Review Articles

Prostate Specific Antigen Kineticsin the Management of Prostate CancerShomik Sengupta,* Christopher Amling,† Anthony V. D’Amico* and Michael L. Blute*,‡From the Department of Urology, Mayo Clinic, Rochester, Minnesota (SS, MLB), Department of Urology, University of Alabama,Birmingham, Alabama (CA), and Department of Radiation Oncology, Brigham and Women’s Hospital, and the Dana Farber CancerInstitute, Harvard Medical School, Boston, Massachusetts (AVD)

Purpose: We review the usefulness of prostate specific antigen kinetics (ie prostate specific antigen velocity and doublingtime) in the treatment of patients with prostate cancer.Materials and Methods: The MEDLINE database was searched to identify studies investigating prostate specific antigenkinetics in patients with prostate cancer.Results: Various techniques are available for estimating prostate specific antigen kinetics, but to minimize the impact of prostatespecific antigen variability on such calculations at least a 90-day period and preferably more than 2 measurements should be used.There is little to suggest which measure of prostate specific antigen kinetics may be superior since both appear to provide usefulprognostic information. Prostate specific antigen velocity is easier to calculate but prostate specific antigen doubling time may havegreater biological justification. Retrospective studies show that before treatment prostate specific antigen kinetics provideprognostic information regarding the risk of treatment failure and subsequent death from cancer. Additionally, in patients treatedsurgically preoperative prostate specific antigen kinetics predict the risk of adverse pathology, while in those undergoingconservative treatment prostate specific antigen kinetics are associated with the risk of progression and need for intervention. Inpatients with biochemical failure after therapy prostate specific antigen kinetics predict the risk and potential site of clinicalrecurrence, the likely response to salvage therapy, and the risk of death from cancer. Preliminary assessments also suggest thatprostate specific antigen kinetics may serve as a surrogate end point to replace cancer specific mortality.Conclusions: Although prospective studies are lacking, the current literature suggests that prostate specific antigen kineticsprovide valuable prognostic information, and should be further evaluated in clinical decision making and as a surrogate endpoint for future trials.

Key Words: prostatic neoplasms, prostate-specific antigen, prostatectomy, radiotherapy, survival

Since its advent in clinical practice in the 1980s, PSAhas had a significant impact on the management ofprostate cancer. Most notably, by allowing the early

detection of asymptomatic disease it has been responsiblefor the stage migration of prostate cancer during the last 2decades. Additionally, serum PSA has become indispensablefor prognostication and surveillance.

However, PSA remains an imperfect test. In the diagnos-tic setting noncancerous causes of PSA increase such asbenign hypertrophy or inflammation of the prostate fre-quently confound the clinical picture. Prostate cancers ofwidely disparate biology may present with comparablePSAs, thus, undermining the prognostic reliability of a sin-gle PSA measurement. Given the prolonged natural historyof prostate cancer, the implications of changes in serum PSAafter therapy are often unclear on short-term followup.

A way to overcome some of the limitations of PSA testinghas been to examine PSA kinetics or the rate of PSA change

Submitted for publication April 22, 2007.* Nothing to disclose.† Financial interest and/or other relationship with Johnson &

Johnson, TAP Pharmaceuticals and Sanofi-Aventis.‡ Correspondence: Department of Urology, Mayo Clinic, 200 First

St. SW, Rochester, Minnesota 55905 (telephone: 507-284-3987;FAX: 507-284-4987; e-mail: Blute.Michael@mayo.edu).

0022-5347/08/1793-0821/0THE JOURNAL OF UROLOGY®

Copyright © 2008 by AMERICAN UROLOGICAL ASSOCIATION

821

with time. PSAV was originally described by Carter et al basedon the Baltimore Longitudinal Study of Aging.1 SubsequentlyPSADT was described by Schmid et al and was studied exten-sively in patients treated with radiotherapy for prostate can-cer.2 An increasing body of work suggests that these 2 mea-sures of PSA kinetics may be useful in managing prostatecancer. We review the available data related to PSAV andPSADT in various clinical settings. It is notable that the liter-ature on PSA kinetics is largely retrospective, resulting fromthe fact that PSA kinetic data are easier to interpret after anumber of measurements have been collected. In contrast, asnew PSA readings are obtained in clinical practice, it may bedifficult to discern changes reflecting cancer behavior fromthose reflecting confounding processes such as inflammation.

PSA KINETICPARAMETERS: ESTIMATION AND ERROR

Because PSAV and PSADT are measures of the rate of PSAchange with time, their estimation relies on the statistical

Editor’s Note: This article is the first of 5 published inthis issue for which category 1 CME credits can beearned. Instructions for obtaining credits are given

with the questions on pages 1208 and 1209.

Vol. 179, 821-826, March 2008Printed in U.S.A.

DOI:10.1016/j.juro.2007.10.023

PROSTATE SPECIFIC ANTIGEN KINETICS AND PROSTATE CANCER822

technique of regression. The calculation of PSAV is basedon the assumption that PSA increases in a linear fashion withtime and, therefore, uses linear regression analysis. In practicethis translates into a fairly simple slope calculation of theformula, PSAV � change in PSA/time between measurements.If more than 2 PSA readings are available PSAV may simplybe averaged. For example, for 3 PSA measurements, PSAV �[{(PSA2 – PSA1)/interval1} � {(PSA3 – PSA2)/interval2}]/2.

In contrast, PSADT assumes an exponential increase inPSA and, therefore, requires a somewhat more complexanalysis for estimation. Regression techniques for the esti-mation of PSADT require the logarithmic transformation ofavailable PSA values by the somewhat daunting formula,PSADT � t � natural log2/natural log (ratio of PSAs). Inpractice the calculation of PSADT requires the use of statis-tical software or online algorithms, although a recently pub-lished graphical tool allows the estimation of PSADT inpractice settings without electronic resources.3

The measurement of serum PSA is susceptible to errorfrom several sources, most notably from interassay and bi-ological variations, and a number of precautions are re-quired to minimize the impact of such errors on estimates ofPSA kinetics. Thus, sequential PSA readings should ideallybe obtained from the same laboratory using a well calibratedassay. Furthermore, the greater the number of PSA mea-surements used and the longer the interval during whichthey are obtained, the more likely it is that the resultingestimate of PSA kinetics truly reflects cancer growth.

However, the need for expedient decision making and thecosts of repeat PSA testing usually place practical limits onthe number and spacing of available PSA measurements.Additionally, since cancer biology may evolve, leading tovariation in PSA kinetics with time, it seems prudent torestrict PSA kinetic assessments to a maximum of 12 to 24months.4 Two PSA readings separated by at least 3 monthsappear to provide a reasonably accurate estimate of PSADT,but a minimum of 3 readings during at least 6 monthsshould be obtained when possible.

PSA KINETICS ANDCONSERVATIVE MANAGEMENT

Men with prostate cancer may be treated conservatively withplanned delayed intervention with curative AS or palliativeWW intent. Although serial PSA data are usually available formen undergoing such conservative treatment, published re-ports are few and somewhat difficult to interpret. Many studiesare retrospective, and include patients with a broad spectrumof disease pathologies treated with AS and WW. While elderly

TABLE 1. Relationship of PSA kinetics to clinical featurwith prostate cancer

Hardie et al5 Stephenson et al6

No. pts 80 94PSA kinetic PSADT PSADTp Value:

Age NS NSPSA 0.039 NST stage 0.036* 0.04Biopsy grade NS NRProgression DRE 0.019, biopsy 0.03Time to treatment

* Multivariate analysis.

or frail patients treated conservatively because of limited lifeexpectancy may have intermediate to high risk tumors,younger patients placed on AS protocols are usually selected onthe basis of low risk features. Not surprisingly, PSA is found toincrease relatively slowly in most of these patients withPSADT longer than 10 years in the majority.

The rate of PSA increase appears to be related to theaggressiveness of the prostate cancer as indicated by anassociation with clinical stage, biopsy grade and initialPSA,5,6 although these findings vary somewhat among se-ries (table 1).7,8 Another difficulty in interpreting the pre-dictive value of PSA kinetics in this setting is that since anincrease in PSA is often a trigger for intervention, men withrapid PSA kinetics are over represented in the treated sub-group, thus leading to selection bias. In fact, some activesurveillance protocols explicitly incorporate PSA kineticmeasures as criteria for intervention. For example, Choo etal offer curative surgery to men under active surveillancewith a PSADT of less than 3 years.7

However, a number of other studies that did not formallyuse PSA kinetics for decisions on intervention have reporteda significant association between PSA kinetics and the needfor intervention or time to intervention, thus providing sup-port for this concept.9–11 Furthermore, PSADT has beenshown to correlate with other measures of progression suchas DRE, repeat biopsy or bone scan.6,9 Finally, PSA kineticmeasures appear to be associated with the risk of death fromprostate cancer in men treated conservatively.12,13

Despite these findings doubts remain regarding the use-fulness of PSA kinetic measures in an individual patientgiven the wide variation and significant overlap.6–8 Addi-tionally, variability in PSA kinetics with time is pronounced,further complicating the application of these measures indecision making during the conservative management ofprostate cancer. 4,8 Currently it seems sensible to incorpo-rate PSA kinetics along with other criteria in AS and WWprotocols, although individualized rather than universalcutoff points may be preferable.5

PSA KINETICS AND RADICAL PROSTATECTOMY

As serial PSA testing becomes more common, preoperativePSA kinetic data are increasingly available for patientstreated surgically. Preoperative PSA kinetic measures areassociated with aggressive pathological features at prosta-tectomy, including local and nodal stage, specimen Gleasonscore, and involvement of surgical margins. Although earlystudies suggested that preoperative PSA kinetics were notpredictive of postoperative outcomes, they were likely lim-

sk of progression and need for intervention in patientsted conservatively

Choo et al7 Gerber et al8 McLaren et al9

134 49 113PSADT PSAV PSADT

NS NSNS NSNS NSNS NS

DRE �0.01, clinical �0.0001�0.0001

es, ritrea

PROSTATE SPECIFIC ANTIGEN KINETICS AND PROSTATE CANCER 823

ited in power, and more recent series have confirmed preop-erative PSA kinetics to be predictive of the risk of progres-sion and death after radical prostatectomy (table 2).

D’Amico et al analyzed a cohort of 1,069 men who hadundergone serial PSA testing at the same laboratory beforesurgery, and PSAV greater than 2.0 ng/ml per year wasassociated with a significantly increased risk of death fromprostate cancer and from all causes.14 Sengupta et al com-pared preoperative PSAV and PSADT, and reported thatPSADT was a stronger predictor of clinical recurrence ordeath from cancer.15 Preoperative PSA kinetics do not ap-pear to simply be surrogate markers for conventional clini-cal or histological indicators of aggressive disease, since theprognostic value of PSA kinetics persists on multivariateanalysis.14–16 The inclusion of PSA kinetics causes preoper-ative PSA to drop out of multivariate models, suggestingthat the rate of increase may be more important prognosti-cally than the absolute PSA value.15

In the postoperative setting, since all PSA producingtissue has been removed, serum PSA is expected to be un-detectable. Although a detectable PSA is taken to be pre-sumptive evidence of cancer recurrence, only about a third ofthese patients have clinically evident recurrence. Definingrecurrence by biochemical criteria remains controversial,with various cut points used in the literature. It has beenreported that a PSA increase of less than 0.4 ng/ml wasrarely associated with further subsequent increases, sug-gesting that this threshold is most appropriate. Stephensonet al also found a PSA cutoff of 0.4 ng/ml to be associatedwith subsequent PSADT and risk of metastasis.17

Even in patients with biochemical recurrence a subgroupcan be identified with slowly increasing PSA, delayed andinfrequent clinical recurrence, and low risk of death fromcancer who may be best treated conservatively.18 When clin-ical recurrence is local, that is at the site of resection, salvagetherapy may be considered. When recurrence is systemic, pal-liative hormonal therapy becomes most appropriate. PSADTafter recurrence is predicted by preoperative risk factors,pathological findings and time to biochemical failure,19 and

TABLE 2. Relationship of preoperative PSA kinetics with p

D’Amico et al14

No. pts 1,069PSA kinetic PSAVp Value:

Stage �0.05Grade �0.05Pos surgical marginSurvival* CSS �0.001, OS �0.01

* Multivariate analysis.

TABLE 3. Relationship of post-recurrence PSA

Dotan et al20 Pou

No. ptsNo. PSA recurrence 239PSA kinetic PSAV PSADTAssociation with survival (p value) Systemic

Comments Bone metastasis(OR 0.93, p �0.003)*

* Multivariate analysis.

these factors may help delineate the risk of metastatic vs localrecurrence.20–23

Early studies suggested that the likely site of recurrencemay be predicted based on PSAV24 or PSADT25 after bio-chemical recurrence. Subsequent reports have confirmed aclear association between rapid PSA kinetics and the subse-quent risk of metastatic recurrence (table 3). Recently anassociation between PSA kinetics after recurrence and sur-vival has been established, with a 5-year cancer specificmortality of 31% vs 1% for patients with PSADT less than 3months vs 3 months or greater,26 suggesting that PSADTmay be useful as a surrogate end point for cancer death.27

Notably, prostate cancer predominates as a cause of death inmen with PSA recurrence after surgery, even up to a PSADTof 15 months.28 Additionally, in patients receiving salvageradiotherapy PSA kinetics before radiation appear to be animportant prognostic factor,29 while high risk patientstreated with adjuvant hormonal therapy are most likely tobenefit if they have a rapid PSADT.30 Furthermore, in theface of an increasing PSA while on hormonal therapy afterinitial local therapy, PSA kinetic measures predict subse-quent survival and may have some use in decision makingregarding additional therapy.31

PSA KINETICS AND RADIOTHERAPY

Data on PSA kinetics in patients treated with external orinterstitial radiation largely parallel those described in sur-gically treated patients (table 4). Thus, Hanks et al foundthe pretreatment PSADT and the radiation dose to be sig-nificant multivariate predictors of biochemical recurrence-free survival.32 Recently D’Amico et al reported that pre-treatment PSAV greater than 2.0 ng/ml per year wasassociated with an increased risk of biochemical recurrence,death from cancer and death from any cause after externalbeam radiation regardless of the risk group.33 Eggener et alobserved that a pretreatment PSAV cutoff of 2.0 ng/ml peryear is also predictive of biochemical outcomes after brachy-therapy.34 D’Amico et al further reported that patients with

logic findings and survival in surgically treated patients

Sengupta et al15 Patel et al16

2,290 202PSAV � PSADT PSAV

�0.0001 0.007�0.0001 0.04

PSAV �0.0001, PSADT NS 0.01CSS �0.0001 RFS 0.03

ics and survival in surgically treated patients

al22 Roberts et al23 Zhou et al26

2,809 NS879 (587 with PSADT) 498PSADT PSADT

(�0.001) Systemic PFS (�0.001),PFS (�0.001)*

PSADT less than 6 mospredicts systemicprogression

PSADT less than 3 mos predictsprostate cancer specificmortality (HR 54.9)

atho

kinet

nd et

1,997315

PFS

PROSTATE SPECIFIC ANTIGEN KINETICS AND PROSTATE CANCER824

a rapid pretreatment PSAV benefit from the addition ofandrogen deprivation to external beam radiation.35

PSA kinetics after radiation are confounded by the con-tinuing production of PSA by the prostate, which is suscep-tible to a number of influences. Thus, the concomitant use ofandrogen deprivation and radiation often causes a with-drawal effect, leading to more rapid PSA increases, which maylead to misclassification of biochemical failure.36,37 In fact,during the initial period after external beam radiotherapy,PSA decreases and the negative PSA kinetics as well as thePSA nadir appear to be prognostic of overall survival.38 Afterinterstitial seed placement the phenomenon of PSA bouncemay occur and cause confusion with early biochemical relapse.In fact, the PSA kinetics for the 2 events appear to be similar,and the time from therapy may be the most reliable distin-guishing characteristic with bounces typically occurring earlierthan recurrence.39 PSADT after recurrence may also be usefulin classifying the likely site of recurrence in patients treatedwith radiotherapy, with those with local relapse amenable tosalvage therapies.40 The success rate of salvage cryotherapy inthis setting also depends on PSADT before salvage.41

Zagars and Pollack,42 and Lee et al43 reported thatPSADT less than 8 months after relapse is associated withthe risk of metastatic disease and death from all causes inpatients treated with external beam radiotherapy alone42 orin combination with androgen deprivation (table 5).43 Using3-dimensional conformal intensity modulated radiationtherapy Zelefsky et al observed a similar relationship betweenPSADT and metastatic recurrence.44 Cancer specific mortalityafter external beam radiation for prostate cancer is stronglyassociated with PSADT after recurrence, and is 75% vs 15% forPSADT less than 3 months vs 3 months or greater in patientswith Gleason score greater than 7, and 35% vs 4%, respec-tively, in those with Gleason score 7 or less.26 Similarly, forpatients treated with interstitial seed placement 10-year CSSis 30% for patients with a PSADT of 6 months or less vs 98% forthose with a PSADT of more than 10 months.45

TABLE 4. Relationship of pretreatment PSA kineti

Hanks et al32

No. pts 99Treatment External beam (dose NS) ExternaPSA kinetic PSADT PSAV

Association with survival bRFS bRFS, CComments 50% bRFS at 18 mos in pts with

PSADT less than 12 mosLow risk

TABLE 5. Relationship of post-recurrence PSA kinet

Zagars and Pollack42 Lee et al43 Zel

No. pts 841 621Treatment 60–78 Gy 76 Gy � ADT 64–81

No. relapse 263 75PSA kinetic PSADT less than 8 mos PSADT PSADAssociation withsurvival

sRFS Clinical RFS p �0.001,OS p � 0.015

sRFS

Comments 93% Vs 46% at 7 yrs PSADmosthan

greater th

CLINICAL USE AND FUTURE DIRECTIONS

The application of PSA kinetic measures in clinical practiceis really no more than a formalization of serial PSA testing.While regular PSA measurements for surveillance aftertreatment have long been routine, it has recently becomemore common in other clinical settings. Thus, many patientsnow undergo repeat PSA screening before prostate cancerdiagnosis. The indication for prostatic biopsy in these patientsis typically an increase in PSA beyond a threshold level. How-ever, controversy persists regarding the optimal PSA cutoff,with emerging data indicating that the risk of prostate cancerremains significant even with a nominally normal PSA.

It remains uncertain whether PSA kinetic measures arelikely to be helpful in this setting. In the prospective Pros-tate Cancer Prevention Trial PSAV based on at least 2 PSAsin 3 years was not a significant predictor of a positive end ofstudy biopsy.46 In contrast, short-term PSAV more than 2months before biopsy was associated with the presence ofmalignancy in patients with an abnormal PSA.47 It is alsonotable that data from the Baltimore Longitudinal Study ofAging revealed differences in PSAV years before prostatecancer diagnosis.1 A problem with retrospective analyses ofthe diagnostic value of PSA kinetics is that not all menundergo biopsy, resulting in assignment bias.48 Further pro-spective study of this issue is awaited, but in younger menrapid unexplained increases in PSA, even in the normalrange, need to be viewed with concern.

For patients with a prostate cancer diagnosis PSA kinet-ics appear to convey prognostic information that should beconsidered with conventional prognostic factors such as ab-solute PSA, clinical stage and biopsy grade. Although PSAkinetics are of the greatest relevance immediately beforediagnosis, PSA velocity can provide prognostic informationyears ahead.13 PSA kinetics are likely to be useful not onlyin counseling patients regarding likely outcomes after treat-ment, but also in suggesting experimental multimodal ther-

th survival in patients treated with radiotherapy

ico et al33 Eggener et al34 Eggener et al34

358 83 47(77 Gy) External beam (dose NR) Brachytherapy

PSAV (greater than2 ng/ml/yr)

PSAV (greater than2 ng/ml/yr)

OS (each p �0.01) bRFS bRFSigh risk groups 80% Vs 55% at 6 yrs 88% Vs 65% at 6 yrs

ith survival in patients treated with radiotherapy

et al44 Sartor et al40 Zhou et al26 Stock et al45

0 399 NS 1,561ADT 70 Gy External beam,

dose NSBrachytherapy �external beam � ADT

1 234 (increasing PSA) 661 131PSADT PSADT PSADTsRFS, local RFS CSS CSS

s than 6eateros, HR

Site of recurrence

cs wi

D’Am

l beam

SS �� h

ics w

efsky

1,65Gy �

38T� OS

T lesvs gr12 m

an 6.6

PROSTATE SPECIFIC ANTIGEN KINETICS AND PROSTATE CANCER 825

apy protocols for those with aggressive disease. Finally, inthe context of AS protocols, PSA kinetic measures may becrucial criteria for intervention, although precise cutoff lev-els require further investigation.

Biochemical recurrence after treatment leads inevitablyto consideration of PSA kinetics in clinical decision making.In this context, along with known pretreatment and post-treatment prognostic factors, PSA kinetics may allow anassessment of the probability of relapse and the likely site,thereby facilitating rational decisions regarding the need forand type of salvage therapy. For instance, it appears that formen with biochemical recurrence after radical prostatec-tomy or external beam radiation and a PSADT of less than3 months, earlier initiation of hormonal therapy delays thedevelopment of bony metastases, thereby improving qualityof life even if survival is not prolonged.27 This group of menmay also be usefully studied in randomized trials of systemicchemotherapy such as docetaxel. Similarly for men withmetastatic prostate cancer PSA kinetic measures may beused to guide the initiation of hormonal therapy and second-ary intervention during the hormone refractory phase.49

In these clinical situations either measure of PSA kinet-ics may be applied. There is empirical evidence that prostatecancer biology does in fact conform to an exponential modeland, therefore, PSADT may be more accurate.2,50 This ispartially supported by the results of a large retrospectivestudy of PSA kinetics before prostatectomy.15 However, dur-ing short intervals the linear and exponential curves closelyapproximate each other and, therefore, PSAV provides asatisfactory alternative that is usually easier to calculate.Finally, it should be noted that PSAV and PSADT are re-lated to each other and to serum PSA by the formula, PSAV �PSA/PSADT. In a linear model (constant PSAV) PSADT isassumed to vary with PSA, and vice versa in an exponentialmodel (constant PSADT). In reality, for a particular cancerin a particular man the exact pattern of PSA increase withtime may not conform exactly to either model, and which-ever model is used ends up being an approximation.

Abbreviations and Acronyms

ADT � androgen deprivation therapyAS � active surveillance

bRFS � biochemical RFSCSS � cancer specific survivalDRE � digital rectal examination

NS � not statistically significantNR � not reportedOS � overall survival

PFS � progression-free survivalPSA � prostate specific antigen

PSADT � PSA doubling timePSAV � PSA velocity

RFS � recurrence-free survivalsRFS � systemic RFSWW � watchful waiting

REFERENCES

1. Carter HB, Pearson JD, Metter EJ, Brant LJ, Chan DW,Andres R et al: Longitudinal evaluation of prostate-specificantigen levels in men with and without prostate disease.

JAMA 1992; 267: 2215.

2. Schmid HP, McNeal JE and Stamey TA: Observations on thedoubling time of prostate cancer. The use of serial prostate-specific antigen in patients with untreated disease as ameasure of increasing cancer volume. Cancer 1993; 71:2031.

3. Sengupta S, Slezak JM, Blute ML and Bergstralh EJ: Simplegraphic method for estimation of prostate-specific antigendoubling time. Urology 2006; 67: 408.

4. Ross PL, Mahmud S, Stephenson AJ, Souhami L, Tanguay Sand Aprikian AG: Variations in PSA doubling time in pa-tients with prostate cancer on “watchful waiting”: value ofshort-term PSADT determinations. Urology 2004; 64: 323.

5. Hardie C, Parker C, Norman A, Eeles R, Horwich A, HuddartR et al: Early outcomes of active surveillance for localizedprostate cancer. BJU Int 2005; 95: 956.

6. Stephenson AJ, Aprikian AG, Souhami L, Behlouli H, JacobsonAI, Begin LR et al: Utility of PSA doubling time in follow-up ofuntreated patients with localized prostate cancer. Urology2002; 59: 652.

7. Choo R, DeBoer G, Klotz L, Danjoux C, Morton GC, RakovitchE et al: PSA doubling time of prostate carcinoma managedwith watchful observation alone. Int J Radiat Oncol BiolPhys 2001; 50: 615.

8. Gerber GS, Gornik HL, Goldfischer ER, Chodak GW andRukstalis DB: Evaluation of changes in prostate specificantigen in clinically localized prostate cancer managedwithout initial therapy. J Urol 1998; 159: 1243.

9. McLaren DB, McKenzie M, Duncan G and Pickles T: Watchfulwaiting or watchful progression? Prostate specific antigendoubling times and clinical behavior in patients with earlyuntreated prostate carcinoma. Cancer 1998; 82: 342.

10. el-Geneidy M, Garzotto M, Panagiotou I, Hsieh YC, Mori M,Peters L et al: Delayed therapy with curative intent in acontemporary prostate cancer watchful-waiting cohort.BJU Int 2004; 93: 510.

11. Carter CA, Donahue T, Sun L, Wu H, McLeod DG, Amling Cet al: Temporarily deferred therapy (watchful waiting) formen younger than 70 years and with low-risk localizedprostate cancer in the prostate-specific antigen era. J ClinOncol 2003; 21: 4001.

12. Fall K, Garmo H, Andren O, Bill-Axelson A, Adolfsson J,Adami HO et al: Prostate-specific antigen levels as a pre-dictor of lethal prostate cancer. J Natl Cancer Inst 2007; 99:526.

13. Carter HB, Ferrucci L, Kettermann A, Landis P, Wright EJ,Epstein JI et al: Detection of life-threatening prostate can-cer with prostate-specific antigen velocity during a windowof curability. J Natl Cancer Inst 2006; 98: 1521.

14. D’Amico AV, Chen MH, Roehl KA and Catalona WJ: Preoper-ative PSA velocity and the risk of death from prostatecancer after radical prostatectomy. N Engl J Med 2004;351: 125.

15. Sengupta S, Myers RP, Slezak JM, Bergstralh EJ, Zincke Hand Blute ML: Preoperative prostate specific antigen dou-bling time and velocity are strong and independent predic-tors of outcomes following radical prostatectomy. J Urol2005; 174: 2191.

16. Patel DA, Presti JC Jr, McNeal JE, Gill H, Brooks JD and KingCR: Preoperative PSA velocity is an independent prognosticfactor for relapse after radical prostatectomy. J Clin Oncol2005; 23: 6157.

17. Stephenson AJ, Kattan MW, Eastham JA, Dotan ZA, BiancoFJ Jr, Lilja H et al: Defining biochemical recurrence ofprostate cancer after radical prostatectomy: a proposal for astandardized definition. J Clin Oncol 2006; 24: 3973.

18. Tollefson MK, Slezak JM, Leibovich BC, Zincke H and BluteML: Stratification of patient risk based on prostate-specificantigen doubling time after radical retropubic prostatec-

tomy. Mayo Clin Proc 2007; 82: 422.

PROSTATE SPECIFIC ANTIGEN KINETICS AND PROSTATE CANCER826

19. Lin DD, Schultz D, Renshaw AA, Rubin MA, Richie JP andD’Amico AV: Predictors of short postoperative prostate-specific antigen doubling time for patients diagnosed dur-ing PSA era. Urology 2005; 65: 528.

20. Dotan ZA, Bianco FJ Jr, Rabbani F, Eastham JA, Fearn P,Scher HI et al: Pattern of prostate-specific antigen (PSA)failure dictates the probability of a positive bone scan inpatients with an increasing PSA after radical prostatec-tomy. J Clin Oncol 2005; 23: 1962.

21. Lee AK and D’Amico AV: Utility of prostate-specific antigenkinetics in addition to clinical factors in the selection ofpatients for salvage local therapy. J Clin Oncol 2005; 23:8192.

22. Pound CR, Partin AW, Eisenberger MA, Chan DW, PearsonJD and Walsh PC: Natural history of progression after PSAelevation following radical prostatectomy. JAMA 1999;281: 1591.

23. Roberts SG, Blute ML, Bergstralh EJ, Slezak JM and ZinckeH: PSA doubling time as a predictor of clinical progressionafter biochemical failure following radical prostatectomyfor prostate cancer. Mayo Clin Proc 2001; 76: 576.

24. Partin AW, Pearson JD, Landis PK, Carter HB, Pound CR,Clemens JQ et al: Evaluation of serum prostate-specificantigen velocity after radical prostatectomy to distinguishlocal recurrence from distant metastases. Urology 1994; 43:649.

25. Trapasso JG, deKernion JB, Smith RB and Dorey F: Theincidence and significance of detectable levels of serumprostate specific antigen after radical prostatectomy. J Urol1994; 152: 1821.

26. Zhou P, Chen MH, McLeod D, Carroll PR, Moul JW andD’Amico AV: Predictors of prostate cancer-specific mortal-ity after radical prostatectomy or radiation therapy. J ClinOncol 2005; 23: 6992.

27. D’Amico AV, Moul JW, Carroll PR, Sun L, Lubeck D and ChenMH: Surrogate end point for prostate cancer-specific mor-tality after radical prostatectomy or radiation therapy.J Natl Cancer Inst 2003; 95: 1376.

28. Freedland SJ, Humphreys EB, Mangold LA, Eisenberger M,Dorey FJ, Walsh PC et al: Death in patients with recur-rent prostate cancer after radical prostatectomy: pros-tate-specific antigen doubling time subgroups and theirassociated contributions to all-cause mortality. J ClinOncol 2007; 25: 1765.

29. Ward JF, Zincke H, Bergstralh EJ, Slezak JM and Blute ML:Prostate specific antigen doubling time subsequent to rad-ical prostatectomy as a prognosticator of outcome followingsalvage radiotherapy. J Urol 2004; 172: 2244.

30. Moul JW, Wu H, Sun L, McLeod DG, Amling C, Donahue Tet al: Early versus delayed hormonal therapy for prostatespecific antigen only recurrence of prostate cancer afterradical prostatectomy. J Urol 2004; 171: 1141.

31. Sengupta S, Blute ML, Bagniewski SM, Myers RP, BergstralhEJ, Leibovich BC et al: Increasing prostate specific antigenfollowing radical prostatectomy and adjuvant hormonaltherapy: doubling time predicts survival. J Urol 2006; 175:1684.

32. Hanks GE, Hanlon AL, Lee WR, Slivjak A and Schultheiss TE:Pretreatment prostate-specific antigen doubling times:clinical utility of this predictor of prostate cancer behavior.Int J Radiat Oncol Biol Phys 1996; 34: 549.

33. D’Amico AV, Renshaw AA, Sussman B and Chen MH: Pre-treatment PSA velocity and risk of death from prostatecancer following external beam radiation therapy. JAMA2005; 294: 440.

34. Eggener SE, Roehl KA, Yossepowitch O and Catalona WJ:

Prediagnosis prostate specific antigen velocity is associated

with risk of prostate cancer progression followingbrachytherapy and external beam radiation therapy. J Urol2006; 176: 1399.

35. D’Amico AV, Loffredo M, Renshaw AA, Loffredo B and ChenMH: Six-month androgen suppression plus radiation ther-apy compared with radiation therapy alone for men withprostate cancer and a rapidly increasing pretreatmentprostate-specific antigen level. J Clin Oncol 2006; 24: 4190.

36. Bates AT, Pickles T and Paltiel C: PSA doubling time kineticsduring prostate cancer biochemical relapse after externalbeam radiation therapy. Int J Radiat Oncol Biol Phys 2005;62: 148.

37. Buyyounouski MK, Hanlon AL, Horwitz EM, Uzzo RG andPollack A: Biochemical failure and the temporal kinetics ofprostate-specific antigen after radiation therapy with an-drogen deprivation. Int J Radiat Oncol Biol Phys 2005; 61:1291.

38. Cheung R, Tucker SL and Kuban DA: First-year PSA kineticsand minima after prostate cancer radiotherapy are predic-tive of overall survival. Int J Radiat Oncol Biol Phys 2006;66: 20.

39. Ciezki JP, Reddy CA, Garcia J, Angermeier K, Ulchaker J,Mahadevan A et al: PSA kinetics after prostate brachyther-apy: PSA bounce phenomenon and its implications for PSAdoubling time. Int J Radiat Oncol Biol Phys 2006; 64: 512.

40. Sartor CI, Strawderman MH, Lin XH, Kish KE, McLaughlinPW and Sandler HM: Rate of PSA rise predicts metastaticversus local recurrence after definitive radiotherapy. Int JRadiat Oncol Biol Phys 1997; 38: 941.

41. Spiess PE, Lee AK, Leibovici D, Wang X, Do KA and PistersLL: Presalvage prostate-specific antigen (PSA) and PSAdoubling time as predictors of biochemical failure of salvagecryotherapy in patients with locally recurrent prostate can-cer after radiotherapy. Cancer 2006; 107: 275.

42. Zagars GK and Pollack A: Kinetics of serum prostate-specificantigen after external beam radiation for clinically local-ized prostate cancer. Radiother Oncol 1997; 44: 213.

43. Lee AK, Levy LB, Cheung R and Kuban D: Prostate-specificantigen doubling time predicts clinical outcome and sur-vival in prostate cancer patients treated with combinedradiation and hormone therapy. Int J Radiat Oncol BiolPhys 2005; 63: 456.

44. Zelefsky MJ, Ben-Porat L, Scher HI, Chan HM, Fearn PA,Fuks ZY et al: Outcome predictors for the increasing PSAstate after definitive external-beam radiotherapy for pros-tate cancer. J Clin Oncol 2005; 23: 826.

45. Stock RG, Cesaretti JA and Stone NN: Disease-specific sur-vival following the brachytherapy management of prostatecancer. Int J Radiat Oncol Biol Phys 2006; 64: 810.

46. Thompson IM, Ankerst DP, Chi C, Goodman PJ, Tangen CM,Lucia MS et al: Assessing prostate cancer risk: results fromthe Prostate Cancer Prevention Trial. J Natl Cancer Inst2006; 98: 529.

47. Lynn NN, Collins GN and O’Reilly PH: The short-term pros-tate-specific antigen velocity before biopsy can be used topredict prostatic histology. BJU Int 2000; 85: 847.

48. Loeb S, Roehl KA, Catalona WJ and Nadler RB: Prostatespecific antigen velocity threshold for predicting prostatecancer in young men. J Urol 2007; 177: 899.

49. Loberg RD, Fielhauer JR, Pienta BA, Dresden S, Christmas P,Kalikin LM et al: Prostate-specific antigen doubling timeand survival in patients with advanced metastatic prostatecancer. Urology, suppl., 2003; 62: 128.

50. Vollmer RT, Egawa S, Kuwao S and Baba S: The dynamics ofprostate specific antigen during watchful waiting of pros-tate carcinoma: a study of 94 Japanese men. Cancer 2002;

94: 1692.