Article Text

Download PDFPDF
Continuing cabazitaxel beyond 10 cycles for metastatic castrate-resistant prostate cancer: is there a benefit?
  1. Loma AL-Mansouri1,
  2. Malmaruha Arasaratnam2,
  3. Howard Gurney1
  1. 1 Clinical Medicine, Macquarie University, Sydney, New South Wales, Australia
  2. 2 Clinical Medicine, Westmead Hospital, Westmead, New South Wales, Australia
  1. Correspondence to Dr Loma AL-Mansouri, Clinical Medicine, Macquarie University, Sydney, NSW 2109, Australia; lametah{at}


Aim Cabazitaxel prolongs survival in patients with metastatic castration-resistant prostate cancer in the postdocetaxel setting. We investigate the benefit of continuing cabazitaxel beyond 10 cycles in patients who are clinically responding without significant toxicity.

Methods A comparison was made between patients who received cabazitaxel for >10 cycles and those who had ≤10 cycles. Overall survival (OS), prostate-specific antigen (PSA) response, alkaline phosphatase (ALP) changes and treatment-associated adverse events were evaluated.

Results The median OS was 9 months (range 0.75–59), with OS significantly higher in patients who received extended duration of treatment: 14 months (range 3–90) vs 7 months (range 1.3–21) in patients treated with 4–10 cycles (HR 0.28, 95% CI 0.1 to 0.74, p=0.01). PSA decline did not show a significant correlation with OS (PSA decline ≥50%, p=0.54). Furthermore, there was no significant difference in OS between patients who had a normal versus high ALP at baseline. There was no clear evidence of cumulative toxicity in those having >10 cycles.

Conclusion A substantial proportion of patients with metastatic castration-resistant prostate cancer were able to receive more than 10 cycles of cabazitaxel without clinically relevant cumulative toxicity.

  • urological tumours
  • cabazitaxel
  • taxane
  • metastatic castration-resistant prostate cancer
  • duration of treatment
  • number of cycles

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Metastatic prostate cancer is the second leading cause of cancer-related morbidity and mortality in developed countries.1 Metastatic prostate cancer is an incurable disease, and its management is an ongoing challenge. Androgen deprivation therapy (ADT) is the standard first-line treatment for metastatic hormone-sensitive prostate cancer, and although not curative helps to delay progression and improve symptom.2 The cancer eventually becomes resistant to low testosterone levels and progresses, a state categorised as castration-resistant prostate cancer (CRPC). CRPC is defined as a clinical or radiological progression with or without rising prostate-specific antigen (PSA) blood level while on ADT, in the presence of castrate levels of testosterone.2 3 Until 2004, the management of metastatic CRPC (mCRPC) was essentially palliative, with the primary goals being pain control and improving quality of life.4 Docetaxel was the first chemotherapy of the taxane group approved in Australia in 2006 for metastatic prostate cancer.4–7 However, most patients develop resistance to docetaxel, and disease progression eventually occurs. Cabazitaxel is a second-generation taxane approved for mCRPC in the postdocetaxel setting, demonstrating improvement in overall survival (OS), quality of life and pain.8 Unlike other taxanes, cabazitaxel has a weak affinity for P-glycoprotein (P-gp), an ATP-dependent drug efflux pump. The expression of P-gp renders the cancer cells resistant to many drugs, including docetaxel.9 10

The most common toxicities with cabazitaxel reported in the TROPIC trial (Cabazitaxel Plus Prednisone Compared to Mitoxantrone Plus Prednisone in Hormone Refractory Metastatic Prostate Cancer) was grade 3 or 4 neutropaenia and febrile neutropaenia, occurring in 82% and 8% of patients, respectively.8 However, data on the efficacy and safety of cabazitaxel from the expanded-access programmes demonstrated significantly lower rates of grade 3 and 4 neutropaenia and febrile neutropaenia.11–14 In general, cabazitaxel-associated adverse events are manageable with early anticipation and implementation of appropriate mitigation strategies, including proper patient selection and education, prophylactic granulocyte colony stimulating factor (G-CSF) and dose modifications in high-risk patients.12 15

Cabazitaxel efficacy used to guide the decision to continue treatment is based on symptomatic improvement, including performance status, quality of life and pain palliation, radiological tumour response, and PSA changes.12 16 17 The prognostic significance of PSA decline is debatable, with one study showing a correlation between postchemotherapy PSA decline and changes in tumour burden,18 and another study reporting early PSA rise after initiation of chemotherapy, also known as a flare response.19 20 Due to these reasons, PSA alone cannot be used to guide the decision on chemotherapy discontinuation.

Patients and methods

Study design

This is a retrospective analysis of treatment details and outcomes for patients with mCRPC receiving cabazitaxel. A unique component is the description of the number of cycles given in the Australian environment which, unlike other international sites, allows more than 10 cycles to be administered. Examination of this data may generate hypotheses for future research.

Patient population

All patients involved in the study had confirmed mCRPC. mCRPC is defined as castration serum testosterone less than 50 ng/dL or 1.7 nmol/L with one of the following types of progression:

  • Biochemical progression: three consecutive rises in PSA 1 week apart, resulting in two 50% increases over the nadir, and PSA >2 ng/mL.

  • Radiological progression: the appearance of new lesions, either two or more new bone lesions, on bone scan or a soft tissue lesion using the Response Evaluation Criteria in Solid Tumours.3

The medical records of 45 patients were examined to extract data. We de-identified the patients during the analysis. All patients received intravenous cabazitaxel (25 mg/m2 every 3 weeks along with 10 mg oral prednisone daily) following disease progression during or after docetaxel therapy. Treatment continued until disease progression, death, unacceptable adverse events, or at the physician’s or the patient’s discretion.

Data extraction

Patients’ medical records were reviewed to extract clinical demographics, treatment history and laboratory results: Gleason score grade, PSA level, histological tumour type, nadir neutrophils after cabazitaxel, and precabazitaxel level of lactate dehydrogenase) LDH, PSA and albumin, and calculation of metastatic prognostic index (Eastern Cooperative Oncology Group Performance Status (ECOG) performance status, disease site, LDH, opioid analgesic use, albumin, haemoglobin, PSA and alkaline phosphatase (ALP)).

An online calculator was used to categorise the patients according to their risk group.21

Outcome measures

The study evaluates the following outcomes: measurement of OS, PSA and ALP changes in response to cabazitaxel and relation to OS, and safety of cabazitaxel measured by the incidence of treatment-related adverse events.

Statistical analysis

We analysed OS using Kaplan-Meier survival analysis and Cox hazard model to estimate the HR. Landmark analysis was done to reduce bias in concluding time to the event. The landmark time point was set to start survival analysis following three cycles of cabazitaxel, excluding patients stopping treatment at or before this point. PSA and ALP changes in relation to OS were analysed and compared. The analysis was done using SPSS V.24.


Pretreatment characteristics

A total of 45 patients from one oncology centre were included in the analysis. Baseline patient and disease characteristics are outlined in table 1. The median age of all patients at the start of the treatment was 71 years, with 16 (35.5%) patients aged ≥75 years. There was no difference in age between patients who received ≤10 cycles and >10cycles of cabazitaxel. In all patients, the ECOG performance status was ≤2. The median baseline PSA was 85.50 (range 0.07–1240), and abnormal baseline ALP and LDH were reported in 24 (55.8%) and 17 (42.5%) patients, respectively. The most common metastatic sites were bone (44, 97.7%), lymph nodes (12, 26.7%), lungs (2, 4.4%) and liver (5, 11.11%). The metastatic prognostic index was calculated for 44 patients to categorise them as high, intermediate and low risk. More patients who were in the high-risk category received 10 cycles or less compared with those patients who received more than 10 cycles: 20 of 24 (83.3%) vs 13 of 20 (65%), respectively. The number of patients with intermediate and low risk was 3 of 24 (12.5%) and 1 of 24 (4.2%), respectively, in the group who received ≤10 cycles, and 6 of 20 (30%) and 1 of 20 (5%), respectively, in the >10 cycles group. All patients had received first-line docetaxel, with a median of 9 cycles (range 2–20). The median time from the last docetaxel treatment to the start of cabazitaxel was 7 months (range 0.75–120). The interval between the last dose of docetaxel treatment and the start of cabazitaxel was ≤6 months in 21 of 43 (48.8%) patients, where 12 patients subsequently received ≤10 cycles of cabazitaxel and 9 patients received more than 10 cycles. Most patients (43/45, 95.5%) had received one or more prior regimens of hormonal therapy. Twenty-one (46.6%) patients received prior androgen axis inhibitors (AAI): nine (20%) patients received abiraterone, eight (17.8%) received enzalutamide, and four (8.9%) received both abiraterone and enzalutamide.

Table 1

Patient characteristics

Cabazitaxel treatment

The median number of cabazitaxel cycles administered in all patients was 9 (range 1–82).

Seven patients had received <4 cycles, 18 patients received a median of 6 cycles (4–10), and 20 patients received a median of 15 (11–82) (figure 1). Most patients (28/45, 62.2%) received the recommended dose of cabazitaxel (25 mg/m2 every 3 weeks) with no dose modification. An upfront dose reduction (20%=20 mg/m2) was done for 10 of 45 (22%) patients due to high risk of neutropaenic sepsis (older age, low starting white cell count, poor performance status), and subsequent dose reductions after the first or second cycles due to toxicity were required in 7 of 45 (15.5%) patients.

Figure 1

Number of cabazitaxel cycles per patient.

The reasons for treatment discontinuation in 42 patients are listed in table 2. Progressive disease and toxicity were the most common causes for treatment discontinuation in both groups of patients. Following progression from cabazitaxel, 18 (40%) patients subsequently received either abiraterone or enzalutamide, and 6 patients (13/3%) received mitoxantrone chemotherapy. Of the 18 patients who received subsequent AAI therapy, 8 patients had received ≤10 cycles of prior cabazitaxel and 10 patients had received >10 cycles.

Table 2

Causes of treatment discontinuation


The median OS for all patients was 9 months (range 0.75–59). Survival landmark analysis was performed, and patients who received <4 cycles of chemotherapy were censored. Kaplan-Meier survival analysis, based on the number of cabazitaxel cycles, showed an OS benefit in favour of extending the duration of cabazitaxel beyond 10 cycles. The median survival for patients who received 4–10 cycles of cabazitaxel was 6.75 months (range 1.25–21), and 13.75 months (range 3–90) in patients who received >10 cycles (HR 0.28, 95% CI 0.1 to 0.74, p=0.01) (figures 2 and 3).

Figure 2

Distribution of number of cycles in relation to overall survival.

Figure 3

Landmark analysis of overall survival according to the number of cycles.

PSA response

PSA response rate was evaluated during the first 6 months of treatment. A ≥30% decline from baseline was seen in 15 (33.3%) patients and ≥50% PSA decline in 9 (20%) (figure 4).

Figure 4

Waterfall plot of prostate-specific antigen (PSA) response.

Patients who had >10 cycles were more likely to have PSA decline in the first 6 months of therapy, where 10 of 20 (50%) patients had a ≥30% PSA decline and 7 of 20 (35%) had a ≥50% PSA decline. In contrast, 5 of 25 (20%) patients had a ≥30% PSA decline and only 2 of 25 (8%) patients had a ≥50% decline. There was no difference in OS between patients who had a ≥30% PSA decline vs <30% decline (HR=1.0, 95% CI 0.44 to 2.27, p=0.98). A similar result was found for a PSA decline ≥50% vs <50% (HR=1.19, 95% CI 0.49 to 2.87, p=0.69) (online supplementary figures 1 and 2).

Supplemental material

ALP changes and survival

Baseline ALP was normal (≤110 U/L) in 17 of 45 (37.8%) patients and elevated (>110 U/L) in 28 of 45 (62.2%) patients. Survival analysis showed no significant difference between normal and elevated baseline ALP in relation to OS (HR=1.73, 95% CI 0.73 to 4.07, p=0.21) (online supplementary figure 3). Changes in ALP with treatment were also assessed, where a normal or falling ALP was observed in 26 of 45 (57.8%) patients, and a persistently elevated or rising ALP was observed in 19 of 45 (42.2%) patients. However, there was no difference in OS in these two groups (HR=0.441, 95% CI 0.175 to 1.113, p=0.083) (online supplementary figure 4).

Changes in pain during treatment

Improvement in pain with decreased requirements or stopping analgesics was reported by 23 (51.1%) patients, and 16 had received >10 cycles of cabazitaxel. Increased doses of narcotics or palliative radiotherapy were needed in 6 (13.3%) patients, and 16 (35.6%) patients had no pain throughout the treatment.

Adverse events

Cabazitaxel in combination with prednisone was well tolerated. Adverse events are listed in online supplementary table 1. The most common adverse effects were haematological, with febrile neutropaenia recorded in 12 (26.7%) patients. None of these patients had prophylactic G-CSF routinely, as it is not funded for that indication in Australia. Blood transfusion was required for symptomatic anaemia in 17 (37.8%). There was no clear evidence of cumulative toxicity, with those in the >10 cycles group experiencing a similar adverse event profile to those having ≤10 cycles. The incidence of neuropathy was numerically higher in the >10 cycles group, although the majority were grade 1 (mild paraesthesia, loss of deep tendon reflexes) and was not an indication for stopping therapy.


The registration study of cabazitaxel, which showed a survival benefit over mitoxantrone in patients with CRPC who had failed docetaxel (TROPIC study), set a maximum limit of 10 cycles for both regimens.10 Since then, subsequent studies have also capped the number of cabazitaxel cycles to 10, with funding agencies in Europe and North America limiting treatment to 10 cycles in clinical practice.22 A post-hoc analysis of the TROPIC study suggested that higher number of cycles of cabazitaxel may be associated with a positive impact on patients’ outcomes, including quality of life and survival. However, in these studies, the number of treatments was limited to 10.20 22 In Australia, there is no cap in the number of cycles that can be delivered, but the utility of continuing beyond 10 cycles has been poorly studied. A recent study has suggested that the continuation of first-generation taxane, docetaxel, beyond the recommended 10 cycles has a survival benefit.23 This retrospective analysis showed those receiving >10 cycles of cabazitaxel seemed to have superior survival compared with those who received 4–10 cycles. The median OS was 13.8 months and 6.8 months, respectively (p=0.006). In our cohort, the median age was comparable between those patients who received >10 and ≤10 cycles. High-risk features at baseline that predict poor prognosis24—defined by ECOG performance status ≥2, visceral disease, opioid analgesic use, elevated levels of LDH, PSA and ALP, and low levels of albumin and haemoglobin—were identified in 65% of patients who received >10 cycles and 83.3% who received ≤10 cycles.24 There was no difference in the two groups regarding the time interval between the last dose of docetaxel and the initiation of cabazitaxel, a clinical marker of taxane sensitivity.25 A major weakness of using retrospective analyses to compare a number of treatment cycles with efficacy such as survival is the possibility of selection bias. Patients who die early in the course of therapy may have done so because of the higher burden of disease, rapid disease progression, comorbidities and other poor prognostic factors. In this case, a correlation between the number of cycles and death is a non-causal association. In order to minimise this selection bias, we performed a landmark analysis whereby patients who stopped cabazitaxel earlier than four cycles were excluded from the survival analysis. This way, we minimised the effect of discontinuation as a result of progressive disease rather than treatment-related factors. However, the suggested beneficial effect of extending cabazitaxel treatment on survival must still be interpreted with caution, as more patients who received ≤10 cycles were in the high-risk category, 83.3% vs 65% who received >10 cycles. Another potential confounding factor is the number of patients receiving postcabazitaxel life-prolonging androgen receptor-targeted therapy. Ten of 20 (50%) patients having >10 cycles of cabazitaxel received post-treatment abiraterone or enzalutamide compared with 8 of 25 (32%) who received <10 cycles.

Our other finding of potential clinical relevance is the lack of correlation between the fall in PSA level within the first 6 months with OS, suggesting that PSA levels should not be used during treatment as a surrogate marker for efficacy. Tumour response evaluation by PSA decline rate was achieved in 20% of all patients. A PSA decline of ≥50% occurred in more patients who received extended treatment than those who received a shorter duration: 35% and 8%, respectively. However, our data do not demonstrate a significant association between PSA decline and OS. Although PSA changes have been widely considered as a measure of treatment effect on tumour burden, the decline rate cannot be regarded as a predictive measure of prognosis.16 It is important to note that the decision on treatment continuation or discontinuation cannot be reliably based on PSA changes alone as some patients were clinically responding to cabazitaxel regardless of PSA decline. PSA production is androgen-driven, and the protein is a sensitive marker for response to hormone therapy.

However, cytotoxic chemotherapy works by means other than through an effect on the androgen receptor.26 Therefore, in the castrate-resistant setting, some prostate cancer cell populations produce little or no PSA.27 For these reasons, PSA change alone should not be used as an indicator of response to chemotherapy, a notion that is supported by the findings in our study. ALP is one of the variables used for predicting prognosis in patients with mCRPC,28 and baseline ALP was associated with shorter survival.20 However, the small sample size of our study failed to show changes in ALP from baseline correlated with survival. The most frequently reported treatment-associated adverse events were neutropaenia, where grade 3/4 neutropaenia occurred in 40% and febrile neutropaenia in 26.7%. Other adverse events were fatigue, diarrhoea and peripheral neuropathy. Other than grade 1 neuropathy, there was no apparent cumulative increase in toxicity. The frequency of adverse events in patients who received >10 cycles of cabazitaxel was similar to those patients who received a lower number of cycles. The reason for this may be in part due to the peak incidence of cabazitaxel adverse effect predominantly occurring within the first few cycles of treatment.29 A higher rate of peripheral neuropathy was detected in a patient who received ≥10 cycles, although the majority were less than grade 2 and was not a reason for discontinuing treatment in the 45 patients. During the course of treatment, seven patients required dose reduction due to toxicity. Of the seven patients, four patients received >10 cycles and three received ≤10 cycles. It is possible that a selection bias still existed, and it is possible that the number of cycles of chemotherapy administered was the result of, not the cause of, the patients living longer with their cancer. For this reason, our study is hypothesis-generating regarding the effect of extended cabazitaxel treatment on survival in patients with mCRPC. However, it does show that prolonged cabazitaxel use is feasible and safe and causes minimal cumulative toxicity, with almost half the patients tolerating treatment beyond 10 cycles. Future prospective studies are required to confirm our results and provide a recommendation regarding extending cabazitaxel in fit and clinically responding patients.

In conclusion, cabazitaxel beyond 10 cycles is associated with better OS in patients with mCRPC without the added risks of toxicity. PSA response rate cannot be used as an independent decision-maker regarding continuity of treatment since it did not correlate with OS.

What this paper adds

What is already known on this subject

  • Cabazitaxel in combination with prednisolone is a generally well-tolerated systemic treatment for metastatic castration-resistant prostate cancer (mCRPC).

  • Cabazitaxel has shown to improve overall survival in patients with mCRPC in postdocetaxel setting.

What this study adds

  • This study focuses on the number of cabazitaxel cycles in mCRPC.

  • The study has shown that cabazitaxel beyond 10 cycles is feasible and associated with a better overall survival in patients with mCRPC without the added risks of toxicity.


Jenny Lee and Dhanusha Sabanathan contributed to the final version of the manuscript and supervised the project.



  • EAHP Statement 6: Education and Research.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests HG: advisory board for Astellas, Sanofi-Aventis and Janssen.

  • Patient consent for publication Not required.

  • Ethics approval Macquarie University Human Research Ethics Committee (HREC (Medical Sciences) Reference No: 5201700474).

  • Provenance and peer review Not commissioned; internally peer reviewed.

  • Data availability statement Data may be obtained from a third party and are not publicly available.