• hormone therapy;
  • metastatic prostate carcinoma;
  • low prostate-specific antigen;
  • response duration


  1. Top of page
  2. Abstract
  6. Acknowledgements


Although < 1% of men present with prostate-specific antigen (PSA)-negative prostate carcinoma, in that they have serum PSA levels much lower than the tumor burden would suggest, such patients represent a management dilemma. To the authors' knowledge, little information exists in the literature regarding patterns of disease and response to treatment. The authors wished to define the clinical features of this patient group.


The British Association of Urological Surgeons Cancer Registry 2000 and 2001 data bases were used to identify the clinical features and outcome of 33 men with metastatic prostate carcinoma who presented with serum PSA levels < 10 ng/mL. Clinical notes and histopathology were reviewed for each patient.


Seventeen patients (51%) presented with urinary symptoms and/or pelvic pain, 6% with cachexia and 21% with bone pain. Characteristic bone metastases were present in 81% of patients, similar to the presentation of men with high serum PSA levels. Hypercalcemia was a feature in 9% of patients. Visceral metastases were present in two patients. The median response duration to first-line hormone manipulation was 7 months. No responses were seen in 11 of 13 patients who received second-line hormones or to any third-line treatment. Three of 5 patients who received chemotherapy responded but developed recurrent disease within 8 weeks of treatment cessation. The median overall survival was 12 months.


The presentation of patients with treatment-naïve PSA-negative metastatic prostate carcinoma is similar to that of patients with high serum PSA levels, but their median survival and response duration to first-line hormone therapy are of much shorter duration. Second-line hormone therapy is ineffective, but early chemotherapy may be beneficial. Hypercalcemia is a particular feature in this group of patients. Cancer 2003. © 2003 American Cancer Society.

Prostate carcinoma is the second most common malignancy in males in the U.K., accounting for 10,000 deaths per annum. Serum levels of prostate-specific antigen (PSA) have been used in the diagnosis and management of patients with prostate carcinoma since 1984.1 The level of serum PSA usually correlates with the volume of tumor present; hence, in the great majority of patients with metastatic disease, PSA levels are increased markedly.2, 3 However, small but significant numbers of patients with metastatic disease, typically < 1%,4 present with abnormally low serum levels of PSA, < 10 ng/mL (“PSA-negative” disease).5–7 To our knowledge, information is sparse concerning this group of tumors, although they have been associated with high tumor grade (Gleason score, 8–10).3, 8

Based on anecdotal evidence, it is believed that overall survival is poorer in these patients who have low PSA levels compared with patients who have PSA-positive metastatic prostate carcinoma.5 The pattern of metastatic disease is considered to be unlike the pattern observed in tumors with high PSA levels, with increased numbers of visceral and soft tissue metastases (up to 55% of patients) and a nontypical pattern of bone disease seen.9 However, these latter data were obtained from patients with androgen independent, clinically progressive disease rather than patients who presented with treatment-naïve metastatic prostate carcinoma with low serum PSA levels. The objective of this study was to provide a better description of the clinical features of patients who present with metastatic prostate carcinoma and serum PSA levels < 10 ng/mL in the largest group of patients characterized to date.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Patients who presented with histologically proven metastatic prostate carcinoma and a serum PSA level < 10 ng/mL were identified from the British Association of Urological Surgeons (BAUS) Cancer Registry. Data were collected from the 2000 and 2001 data bases. All data base information was anonymized, with the patient identifiable by hospital identification and date of birth, referring consultant, and referring center. Details on disease stage, tumor grade, presenting PSA level, and treatment decision were extracted. There was no policy of exclusion of patients who had tumors with neuroendocrine histology. Referring consultants were contacted by mail and were asked to forward consent forms and study information sheets on to patients. Agreement had been obtained previously from the Multicenter Research Ethics Committee that consent was not needed from next of kin for patients who had died. Patients also were also identified from local referrals to urooncologists at the Meyerstein Institute of Oncology at the Middlesex Hospital.

One investigator then reviewed the patient's notes at centers nationwide. Presentation, pattern of metastases, initial PSA level, histology, treatment and treatment response times, and overall survival were reviewed. The definition by Hall et al. of clinical progression as new symptoms or signs, or worsening of already existing symptoms or signs, clearly attributable to prostate carcinoma was used,10 together with the time to first subjective progression, which was evaluated using response criteria similar to those used by the European Organization for Research and Treatment of Cancer in their Phase III 30853 study of hormone therapy in patients with metastatic prostate carcinoma.11 This was defined as a progression of urological symptoms with the appearance of severe symptoms requiring surgical relief or catheterization, weight loss > 10% within 1 year, or deterioration in World Health Organization (WHO) performance status. An increase in pain score, as reflected by a step-up in the WHO analgesia classification and ladder,12 also was considered to reflect progression. If sequential radiologic assessments were performed, then disease progression was defined as the appearance of new hotspots on bone scan, or new lesions on X-ray film, or an increase > 25% in index lesions measured as the sum of products of the largest perpendicular dimensions. PSA response, which often is used as a surrogate marker for treatment response, was inappropriate for this study. Survival curves were plotted using the Kaplan–Meier method. Archival prostate blocks were obtained for a pathologic and morphologic review at the Department of Histopathology, University College London Hospital. Immunohistochemical and molecular biology studies form the basis of an ongoing investigation.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Sixty patients were identified from the BAUS Cancer Registry from a total of 2708 patients with metastatic prostate carcinoma who were diagnosed during the 2000 and 2001 data base collection period, with the PSA-negative group representing 2.21% of the total number of registered patients with metastatic disease. Forty-one replies were elicited for those 60 patients. Of these, 14 patients were excluded either because review of the clinical notes revealed that they had serum PSA levels > 10 ng/mL (range, 22.1–121.0 ng/mL) at presentation, lack of evidence of metastatic disease, an equivocal diagnosis of prostate carcinoma, or refusal by the consultant to approach the patient for consent (due to factors such as terminal illness). Six other patients were collected from referrals to the local cancer center at the Meyerstein Institute of Oncology. Thirty-three sets of clinical case notes were reviewed. Three additional patients who were registered with neuroendocrine differentiation were omitted, because clinical case notes were missing and/or because no pathology blocks were available.

Patient Demographics

The median patient age was 67 years (range, 48–81 years), and the median serum PSA at presentation was 3.5 ng/mL (range, 0.1–9.8 ng/mL) (Table 1). The median age of the group was slightly younger compared with the median age of the overall group of patients with prostate carcinoma (72.3 years).13 The presenting features, pattern of metastases, and extent of local disease are summarized in Table 1. The pattern of bone disease was typical of prostate carcinoma, predominantly affecting the axial skeleton, pelvis, and long bones.

Table 1. Patient Demographics
CharacteristicNo. of patients (%)
  • PSA: prostate-specific antigen; TURP: transurethral resection of the prostate; BHP: benign hyperplasia.

  • a

    In addition to bone metastases.

Age (yrs) 
PSA at diagnosis (ng/mL) 
Previous TURP for BHP 5 (15)
Previous urinary investigations 6 (18)
Presenting symptoms 
 Urinary symptoms/pelvic pain17 (51)
 Acute retention 6 (18)
 Cachexia/malaise 3 (10)
 Bone pain 7 (21)
Bone metastases27 (81)
 Axial skeleton16 (48)
 Pelvis 9 (27)
 Femur/humerus 8 (24)
Soft tissuea 
 Lung 2
 Liver 1
 Glans penis (cutaneous) 1
Tumor status 
 T2 1
Abdominal lymph node metastases 6 (18)
Bone marrow infiltration 2 (6)


Histopathologic confirmation of disease was obtained in all patients, with 57% of patients diagnosed using core needle biopsy; the other patients underwent transurethral resection of the prostate for symptoms and diagnosis (Table 2). Referral pathology reports suggested that Gleason scores for the PSA-negative group were only slightly higher compared with Gleason scores for all patients with metastatic prostate carcinoma in the BAUS cancer registry (Fig. 1). On review by a single urologic histopathologist, 25 patients (76%) had high-grade tumors with a mean Gleason score of 9 (Table 2). No tumors showed neuroendocrine differentiation.

Table 2. Histopathology
VariableNo. of patients (%)
  1. TURP: transurethral resection of the prostate; TUR: transurethral resection.

Source of material 
 TURP13 (39)
 Prostatic biopsy19 (57
 TUR bladder neck 1
Histology at local center 
 Gleason score 8–1020 (60)
 Gleason score 6–713 (40)
Histology review 
 Gleason score 8–1025 (76)
 Gleason score 6–7 8 (24)
thumbnail image

Figure 1. Histopathologic grading of prostate-specific antigen (PSA)-negative metastatic prostate carcinoma and metastatic prostate carcinoma overall (data from local centers).

Download figure to PowerPoint


All patients received hormone manipulation as primary therapy (Table 3), with a subjective response rate of 73%. Benefit was symptomatic (80%), with either a reduction in the amount of analgesia required or improvement of performance status. One patient received initial hormones as part of an intermittent hormone study; these were discontinued until the patient achieved a rising PSA level.

Table 3. Treatment and Response
TreatmentNo. of patientsNo. of responses
  • LHRH: luteinizing hormone-releasing hormone; NSAA: nonsteroidal antiandrogen monotherapy; MAB: maximal androgen blockade.

  • a

    One patient was treated with intermittent hormone therapy.

First-line therapy3324
 LHRH agonistsa2721
Second-line therapy  
 Zoladex restarteda11
Third-line therapy  
  Single agent31

Symptomatic response durations to primary treatment were 0–14 months, with a median time to disease progression of 7 months (Fig. 2). Subsequent treatment most commonly (36%) was the combination of a luteinizing hormone-releasing hormone (LHRH) agonist and nonsteroidal antiandrogen (maximal androgen blockade). Third-line hormone therapy was commenced in 12% of patients. Five of 33 patients (15%) received chemotherapy (Table 3).

thumbnail image

Figure 2. Progression-free survival in patients with prostate-specific antigen-negative metastatic prostate carcinoma.

Download figure to PowerPoint

No evidence of symptomatic benefit was seen in 83% of patients who were placed on maximal androgen blockade or in any patients who started third-line hormones. Three symptomatic and radiologic responses were seen in 5 patients who were treated with chemotherapy (1 patient received single-agent mitozantrone; 2 patients received epirubicin, carboplatin, and 5-fluorouracil), but recurrences occurred shortly after those patients completed treatment (range, 3–7 weeks). The number of chemotherapy cycles ranged from three to six. Sixty-five percent of patients received palliative radiotherapy for symptomatic bone disease, and 12% of patients received palliative radiotherapy for symptoms relating to the primary tumor, such as hematuria.

The frequency of serum PSA estimation subsequent to the initiation of treatment varied considerably. PSA levels fell from baseline values in 6 patients, but an association between a fall in PSA level (i.e., within the 0–10 ng/mL range) and improvement in clinical symptoms was noted in only 2 patients. Only 2 patients had rises in PSA values above 10 ng/mL (27.8 ng/mL and 43.1 ng/mL, respectively) at any time during the course of their disease. Hypercalcemia was observed in 3 patients (9%), either at initial presentation or during disease progression.

The median overall survival was 12 months, with a range of 1–36 months (Fig. 3). All 22 deceased patients died of their disease. With a median follow-up of 19 months, 11 of 33 patients (33%) remain alive: 7 patients who have progressive disease and 4 patients (12%) who are alive and well.

thumbnail image

Figure 3. Overall survival of patients with prostate-specific antigen–negative metastatic prostate carcinoma.

Download figure to PowerPoint


  1. Top of page
  2. Abstract
  6. Acknowledgements

Metastatic prostate carcinoma that is PSA-negative, presenting with abnormally low levels of PSA, accounts for only a small proportion of patients but presents a significant management dilemma. To our knowledge, this study has described the clinical features from the largest series documented to date in this group of PSA negative patients. The median age of the group was slightly younger compared with the age of patients with prostate carcinoma overall (67 years vs. 72 years, respectively).13 The presenting symptoms were typical of metastatic prostate carcinoma in general, and most patients presented with symptoms of advanced local disease or symptomatic bone metastases. Despite reported rates of soft tissue metastases of up to 55%,9 there was a relatively low rate (30%) of non bone disease (either visceral metastases or nonregional lymph node metastases) in this study.

In the staging of newly diagnosed patients, various studies have recommended a threshold below which imaging investigations are unlikely to reveal metastases. In a study of 861 men with PSA levels < 20 ng/mL, 8 patients (0.9%) had positive bone scans, and 13 patients (1.5%) had computed tomography-defined lymph node disease.14 In reviewing 167 patients with M1 prostate carcinoma who presented over an 11-year period, only 8 patients with a serum PSA levels < 10 ng/mL were identified.5

It has been proposed that an undifferentiated component correlates with nonresponsiveness to hormone therapy,15 and other series have been comprised almost exclusively of poorly differentiated and undifferentiated tumors.5 In 3 patients with metastatic prostate carcinoma who had serum PSA levels < 10 ng/mL, all tumors were of high Gleason grade, and the authors postulated that the tumors had lost the characteristics of the original prostate tissue itself, which affected PSA release into the blood stream.16 Other authors also have described the association between advanced prostate carcinoma stage with a low PSA level and poorly differentiated tumors.8 However, although the current series included a high proportion of poorly differentiated tumors with Gleason scores of 8–10 (76% on histology review), the remaining tumors were moderately differentiated. It is possible, of course, that, in these latter tumors, the bulk of the tumor had a different morphology from that sampled. None of the 33 tumors contained morphologic evidence of neuroendocrine differentiation. In three further tumors that were identified from the BAUS cancer registry and were not studied, either because of lack of case notes or because no tissue remained on the pathology blocks, there were features of neuroendocrine differentiation.

Response to treatment in this group of patients is difficult to measure: PSA levels were of limited value, because their baseline values were < 10 ng/mL. There was a fall from baseline PSA levels in six patients, but this was correlated with improvement in clinical symptoms or signs in only two patients. In those two patients, it is likely that the dominant clone produced small amounts of PSA, in contrast to the other four patients, who demonstrated a fall in PSA levels from baseline but had clinically progressive disease. In all but two patients, PSA values never rose above 10 ng/mL, a remarkable observation in the context of widespread, symptomatic, progressive disease. Only in 2 patients were rises above 10 ng/mL seen, reflecting the likely presence of a mixed PSA negative and PSA positive clonal population.

Hypercalcemia is uncommon in patients with prostate carcinoma. Recent data from a bisphosphonate trial suggest < 1% incidence in the placebo arm in men with hormone-refractory metastatic prostate carcinoma.17 Indeed, when reviewed, the rare instances of hypercalcemia in patients with prostate carcinoma have been associated largely with neuroendocrine carcinomas.18 The high rate (9%) of hypercalcemia in the current study reinforces the differing biologic nature of the PSA-negative tumor despite the absence of neuroendocrine differentiation.

The median response duration to first-line therapy in the current study was 7 months, in marked contrast to the response duration of 12–18 months noted in randomized studies of previously untreated patients with metastatic prostate carcinoma.12, 19 Our response rate is similar to that seen in smaller case series of patients with PSA-negative tumors, in which responses to first-line hormone therapy were observed in two of four patients.5 Nonetheless, the majority of patients did demonstrate a symptomatic response to androgen-deprivation therapy, and this remains the recommended initial treatment.

For patients with metastatic prostate carcinoma in recurrence, further responses to subsequent endocrine treatments have been re seen in 35–45%,20, 21 and a median survival of 9 months after the initation of second-line treatment has been documented well.22 In the current study, no responses were noted in 83% of patients who had a nonsteroidal antiandrogen added to their initial LHRH agonist or in any patients in whom treatment with third-line hormones was initiated.

Of five patients who received chemotherapy, three responses were seen and were maintained throughout the chemotherapy course, but those responses were not durable once treatment was completed. Response rates of 10–15% to chemotherapy have been documented in patients with metastatic prostate carcinoma,23 with responses maintained for up to 9 months. More recently, Phase II studies have proved encouraging, with response rates > 40% in selected patients with hormone-refractory disease, for combinations such as 5-fluoruracil, epirubicin, and cisplatin24 or various taxane-containing regimes.25 Given the poor response to second-line or third-line hormone therapy in the PSA-negative group, early chemotherapy may be worth considering in patients with good performance status.

The median overall survival of 12 months for the PSA-negative patients studied was much shorter compared with the survival of patients who were treated for metastatic prostate carcinoma overall, with reports of a median survival of 27 months in patients with metastatic disease19 or 1 year after the first disease progression.11 In another, smaller series, a 3-year survival rate of 33.3% was observed in a group of patients with PSA-negative metastatic prostate carcinoma.5

The current study describes the clinical features in what to our knowledge is the largest series to date of patients who presented with untreated metastatic prostate carcinoma and serum PSA levels < 10 ng/mL: the PSA-negative metastatic group. Although the presentation and pattern of metastases in these patients are similar to those in patients with metastatic disease who have high PSA levels, their response to hormone therapy is poor, with short response durations to first-line treatment and minimal response to subsequent management other than chemotherapy. Hypercalcemia appears to be a prominent feature. Overall survival rates are much lower compared with patients who have high PSA levels, and alternative strategies for the management of patients in the PSA-negative metastatic group, including early chemotherapy in appropriate patients, should be considered.


  1. Top of page
  2. Abstract
  6. Acknowledgements

The authors acknowledge the following individuals for their support: Sarah Fowler, British Association of Urological Surgeons (BAUS) Cancer Registry Data Base Manager; and the contributors to the BAUS Section of Oncology Cancer Registry, in particular, E. Ahiaku, R. A. Blades, F. J. Bramble, S. P. Bramwell, N. R. Boucher, W. G. Bowsher, T. E. Briggs, C. Bunce, C. Carter, D. Chadwick, R. A. Corfield, B. Ellis, J. G. W. Feggetter, M. J. James, W. G. Jones, R. P. Kulkarni, H. G. Kynaston, S. S. Matanhelia, G. S. McIntosh, G. E. Mobb, M. L. Pantelides, A. Paracha, R. D. Pocock, K. K. Prasad, G. Sole, and B. Waymont.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Stamey TA, Yang N, Hay AR, McNeal JE, Freiha FS, Redwine E. Prostate-specific antigen as a serum marker for adenocarcinoma of the prostate. N Engl J Med. 1987; 317: 909916.
  • 2
    Pantelides ML, Bowman SP, George NJ. Levels of prostate specific antigen that predict skeletal spread in prostate cancer. Br J Urol. 1992; 70: 299303.
  • 3
    Daver A, Soret JY, Coblentz Y, Allain YM, Cellier P, Chauveau P. The usefulness of prostate-specific antigen and prostatic acid phosphatase in clinical practice. Am J Clin Oncol. 1988; 11 (Suppl 2): S53S60.
  • 4
    Oesterling JE. Using PSA to eliminate the staging radionuclide bone scan. Significant economic implications. Urol Clin North Am. 1993; 20: 705771.
  • 5
    Yamamoto S, Ito T, Akiyama A, Aizawa T, Miki M, Tachibana M. M1 prostate cancer with a serum level of prostate-specific antigen less than 10 ng/mL. Int J Urol. 2001; 8: 374379.
  • 6
    Oommen R, Geethanjali FS, Gopalakrishnan G, et al. Correlation of serum prostate specific antigen levels and bone scintigraphy in carcinoma prostate. Br J Radiol. 1994; 67: 469471.
  • 7
    Wolff JM, Zimny M, Borchers H, Wildberger J, Buell U, Jakse G. Is prostate-specific antigen a reliable marker of bone metastasis in patients with newly diagnosed cancer of the prostate? Eur Urol. 1998; 33: 376381.
  • 8
    Partin AW, Carter HB, Chan DW, et al. Prostate specific antigen in the staging of localized prostate cancer: influence of tumor differentiation, tumor volume and benign hyperplasia. J Urol. 1990; 143: 747752.
  • 9
    Sella A, Konichezky M, Flex D, Sulkes A, Baniel J. Low PSA metastatic androgen-independent prostate cancer. Eur Urol. 2000; 38: 250254.
  • 10
    Hall R, Hedlund PO, Ackermann R, et al. Evaluation and follow-up of patients with N1–3 M0 or NXM1 prostate cancer in Phase III trials. Urology. 1997; 49: 3945.
  • 11
    Denis LJ, Carnelro de Moura JL, Bono A, et al. Goserelin acetate and flutamide versus bilateral orchiectomy: a Phase III EORTC trial (30853). EORTC GU Group and EORTC Data Center. Urology. 1993; 42: 119129.
  • 12
    Ventafridda V, Saita L, Ripamonti C, De Conno F. WHO guidelines for the use of analgesics in cancer pain. Int J Tissue React. 1985; 7: 9396.
  • 13
    BAUS Section of Oncology. Analysis of minimum data set for urological cancers, January 31 to December 31, 2001. British Association of Urological Surgeons, 2001.
  • 14
    Levran Z, Gonzalez JA, Diokno AC, Jafri SZ, Steinert BW. Are pelvic computed tomography, bone scan and pelvic lymphadenectomy necessary in the staging of prostatic cancer? Br J Urol. 1995; 75: 778781.
  • 15
    Gaffney EF, O'Sullivan SN, O'Brien A. A major solid undifferentiated carcinoma pattern correlates with tumour progression in locally advanced prostatic carcinoma. Histopathology. 1992; 21: 249255.
  • 16
    Kageyama Y, Kihara K, Kamata S, et al. Relationship between pretreatment serum levels of prostate specific antigen and bone metastasis in prostate cancer. Hinyokika Kiyo. 1996; 42: 197199.
  • 17
    Saad F, Bertrand P, Paquin JM, Peloquin F. Usefulness of prostate-specific antigen in the diagnosis of lymphatic metastases in cancer of the prostate. Ann Chir. 1995; 49: 680684.
  • 18
    Smith DC, Tucker JA, Trump DL. Hypercalcemia and neuroendocrine carcinoma of the prostate: a report of three cases and a review of the literature. J Clin Oncol. 1992; 10: 499505.
  • 19
    Eisenberger MA, Blumenstein BA, Crawford ED, et al. Bilateral orchiectomy with or without flutamide for metastatic prostate cancer. N Engl J Med. 1998; 33: 10361042.
  • 20
    Labrie F, Dupont A, Giguere M, et al. Benefits of combination therapy with flutamide in patients relapsing after castration. Br J Urol. 1988; 61: 341346.
  • 21
    Small EJ, Vogelzang NJ. Second-line hormonal therapy for advanced prostate cancer: a shifting paradigm. J Clin Oncol. 1997; 15: 382388.
  • 22
    Hussain M, Wolf M, Marshall E, Crawford ED, Eisenberger M. Effects of continued androgen-deprivation therapy and other prognostic factors on response and survival in Phase II chemotherapy trials for hormone-refractory prostate cancer: a Southwest Oncology Group report. J Clin Oncol. 1994; 12: 18681875.
  • 23
    Brausi M, Jones WG, Fossa SD, et al. A high dose epirubicin is effective in measurable metastatic prostate cancer: a Phase II study of the EORTC Genitourinary Group. Eur J Cancer. 1995; 31A: 16221626.
  • 24
    Chao D, von Schlippe M, Harland SJ. A Phase II study of continuous infusion 5-fluorouracil (5-FU) with epirubicin and cisplatin in metastatic, hormone-resistant prostate cancer: an active new regimen. Eur J Cancer. 1997; 33: 12301233.
  • 25
    Petrylak DP. Chemotherapy for androgen-independent prostate cancer. Semin Urol Oncol. 2002; 20: 3135.