ANDROGEN DEPRIVATION PRIOR TO RADICAL PROSTATECTOMY FOR T2b AND T3 PROSTATE CANCER

MARK S. SOLOWAY, M.D. WILLIAM M. MURPHY, M.D TAKAHIKO HACHIYA, M.D. COSME C. GOMEZ, M.D.

FRANCISCO CIVANTOS, M.D. HENRY E. RUIZ, M.D.

From the Department of Urology, the Department of Pathology (EC.), University of Miami School of Medicine, Miami, Florida, and the Department of Pathology

(W.M.M.), Tulane University School of Medicine, New Orleans, Louisiana

ABSTRACT-Objective, In an effort to improve on the results of radical prostatectomy for clinical stages T2b and T3 prostate cancer, a selected group of patients received andro- gen deprivation for three to sixteen months prior to surgery.

Methods. Fifteen men with clinical T2b and 22 with small T3 tumors received a luteinizing hormone-releasing hormone analog (n = 34) or a bilateral orchiectomy (n = 3) three to sixteen months prior to radical retropubic prostatectomy. The prostate was evaluated with particular attention to tumor grade, presence of extracapsular extension, tumor at the inked margin, seminal vesicle invasion, and tumor in the lymph nodes.

Results. No patient had clinical or chemical (prostate-specific antigen [PSA]) progres- sion during androgen deprivation. The PSA level declined a mean 90 percent and re- mained above 4 ng/mL in only two patients. The prostate volume decreased an esti- mated 30-50 percent. Prostate cancer at the inked margin was found in 15 (41%) and seminal vesicle involvement in 11 (30%) patients. Five (14%) had tumor in regional lymph nodes. There was no difference in regard to positive margins or lymph node metas- tases between those clinically staged as T2b and those preoperatively staged as T3. Fourteen patients have received adjuvant therapy (13 androgen deprivation, one radia- tion therapy). None has progressed (mean follow-up, 38.4 months). Of 23 who did not receive immediate additional therapy, six (26%) had progression, as was evident from an increase in PSA and have since been treated. Only one continued to progress. Thirty-five of the 37 patients are alive. Seventeen (46%) are tumor free (PSA < 0.4 ng/mL) without further androgen deprivation.

Conclusions. Only a prospective randomized trial can determine whether androgen deprivation prior to radical prostatectomy has a role. The results from this trial are en- couraging for several reasons. The prostate is much smaller as a result of androgen dep- rivation and this may facilitate surgery. Although the great majority of these patients were expected to be margin positive, 60 percent had negative margins and only 14 per- cent had positive lymph nodes.

There has been a dramatic increase in the use of radical prostatectomy for the treatment of appar- ent organ-confined prostate cancer. The primary reasons for this include: (1) a dramatic decrease in the morbidity associated with radical prostatec- tomy, notably a reduction in the likelihood of uri- nary incontinence; (2) potential to preserve po- tency with preservation of the neurovascular bundles’,>; (3) disappointing results with the use

of radiation therapy and, in particular, the high incidence of positive biopsies obtained two to five years following external beam radiation therapy3; and (4) the increase in potentially curable patients discovered by the combined use of the serum prostate-specific antigen (PSA) digital rectal ex- amination (DRE), and transrectal ultrasonography

Accompanying this resurgence in radical prosta- tectomy is an increase in the number of cases in

52 SUPPLEMENT TO UROLOGY / F~;RKI’AK~ 194 / Vor IW 4.3, NIWU I( 2

which pathologic evaluation of the prostate indi- cates that the operation may not have been curative, i.e., capsular penetration and/or positive surgical margins. There is no consensus on the impact of lhcsc pathologic findings on survival nor on the optimal approach toward these patients. Among the reasons for the number of patients with a pos- itive margin are the inability to accurately stage prostate cancer preoperatively and the expanded indication for this procedure. Paulson ot ~1.’ indi- cated that the ten-year survival for those with a positive margin was 68 percent in contrast to 94 percent when the malignancy was organ confined.

In one series the incidence of positive margins in patients \vith clinical stage B disease was 9 per- cent, 2-t percent, and 40 percent in those who were clinicall>, staged as Bl nodule, Bl, and B2 tumor, respectively.’ The incidence of positive lymph nodes in these same categories was 9 per- cent, 4 percent, and 21 percent, respectively.

One approach is to consider using androgen deprivation prior to radical prostatectomy for pa- tients at high risk for having a positive margin, i.e., clinical stages T2b and T3a, with the hope of reducing the incidence of radical prostatectomy specimens with extracapsular extension, positive seminal vesicles, and/or nodal metastases.

The use of androgen deprivation prior to radical prostatcctomy is not a new concept. ‘In the late 1940s there were several encouraging reports of androgrn deprivation prior to radical prostatec- tomy in an attempt to reduce the size of both nor- mal and malignant prostates.h-” There arc various reasons why this approach to the patient with clinical stage T3 prostate cancer never gained wide acceptance. These reasons include side ef- fects associated with permanent androgen depri- vation, the morbidity associated with total prosta- tectomy in men who were not likely to be cured by this procedure, and the doubt that this ap- proach altered the course of the disease.

Probably the majority of clinicians believe that patients with small T3 prostate cancer should re- ceive either androgen deprivation alone or exter- nal beam radiation therapy. They are unlikely to lx cured by surgery. The rationale for considering total prostatectom)’ in this group of patients is that continued growth of the prostate in those not effecti\,ely treated by either androgen deprivation or radiation therapy may require treatment. If transurethral prostatectomy becomes necessary, the likelihood of urinary incontinence is high. If these patients fail their initial form of treatment, hematuria and ureteral obstruction are additional causes of morbidity that can be obviated if the

prostate is removed. In addition, c,omc of those who are clinically staged as T3a ma.y be cured by total prostatectomy (pathologic stage, -1‘2, N0).4 The morbidity of total prostatectomy is now ac- ceptable and, importantly, the psvchologic benefit is not insignificant.

One of the reasons for the resurgence in the in- terest of the use of androgen depri-Gatron prior to radical prostatectomy was the introduction of a reversible form of androgen deprivation, notably luteinizing hormone-releasing hormone (LfIRH) analogs. Approximately eight years ago, one of the authors (M.S.S.) initiated the USC c>f androgen deprivation prior to radical prostatectomy in those patients with clinical stage T2b or T-‘I prostate cancer who were thought likely to have capsular penetration, positive seminal vesicles, and/or posi- tive lymph nodes. This report details this experi- ence with this select group of patient>.

METHC>DS

Thirty-seven patients with clinic,~l stage T2b (palpable induration involving > 1 .‘i cm of one lobe of the prostate, but thought to be confined to the prostate) or T3a (palpable extracapsular ex- tension) were selected. Fifteen cac.es were clinical stage T2b and 22 cases, T3. The prostate was mo- bile. The ages of the patients ranged from forty- eight to seventy-seven years old, with a mean of sixty-four years. The first patient was treated in 1986 and the last in 1992.

Each patient underwent a standard evaluation that included a careful history, physical examination. serum PSA, prostate biopsy, and a raclionuclide bone scan. Most, but not all, patients had a computer- ized tomogram or magnetic resonance imaging (MRI) scan of the pelvis. No patient received prior radiation therapy or hormonal therapy. Digital rec- tal examination was an integral part of the preoper- ative staging. The serum PSA was obtained preop- cratively in 26 patients and a radioimmunoassay for prostatic acid phosphatase (PAP) m 12 patients. A PSA was obtained following anclrogen depriva- tion and prior to surgery in 19 of the 37 patients.

The duration of androgen deprivdtion prior to radical prostatectomy varied from thrrc to sixteen months with 31 patients undergolng induction therapy for three to six months. Three patients re- ceived androgen deprivation fo’r more than one year prior to radical prostatectomy. An LHRH ana- log alone was administered to 15 individuals; 19 received an LHRH analog with flutamide. Three patients underwent bilateral orchirctomy instead of chemical castration, since they wcrc not potent,

and it was decided that androgen deprivation would be indefinite. Flutamide was given to all patients once it received Food and Drug Adminis- tration approval.

A bilateral modified pelvic lymphadenectomy and standard retropubic prostatectomy were per- formed without an attempt at nerve-sparing modi- fication in each of the 37 patients. The prostatec- tomy specimen was evaluated with respect to tumor grade, pathologic stage, seminal vesicle in- vasion, and lymph node metastasis. Ploidy was not determined. Careful attention was directed to tumor at the inked margin. The location of tumor was marked on a form that became a part of the pathology report. The follow-up ranged from four to eighty-one months following surgery, with a mean of thirty-three months. Follow-up consisted of a PSA and digital rectal examination at six- month intervals. A postoperative PSA level more than 0.4 ng/mL was considered indicative of re- current prostate cancer.

RESULTS

No patient progressed during the period of an- drogen deprivation as determined by the DRE or PSA. Twenty-five of 26 (96%) who had a PSA prior to androgen deprivation had a pretreatment PSA 4.0 ng/mL or higher, with a mean of 33.7 ng/mL (4.6-141.9). Twenty of the 26 (76.9%) had a PSA level 10.0 ng/mL or higher.

Analysis of the results by separating the patients into clinical stage T2b and T3 did not reveal any difference between the two groups in terms of their pathologic stage and thus for the purpose of this report they will be combined.

As expected, the PSA levels dramatically de- creased following androgen deprivation. In the 19 incidences in which both the pre- and post-depri- vation levels were available, the post-deprivation levels ranged from 0.1-8.0 ng/mL, which repre- sents a 75-99.9 percent decline (mean 89.9%). Only two patients’ PSA levels were still more than 4.0 ng/mL following induction therapy. Three pa- tients had an initial PSA above 100 ng/mL and in all cases there was a dramatic decline following induction therapy. Turelve patients had a pretreat- ment PAP available and this decreased an average of 85 percent in the nine cases in which this infor- mation was available.

Although a transrectal ultrasound (TRUS) was performed prior to induction therapy in all pa- tients, it was not uniformly repeated and thus pre- cise data of the reduction of the size of the prostate is not available. Based on the seven in- stances in which the TRUS was performed before

and after induction therapy, as well as the estima- tion of the size of the prostate based on the DRE, it is estimated that the prostate diminished 30-50 percent during the period of induction androgen deprivation therapy.

The radical prostatectomy was not significantly altered by the induction therapy, In some instances there was the advantage of performing a total pros- tatectomy on a smaller prostate than would have been present prior to androgen deprivation. The operative mortality was zero. There were no post- operative complications. One patient has signifi- cant stress urinary incontinence five months post- operatively and requires four pads per day.

The initial prostate biopsy was well differenti- ated in 11 patients, moderately differentiated in 22, and poorly differentiated in 4. Since the initial tumor grade is based on the samples obtained by needle biopsy, the predominant grade in the radi- cal prostatectomy specimen provides superior prognostic information and correlates better with pathologic stage. Androgen deprivation often has a dramatic effect on the prostate gland and the pathologist may misinterpret the tumor grade if not familiar with these alterations.‘” The primary tumor grade in the prostatectomy specimen was well differentiated in 6 patients, moderate in 22, and 9 had a poorly differentiated carcinoma. In one case only a few tumor cells remained.

The definition of a “positive margin” was tumor at the ink. Fifteen (41%) of the 37 cases had tumor at the inked margin. An additional 8 (22%) had extracapsular extension but there was no tumor at the inked margin. Seminal vesicle involvement was observed in 11 (30%). Only 5 (14%) of the pa- tients had metastasis to regional lymph nodes. The inaccuracy of clinical staging was evident in that a positive margin was observed in 6 of 15 (40%) who were clinical stage T2b and 9 of 22 (41%) who were clinical stage T3.

There were 15 cases with one or more positive margins; nine were located at the apex, 12 were at the lateral margin, and only four at the bladder neck. In no instance was the bladder neck the only location of a positive margin.

ADJWANT THERAPY Fourteen of 37 (38%) received adjuvant ther-

apy: 3 clinical stage T2b and 11 T3 patients. An- drogen deprivation was the form of additional treatment in 13 patients, which included the 3 tnen who underwent bilateral orchiectomy prior to radical prostatectomy. One patient received radiation therapy because of a positive margin.

54 SUPPLEMENT TO UROLOGY / F~I~KI’,\R). IO’)+ / vor I ,\JI 4.1. .vi \JRI 18 2

“cone 01 the 1-f patients who ruxivccl acljuvant 1 herapy has prngresd. with a hollow-up ranging tram sis to sc\:tnt),-five months (mean. 38.-f months). All have a PSA below 0.4 ng/mL. One patient died f’rom a myocardial infarction thirty- three months following surgery kvithout cvidcncc ol rccurrcncc.

Sis of 23 palicnts (26%) who did not rcccive ad- ju\,ant therap!, progressed. Three of 15 (20%) with a positive margin progressed (2 with 311 clc-

vated PSA only) within seven months. Two ol these 3 patients had seminal vesicle involvement and a poorly differentiated pattern; none had posi- ti1.t lymph nocles. One died of metastatic prostatc cancer fort!~-fi\‘e months following radical prosta- tectomy dcspitc androgen deprivation. Three pa- tients with a negative margin progressed (clcvatcd PSA onl! ). ,411 3 have a normal PSA and arc clini- call!, fret ol disease thirteen, fourteen, and twent\ months following androgen deprivation. One (;I thcsc patients had seminal vesicle involvement and lymph node metastasis. A pretreatment PSA uas available in 3 of these 6 patients who pro- gressed, and all values were more than 30 ng/mL.

With a mean follow-up of thirty-three months (range. 4-81 months), 35 patients are alive. Thirty-three of thcsc 35 are free 01. tumor. Seven- trcn of 37 I-+6%) arc tumor free without further androgcn deprivation.

COMMENT

Radical pro5tatcctorn)’ is an excellent operation with acceptable morbidity and a high likelihood of achieving J. curt for the man whose tumor is confinrd to the prostatc. The ideal candidate has a palpable nodule that is less than 1.5 cm and is surrounded b). normal prostate. Unfortunately, a minority of patients discovered with apparent lo- calizcd prostate cancer have this B 1 (T2a) nodule. Marc often, the firm area palpated on rectal exam- ination is larger than 1.5 cm and extends to tbc so-called capsulc of the prostate and these arc des- ignated as B2 (T2h). Unfortunately? the accuraq of clinical staging is not high. Since there is a dis- tinct possibility that radical prostatectomy will render the patient with a T2b lesion tumor free, man)~ belicvc that this is the treatment of choice. Once the palpable tumor is thought to have cx- tended beyond the capsule, stage C or T3, the in- cidcnc,e of a pnsitivc margin or I>,mph node metastasts is quite high. There is no consensus on

the treatment for T3 prostate cancer; however, the most commonly used forms of therap), are cxtrr- nal beam radiation therapy, androgen deprivation, or a combination. Some belicvc, ho\\ cvcr, that

radical prcjstatcctom), i\ indlcatcd. :Idvocatc\ 01 surgery cmphasizc local control. ~hcx inacc~ttrac~ 01 clinical staging, and the acccptablc 11101 bidit;, 01 the operation.

To illustrate the trcqucnq 01 pobitivt margins following radical pi-ostatcctorn!: Jones” rc\;ic\vcd 199 radical rctropubic prostatcctoml~~~ pcrlornicd between 1080 and 1987. I:c,lrt>.-si K pcrccnt had positive margins. The likelihood 01 ;I pl>sltlvc mar- gin was rclatcd to ~utnor volume and a high tumor grade. Other scrics ol paticntb ~8ut,jcctcd to radical prostatcctomy indicate that approxim,ltcl) 3040 percent of thnsc with clinical BJ prostate cancer have a tumor at the Ink4 Ill,irgin.” ’ Lo- cation of the positive margln IS rn<,sl c ommonl) at the apex and is noted in lumors with ;I volume g-cater than -t cc or with higher grade. I hc opti- mal means 01 managing the patic>nt with a pohitlvc margin is unclear.

The concept of androgen dcprix, at Ion prior to radical prostatcctorny was suggcstcd 1,) \cotl and

Boyd” with their puhlicatic>n in IC)W) rcvicwing their fifteen-year follobv-up of a \‘cr\’ sctcct group of 31 patients who recei\~ed dicth~,~lstilbcstrol and/or bilateral orchiectom!, Mith ~,uhscqutnt rad- ical pcrincal prostatectomy. Prior 10 androgen deprivation, all of these patlent\ IIC‘I’C‘ thought to have cxtcnsion beyond the capsr,Ic 1 hey had a normal prostatic acid phosphatasc and rht tllmors wcrc not fixed. I-hc fifteen-year lol’o~, -up indi- cated that 29 percent of these mcll M’crc ali\c and free of’ cancer; 32 percent died of prij\tate (ancci- and 39 percent died of other caukt‘\ ‘1 hc authors compared their results to Barnes’ who trc‘ltcd a group of 75 patients with clinicalI~,. c)rgail-con- fined discasc with only androgcn deprivation and a scrics by Jcwctt”’ who uwd radlc,lt pcrincal prostatcctom)’ for organ-cc.jnfincd prostate C~;~IIC‘CI

and noted that 33 percenl were Irct of caljcer at fifteen j’cars. Barnes” obstrvccl 22 I)crccnt free 01 cancer at fifteen years. The sirnilaric\, among the results pro\,idcd sornc support lor c’( Imbinecl thcr- spy for this select group 01 patients who might otherwise not he candidates for radic,al surgery

Androgen deprivation prior to radical prosta- tcctomy has had a resurgence due to the intro- duction of a rcvcrsiblr and salt Ic>rin 01 androgcn deprivation, notahljr the LHKl1 analog\. Estrogens prior to radical prostatcc tom\’ 3 t-t‘ Its\ dcsirablc due to the ad\rcrse cardiovascular ( \t’nt> asso- ciated with this compound. Orchicc.tomy Icduccs tcstostcrone but the effect is ptrniatlr’nt. ;\II al- ternativc would bc androgtn clcpri\ ariot as an ad,juvant following surgc’r) Ior th.)\c who have adverse prognostic factors-. c-g,. ;I pc>blti\‘c margin.

However, the length of lreatrnenl is arbitrary and once initiated is usually indefinite.

Oesterling et al. I7 evaluated 21 patients with stage 82 (n = 4) or C (n = 17) prostate cancer who received androgen deprivation prior to radical prostatectomy. Despite a dramatic decline in the PSA value, the pathologic findings indicated that only 3 of the patients had negative surgical mar- gins. MacFarlane et al. Ix similarly were unable to demonstrate any significant shift in stage in a rela- tively small group of 20 patients with stages B2 (n = 8) or C (n = 12) prostate cancer.

We have reviewed our results in 37 men with clinical T2b or T3 prostate cancer who received androgen deprivation prior to radical prostatec- tomy. The positive margin rate was 41 percent, which compares favorably with the positive mar- gin rate in patients with clinical T2 or T3 prostate cancer (3%50%). Only 14 percent had positive lymph nodes, which is lower than in previous re- ports of T2b and T3 prostate cancer. The low inci- dence of positive lymph nodes suggests two possi- ble explanations: (1) absence of tumor or (2) failure to identify tumor due to the histologic changes resulting from androgen deprivation.

The follow-up is too short to comment on sur- vival benefits for this combined approach. Since preoperative staging by DRE, TRUS. and com- puted tomography or MRI is relatively inaccurate, it is difficult, if not impossible, to indicate if an- drogen deprivation has altered the tumor stage. The low incidence of positive lymph nodes, the decrease in PSA, and the apparent shrinkage of the prostate are encouraging. Only a well-designed prospective randomized trial can evaluate the im- pact of this approach. A protocol is in progress. El- igible patients for this protocol are less than sev- enty-five years of age and have clinical stage T2b prostate cancer with a PSA of less than 50 ng/mL. The patients are randomized to either three full months of androgen deprivation (leuprolide ac- etate plus flutamide) followed by radical prostatec- tomy or radical prostatectomy alone. The first end- point is the pathologic status: tumor grade, presence or absence of capsular penetration, posi- tive margins, tumor in the seminal vesicles, and lymph node metastasis. The time to recurrence, based on PSA values, as well as overall survival, will also be compared.

REFERENCES 1 Walsh PC, Lepor H, and Eggleston JC.: Radical prosta-

tcctomy with preservation or sexual function: anatomical and pathological considerattons. Prostate 4: 473-485. 1983.

2. Catalona WJ, and Bigg SW: Nerve sparing radical pros- tatectomy: evaluation of results after 250 patients. J Ural 143: 53%543,109o.

3. Scardino PT. and Wheeler TM: Prostatic biopsy after ir- radiation therapy for prostatic cancer. Urology (Suppl) 25: 3%46,lYK5.

4. Paulson DF, Maul JW, Robertson ]E, and Walther PJ. Postoperative radiotherapy of’the prostate for patients undcr- going radical prostatcctomy with positive margins, \ctninal vesicle involvement and/or penetration through the capsule. J Ural 143: 1178-l 182. lYY0.

5. Rosen MA, Goldstone I.. Lapin S. Wheeler 1.. and Scardino PT: Frequency and location of extracapsular cxten- sion and positive surgical margins in radical prostatectomy specimens. J Ural 148: 331-337, 1992.

6. Colston JAC, and Brendler H: Endocrine therapy in carcinoma of the prostate; preparation oC patients Jar radical perineal prostatectomy. JAMA 134: 848-853. 1947.

7. Parlow AL, and Scott WW Hormone control therap) as a preparation for radical perineal prostatectomy in ad- vanced carcinoma of the prostate. NY J Med 49: 620-634, lY49.

8. Sullivan JJ. and Hartwig CH: The use of estrogen ther- apy preliminary to radical perineal prostatoseminal vesiculec- tomy in carcinoma of the prostate. J Ural 70: 494-502, 1953.

9. Schrodcr FH, van den Ouden D, Davidson PJT, and Hop WCJ: Radical prostatcctomy in P3 carcinoma of the prostate-a prospective study of 100 consecutive cases. J Ural 149: 348A, lYY3.

10. Murphy WM. Soloway MS, and Barrows CH: Patho- logic changes associated with androgen deprivation therap) for prostate cancer. Cancer 61: 821-828, 100 I.

11. Jones EC: Resection margin status and radical retro- pubic prostatcctomy specimens: relationship to type of opera- tion, tumor size, tumor grade and local tumor extension. J Urol 144: 8Y-Y3, 1990.

12. Stamey TA. Villers AA, McNcal JE. Link PC., and Freiha FS: Positive surgical margins on radical prostatec- tomy: importance of’ the apical dissectlon. J Ural 143: 1166-I 172. lYY0.

13. Bigg SW, Kavoussi LR, and Catalona WJ: Role of nerve sparing radical prostatcctomp for clinical stage B2 prostate cancer. J Ural 144: 1420-1424, 1990.

14. Scott w, and Boyd HL: Combined hormone con- trolled therapy and radical prostatectomy in the treatment of selected cases of advanced carcinoma of the prostate: a retro- spective study based upon 25 years experience. J Ural 101: 86-92,196Y.

15. Barnes RW. Results of palliative treatment of earl) carcinoma of the prostate. J Ural 70: 489-4Y0, 1953.

16. Jewett HJ: Radical perineal prostatectomy l’or carcl- noma: an analysis of cases at the Johns Hopkins Hospital, 1904-1954. JAMA 156: 1039-1041, 1954.

17. Oesterling JE, Andrews PE, Suman VJ, Zincke H. and Myers RP: Preoperative androgen deprivation therapy: artlh- cial lowering of serum prostate specific antigen without downstaging the tumor. J Urol 149: 77Y-782, 1993.

18. MacFarlane MT, Abi-Aad A, Stein A, Danella J, Bellde- grun A, and DeKernion JB: Neoadjuvant hormonal depriva- tion in patients with locally advanced prostate cancer. J LJrol 150: 132-134. 1993.

56