Quality of life and pain relief during treatment with calcitriol and docetaxel in symptomatic metastatic androgen-independent prostate carcinoma

Authors

  • Tomasz M. Beer M.D.,

    Corresponding author
    1. Division of Hematology and Medical Oncology, Department of Medicine, Oregon Health and Science University, Portland, Oregon
    • Department of Medicine, Oregon Health and Science University, Mail Code CR145, 3181 SW Sam Jackson Park Road, Portland, OR 97239
    Search for more papers by this author
    • Fax: (503) 494-6197

  • Kristine M. Eilers B.S.,

    1. Division of Hematology and Medical Oncology, Department of Medicine, Oregon Health and Science University, Portland, Oregon
    Search for more papers by this author
  • Mark Garzotto M.D.,

    1. Division of Urology, Department of Surgery, Oregon Health and Science University, Portland, Oregon
    2. Urology Section, Department of Surgery, Portland Veterans Administration Medical Center, Portland, Oregon
    Search for more papers by this author
  • Yi-Ching Hsieh M.S.,

    1. Biostatistics and Bioinformatics Shared Resource, Oregon Health and Science University Cancer Institute, Portland, Oregon
    Search for more papers by this author
  • Motomi Mori Ph.D.

    1. Biostatistics and Bioinformatics Shared Resource, Oregon Health and Science University Cancer Institute, Portland, Oregon
    Search for more papers by this author

Abstract

BACKGROUND

The current study evaluated the analgesic activity and impact on quality of life (QOL) of a new chemotherapy regimen of calcitriol and docetaxel in men with androgen-independent prostate carcinoma.

METHODS

Analgesic response was defined as a 2-point reduction on the Present Pain Intensity (PPI) scale (or compete relief if baseline PPI was 1) without an increase in analgesic use or a 50% decrease in analgesic medication use without an increase in pain, maintained for ≥ 4 weeks. Pain, pain medication consumption, and QOL (measured by the European Organization for Research and Treatment of Cancer QLQ-C30) were evaluated every 4 weeks.

RESULTS

Treatment resulted in an analgesic response in 14 of 29 evaluable patients (48%; 95% confidence interval [95% CI], 30–67%). The median time to symptomatic progression in the 14 patients who met criteria for analgesic response was 41 weeks (95% CI, 26–56 weeks). Worsening in physical and role functioning, fatigue, appetite, and global health status and improvement in constipation were detected using the QLQ-C30 QOL questionnaire.

CONCLUSIONS

Significant analgesic activity was demonstrated, although worsening in several QOL domains was observed in a patient population with relatively low pain intensity (median PPI, 2). Cancer 2004;100:758–63. © 2004 American Cancer Society.

Patients with cancer frequently experience disease-related pain. Approximately 75–90% of patients report pain in the advanced stages of cancer. Optimization of pain management can increase a patient's quality of life (QOL).1–3 Bone pain has been reported in 60–90% of patients with bony metastases4 and is the most common symptom reported by men with metastatic androgen-independent prostate carcinoma (AIPC).5 The importance of effective pain management in cancer has gained recognition in recent years.6 This trend is important in the treatment of men with AIPC, for whom pain reduction and QOL improvement are important goals of therapy. Indeed, the only modern cytotoxic chemotherapy regimen to gain approval from the Food and Drug Administration to treat men with AIPC is mitoxantrone and prednisone, which was approved based on its analgesic activity despite the absence of a survival benefit.7, 8

Our group recently reported the development of a regimen of high-dose calcitriol plus docetaxel for the treatment of AIPC.9 Docetaxel is one of the most active cytotoxic chemotherapy agents in the treatment of advanced prostate carcinoma. Its principal mechanism of antineoplastic activity is believed to be inhibition of microtubule depolymerization with associated G2M arrest. In a previous Phase II study, weekly docetaxel had significant analgesic activity in men with symptomatic metastatic prostate carcinoma.10 Calcitriol is the natural ligand for the vitamin D receptor that, in preclinical models of prostate carcinoma, enhances the activity of docetaxel and other cytotoxic agents.11 The Phase II activity of this regimen, as measured by prostate-specific antigen (PSA) reduction, measurable disease response, and time to progression, is encouraging.9 Although defining clinical activity by conventional criteria was the primary aim of the current Phase II trial, evaluation of the impact of calcitriol plus docetaxel on pain and QOL was a prospectively planned secondary goal. This analysis is presented here.

MATERIALS AND METHODS

Eligibility

Men with pathologically proven adenocarcinoma of the prostate, metastases, and evidence of progression (development of new metastatic lesions or an increase in cancer-related pain or a 50% increase in the PSA confirmed by a second measurement after 2 weeks) despite standard hormonal management (including antiandrogen withdrawal: 6 weeks for bicalutamide, 4 weeks for flutamide or nilutamide) were eligible. Detailed eligibility criteria for entry into the trial were reported previously.9 Briefly, patients had Eastern Cooperative Oncology Group performance status ≤ 3; adequate cardiac, hepatic, renal, and bone marrow function; no history of cancer-related hypercalcemia; no kidney stones within the past 5 years; and no previous chemotherapy. Pain was not specifically required for entry into the study. However, the protocol prespecified that pain relief and QOL would be analyzed in patients who reported pain at the time of study entry. The study was approved by the institutional review boards of the Oregon Health and Science University and the Portland VA Medical Center. Signed, informed patient consent was obtained before any procedures were performed.

Treatment

Study treatment was comprised of calcitriol at a dose of 0.5 μg/kg, given orally in 4 divided doses over 4 hours on Day 1, then docetaxel at a dose of 36 mg/m2 given intravenously over 15–30 minutes on Day 2 of each treatment week. Treatment was administered weekly for 6 consecutive weeks on an 8-week cycle. Subjects received dexamethasone at a dose of 8 mg orally 12 hours and 1 hour before docetaxel infusion and 12 hours after docetaxel infusion (required to reduce allergic reactions and edema). Treatment was continued until disease progression, unacceptable toxicity, or, in responders, until the PSA was < 4 ng/mL.

Monitoring of Pain and Quality of Life

QOL was assessed every 4 weeks by the European Organization for Research and Treatment of Cancer (EORTC) core questionnaire QLQ-C3012, 13 obtained with permission from the EORTC (Brussels, Belgium). The Present Pain Intensity (PPI) scale of the McGill-Melzack Pain Questionnaire14, 15 was used to assess pain level every 4 weeks and analgesic consumption was quantified as described by Tannock et al.7 on the same schedule.

Assessment of Outcome

The analgesic end points were modeled after those reported by Tannock et al.7, 14 in their investigations of prednisone and mitoxantrone plus prednisone in men with AIPC. The PPI scale of the McGill-Melzack Pain Questionnaire asks the patient to characterize their pain level by assigning one of the following six categories: 0: no pain, 1: mild pain, 2: discomforting pain, 3: distressing pain, 4: horrible pain, and 5: excruciating pain.15 Pain response was defined as a 2-point reduction in the PPI score (or compete relief if the initial PPI score was 1)14, 15 maintained on two consecutive evaluations ≥ 4 weeks apart without an increase in analgesic medication consumption. Analgesic response was defined as a pain response as described earlier or a 50% decrease in analgesic medication use without an increase in pain maintained on 2 consecutive evaluations ≥ 4 weeks apart. Pain progression was defined as an increase in the PPI score by ≥ 1 or an increase in analgesic medication use by ≥ 50% confirmed 4 weeks later.7

Statistical Analysis

Descriptive statistics were used to summarize patient characteristics. The Fisher exact test was used to compare baseline characteristics between patients who reported any pain (PPI score ≥ 1) and provided analgesic consumption information on study entry and those who did not. Kaplan–Meier analysis was used to estimate the median time to pain progression.

Baseline QLQ-C30 scores were calculated by averaging the scores obtained at screening and on Day 1 of treatment (median of 7 days apart). For each domain, we determined whether patients had a clinically meaningful change, defined by a change of ≥ 10 or ≤ −10 in two consecutive evaluations ≥ 4 weeks apart. This degree of change is believed to be clinically important because it is easily perceived by patients.16, 17 The sign test was used to determine whether the proportion of patients who improved is significantly higher than those who worsened.

Changes over time in each domain score were also evaluated. Each domain score was analyzed separately using the unbalanced repeated-measures analysis.18 The variance–covariance matrix was assumed to have a toeplitz structure, which assumes the same correlation between observations taken at a certain week apart. The repeated-measures analysis included data up to 28 weeks as a significant number of patients were removed from the study due to tumor progression beyond this time point. We evaluated the overall effect of time as well as specific comparisons of the 12 weeks, 16 weeks, and 28 weeks of treatment to baseline. The specific comparisons were performed using appropriate contrast statements. Bonferroni adjustment was used to control the overall Type I error for specific comparisons. QOL analyses were performed using SAS PROC MIXED procedure (SAS/STAT User's Guide, version 8; SAS Institute Inc., Cary, NC).

RESULTS

Patient Characteristics and Evaluable Status

Thirty-seven eligible patients were registered between May 2000 and May 2001. Thirty of these patients reported a PPI score ≥ 1 at baseline and analgesic consumption information was available for all but 1 patient. Therefore, 29 patients were evaluable for pain and analgesic response. The baseline characteristics of these patients are summarized in Table 1.

Table 1. Patient Characteristics on Study Entry
CharacteristicsNo. of patients (%)
  1. ECOG: Eastern Cooperative Oncology Group; PSA: prostate-specific antigen; PPI: Present Pain Intensity scale.

No. of patients29
Median age (yrs) (range)72 (46–83)
ECOG performance status 
 09
 112
 28
 30
Median PSA level (ng/mL) (range)95 (6–921)
Site of metastases 
 Bone only20 (69)
 Bone and lymph nodes8 (28)
 Lymph nodes only1 (3)
Previous radiation 
 External beam radiation to primary12 (41)
 External beam radiation to metastases7 (24)
 Radiopharmaceuticals1 (3)
Median no. of previous systemic agents for prostate carcinoma (range)2 (1–3)
PPI score at baseline (range, 0–5) 
 113
 29
 37
Median analgesic score (range)3 (0–30.9)

Analgesic Response

Of the 29 patients, 14 (48%; 95% confidence interval [95% CI], 30–67%) met criteria for analgesic response. Seven of these patients met this criterion by a reduction in the PPI score with stable analgesic medication consumption; four patients met this criterion by having a 50% reduction in analgesic medication consumption without an increase in the PPI score; and 3 patients met this criterion by achieving the required reduction in both the PPI score and analgesic medication use.

The median time to pain progression for all 29 patients was 33 weeks (95% CI, 28–38 weeks). The median time to pain progression for the 14 patients who met criteria for analgesic response was 41 weeks (95% CI, 26–56 weeks).

QLQ-C30 Results

The proportion of patients who experienced a ≥ 10-point improvement or a 10-point worsening is shown in Table 2. Briefly, there was a statistically significant decrease in Physical Functioning, Role Functioning, and Global Health Status/Quality of Life. Among the symptom scales, there was a statistically significant increase in Fatigue and Appetite Loss and a statistically significant decrease in Constipation.

Table 2. Proportion of Patients (n = 29) with a Clinically Meaningful Change from Baseline (Confirmed Change of ≥ 10)
QOL domainBetterWorseP valuea
  • QOL: Quality of life.

  • a

    Sign test.

Functional scales   
 Physical Functioning14%55%0.012
 Role Functioning7%66%0.0002
 Emotional Functioning17%24%0.77
 Cognitive Functioning14%28%0.39
 Social Functioning17%31%0.42
 Global Health Status/QOL7%38%0.023
Symptom scales   
 Fatigue10%62%0.002
 Nausea and Vomiting10%10%1.00
 Pain52%24%0.13
 Dyspnea14%34%0.18
 Insomnia34%31%1.00
 Appetite Loss14%48%0.031
 Constipation38%7%0.026
 Diarrhea21%28%0.79
 Financial Problems24%17%0.77

We evaluated the impact of time on changes in QOL domains. A significant time effect was observed for Physical Functioning, Role Functioning, Social Functioning, Global Health Status/Quality of Life, Fatigue, and Appetite Loss. A significant decline was observed at Week 12 for Physical Functioning, Role Functioning, and Global Health Status/Quality of Life. The decline persisted at Weeks 16 and 28. A significant increase in Fatigue was observed at Weeks 16 and Week 28, whereas Appetite Loss increased significantly at Week 28 (Table 3, Fig. 1).

Table 3. Changes in Domains for which a Time Effect Was Significant
ScoresTime versus baseline: mean change (SE)
Week 12Week 16Week 28
  • SE: standard error; QOL: quality of life.

  • a

    Bonferroni-adjusted (P < 0.01).

  • b

    Bonferroni-adjusted (P < 0.05).

Physical Functioning−10.5 (3.4)a−11.8 (4.0)a−12.8 (3.9)a
Role Functioning−14.8 (5.1)b−16.5 (6.0)b−21.6 (5.5)a
Social Functioning−14.8 (4.4)a−13.1 (4.9)b−11.2 (3.8)b
Global Health Status/QOL−10.7 (3.8)b−10.0 (4.0)b−12.6 (3.7)a
Fatigue15.2 (4.6)a17.5 (5.2)a16.3 (5.0)a
Appetite Loss15.0 (6.4)16.0 (7.4)19.6 (5.7)a
Figure 1.

Mean change from baseline for all QLQ-C30 domains for which a time effect was observed (n = 29).

DISCUSSION

Treatment with calcitriol and docetaxel was found to produce significant reductions in pain and/or analgesic consumption in 48% of the patients. The analgesic response rate observed in the current study is relatively similar to that reported by Tannock et al. with mitoxantrone and prednisone7 and by our group for single-agent docetaxel.10 The current study provides further evidence of the potential of chemotherapy to palliate pain related to prostate carcinoma.

The contribution of corticosteroids to the observed analgesic activity of this regimen is difficult to determine. Steroids administered on a daily schedule have important clinical activity in prostate carcinoma7, 8, 14, 19–22 and have well described analgesic properties in this patient population.7, 14 A randomized trial showed that the analgesic response rate of prednisone (using the same definition) is approximately 20% and is significantly improved with the addition of mitoxantrone.7 Brief administration of steroids routinely accompanies docetaxel administration as a means of reducing the adverse effects of docetaxel.23 Weitzman et al.24 evaluated 20 mg dexamethasone given orally every 6 hours for three doses administered every 3 weeks and showed no activity in 12 patients. To our knowledge, a similar study with the regimen of dexamethasone that routinely accompanies weekly docetaxel (3 doses over 24 hours once per week) has not been conducted.

A comparison of these data to the published experience with docetaxel alone10 shows that a similar proportion of patients achieved an analgesic response. Although the duration of analgesic response for docetaxel alone was not reported, the median time to clinical progression was 20 weeks. The duration of analgesic response observed in the current study is considerably longer. Without a direct randomized comparison, it is not possible to determine if calcitriol adds to the analgesic activity of docetaxel.

The current analysis was handicapped somewhat by the patient population and the limitations of the methodology utilized. The PPI and analgesic scores are limited by their inability to differentiate between malignant and nonmalignant pain. The prevalence of nonmalignant pain ranges from 25–50% in populations of similar age.25, 26 The presence of nonmalignant pain could confound results in two ways. First, observed pain relief may be due to relief of nonmalignant pain. Second, persistent nonmalignant pain could preclude a patient from meeting criteria for pain relief even when cancer pain is relieved. Although the PPI score provides a pragmatic tool to assess overall pain in patients with prostate carcinoma, its utility to compare the antineoplastic activity of different regimens may be blunted by its lack of specificity for antineoplastic effects. The lack of specificity would be expected to be greatest in populations with relatively low levels of cancer-related pain.

Not surprisingly, the evidence in the literature suggests that analgesic benefits are difficult to detect in populations with a relatively low level of pain. For example, the regimen of docetaxel, estramustine, and hydrocortisone, which has substantial antineoplastic activity, did not have demonstrable palliative activity in a group of patients with AIPC with relatively low levels of pain.27 Similarly, in a randomized study that compared clodronate/mitoxantrone/prednisone with mitoxantrone and prednisone, Ernst et al.28 were not able to show a palliative advantage with the addition of the bisphosphonate. In a post-hoc subset analysis of patients with a PPI score > 2, a significant difference emerged. In the current study, the levels of pain were relatively low. Approximately 76% of evaluable patients had PPI scores ≤ 2 and 38% of patients who reported pain were not taking analgesic medications.

The QOL analysis shows changes in several function and symptom domains and a reduction in Global Health Status/Quality of Life. Available evidence suggests that these changes represent toxicity of therapy rather than tumor progression. Only 1 of 37 patients enrolled in the trial had initial tumor progression. Of the 37 patients, 30 met criteria for response and the remaining 6 patients had stable disease lasting ≥ 16 weeks. Further, fatigue and anorexia are the expected adverse effects with this regimen.

It is not surprising that QOL declined in a population of patients with a relatively low intensity of pain at study entry. Pain at study entry was significantly and negatively correlated with baseline QOL (Spearman's r = −0.67; P < 0.0001). Therefore, patients with low levels of pain and high QOL at baseline had little opportunity for QOL gains from pain relief. The results reported in the current study should be viewed as hypothesis generating as the absence of a control arm (standard in Phase II studies of cancer chemotherapy) makes it impossible to definitively determine the palliative value of calcitriol and docetaxel. The results are consistent with those of other investigators,27, 28 suggesting that future investigations of the palliative value of chemotherapy in AIPC should focus on patients with significant levels of pain or other cancer-related symptoms. The results also illustrate the importance of analysis of palliative end points and QOL in clinical trials of cancer chemotherapy. The inclusion of a QOL analysis puts the relief of specific symptoms, i.e., pain, in a broader context that likely better reflects the patient's overall experience.

Treatment with calcitriol and docetaxel produced significant reductions in pain and analgesic consumption in approximately 50% of the patients in the current study. However, in this population of patients with relatively low levels of pain, treatment was also associated with worsening of several QOL domains. The palliative activity and effects on QOL of this regimen should be evaluated further in a controlled study that targets patients with higher levels of cancer-related pain.

Acknowledgements

Calcitriol was supplied by Roche Laboratories, Inc.

Ancillary