96
`
`Prostate Specific Antigen Levels and
`Clinical Response to Low Dose
`Dexamethasone for Hormone-Refractory
`Metastatic Prostate Carcinoma
`James A. Storlie, M.D.,* Jan C. Buckner, M.D.,t Gregory A. Wiseman, M.D.,$
`Patrick A. Burch, M.D.,t Lynn C. Hartmann, M.D.,t and Ronald L. Richardson, M.D.t
`
`Background. It has been suggested that suppression
`of adrenal androgens may provide benefit to patients with
`metastatic prostate cancer refractory to initial hormonal
`therapy (e.g., orchiectomy).
`Methods. The records of 38 patients with metastatic
`prostate cancer that had progressed after orchiectomy
`who were placed subsequently on low dose dexametha-
`sone (DXM) with no other concurrent therapy (36 patients
`received 0.75 mg twice daily and two received 0.75 mg
`three times daily) were reviewed. Symptomatic status,
`prostate specific antigen (PSA) measurements, and avail-
`able radiographic assessments were recorded. Bone scans
`were reviewed by an independent, blinded evaluator.
`Results. Symptomatic improvement was experienced
`by 24 patients (63%), 20 (83%) of whom also had decreases
`in PSA. Prostate specific antigen values decreased in 30
`patients (79%) with decreases 50% or greater and 80% or
`greater in 23 (61%) and 13 (34%) patients, respectively. Of
`the 23 patients with PSA decreases 50% or greater, 8
`(35%) had radiographic evidence of disease regression, 5
`(22%) were stable, 7 (30%) had disease progression, and 3
`(13%) did not have serial radiographic exams. Flutamide
`was discontinued shortly before DXM treatment for 2 of
`the 23 patients.
`Conclusions. Low dose DXM may produce important
`symptomatic improvement and decreased PSA levels in
`the majority of patients with hormone-refractory pros-
`tate cancer. In addition, a substantial percentage of those
`patients with decreases in PSA also will have radio-
`graphic evidence of disease regression. These results sug-
`gest the need for additional prospective controlled studies
`of DXM as a therapy for hormone-refractory prostate
`cancer. Cancer 1995; 76:96-100.
`
`Presented in abstract form at the Annual Meeting of the Ameri-
`can Society of Clinical Oncology, Dallas, Texas, May 1994.
`From the Departments of *Family Medicine, tOncology, and
`$Diagnostic Radiology, Mayo Clinic, Rochester, Minnesota.
`Address for reprints: Jan C. Buckner, M.D., Division of Medical
`Oncology, Mayo Clinic, 200 1st Street SW, Rochester, MN 55905.
`Received December 7, 1994; revision received March 8, 1995;
`accepted March 8, 1995.
`
`Key words: prostatic neoplasms, prostate specific antigen,
`dexamethasone, orchiectomy, metastatic neoplasms.
`
`Most patients with metastatic prostate carcinoma will
`show some response to first-line hormonal therapy in
`the form of surgical ablation of testicular androgens (or-
`chiectomy) or gonadal suppression (estrogens, proges-
`tins, luteinizing hormone-releasing hormone agonists,
`or antiandrogens). ’,’ However, the majority of patients
`with metastases will experience disease progression
`within 2 years after initial hormonal therapy. The mean
`survival of these patients is less than 1 year after re-
`lapse.’
`The biologic significance of adrenal androgens in
`prostate cancer refractory to primary hormonal ablation
`was recognized as early as 1945, when H u g p s et al.3
`reported the effects of bilateral adrenalectomy in four
`patients whose disease had progressed after orchiec-
`tomy. Harper et al.,4 in 1974, examined the conversion
`of adrenal androgens to dihydrotestosterone in the
`prostate. Based on that study, it has been estimated that,
`after ablation of testicular androgens, adrenal andro-
`gens could provide as much as one-fifth the original
`stimulus for prostatic cell growth.’ The role of adrenal
`androgen production has been a major focus of second-
`line hormonal therapies directed at inhibition of adrenal
`androgen biosynthesis (e.g., aminoglutethimide, keto-
`conazole, spironolactone), as well as peripheral block-
`ade of adrenal androgen effects (e.g., flutamide and cy-
`proterone acetate).’s5f6
`The use of low dose corticosteroids to suppress the
`pituitary-adrenal axis in advanced prostate cancer has
`been another approach to limit adrenal androgen pro-
`duction. The concept of “medical adrenalectomy” with
`high dose cortisol for patients with advanced prostate
`cancer was introduced by Miller and Hinman in 1954.7
`Several studies have demonstrated that suppressive
`doses of corticosteroids combined with testicular andro-
`
`1
`
`

`

`DXM and PSA in Refractory Prostate Cancer/Storlie et al.
`
`97
`
`gen ablation produces lower serum androgen levels and
`better clinical response thlan testicular androgen abla-
`
`tion a l ~ n e . ~ , ~ Although thlere are also reported studies
`of low dose corticosteroids used alone as second-line
`therapy after failure of primary hormonal therapy, we
`are not aware of previously published work that sys-
`tematically examines prostate specific antigen (PSA)
`levels in men with hormalne-refractory prostate cancer
`receiving low dose corticosteroids as single agent ther-
`apy. We present our results with 38 patients with meta-
`static prostate cancer refractory to orchiectomy who re-
`ceived low dose dexamethasone (DXM) as single-agent
`therapy with PSA levels as a marker of disease status.
`
`Methods
`We retrospectively reviewed 38 patients treated at the
`Mayo Clinic in Rochester, Minnesota, between 1987
`and 1992 who fulfilled the following criteria: (1) prior
`orchiectomy for metastatic prostate cancer; (2) progres-
`sion of disease after initial response to orchiectomy; (3)
`secondary hormonal therapy with low dose DXM (usu-
`ally 0.75 mg twice daily); (4) no other therapy concur-
`rent with DXM during the period of response (including
`radiation); (5) initial PSA 210 ng/ml and on a rising
`trend at start of DXM; andl (6) serial PSA measurements
`obtained during DXM therapy (usually 4-6 month in-
`tervals). Patients who had received DXM but who did
`not meet the above criteria were excluded.
`Prostate specific antigen measurements were used
`to determine response to therapy, with best response
`being the lowest PSA level recorded during therapy and
`progression being the first sustained rise in PSA above
`the lowest measurement. Prostate specific antigen re-
`sponses were categorized into three groups: (1) increase
`above initial value, (2) decrease of <50% of initial
`value, and (3) decrease of 250% of initial value (regres-
`sion). Symptomatic status and available imaging corrc-
`sponding to the date of the lowest PSA were used for
`comparison of biochemical to symptomatic and radio-
`graphic response. Symptomatic response as recorded in
`the patient history by the examining physician was cat-
`egorized as improved, unchanged, or worse than at the
`beginning of DXM therapy. Serial bone scans were re-
`viewed by a blinded nuclear medicine physician and
`graded on the same scale. In cases of heterogeneous re-
`sponse, bone scans were graded based on overall level
`of tumor activity. Because of the limitations of the ret-
`rospective study design, no attempt was made statisti-
`cally to determine any specific correlations, and data are
`reported as descriptive results only.
`
`Results
`
`Twenty-seven (71%) of the patients in this study had
`metastatic prostate cancler at the time of diagnosis,
`
`I
`
`-1 ’
`
`Reduction 4 0 %
`Best PSA Response
`Figure 1. PSA response to low dose dexamethasone in 38 patients
`with progressive metastatic prostate cancer despite orchiectomy.
`
`Increase
`
`-87%
`Reduction >=SO%
`
`whereas in four cases metastases were detected at 15,
`20, 26, and 122 months after initial diagnosis. In seven
`cases, information regarding initial staging was unavail-
`able. The mean age at time of diagnosis was 68 years
`(range, 49-85 years). All of the patients had orchiec-
`tomy as primary therapy for metastatic disease. Nine-
`teen had failed at least one other form of pharmacother-
`apy, and 25 had had radiotherapy before a trial of low
`dose DXM. The range of previous pharmacotherapies
`included megesterol acetate, flutamide, diethylstilbes-
`terol, and leukocyte A human interferon. All patients
`had progressive disease clinically and biochemically
`when DXM was begun. The mean PSA at the start of
`therapy was 238 ng/ml (range, 14.4 to 1430 ng/ml),
`and only two patients had PSA levels below 20 ng/ml
`(17.5 and 14.4 ng/ml). Thirty-six patients received
`DXM 0.75 mg twice daily, and two received 0.75 mg
`three times daily.
`Biochemical responses are shown in Figure 1. After
`initiation of DXM, 30 (79%) of the patients in the study
`had lowered PSA Ievels. In 23 patients (61%), the PSA
`level dropped to less than 50% of the original value,
`with a mean time to best response of 90 days (range,
`39-309 days). Both of the patients receiving DXM 0.75
`mg three times daily were in this group. Thirteen pa-
`tients (34%) had a greater than 80% decrease in PSA.
`For eight patients (21%), the PSA level rose.
`Of the 38 patients studied, 24 (63%) had symptom-
`atic improvement; 20 of these 24 (83%) had concomi-
`tant lowering of PSA levels. Overall, 67% of patients
`with lowered PSA levels also had symptomatic im-
`provement, and 93% had either improvement or stabil-
`ity of symptoms. Of the eight patients with PSA in-
`creases, four (50%) had improvement of symptoms,
`two (25%) had stable symptoms, and two (25%) had
`progressive symptoms (Table 1).
`The mean time to biochemical progression for those
`patients with a greater than 50% decrease in PSA was
`245 days (range, 99-660 days). In the subgroup with a
`>8O% reduction, the mean time to progression was 327
`
`2
`
`

`

`98
`
`CANCER July 2,2995, Volume 76, No. 1
`
`Table 1. Comparison of Symptomatic to Prostate Specific
`Antigen Response to Dexamethasone in Refractory
`Metastatic Prostate Cancer
`
`PSA resuonse*
`
`Imvroved
`
`Stable Worse
`
`Symptomatic response
`
`Increase
`4
`7
`Reduction <50°/o
`Reduction 50-7996
`4
`9
`Reduction z8W0
`Total (90)
`24(63)
`PSA: prostate specific antigen.
`* A total of 67% of patients with decreases in PSA had improvement of symp-
`toms; 93% of uatients with decreases in PSA had stable or imuroved svmutoms.
`
`2
`0
`6
`2
`lO(26)
`
`2
`0
`0
`2
`4(11)
`
`Total (YO)
`8 (21)
`7 (18)
`10 (26)
`13 (34)
`
`days (range, 151-659 days) versus 165 days (range, 99-
`110 days) for those with a 50%-79% reduction, One
`patient in the responder group was lost to follow-up on
`day 123 before progressing biochemically. Two patients
`died of unspecified causes before biochemical progres-
`sion at days 161 and 108. For two patients, DXM was
`discontinued despite biochemical response on days 78
`and 219, because these patients were judged to have
`progressive disease based on other parameters (one had
`worsening symptoms and the other had a worsening
`bone scan).
`Of the 23 patients with decreases in PSA 250%, 8
`(35%) had radiographic evidence of disease regression
`(of whom 6 were evaluated by bone scan and two by
`radiograph only), 5 (22%) were stable, and 7 (30%) had
`disease progression (Table 2). Three (13%) of these pa-
`tients did not have serial radiographic exams. Among
`the seven patients with PSA decrease <50%, none had
`radiographic evidence of disease regression, two (29%)
`
`were stable, and five (71%) had disease progression. Of
`the eight patients with increases in PSA, none had ra-
`diographic evidence of regression, three (38%) were
`stable, and five (62%) had disease progression. In sum-
`mary, 67% of patients with PSA increase or only partial
`response had radiographic evidence of disease progres-
`sion, and none had regression. However, 35% of pa-
`tients with 250% decrease in PSA had radiographic ev-
`idence of disease regression, and 5 7% had either stabil-
`ity or regression of disease (Fig. 2). In light of recent
`reports of PSA decreases after flutamide withdrawal, it
`should be noted that four patients had had flutamide
`discontinued shortly before starting DXM." Two of the
`patients had PSA decreases 250% (with flutamide
`stopped within 1 day of the start of DXM), and two pa-
`tients had PSA decreases t50% of initial PSA value
`(one with flutamide stopped 1 day and the other 9 days
`before the start of DXM). Five other patients had had
`flutamide discontinued before other pharmacothera-
`pies before DXM, of whom three had progression, one
`stability, and one regression of disease after flutamide
`was discontinued.
`
`Discussion
`
`Several studies reported use of low dose corticosteroids
`alone as second-line therapy after failure of primary
`hormonal therapy. Plowman et al." in 1987 demon-
`strated that hydrocortisone (30 mg daily) alone pro-
`duced more suppression of adrenal androgen secretion
`than hydrocortisone plus aminoglutethimide. Similar
`results with hydrocortisone (20 mg twice daily) were re-
`ported by Dowsett et al." in 1988. Tannock et aI.l3 in
`1989 demonstrated a correlation of symptomatic relief
`with lowered serum adrenal androgens in 37 patients
`
`Table 2. Comparison of Radiographic Assessment of Disease Status to Prostate
`Specific Antigen Response to Dexamethasone in Patients With Refractory
`Metastatic Prostate Cancer
`
`Radiographic response
`
`Imaging
`
`Improved
`
`Stable Worse
`
`Unknown Bone scan
`
`X-ray
`
`PSA response*
`Nonresponder group (n = 15)
`Increase
`Reduction <50%
`Total (90)
`Responder group (n = 23)t
`Reduction 50-79%
`Reduction 280%
`Total (90)
`PSA: prostate specific antigen.
`* A total of 67% of nonresponder group had radiographic evidence of disease progression and none had evidence of
`disease regression; 35% of responder group had radiographic evidence of disease regression.
`t Two patients in responder group had flutamide discontinued shortly before dexamethasone therapy.
`
`0
`0
`0
`
`4
`4
`8(35)
`
`3
`2
`5 (33)
`
`1
`4
`5 (22)
`
`5
`5
`lO(67)
`
`2
`5
`7(30)
`
`0
`0
`0
`
`3
`0
`3(13)
`
`6
`3
`
`3
`10
`
`2
`4
`
`4
`3
`
`3
`
`

`

`DXM and PSA in Refractory Prostate Cancer/Storlie et a1.
`
`99
`
`Figure 2. Bone scans (anterior and posterior views) of a patient with
`progressive metastatic prostate cancer despite orchiectomy. (Top
`panels) Day 1 of dexamethasone, when PSA measured 1430 ng/ml.
`(Bottom panels) Day 309, when PSA reached a nadir of 1.2 ng/ml.
`The apparently new rib lesion actually corresponds to a rib contusion
`sustained just before the follow-up bone scan.
`
`with advanced prostate cancer treated with low dose
`prednisone (7.5-10 mg daily) who were no longer re-
`sponding to primary hormone ablation. However, these
`investigators were unable to verify objective disease re-
`gression with prostatic acid phosphatase levels, X-rays,
`or bone scans. Patel et ah'' reported in 1990 that of 23
`patients with hormone-refractory prostate cancer re-
`ceiving low dose DXM (0.75 mg by mouth twice daily),
`2 had objective disease regression and the remaining 21
`had clinically stable disease for a mean duration of 86
`days. None of these studies examined clinical response
`to DXM in the context of PSA levels.
`Although objective assessment of disease status in
`metastatic prostate cancer presents some difficulty, se-
`rum PSA is elevated in 90% of patients with hormone-
`refractory disease.I5 Miller et a1.I6 in 1992 demonstrated
`that serum PSA is a sensitive monitor of objective re-
`sponse to primary hormonal ablation. In addition, Kelly
`et al.I7 demonstrated that decreasing PSA levels in re-
`sponse to treatment correlate with better survival.
`Gerber and Chodak" in 1990 demonstrated a signifi-
`cant correlation between symptom improvement and
`decreased serum levels of PSA in response to ketocona-
`zole and prednisone in 15 patients with hormone-re-
`fractory metastatic prostate cancer. Of 12 patients with
`bone pain and/or other symptoms of progressive &s-
`ease, 9 (75%) experienced symptomatic relief, which,
`with one exception, was accompanied by decreased
`PSA levels.
`Although ours is a retrospective study with inher-
`ent limitations, numerous conclusions can be drawn
`from our data. First, it is clear that low dose DXM may
`produce important symptomatic improvement in pa-
`tients with hormone-refractory prostate cancer, an un-
`fortunately large and challenging patient population.
`Whether this effect is produced by actuaI regression of
`disease or by independent actions of systemic steroids is
`unclear. That most symptomatically improved patients
`also had significant decreases in PSA in the absence of
`any other form of antitumor therapy suggests some
`effect on their underlying disease. Although it is unclear
`whether the decrease in PSA represents disease regres-
`sion or some independent effect of DXM on PSA pro-
`duction, a substantial proportion of patients who re-
`sponded to DXM symptomatically and biochemically
`also had radiographic evidence of disease regression.
`These results are in agreement with other investigators'
`
`4
`
`

`

`100
`
`CANCER July 1,1995, Volume 76, No. 1
`
`findings that substantial PSA reductions do correlate
`with tumor response.
`The design of the current study had numerous lim-
`itations. As a retrospective chart review, it lacks any
`form of control group. Also, the large percentage of re-
`sponders (symptomatic and biochemical) may repre-
`sent a selection bias in that some patients who did not
`respond well to DXM might not have returned for initial
`follow-up and thus been excluded from the study. In
`addition, because there were no consistent objective
`symptomatic criteria, determination of symptomatic re-
`sponse was dependent on the subjective impression of
`the examining physician as recorded. The radiographic
`assessment of disease status was also not uniform
`among these patients. Twenty-two (58%) patients had
`serial bone scans, most of whom were in the responder
`group. However, 13 (34%) had serial X-rays, usually to
`view a particular region that was producing symptoms
`and not to assess overall tumor status. Last, although
`serial PSA measurements were monitored over rela-
`tively consistent intervals, follow-up intervals were
`generally in the range of several months rather than
`weeks, which makes determination of time to regres-
`sion and time to progression less accurate. Despite these
`limitations, the results of the current study suggest that
`low dose DXM may effect tumor regression, as evi-
`denced by symptomatic improvement, improvement in
`radionuclide bone scans, and substantial decreases in
`PSA levels. Additional, carefully controlled prospective
`studies of DXM for patients with hormone-refractory
`prostate cancer are warranted.
`
`References
`
`1.
`
`2.
`
`3.
`
`4.
`
`Geller J, Albert JD. Adrenal androgen blockade in relapsed pros-
`tate cancer. EurJCancerClin Oncol 1985;21:1127-31.
`Grayhack JT, Keeler TC, Kozlowski JM. Carcinoma of the pros-
`tate: hormonal therapy. Cancer 1987;60:589-601.
`Huggins C , Scott WW. Bilateral adrenalectomy in prostatic can-
`cer; clinical features and urinary excretion of 17-ketosteroids and
`estrogens. Ann Surg 1945;122:1031-41.
`Harper ME, Pike A, Peeling WB, Griffiths K. Steroids of adrenal
`
`7.
`
`8.
`
`10.
`
`origin metabolized by human prostatic tissue both in vivo and in
`vitro. J Endocrinol 1974; 60:117-25.
`5. Frank IN, Graham SD, Nabors WL. Urologic and male genital
`cancers. In: Hollab AI, editor. Textbook of clinical oncology. At-
`lanta: The American Cancer Society, 1991:271-89.
`6. Trachtenberg J. Hormonal therapy in metastatic prostatic cancer.
`In: Bruce AW, Trachtenberg J, editors. Adenocarcinoma of the
`prostate. New York: Springer-Verlag, 1987:173-84.
`Miller GM, Hinman F. Cortisone treatment in advanced carci-
`noma of the prostate. 1 Urol 1954;72:485-96.
`Stahl F, Schnorr D, Bar C, Frohlich G, Dorner G. Suppression of
`plasma androgen levels with a combination therapy of depo-
`estrogen (Turisterona) and Dexamethasonem in patients with
`prostate cancer. Exp Clin Endocrind 1989;94:239-43.
`9. Williams G, Asopa R, Abel PD, Smith C. Pituitary adrenal and
`gonadal endocrine suppression for primary treatment of pros-
`tate cancer. RrJ Urol 1990;65:504-8.
`Kelly WK, Scher HI. Prostate specific antigen decline after anti-
`androgen withdrawal: the flutamide withdrawal syndrome. 1
`Urol 1993; 149:607-9.
`11. Plowman PN, Perry LA, Chard T. Androgen suppression by hy-
`drocortisone without aminoglutethimide in orchiectomised men
`with prostatic cancer. Br J Urol 1987;59:255-7.
`Dowsett M, Shearer RJ, Ponder BAJ, Malone P, Jeffcoate SL. The
`effects of aminoglutethimide and hydrocortisone, alone and
`combined, on androgen levels in post-orchiectomy prostatic
`cancer patients. BrJ Cancer 1988;57:190-2.
`Tannock I, Gospodarowicz M, Meakin W, Panzarella T, Stewart
`L, Rider W. Treatment of metastatic prostatic cancer with low-
`dose prednisone: evaluation of pain and quality of life as prag-
`matic indices of response. J Clin Otzcol 1989;7:590-7.
`Pate1 SR, Kvols LK, Hahn RG, Windschitl H, Levitt R, Therneau
`T. A phase I1 randomized trial of megestrol acetate or dexameth-
`asone in the treatment of hormonally refractory advanced carci-
`noma of the prostate. Cancer 1990;66:655-8.
`Seidman AD, Scher HI, Petrylak D, Dershaw DD, Curley T. Es-
`tramustine and vinblastine: use of prostate specific antigen as a
`clinical trial end point for hormone refractory prostatic cancer. J
`Urol 1992; 147:931-4.
`Miller JI, Ahmann FR, Drach GW, Emerson SS, Bottaccini MR.
`The clinical usefulness of serum prostate specific antigen after
`hormonal therapy of metastatic prostate cancer. ] Urol 1992; 147:
`956-61.
`Kelly WK, Scher H, Mazumdar M, Vlamis V, Schwartz M, Fossa
`F. Prostate-specific antigen as a measure of disease outcome in
`metastatic hormone refractory prostate cancer. ] Clin Oncol
`1993; 11:607.
`Gerber GS, Chodak GW. Prostate specific antigen for assessing
`response to ketoconazole and prednisone in patients with hor-
`mone refractory metastatic prostate cancer. J Urol 1990; 144:
`11 77-9.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`5
`
`