Follow-Up Evaluation of Prostate CancerPatients Infused With Autologous Dendritic

Cells Pulsed With PSMA Peptides

B.A. Tjoa,1* S.J. Erickson,1 V.A. Bowes,1 H. Ragde,1 G.M. Kenny,1 O.E. Cobb,2R.C. Ireton,2 M.J. Troychak,2 A.L. Boynton,3 and G.P. Murphy1

1Pacific Northwest Cancer Foundation, Cancer Research Division, Northwest Hospital,Seattle, Washington

2Northwest Hospital, Seattle, Washington3Department of Molecular Medicine, Northwest Hospital, Seattle, Washington

BACKGROUND. We recently conducted a phase I clinical trial administering autologousdendritic cells pulsed with prostate-specific membrane antigen (PSMA) peptides to advancedprostate cancer patients. Participants were divided into 5 groups receiving 4 or 5 infusions ofpeptides alone (PSM-P1 or -P2; groups 1 and 2, respectively), autologous DC (group 3), or DCpulsed with PSM-P1 or -P2 (groups 4 and 5, respectively). Seven partial responders wereobserved. Follow-up evaluation of these responders is presented in this report.METHODS. Clinical monitoring for hematological studies and prostate markers was con-ducted up to 370 days from the start of the phase I study. Data collected include: lympho-cyte, hematocrit, alkaline phosphatase, prostate-specific antigen (PSA), free PSA, and PSMAlevels.RESULTS. Groups 4 and 5 (patients infused with DC pulsed with PSM-P1 or -P2) represen-ted 5/7 responders. The length of response was between 100 days (1 patient) to 200 days orabove (6 patients). Four patients still remained responsive at the end of the period of obser-vation.CONCLUSIONS. The responses observed in this phase I clinical trial are significant and oflong duration. Most of the responders were in treatment groups infused with DC pulsed withPSM-P1 or -P2, suggesting the requirement of both components for effective immunotherapy.Prostate 32:272–278, 1997. © 1997 Wiley-Liss, Inc.

KEY WORDS: dendritic cells; T cell; immunotherapy; prostate cancer; phase I trial;follow-up observation

INTRODUCTION

Prostate cancer is the most common cancer in adultmen and the second leading cause of male cancerdeath in the United States [1]. Conventional treatmentfor advanced prostatic cancer has failed to demon-strate curative potential in all patients [2]. Recent stud-ies suggest that immunotherapy may provide an ef-fective treatment for various cancers, including mela-noma, breast, and ovarian cancers [3–7]. We recentlyreported our phase I clinical trial to assess the safety ofadministration of autologous dendritic cells (DC) in

combination with HLA-A0201-specific peptides fromprostate-specific membrane antigen (PSMA) [8].PSMA is a specific prostate antigen recognized bymonoclonal antibody 7E11.C5 [9]. Levels of PSMA areelevated in the serum of hormone-refractory advanced

*Correspondence to: Benjamin A. Tjoa, Ph.D., Pacific NorthwestCancer Foundation, Northwest Hospital, 120 Northgate Plaza, Suite205, Seattle, WA 98125.Received 1 February 1997; Accepted 1 June 1997

The Prostate 32:272–278 (1997)

© 1997 Wiley-Liss, Inc.

prostate cancer patients [10]. DC, known as the mostefficient antigen-presenting cell type, were culturedfrom patients’ peripheral blood, pulsed with PSMApeptides, and infused back into the patients’ bodies[8]. Patients received 4 or 5 dosages of infusion duringthe study, which spanned 6–8 months. Infusions oftest substances were well-tolerated by all study par-ticipants. Cellular response and decrease in PSA levelwere observed in some patients. Seven partial re-sponders were identified based on National ProstateCancer Project (NPCP) criteria + PSA. In this report,we present the follow-up evaluation of the 7 partialresponders. Comparisons of prostate marker levels, aswell as clinical monitoring data between responderand nonresponder populations, are discussed.

MATERIALS AND METHODS

Reagents and Cytokines

PSMA peptides with HLA-A0201-specific motif(PSM-P1, LLHETDSAV; PSM-P2, ALFDIESKV) weresynthesized and purified (>95% purity) by PeninsulaLaboratories, Inc. (Belmont, CA) and obtained asa lyophilized powder. The powder was dissolved in0.9% saline (USP 0.9%, sodium chloride injection,American Reagent Laboratories, Shirley, NY) to aconcentration of 2 mg/ml. The peptide solution wassterilized using a 0.2-mm filtration. Sterile 0.9% salinewas added to achieve concentrations of 200, 20, and 2mg/ml. One hundred-microliter aliquots of the threepeptide concentrations containing 20, 2, or 0.2 mg pep-tide were transferred into sterile containers and frozenuntil use. Granulocyte-macrophage colony-stimu-

lating factor (GM-CSF), approved for human use, wasprovided by Immunex Corp. (Seattle, WA). Interleu-kin-4 (IL-4), approved for human use, was providedby the Schering-Plough Research Institute (Kenil-worth, NY).

DC Culture

DC were cultured as described previously [11].In short, peripheral blood was drawn, and peripheralblood mononuclear cells (PBMC) were isolated us-ing Lymphoprep (GIBCO-BRL, Gaithersburg, MD)density gradient centrifugation. PBMC were resus-pended in complete medium (OPTIMEM medium[GIBCO-BRL] and 5% heat-inactivated autologousplasma) and plated in a 75-cm2 tissue culture flask(2–3 × 107 cells/flask). Cell suspensions were in-cubated in a humidified incubator (37°C, 5% CO2) for60 min. Nonadherent cells were removed and adher-ent cells were washed gently with warm (37°C) com-plete medium. Dendritic cell propagation medium(DCPM: complete medium, 1,000 units/ml GM-CSFand 1,000 units/ml IL-4) was added to the adherentcells (10 ml/flask). These cells were cultured for 4–6days.

Patient Population

Fifty-one patients with advanced, hormone-resistant prostate cancer were selected for the study,which included a signed informed consent. In addi-tion, patients were tested for HLA-A2 expression us-ing flow cytometric analysis with a monoclonal anti-body specific for HLA-A2 (BB7.2) [12], as well as poly-merase chain reaction.

Treatment Groups

Participants were divided into five treatmentgroups. In the first group, the patients were infusedwith PSM-P1 (LLHETDSAV) peptide at 0.2, 2.0,and 20.0 mg; group 2 received PSM-P2 peptide(ALFDIESKV) at the same concentrations as in group1; group 3 received autologous DC at 1, 5, 10, or 20 ×106 per infusion; groups 4 and 5 received up to 2 × 107

autologous DC pulsed for 2 hr with 1 mg/ml PSM-P1or PSM-P2, respectively. Every group received 4 or 5dosages of test substance at 6–8-week intervals duringthe study period.

Clinical Monitoring

Patients were followed before, during, and aftertreatment with periodic PSA (prostate-specific anti-

TABLE I. Five Treatment Arms of the Phase IClinical Trial

Treatment arm Dose

Group 1: PSM-P1 0.2 mg2 mg

20 mgGroup 2: PSM-P2 0.2 mg

2 mg20 mg

Group 3: Autologous DC 1 × 106

5 × 106

1 × 107

2 × 107

Group 4: DC/PSM-P1a Up to 2 × 107

Group 5: DC/PSM-P2a Up to 2 × 107

aPSM-P1 or -P2 (1 mg/ml/106 DC) was incubated in the pres-ence of autologous DC, as described in Materials and Methods.

Follow-Up of Prostate Cancer Immunotherapy 273

gen; Tandem-E PSA kit, Hybritech Inc., San Diego,CA), free PSA (Tandem-R PSA kit, Hybritech, Inc.),PSMA, complete blood counts, CHEM-22, initial chestX-ray, bone scan, and CAT-scan of the pelvis. All test-ing was conducted on an outpatient basis at North-west Hospital. Patients were also evaluated every in-fusion cycle by one of the study physicians.

RESULTS

Phase I Clinical Trial (General)

The study examined the administration of HLA-A0201-specific PSMA peptides (PSM-P1 and PSM-P2),

autologous DC, and PSM-P1 and -P2 pulsed autolo-gous DC to 51 patients with advanced hormone-refractory prostate cancer. The majority of these pa-tients (39/51) were in stage D2. Many of them wereanemic and had undergone various treatments whichresulted in impaired immune competency. Less than25% of this population was considered fully immuno-competent at the start of the study, as assessed bydelayed-type hypersensitivity (DTH) skin tests. Studyparticipants were divided into 5 treatment groups, asdescribed in Materials and Methods and summarizedin Table I. At the completion of 4 cycles of infusions,maximum tolerated dose was not yet achieved. Nei-ther significant acute nor chronic toxicity was ob-served in all doses of test substances, except for mild-to-moderate cases of hypotension during time of infu-sion. In addition, no significant increase in serumtumor necrosis factor a (TNF-a) or interferon g (IFN-g) was measured during the course of the study.

Patients were monitored for cellular immunemodulation to the appropriate PSMA peptides(PSM-P1 or -P2). An increase of cellular responsewas observed within the HLA-A2-positive subpopu-lation of groups 4 and 5. Patients’ clinical respon-ses were analyzed based on NPCP criteria + PSA.Seven partial responders were observed (NPCP plus>50% decrease in PSA).

Comparison of Clinical Monitoring and ProstateMarkers Between Responder and

Nonresponder Groups

Average values of clinical monitoring studies, i.e.,total lymphocytes, hematocrit and alkaline phospha-tase, and prostate markers, i.e., PSA, PSMA, and per-cent free PSA, are computed and compared betweenthe 7 partial responders (designated as responders)and the rest of the study population (designated asnonresponders). No significant change in levels of to-

Fig. 1. Total alkaline phosphatase levels of responder and non-responder populations. Normal ranges for total alkaline phospha-tase are 25–125 IU/l. Mean values and standard error of the mean(SEM) were obtained from patient sera drawn preinfusion and 7days postinfusion.

TABLE II. Summary of Clinical Monitoring for Hematological Studies*

Category Category

Preinfusion Postinfusion 1 Postinfusion 2 Postinfusion 3 Postinfusion 4

Mean SEM Mean SEM Mean SEM Mean SEM Mean SEM

Hematocrit Nonresponder 37.49 1.46 36.65 1.47 36.75 1.76 37.11 1.69 37.55 1.72Responder 34.87 1.39 35.07 1.38 35.9 1.1 35.16 1.76 36.11 1.33

Totallymphocytes Nonresponder 1,170 202 1,066 187 1,132 193 1,115 192 1,173 276

Responder 811 222 900 204 932 169 851 142 793 119

*Normal ranges are: absolute lymphocyte = 1,000–3,500 × 10V3/ml; hematocrit = 40–52%. Mean values and standard error of the mean(SEM) are obtained from patient sera obtained preinfusion and 7 days postinfusion.

274 Tjoa et al.

tal lymphocytes or hematocrit was observed in re-sponder or nonresponder populations throughout thestudy period (Table II). Average alkaline phosphataselevel decreased after the first infusion among the re-sponder group, while it showed a slight increasewithin the nonresponder population (Fig. 1). AveragePSA levels showed an increase in the nonrespondergroup, both those with low (0–19) and high (>19) pre-infusion values, while a decrease was observed in the7 partial responders (Fig. 2). No significant change inPSMA level was observed throughout the 4 infusionsin both populations (Fig. 3). Levels were higher inresponders compared to nonresponders. Free PSA lev-els in the responder group with low PSA generallyrose toward normal (Table III). In both the nonre-sponder and responder groups, when PSA levels werehigh, percent free PSA levels were of no diagnostic orprognostic value (Table III).

Long-Term Follow-Up of the SevenPartial Responders

The seven partial responders were monitored up to370 days from the start of the study. Clinical data col-

Fig. 3. Total alkaline phosphatase levels of responder and non-responder populations. Normal ranges for PSMA are 0.08–0.21.Mean values and standard error of the mean (SEM) were obtainedfrom patient sera drawn preinfusion and 7 days postinfusion.

Fig. 2. PSA levels of responder and nonresponder populations. Each group is divided into low (initial PSA values 0–19) and high (initialPSA values >19) PSA categories. Normal ranges for PSA are 0–4 ng/ml. Mean values and standard error of the mean (SEM) were obtainedfrom patient sera drawn preinfusion and 7 days postinfusion. There were 24 patients in the nonresponder and 2 patients in the responderlow PSA groups. There were 20 patients in the nonresponder and 5 patients in the responder high PSA groups. In the latter, the mean PSApreinfusion for the nonresponder group was 175.0 ng/ml. At the end of four infusions, it was 583.4 ng/ml. For the responder group, it was60.0 ng/ml at the preinfusion level and 24.4 ng/ml at completion. In both groups, significant differences in pre- vs. postinfusion values (P <0.05) were observed.

Follow-Up of Prostate Cancer Immunotherapy 275

lected include CBC (complete blood count), SMA-12,and prostate marker levels. Table IV summarizes eachpatient’s length of response time. Four of the 7 partialresponders continued to respond beyond the follow-up period, at which date these patients were enteredinto a phase II study. The length of responses are over220, 220, 260, or 230 days for patients 17, 35, 38, and 47,respectively. Two patients responded up to 200 days(patients 10 and 39), while one patient stopped re-sponding after 100 days (patient 37). Figures 4–10show prostate marker levels (PSA, percent free PSA,and PSMA), as well as total alkaline phosphatase lev-

Fig. 4. A partial responder from group 1 who responded for200 days.

TABLE IV. Partial Responders: Length of Response*

Patient Group HLA-A2

Length ofResponse

(days) Figure

10 1 Positive 200 417 2 Positive >200a 535 4 Negative >200a 637 4 Positive 100 738 4 Negative >260a 839 4 Negative 200 947 5 Positive >230a 10

*Responses were determined based on NPCP criteria + PSA.aPatient continued to respond beyond end of observation period.After this date, the patient was entered into a phase II study.

Fig. 5. A partial responder from group 2 whose hormone treat-ment was changed to 1 mg/qd DES. The initial bone scan was ofsuperscan type. There are no major improvements in the bone scan,although the bone and total alkaline phosphatase are now normal.

TABLE III. Comparison of Percent Free PSA Levels*

Category Subgroup Group

Preinfusion Postinfusion 1 Postinfusion 2 Postinfusion 3 Postinfusion 4

Mean SEM Mean SEM Mean SEM Mean SEM Mean SEM

Percentfree PSA Low PSA Nonresponder 10.34 6.43 12.96 8.11 13.06 8.35 11.4 6.04 13.68 5.97

Responder 8.6 1.27 11.75 2.28 12.45 4.28 9.5 1.2 11.95 1.33High PSA Nonresponder 22.67 7.24 19.46 5.99 19.93 5.32 19.73 5.66 16.12 4.71

Responder 14.4 1.34 17.05 4.81 26.58 8.92 20.73 2.48 10.45 3.1

*Each group is divided into low (initial PSA values 0–19) and high (initial PSA values >19) PSA categories. Normal ranges for thepercent free PSA marker are 17.86 ± 1.89 [26]. Mean values and standard error of the mean (SEM) are obtained from patient seraobtained preinfusion and 7 days postinfusion.

276 Tjoa et al.

Fig. 6. A partial responder from group 4 who persisted to re-spond >220 days.

Fig. 7. A partial responder from group 4 with a short (100-day)period of response.

Fig. 8. A partial responder from group 4 who persisted to re-spond >260 days.

Fig. 9. A partial responder from group 4 with a period of re-sponse of 200 days.

Follow-Up of Prostate Cancer Immunotherapy 277

els, of all 7 partial responders throughout the course ofthe study and follow-up.

DISCUSSION

Three of the initial patients (Table IV, and Figs. 4, 7,and 9) did not continue to respond over the postinfu-sion observation period. Patient 10 (Fig. 4) was HLA-A2-positive and received only PSM-P1. Patient 37 wasHLA-A2-positive and received DC pulsed with PSM-P1. Patient 39 (HLA-A2-negative) received the sametreatment as patient 37. There appears to be no appar-ent basis for the duration of response (100–200 days)and type of treatment. In contrast, the two longest pa-tient responders were patient 38 (Fig. 8) and patient 47(Fig. 10). There is no detectable treatment-related reasonfor this response duration. Overall, however, the major-ity of responders were in groups 4 and 5, who weregiven autologous DC pulsed with PSMA peptides.

It is unknown how long the entire duration ofthese measurable responses could be, as at one point(Table IV), all patients were entered into a phase IItrial and given additional treatment which includedautologous DC pulsed with both PSM-P1 and -P2.In general, these responses must be considered asvalid as they have existed for prolonged periods oftime, and there have been multiple evaluations. Thetype of response using these four serum markers mustbe considered as relatively objective. All responders also

had improvements in secondary signs including weightgain, relief of pain, and good performance status.

Dendritic cell therapy using PSMA peptides will befurther evaluated in an ongoing phase II trial. Thisappears to be an acceptable new modality for prostatecancer treatment.

ACKNOWLEDGMENTS

We acknowledge the help and assistance of YvonneVowell, R.N., B.S.N., Northwest Hospital Administra-tion/Development Division, Tana Olson, NorthwestHospital Cancer Registrar, and the nursing staff ofNorthwest Hospital Day Surgery. This work was sup-ported by grants from the CaPCURE Foundation andthe Phi Beta Psi Sorority.

REFERENCES

1. Parker SL, Tong T, Bolden S, Wingo PA: Cancer statistics, 1996.CA 1996;46:5–27.

2. Jones G, Mettlin C, Murphy G, Guinan P, Herr H, Hussey D,Chmiel J, Fremgen A, Clive R, Zuber-Ocwieja K, Winchester D:Patterns of care for carcinoma of the prostate gland: Results of anational survey of 1984 and 1990. J Am Coll Surg 1995;180:545–554.

3. Traversari C, van der Bruggen P, Luescher IF, Lurquin C,Chomez P, Van Pel A, De Plaen E, Amar-Costesec A, Boon T: Anonapeptide encoded by human gene MAGE-1 is recognized onHLA-A1 by cytolytic T lymphocytes directed against tumor an-tigen MZ2-E. J Exp Med 1992;176:1453–1457.

4. Peoples GE, Goedegebuure PS, Smith R, Linehan DC, Yoshino I,Eberlein TJ: Breast and ovarian cancer-specific cytotoxic T lym-phocytes recognize the same HER2/neu-derived peptide. ProcNatl Acad Sci USA 1995;92:432–436.

5. Ioannides CG, Fisk B, Pollack M, Frazier ML, Wharton JT,Freedman RS: Cytotoxic T cell clones isolated from ovarian tu-mor-infiltrating lymphocytes recognize common determinantson non-ovarian tumour clones. Scand J Immunol 1993;37:413–424.

6. Morton DL, Foshag LJ, Hoon DSB: Prolongation of survival inmetastatic melanoma after active specific immunotherapy with anew polyvalent melanoma vaccine. Ann Surg 1992;216:463–482.

7. Pardoll D: New strategies for active immunotherapy with geneti-cally engineered tumor cells. Curr Opin Immunol 1992;4:619–623.

8. Murphy G, Tjoa B, Ragde H, Kenny G, Boynton A: Phase Iclinical trial: T-cell therapy of prostate cancer using autologousdendritic cells pulsed with HLA-A0201-specific peptides fromprostate-specific membrane antigen. Prostate 1996;29:371–380.

9. Horoszewicz JS, Kawinski E, Murphy GP: Monoclonal antibodiesto a new antigenic marker in epithelial prostate cells and serum ofprostatic cancer patients. Anticancer Res 1987;7:927–936.

10. Murphy G, Ragde H, Kenny G, Barren R III, Erickson S, Tjoa B,Boynton A, Holmes E, Gilbauch J, Douglas T: Comparison of pros-tate specific membrane antigen, and prostate specific antigen lev-els in prostatic cancer patients. Anticancer Res 1995;15:1473–1480.

11. Tjoa B, Erickson S, Barren R III, Ragde H, Kenny G, Boynton A,Murphy G: In vitro propagated dendritic cells from prostatecancer patients as a component of prostate cancer immuno-therapy. Prostate 1995;27:63–39.

12. Parham P, Brodsky FM: Partial purification and some properties ofBB7.2. A cytotoxic monoclonal antibody with specificity for HLA-A2 and a variant of HLA-A28. Hum Immunol 1981;3:277–299.

Fig. 10. A partial responder from group 5 who persisted torespond >230 days. He underwent palliative radiation therapy tohis right shoulder and lower spine (LS) 3 weeks prior to the firstinfusion. Repeat bone scans show continued improvement.

278 Tjoa et al.