Castrate-resistant prostate cancer (CRPC) is disease progression despite castrate testosterone (<50 ng/dL), shown by a rising PSA, progression of existing disease, and/or new metastases. ADT almost always eventually fails this way, but the androgen receptor stays active — so ADT is continued for life and layered with successive agents. The field has expanded rapidly: chemotherapy, AR-pathway inhibitors, a radiopharmaceutical, a radioligand, immunotherapy, and PARP inhibitors all extend survival in the metastatic setting.
Definition and Workup
Men on ADT are monitored with PSA (the first sign of progression, preceding bone-scan changes while the patient is still asymptomatic), imaging (bone scan and CT), and serum testosterone (to confirm castrate levels — useful if non-compliance is suspected or if prior therapy was non-steroidal anti-androgen monotherapy, which may not sustain castration). Consider neuroendocrine differentiation in a man who progresses clinically or radiologically without a matching PSA rise; biopsy an accessible lesion and treat with platinum-based chemotherapy (cisplatin or carboplatin with etoposide).
The most important urologic sequela of advanced disease is obstructive uropathy, and paraneoplastic effects include anaemia, weight loss, fatigue, hypercoagulability, and increased infection risk. Eight prognostic factors are PSA, PSA doubling time, performance status, visceral metastases, bone pain, extent of disease on bone scan, LDH, and alkaline phosphatase; a PSADT under 3 months signals a rapid course warranting aggressive management.
General Principles
- Continue ADT for life — the AR remains active in most CRPC, and trials of newer agents mandate ongoing castration.
- Anti-androgen withdrawal — in men on combined AR-antagonist + LHRH-agonist therapy, simply stopping the anti-androgen produces a >50% PSA decline in 15–30% (median 3.5–5 months), because the antagonist has begun acting as an agonist; overall survival is not improved.
- Changing the anti-androgen or adding corticosteroids ± ketoconazole gives transient PSA responses in ~30% but no meaningful outcome benefit.
Non-Metastatic CRPC
Stop any first-generation anti-androgen, then stratify by PSA doubling time:
- PSADT ≥10 months — observation or secondary hormonal manipulation.
- PSADT <10 months (high-risk) with life expectancy >5 years — apalutamide, enzalutamide, or darolutamide, each with continuous ADT.
Three practice-changing trials (all enrolled PSADT ≤10 months, all used metastasis-free survival as the primary endpoint, and all improved MFS by ~22 months and OS by ~12 months):
| Trial | Agent | Metastasis-free survival | Overall survival |
|---|---|---|---|
| SPARTAN | Apalutamide 240 mg/day | 40 vs 16 mo (HR 0.28) | 74 vs 60 mo (HR 0.78) |
| PROSPER | Enzalutamide 160 mg/day | 37 vs 15 mo (HR 0.28) | 67 vs 56 mo (HR 0.73) |
| ARAMIS | Darolutamide 600 mg BID | 40 vs 18 mo | 83% vs 77% at 3 yr |
Monitor for metastases with bone scan and CT chest/abdomen/pelvis — every 3–6 months if PSADT <10 months or PSA >20, every 6–12 months if PSADT >10 months (the role of PSMA-PET is not yet defined).
Metastatic CRPC
About 90% of men with CRPC develop bone metastases (causing pain, pathological fracture, spinal cord compression — an emergency — and marrow failure); visceral metastases are much less common and rarely occur without bone disease. Because most tumour burden is in bone, radiographic progression-free survival is preferred over response rate as a trial endpoint.
Approved Systemic Therapies
| Class / Agent | Setting | Pivotal trial(s) | Key result |
|---|---|---|---|
| Docetaxel (taxane) | First-line | TAX 327 | OS 18.9 (3-weekly) vs 16.4 mo (mitoxantrone); standard first-line chemo |
| Cabazitaxel (taxane) | Post-docetaxel | TROPIC | OS 15.1 vs 12.7 mo vs mitoxantrone; 82% grade 3–4 neutropenia — use growth-factor support |
| Enzalutamide (AR antagonist) | Pre- or post-docetaxel | AFFIRM (post), PREVAIL (pre) | OS 18.4 vs 13.6 mo (AFFIRM); OS HR 0.71 (PREVAIL) |
| Abiraterone + prednisone (CYP17 inhibitor) | Pre- or post-docetaxel | COU-AA-301 (post), COU-AA-302 (pre) | OS 14.8 vs 10.9 mo (301); rPFS 16 vs 8 mo (302) |
| Radium-223 (α-emitter) | Bone pain, no visceral mets, nodes ≤3 cm | ALSYMPCA | OS 14.9 vs 11.3 mo; do not combine with abiraterone (ERA 223: more fractures) |
| ¹⁷⁷Lu-PSMA-617 (radioligand) | PSMA-positive after an ARPI + taxane | VISION, TheraP | Improved PFS and OS (VISION); better PSA response than cabazitaxel (TheraP) |
| Sipuleucel-T (vaccine) | Asymptomatic/minimally symptomatic, no visceral mets | IMPACT | OS 26 vs 22 mo; no change in PFS or PSA |
| Olaparib (PARP inhibitor) | HRR mutation after an ARPI | PROfound | rPFS 7.4 vs 3.6 mo; OS 19.1 vs 15.7 mo (BRCA/ATM cohort) |
| Rucaparib (PARP inhibitor) | BRCA1/2 after an ARPI + taxane | TRITON-2 (phase II) | Activity in BRCA-altered disease |
Key nuances:
- Mitoxantrone improves pain and quality of life but not survival — a palliative option only.
- Docetaxel is dosed 75 mg/m² every 3 weeks with prednisone; weekly dosing did not improve survival (TAX 327). Toxicities include neutropenia, fatigue, neuropathy, peripheral oedema, hyperlacrimation, and nail dystrophy. FIRSTANA showed cabazitaxel is not superior to docetaxel first-line.
- Abiraterone blocks adrenal and intratumoral androgen synthesis (sources LHRH analogues miss); on radium-223, alkaline phosphatase tracks activity better than PSA.
- HRR mutations occur in 20–30% of metastatic disease (most commonly BRCA2); defective repair makes the tumour susceptible to PARP inhibition (synthetic lethality).
- Checkpoint inhibitors have disappointed — pembrolizumab gave objective responses in only 3–5% (KEYNOTE-199), and ipilimumab did not improve overall survival (CA184-043). Cabozantinib (COMET-1/-2) failed its survival and pain endpoints despite bone-scan activity.
Treatment Sequence
- Chemotherapy-naïve, asymptomatic or minimally symptomatic — first-line abiraterone 1000 mg/day + prednisone or enzalutamide 160 mg/day; reserve docetaxel for second-line (or earlier with poor hormonal response or visceral disease).
- Moderate or severe symptoms — docetaxel; radium-223 (every 4 weeks × 6) for bone pain without visceral disease; abiraterone or enzalutamide if docetaxel is refused or not tolerated.
- After docetaxel — five options improve survival: cabazitaxel, radium-223, and (if not used earlier) abiraterone or enzalutamide. Docetaxel re-challenge and mitoxantrone offer palliation without a proven survival gain.
The optimal sequence is unknown, but changing the mechanism of action with each line is thought to give better, more durable responses.
Bone-Targeted Therapy and Supportive Care
In mCRPC with bone metastases, give denosumab 120 mg SC or zoledronic acid 4 mg IV every 4 weeks, plus daily calcium and vitamin D, to prevent skeletal-related events (pathological fracture, spinal cord compression, bone surgery, or radiation to bone):
- Zoledronic acid is the only bisphosphonate proven to reduce disease-related SREs (Saad 2004); avoid it if creatinine clearance is <30 mL/min, and watch for osteonecrosis of the jaw and hypocalcaemia.
- Denosumab (anti-RANK-ligand) beat zoledronic acid on time to first SRE (20.7 vs 17.1 months) with no survival difference, and needs no renal dose adjustment; ONJ occurs in 2–4%.
- ONJ risk rises after ~2 years — encourage oral hygiene, a baseline dental review, and avoidance of invasive dental work on therapy. Neither agent is approved for castrate-sensitive disease or for preventing bone metastases.
Palliative radiotherapy controls focal bone pain (exclude a pathological fracture first). Malignant spinal cord compression is an emergency — image with MRI and treat with radiation plus steroids, debulking surgery, or vertebrectomy with stabilisation.
Self-Test
-
What is the mechanism of action of docetaxel, and what are its main toxicities? It inhibits microtubule assembly and disassembly (a taxane); toxicities include myelosuppression, neurotoxicity, fatigue, peripheral oedema, hyperlacrimation, and nail dystrophy.
-
How is non-metastatic CRPC risk-stratified, and how is high-risk disease treated? By PSA doubling time — under 10 months defines high-risk; with a life expectancy >5 years, treat with apalutamide, enzalutamide, or darolutamide plus continuous ADT (SPARTAN, PROSPER, ARAMIS).
-
Which radiopharmaceutical must not be combined with abiraterone, and when is it indicated? Radium-223 — indicated for symptomatic bone metastases without visceral disease or bulky (>3 cm) nodes; ERA 223 showed more fractures when it was combined with abiraterone.