The experience with chemotherapy in the pre-taxane era

The taxane agents (docetaxel and cabazitaxel) represent the current standard chemotherapy drugs used for the treatment of advanced castration-resistant prostate cancer (CRPC). However, in the years preceding the Food and Drug Administration (FDA) approval of docetaxel (which occurred in 2004) and cabazitaxel (in 2010), various other chemotherapeutic drugs were being used, although none had demonstrated evidence of improved survival. Importantly, the historical data reported in the literature should be evaluated in the context of when it was published. More specifically, clinical trials conducted before the late 1980s should be evaluated separately from those reported subsequently, because of the introduction and availability of serum prostate-specific antigen (PSA) measurement in 1987. In addition, the evolving data with chemotherapy during the past 3 decades suggest that the survival of patients with metastatic castration-resistant prostate cancer (mCRPC) is substantially longer than previously described, which is probably also a reflection of lead-time effect caused by PSA-detecting treatment failure earlier in the disease process. Therefore, patients in contemporary chemotherapy trials have lesser tumor burden. In addition, the incidence and severity of critical clinical parameters associated with poor prognosis (such as anemia, renal dysfunction, symptoms, and impaired functional status) are much less frequent in contemporary series, which, in part, explains the differences in clinical outcomes since PSA was introduced. A historical perspective of chemotherapy development in the pre-docetaxel era is outlined later in this article.

There is modest single-agent activity with the alkylating drug, cyclophosphamide, with reported responses in only about 10% to 20% of patients based on the older literature. In more recent reports, cyclophosphamide given in standard oral doses or high intravenous doses with hematopoietic growth factors support has been reported to have a higher order of activity. Doxorubicin, 5-fluorouracil, and cisplatin, agents all of which have demonstrable antitumor activity in other various tumor types, have shown only modest single-agent benefits in patients with mCRPC. None of these agents are used with any frequency in the current era.

Estramustine phosphate is a nitrogen mustard derivative of estradiol, which has demonstrated limited single-agent activity in advanced prostate cancer. This drug has been shown to exert its cytotoxic activity through microtubule inhibition and binding to nuclear matrix, which provided the rationale for testing this compound in combination with other agents targeting the microtubule, such as vinblastine, paclitaxel, docetaxel, and vinorelbine. However, in a prospective randomized study conducted in patients with castration-resistant disease, estramustine phosphate given at an oral dosage of 560 mg/d was not shown to be superior to placebo in terms of palliative effects or survival. In addition, venous thromboembolic disease was a significant toxicity with this agent. It is considered that estramustine has only marginal activity in mCRPC, and because of its unfavorable therapeutic index (safety/efficacy) compared with more modern chemotherapeutics, its use in mCRPC has virtually been abandoned.

The treatment with the best response in the pre-taxane era is mitoxantrone. Mitoxantrone is a semisynthetic anthracenedione derivative that has shown modest subjective benefits with otherwise minimal evidence of objective antitumor activity or survival improvement. A prospective randomized comparison of mitoxantrone plus prednisone with prednisone alone resulted in significant improvements of various quality-of-life issues including bone pain response. However, survival was not significantly different between the treatment arms. This study provided the justification for the FDA approval of the combination of mitoxantrone and prednisone for the symptomatic treatment of patients with metastatic prostate cancer in 1997. In current practice, mitoxantrone is rarely used, although it can sometimes be a reasonable agent used in third-line chemotherapy and beyond, if other life-prolonging chemotherapy drugs have already been exhausted.

Taxanes in mCRPC

Both paclitaxel (extracted from the bark of the Pacific yew tree, Taxus brevifolia ) and docetaxel (from the leaves of the European yew tree) disrupt microtubule function that is required for chromosome segregation at mitosis, thereby promoting apoptosis in eukaryotic cells. Nonetheless, the exact mechanism by which these drugs act against prostate cancer cells is not entirely clear. Prostate cancers tend to have low growth fractions, as compared with many normal cells and most other cancers; yet prostate cancer cells remain sensitive to taxane drugs. One possibility is that microtubule function may be needed for other critical cell processes, such as nuclear-cytoplasmic shuttling of critical regulatory proteins essential for prostate cancer cell viability, including translocation of the androgen receptor from the cytoplasm into the nucleus.

Taxanes in mCRPC

Both paclitaxel (extracted from the bark of the Pacific yew tree, Taxus brevifolia ) and docetaxel (from the leaves of the European yew tree) disrupt microtubule function that is required for chromosome segregation at mitosis, thereby promoting apoptosis in eukaryotic cells. Nonetheless, the exact mechanism by which these drugs act against prostate cancer cells is not entirely clear. Prostate cancers tend to have low growth fractions, as compared with many normal cells and most other cancers; yet prostate cancer cells remain sensitive to taxane drugs. One possibility is that microtubule function may be needed for other critical cell processes, such as nuclear-cytoplasmic shuttling of critical regulatory proteins essential for prostate cancer cell viability, including translocation of the androgen receptor from the cytoplasm into the nucleus.

Docetaxel

Although paclitaxel and docetaxel showed promising activity against prostate cancer in early clinical trials, only docetaxel has been subjected to large-scale trials testing effects on survival in patients with mCRPC. Docetaxel, given intravenously every 3 weeks (75 mg/m ² ) to men with prostate cancer in an initial small trial (n = 35), was associated with a 46% serum PSA response rate. In a series of clinical trials featuring intravenous weekly docetaxel (35–40 mg/m ² ), serum PSA response rates ranged from 41% to 64%. The addition of estramustine (see earlier discussion) to docetaxel appeared to produce significant benefits in early studies, both for serum PSA responses (45%–74%) and for measurable disease responses (11%–57%).

These initial promising studies conducted in the late 1990s eventually resulted in the design and execution of 2 definitive randomized phase 3 studies, known as TAX-327 and SWOG-9916. The TAX-327 study (considered the pivotal trial for FDA approval of docetaxel in prostate cancer) included 3 treatment arms: docetaxel, 75 mg/m ² , every 3 weeks plus prednisone, 10 mg/d; weekly docetaxel, 30 mg/m ² , (5 of 6 weeks) plus prednisone; and mitoxantrone, 12 mg/m ² , every 3 weeks plus prednisone. A total of 1006 patients were randomized into the 3 arms. The docetaxel every-3-weeks regimen resulted in significantly superior survival and higher PSA and pain response rates compared with the mitoxantrone arm. The overall survival was 18.9 versus 16.5 months; the hazard ratio (HR) for risk of death was 0.76 (confidence interval [CI], 0.62–0.92) ( Fig. 1 ). The weekly docetaxel regimen showed a trend toward survival benefit compared with mitoxantrone, but this did not reach statistical significance. A significantly greater proportion of patients who were treated with the 3-weekly docetaxel regimen experienced a reduction in pain (35% vs 22%, P = .01), a greater than or equal to 50% reduction in PSA (45% vs 32%, P <.001), and an improvement in quality of life (22% vs 13%, P = .009) compared with patients who received mitoxantrone, which itself had previously shown a quality-of-life improvement. Grade 3/4 toxicity was infrequent, other than neutropenia (32%), but the incidence of febrile neutropenia during the entire course of chemotherapy was 3% or less.

Updated TAX-327 survival figures. A comparison of docetaxel plus prednisone (weekly vs 3-weekly) versus mitoxantrone plus prednisone.

( Data from Berthold DR, Pond GR, de Wit R, et al. Survival and PSA response of patients in the TAX 327 study who crossed over to receive docetaxel after mitoxantrone or vice versa. Ann Oncol 2008;19:1749–53.)

The second study, SWOG-9916, evaluated the combination of docetaxel plus estramustine versus mitoxantrone plus prednisone. The median overall survival was superior in the group receiving the docetaxel regimen: 17.5 versus 15.6 months, HR 0.80 (CI, 0.67–0.90). The patient characteristics in this trial were similar to those of the TAX-327 study, as were the survival outcomes. The incorporation of estramustine in the docetaxel regimen, however, was characterized by increased gastrointestinal and vascular toxicities (mostly thromboembolic complications). Because of the high rate of venous thromboses with estramustine, prophylactic low-dose warfarin and aspirin were added to that arm, although did not reduce the incidence of thromboembolism. Similarly, 20% and 15%, respectively, of patients in the docetaxel-estramustine arm had grade 3/4 gastrointestinal and cardiovascular toxicities. Although comparisons between the docetaxel arms across these 2 seminal trials may not be entirely appropriate because of differences in treatment schedules, patient populations, and docetaxel dosing (60 mg/m ² in SWOG-9916 and 75 mg/m ² in TAX-327), it may be concluded that estramustine is unlikely to add significantly to the activity of single-agent docetaxel. Consequently, estramustine is not currently used in this disease.

Based on these 2 pivotal studies, docetaxel was approved by the FDA in 2004 as the first life-prolonging chemotherapy for use as front-line treatment of mCRPC. TAX-327 was designed to include 10 cycles of docetaxel at 3-week intervals, although there were no preexisting data to justify this approach. Therefore, in clinical practice, it is reasonable to administer docetaxel for a total of up to 10 doses. However, another commonly used approach has been to treat patients with a shorter number of chemotherapy cycles (usually between 5 and 7) until stabilization of disease is observed and then to hold treatment until subsequent progression (PSA progression, clinical progression, or both). This approach provides a rationale for docetaxel retreatment before moving to second-line chemotherapeutic options. Indeed, this pattern of chemotherapy administration can be viewed as intermittent docetaxel treatment. Preliminary published reports with intermittent chemotherapy indicate that this approach is feasible and has potential advantages in terms of extending disease control without increased toxicity.

Crossover from mitoxantrone to docetaxel or vice versa

The TAX-327 study did not include a crossover design allowing crossover from mitoxantrone to docetaxel and vice versa. As reported, treatment with docetaxel on the TAX-327 trial benefited all patient subgroups (pain vs no pain, ECOG performance status of 0/1 vs 2, and most other prognostic groups). Because the toxicity picture of mitoxantrone was more favorable than that of docetaxel, one consideration was whether patients could be treated initially with mitoxantrone, derive a reasonable quality-of-life benefit (delay of symptoms and pain), and then be treated with docetaxel at the time of subsequent progression. To investigate this issue, Berthold and colleagues reported retrospective data addressing the experience with crossover treatments from mitoxantrone to docetaxel and vice versa. In that analysis, 232 crossover patients were investigated, and the median survival after crossover was 10 months and did not depend on direction of crossover. Data on PSA response were available for 96 patients: PSA response (≥50% reduction) occurred in 15% men receiving mitoxantrone after docetaxel and 28% of men receiving docetaxel after mitoxantrone. Median PSA progression-free survival was 3.4 months for mitoxantrone after docetaxel and 5.9 months for docetaxel after mitoxantrone. In that analysis, response to first-line chemotherapy did not predict outcome of second-line treatment. While retrospective, these data indicate that the response to second-line docetaxel (after mitoxantrone) is substantially lower than reported in the first-line setting, and although survival after crossover was similar, this only included a small number of selected patients who were considered candidates for second-line treatment. In summary, the totality of the available evidence strongly supports the use of docetaxel as the first-line treatment of choice for patients with mCRPC.