSION

Expectedly, we found only weak correlations between patient age and obstructive or irritative symptoms, and the weight of the surgical specimen. Furthermore, it seems that only patients who are in urinary retention have a significantly larger gland,

whereas otherwise no correlation between post-void residual and prostatic size could be found. The preoperative flow rate apparently allows limited conclusions about the expected size of the prostate. While virtually all of the patients studied with this modality had an improved flow rate postoperatively, the degree of improvement did not correlate with the removed tissue weight. The most accurate prediction of the actual prostatic size was obtained by transabdominal sonography. If one would rely on sonography alone in the preoperative evaluation of patients with benign prostatic hypertrophy, undesirable intraoperative surprises (difficult enucleation of small glands or prolonged resection of large glands) could be avoided completely. On the other hand, the extremely good correlation between predicted weight and a surgical specimen permits the hypothesis that whenever the resected specimen weighs significantly less than predicted, the resection most likely was incomplete. Like other investigators8 we were unable to identify stage A cancer prospectively (3 patients in this series had stage A2 cancer). In summary, we recommend preoperative transabdominal sonography of the prostate highly because it requires no preparation except natural bladder filling, is noninvasive (compared to cystoscopy or transrectal/transurethral sonography), provides an accurate estimation of prostatic size, gives additional information (that is intravesical growth of the prostate, amount of post-void residuaP 2 and presence of bladder abnormalities, such as diverticula) without additional costs or loss of time and allows clear separation of patients requiring an open operation for relief of bladder neck obstruction. REFERENCES

1. Kidd, E. E. and Burnside, K.: Bacteraemia, septicaemia and intravascular haemolysis during transurethral resection of the prostate gland. Brit. J. Urol., 37: 551, 1965. 2. O'Flynn, J. D.: The management of simple prostatic hyperplasia. Brit. J. Hosp. Med., 2: 562, 1969. 3. Thumann, R. C., Jr.: Estimation of the weight of the hyperplastic prostate from the cystourethrogram. Amer. J. Roentgen., 65: 593, 1951. 4. Frimodt-M~ller, P. C., Jensen, K. M.-E., Iversen, P., Madsen, P. 0. and Bruskewitz, R. C.: Analysis of presenting symptoms in prostatism. J. Urol., 132: 272, 1984. 5. Peeling, W. B. and Griffiths, G. J.: Imaging of the prostate by ultrasound. J. Urol., 132: 217, 1984. 6. Watanabe, H., Saitoh, M., Mishina, T., Igari, D., Tanahashi, Y., Harada, K. and Hisamichi, S.: Mass screening program for prostatic diseases with transrectal ultrasonotomography. J. Urol., 117: 746, 1977.

1193

PREOPERATIVE ULTRASOUND IN BENIGN PROSTATIC HYPERTROPHY

A

130

C'

D
D

0

E 100

Cl'

..... .c

0,

.S!'

·w

Q)

5

<.> C!.>

a.

(/)

OiO

C

TURP Proslalec1omy : o Retropubic o Suprapubic

50

0

s:

.

C:

C!.>

r

= -5,9+ 1,38 · X = 0,604

r

1:

_!:;

y = -7,9+1,32 · = 0,730

130 y

100

B

0

D


E

Rectal Examination

50

100

(1)

E

50

u

(1)

a.

(/)

0

130

10

50

Rectal (gms)

100

130

Cysto (gms)

C

Sonogrophy

130 y=-1,1+0,92·x = 0,975

r If>

E

O>

100

.,._ .i:::. O>

"iii

5

C: (1)

E ·c3

50

(1)

a.

(/)

0-~~--~---~~-~~~~I~

0 10

50

100

130

Sono (qms) FIG. 4. Correlation between estimated weight of prostate by 3 different modalities and weight of specimen. Dotted line (A and B) indicates ideal regression line. Solid line indicates actual regression line. Sonographic estimation gives best result, all values being within residual variance lines ( C). 7. Miller, S.S., Garvie, W. H. H. and Christie, A. D.: The evaluation of prostate size by ultrasonic scanning: a preliminary report. Brit. J. UroL, 45: 187, 1973. 8. Henneberry, M., Carter, M. F. and Neiman, H. L.: Estimation of prostatic size by suprapubic ultrasonography. J. Urol., 121: 615, 1979. 9. Abu-Yousef, M. M. and Narayana, A. S.: Transabdominal ultrasound in the evaluation of prostate size. J. Clin. Ultrasound, l 0: 275, 1982.

10. Whisnand, J.M. and Moses, J. J.: Weight loss of the prostate gland during transurethral smgery, with a note on the effect of formalin on the fragments. J. Urol., 83: 718, 1960. 11. Riihrbom, C. and Rugendorff, E.W.: Uroflow-Index-Eine Formel zur Vereinfachung von Befundung und Vergleich von Uroflowmetriekurven fiir die Praxis. Z. Urol. Nephrol., 76: 721, 1983. 12. Roehrborn, C. G. and Peters, P. C.: Can transabdominal ultrasound estimation of postvoiding residual (PVR) replace catheterization? J. Urol., part 2, 133: 169A, 1985.