Long-term results of a phase II study with neoadjuvant docetaxel chemotherapy and complete androgen blockade in locally advanced and high-risk prostate cancer

The clinical patient characteristics are outlined in Table 1. Between July 2005 and February 2010, 30 patients were enrolled. Median pretreatment values were 25.8 ng/ml for PSA, 7 for biopsy Gleason score, cT3b for clinical stage and 68 years for age. According to Kattan’s nomogram, the median probability of bRFS was 10% (range 0–55) [1].

Overall 86 cycles of docetaxel were administered and 27 (90%) patients completed all 3 cycles. Three patients (10%) discontinued NCHT; one withdrew consent after one cycle, and two interrupted due to NCHT-unrelated severe AE (lumbar disk herniation; exacerbating peripheral arterial occlusion disease (PAOD)) after two cycles of docetaxel. Bicalutamide was continued up to RP except for the patient with PAOD. Dose reductions of docetaxel were necessary in two patients (6.7%) and a treatment deferral in one patient (3.3%). Adverse events occurring in ≥10% of patients are presented in Table 2. Grade 3/4 non-hematologic toxicities were pneumonia (3.3%) and hyperglycemia (3.3%). Grade 3/4 hematologic toxicities were leucopenia in 53.8% (n = 14) and neutropenia in 90% (n = 18), in patients amenable for hematologic analyses one week after docetaxel administration. Febrile neutropenia was observed in 13.3%. Late toxicity assessments after a median follow up (FU) of 48.6 months (range 14.1-87.8) revealed one case (3.3%) of lower neuropathy (Grade 1). Second malignancies were observed in 2 cases (6.7%), one patient with bronchial carcinoma and another patient with bladder cancer, who received adjuvant radiation therapy.

All patients (n = 30) underwent RP with histopathological analyses after a median interval of 70 days from day 1 (range 61–143). Median removed lymph nodes were 16 (range 5–39). Surgical variables are depicted in Table 3 and pathological results in Table 4. Resectability was hindered due to periprostatic fibrosis in 23.3% and to increased vulnerability with increased diffuse bleeding in 3.3%. In case of periprostatic fibrosis the identification and dissection of the appropriate surgical layers was more difficult, when compared to not pretreated patients. Early re-interventions were required due to symptomatic lymphoceles (20%), hydronephrosis (3.3%) and bleeding (3.3%). Late events were urethral strictures (6.7%). Pathological analyses revealed extracapsular extension in 56.5% with positive surgical margins (R1) in 33.3% (n = 10) and lymph node involvement (pN1) in 36.7% of patients (n = 11).

Long time functional outcome, following a median of 48.6 months (range 14.1-87.8), revealed continence in 96.7%. Before treatment 86.7% (n = 26) patients were potent. After RP 15.4% had spontaneous erections (E3-E4), 11.5% were potent with erection aids and 19.2% had tumescence sufficient for sexual activity but not for vaginal penetration (E1-E2).

All patients (n = 30) demonstrated a PSA reduction following NCHT with a median decline of 97.3% (range 81.3-99.9%; p < 0.001) (Figure 1). Partial PSA response was observed in 100% and a complete response in 13.3%. After RP 96.7% of patients were therapy responders with undetectable PSA-values.

Downstaging of the clinical T-stage, indicated by MRI, was observed in 32% of 25 evaluable patients (cT3a to cT2 n = 3; cT3b to cT2 n = 2; cT3b to cT3a n = 2; cT4 to cT3a n = 1), without upstaging. Prostate volume was reduced in all cases with a significant median decrease of 37.1% (range 17.2-68.5%; p < 0.001). Similarly a significant median tumor volume reduction of 46.4% (range −31.3-82.8%; p < 0.001) was observed in 95.7% (n = 22) of patients harboring measurable tumors, while one progressed (4.3%). Pathological downstaging, comparing initial T-stages on MRI (n = 29) with histopathological stages revealed downstaging in 48.3% (cT3a to pT2 n = 2; cT3b to pT2 n = 10; cT3b to pT3a n = 1; cT4 to pT3b n = 1) and upstaging in 13.8% (cT2 to pT3b n = 1; cT3a to pT3b n = 3). Histopathological analyses revealed no pathological complete response (pCR), and only one (3.3%) pathological minimal residual disease (pMRD) with a tumor volume of <5%.

After a median FU of 48.6 months (range 19.9-87.8), of the 29 treatment responders 55.2% experienced a biochemical progression. The estimated median time to PSA progression was 38.6 months (95% CI 30.9-46.4) with a 5-year biochemical recurrence free survival of 40% (Figure 2a).

Adjuvant ADT was administered in 7 (23.3%) patients over a mean of 19 months (range 11–28) due to pathological R1N1 (n = 2) or R0N1 (n = 5) stages. Adjuvant radiation therapy (RT) was performed in 3 (10%) patients due to positive surgical margins without lymph node involvement (R1N0). No adjuvant RT was performed in 3 patients despite a R1N0 stage in favor a deferred RT in case of biochemical recurrence and in 4 patients with a R1N1 stage.

At the time of this analysis 26 (86.7%) patients were still alive. Two patients (6.7%) experienced a PC related death. The estimated median OS in the complete cohort was 85.3 months (95%CI 39.3–131.3) (Figure 2b).

Presurgical treatment consisted of CAB comprising bicalutamide 50 mg/day p.o. starting day 1 until the day before RP with trimestral buserelin 9.45 mg s.c. on day 3 and three cycles of docetaxel chemotherapy (75 mg/m²) on days 8, 29 and 50 in combination with prednisolone 10 mg/day p.o. starting on day 9. Dexamethasone (8 mg p.o.) was administered 12 hours before and after docetaxel as well as 60 min before. A dose reduction of docetaxel to 60 mg/m² and a treatment deferral of 21 days were allowed. All patients underwent open retropubic RP including extended bilateral pelvic lymph node dissection up to the aorta. Adjuvant radiation and hormonal therapies were individualized according to pathological results [13,23]. NCHT was aborted in case of unacceptable toxicities, progression or on patients’ request.

Tumor staging included digital rectal examination, abdominal and transrectal ultrasound, chest X-ray, bone- and computed tomography scans (CT) or magnetic resonance imaging (MRI) of the pelvis. Pelvic lymph nodes were considered non-metastatic if the maximum diameter did not exceed 1.5 cm. Local clinical downstaging was assessed by MRI of the prostate with endorectal coil (1.5 T, Magnetom Avanto, Siemens, Germany) before and after NCHT. T2-weighted sequences in three planes, axial T1-weighted and a dynamic-contrast-enhanced sequence were analyzed independently by a board certified radiologist. Median prostate and tumor volumes were calculated [13, 25‐27]. Tumor volume was determined by measuring the maximum tumor area in the axial plane and the longest cranio-caudal diameter. In case of multiple tumor nodules, each was measured separately. Pathologic specimens were classified according to the “Union international contre le cancer” version 2002. Gleason grading is not valid following NCHT, consequently it was not determined [7, 15]. The absence of PC in the histo-pathological sections was defined as pCR and tumor volumes <5% as pMRD [15]. In addition, pathologic downstaging, comparing initial stages on MRI with histopathological stages, was analyzed [7, 13, 22]. PSA response was assessed before RP and latest 3 months post-RP. Before RP, a PSA reduction of >50% indicated partial response and PSA-levels <0.2 ng/ml a complete response [15]. After RP, patients with PSA-values <0.07 ng/ml were deemed therapy responders. Biochemical recurrence was defined by two PSA increases >0.2 ng/ml [28].

Surgical variables were recorded implying the functional outcome. Continence was defined as the use of no pads and erectile function using the erection hardness score (E0-E4) [7, 29]. Adverse events (AE) were registered according to Common Terminology Criteria (CTCAE 3.0), including dose reductions and treatment deferrals. Additional hematologic analyses were performed one week after docetaxel application.

This trial was conducted as an intent to treat approach [30]. Sample sizes were calculated using the software N-Query version 7.0. With a sample size of 24 patients, a one-sided 97.5% confidence interval for the downstaging proportion, using the large sample normal approximation, will extend 20% from the observed proportion for an expected proportion of 40%. Analyses were performed using SPSS version 20 (SPSS Inc., Chicago, IL, USA). Changes in continuous measures over time were determined using the Wilcoxon signed-rank test and linear trend analyses using the Kruskal-Wallis test. Applying a Bonferroni correction, a p-value of <0.025 was considered statistically significant. Estimates of survival were analyzed using the Kaplan-Meier method. OS was defined as the elapsed time from screening to death and time to PSA progression as time from the first study treatment to biochemical progression [30].