1 Massey Cancer Center of Virginia Commonwealth University and other institutions, Richmond, VA; 2 Baltimore, MD; 3 Charleston, WV; 4 Phoenix, AZ; 5 Winston-Salem, NC; 6 Minneapolis, MN, USA; 7 Morges, Switzerland; 8 Barcelona, Spain; 9 Marshfield, WI, USA; 10 Sydney, Australia
* Correspondence to: Dr T.J. Smith, Virginia Commonwealth University, Division of Hematology/Oncology and Palliative Care, MCV Box 980230, Richmond, VA 23298-0230, USA. Tel: +1-804-828-9723; Fax: +1-804-828-8079; Email: tsmith{at}hsc.vcu.edu
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Patients and methods:: We compared those who received IDDS with those who did not receive IDDS (non-IDDS). All patients had Visual Analogue Scores (VAS) for pain 5/10 on at least 200 mg morphine or equivalent daily.
Results:: At 4 weeks, 46 of 52 (88.5%) IDDS patients achieved clinical success compared with 65 of 91 (71.4%; P=0.02) non-IDDS patients, and more often achieved 20% reduction in both pain VAS and toxicity [35 of 52 (67.3%) versus 33 of 91 patients (36.3%); P=0.0003]. By 12 weeks, 47 of 57 (82.5%) IDDS patients had clinical success compared with 35 of 45 (77.8%; P=0.55) non-IDDS patients, and more often had a
20% reduction in both pain VAS and toxicity [33 of 57 (57.9%) versus 15 of 45 patients (33.3%); P=0.01]. At 12 weeks the IDDS VAS pain scores decreased from 7.81 to 3.89 (47% reduction) compared with 7.21 to 4.53 for non-IDDS patients (42% reduction; P=0.23). The 12 week drug toxicity scores for IDDS patients decreased from 6.68 to 2.30 (66% reduction), and for non-IDDS patients from 6.73 to 4.13 (37% reduction; P=0.01). All individual drug toxicities improved with IDDS at both 4 and 12 weeks. At 6 months, only 32% of the group randomized to CMM and who did not cross over to IDDS were alive, compared with 52%59% for patients in those groups who received IDDS.
Conclusions:: IDDS improved clinical success, reduced pain scores, relieved most toxicity of pain control drugs, and was associated with increased survival for the duration of this 6 month trial.
Key words: cancer, implantable devices, intraspinal therapy, intrathecal therapy, opioids, pain
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
This pre-planned analysis was conducted to determine if relief from pain and drug toxicity was sustained after the first 4 weeks, how long it was sustained, and to define the impact of the interventions as treated.
![]() |
Patients and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The primary evaluation of clinical effectiveness was done 4 weeks after randomization by comparing the Visual Analog Scale (pain today on a 010 continuous scale ranging from no pain to the worst pain imaginable) and the composite drug toxicity score, the sum of 15 individual drug toxicity scores (04) selected before the trial. These were measured by the Common Toxicity Criteria (CTC) used by the National Cancer Institute (NCI) for all cooperative group clinical trials for the measurement of drug-related side-effects, available at the NCI website (http://ctep.info.nih.gov/CTC3/default.htm). Comprehensive medical management of pain and toxicity in both study arms was done by the algorithms described in Management of Cancer Pain: A Quick Reference Guide for Clinicians [3]. Those assigned to the CMM group received all pain therapy except spinally administered drugs, cordotomy, or other similar neurosurgical interventions. Those who received IDDS started with morphine but could receive other analgesics if morphine proved to be inadequate for pain relief, using specified algorithms. Crossover was allowed for clinical failure (VAS pain score persistently >5 despite maximum tolerated drug dosages) after consultation with one of the lead investigators (T.J.S. or P.S.) to ensure consistent determination that pain was not being adequately controlled and that reasonable modalities had been tried. Other treatments such as radiation for palliation, chemotherapy for palliation, bisphosphonates, etc. were allowed, and were equally distributed across the study arms.
To analyze the results of sustained treatment, as in the primary study analysis, the primary outcome was the difference in number of patients receiving clinically successful and effective pain treatment [20% improvement in either pain VAS or composite drug-related toxicity using NCI CTC]. The as treated analyses presented here were planned in the original protocol. As randomized comparisons are not presented as the crossover rate and the reduction in sample size due to deaths from advanced disease make them of little practical use.
Results were analyzed by two main groups. The IDDS group included patients implanted after randomization to the IDDS arm (IDDS/implant, n=73), and those assigned to the CMM arm who were implanted after failure of assigned therapy (CMM/implant, n=30). The non-IDDS group included both patients assigned to the CMM arm (CMM/no implant, n=69), and those who were randomized to IDDS but did not receive an implant (IDDS/no implant, n=28).
Crossover was explicitly allowed in the trial, as IDDS is a Food and Drug Administration (FDA)-approved device with multiple cohort studies documenting efficacy [4] and its use could not be ethically withheld. Alternative study designs, such as placement of a sham pump, or placebo in the IDDS after the patient had already experienced relief by an intraspinal trial prior to IDDS implant, were also rejected as unethical in this vulnerable population.
The pain and VAS scores were adjusted using multiple linear regression models that included adjustments for candidate baseline factors to control for differences in these patient groups, after reassignment from the well balanced randomization groups to the new groups based on treatment actually received. The final model used to evaluate VAS scores included the baseline score, and the treatment received during the first 4 weeks. The model for the 12-week analysis also included whether chemotherapy had been administered previously. The final model for drug toxicity scores at both the 4-week and 12-week intervals included only the baseline score and the treatment actually received. The reduction in scores for both VAS and drug toxicity are from the adjusted linear models and not an exact subtraction of the baseline and follow-up values.
A P-value of 0.05 was predetermined as significant, and all tests are two-tailed. The authors had full access to all the data in the study and T.S. takes responsibility for the integrity of the data and the accuracy of the data analysis.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
|
|
|
|
As reported in the paper describing early results [2], complications were infrequent and balanced between the two groups (data not shown), and did not represent a clinical burden.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
This as treated analysis more fully illustrates the impact of IDDS. At 4 weeks, 88.5% of IDDS patients achieved clinical success compared with 71.4% (P=0.02) of non-IDDS patients, and IDDS patients more often achieved 20% reduction in both pain VAS and toxicity (67.3% versus 36.3%; P=0.0003). By 12 weeks, 22 CMM patients had crossed over to IDDS due to refractory pain or toxicity, improving the values of the remaining non-IDDS group and lowering the values of the IDDS group: the Will Rogers phenomenon [10
]. Despite this, by 12 weeks 82.5% of IDDS patients achieved clinical success compared with 77.8% (P=0.55) of non-IDDS patients, and more often achieved
20% reduction in both pain VAS and toxicity (57.9% versus 33.3%; P=0.01). The groups became more similar as the most refractory CMM patients crossed over to IDDS and achieved pain and drug toxicity relief.
At 4 weeks the IDDS VAS pain scores decreased by 60% compared with 37% in non-IDDS patients (P=0.002). At 12 weeks the IDDS VAS pain scores had decreased by 47%, compared with 42% in the non-IDDS group (P=0.23). The most refractory group, CMM patients who crossed over and received IDDS implants, had pain VAS reductions of 27%, which was both clinically and statistically significant.
The reduction in drug side-effects in IDDS patients was remarkable. The 4-week drug toxicity scores for IDDS decreased 55%, compared with 20% for the non-IDDS group (P=0.0003). The 12-week drug toxicity scores for IDDS patients decreased 66%, and for the non-IDDS group by 37% (P=0.01). All of the individual toxicities were reduced with IDDS. One reason that toxicities were reduced was that the pumps were used to deliver a variety of agents. While morphine is the only drug approved by the FDA for intrathecal delivery, by 4 weeks after implant two-thirds of patients had other drugs in the IDDS, including more lipophilic opioids, local anesthetics, and clonidine. Local anesthetics and clonidine do not have a similar effect when delivered systemically. The most refractory group, CMM patients who crossed over and received IDDS implants, had drug toxicity reductions of 50%, which was both clinically and statistically significant.
Toxicities are one of the most important reasons for non-compliance with pain medicine regimens and poor pain control [9], and are significant sources of disability and suffering for patients and their families. There is even indirect evidence that opioid toxicities may reduce survival. As early as 1983, animal studies had shown that common side-effects of high-dose opioids in clinically relevant concentrations included immunosuppression and tumor enhancement [11
13
]. More recent animal studies have demonstrated tumor angiogenesis secondary to high-dose opioids [14
] with drug levels similar to those obtained with the drug doses used in this trial. Clearly, further research is needed before any definitive statements can be made about causation.
The improved survival observed in the earlier as randomized analysis was associated with assignment to the IDDS study arm and, independently, to reduction in opioid side-effects within 4 weeks of randomization and improvement in performance status [2]. In this analysis, survival was examined by the treatment actually received, as shown in Figure 6. It is important to note that survival was not a primary end point of the trial, so all observations must be taken with appropriate caution. Patients who received IDDS therapy had improved survival, with 52%54% alive at 6 months compared with 32% of the non-IDDS group. The patients who were failed by CMM therapy after randomization and who received IDDS appeared to have similar survival to those who started with IDDS. Explanations for the worse survival of the CMM/no implant group could include pain or drug toxicity, poor prognosis precluding an implant (although performance status does not appear to be different), chance alone, or some combination of the above. Explanations for the relatively good survival of the IDDS/no implant group could include a relatively higher proportion of nociceptive pain, which responds well to CMM, and relatively lower pain scores, as seen in Table 2; however, none of these differences were statistically significant. Those who were randomized to IDDS but never received it because their pain and toxicity was controlled had survival equal to those who received IDDS, suggesting that control of symptoms or maintenance of performance status may be one key to better survival. The data are not sufficient to resolve these important questions; unfortunately, the planned replication trial of the North Center Cancer Treatment Group has been cancelled due to concern about low accrual (personal communication, Dr Charles Loprinzi, 4 October 2004).
Pain control may be just one type of supportive care that has the potential to improve survival. The results in this trial are similar to preliminary results [15] showing better preserved quality of life, better symptom control, and improved survival of patients randomized to standard oncology care versus standard oncology care with ongoing palliative care consultation (personal communication, Dr John Fenn, 30 January 2003). For pancreas cancer patients with pain undergoing the Whipple procedure, a splanchnic block with alcohol versus placebo significantly improved pain control and survival [16
]. A more recent study of similar patients showed that splanchnic block improved pain relief in patients with pancreatic cancer versus optimized systemic analgesic therapy alone. At 1 year, 16% of splanchnic block patients and 6% of opioid-only patients were alive, but the change in survival was not significantly different (P=0.26, proportional hazards regression) [17
]. In a randomized trial of erythropoietin versus placebo for chemotherapy-associated severe anemia, survival was improved with erythropoietin through most time on treatment, with a P-value of 0.13 [18
]. In a randomized trial of darbepoetin versus placebo, survival was increased in the darbepoetin group to 46 weeks [95% confidence interval (CI)=39 to 53 weeks] compared with 34 weeks in the placebo group (95% CI = 29 to 39 weeks) [19
].
The median survival of >100 days for patients who received IDDS would make the IDDS system cost-neutral compared with alternative methods such as an external epidural catheter [20] or home intravenous patient-controlled analgesia, with costs estimated at $4000/month [3
]. As reported in preliminary form [5
], the cost-effectiveness analysis ratio of $239 000 would be in the same realm as other expensive supportive care interventions such as bisphosphonates for bone metastases [21
] or expensive anti-emetics (reviewed in [22
]). A full cost analysis of the final trial results is in progress.
These data suggest that current oncology practice, even at expert centers, is not optimal, and that such patients can be helped by referral to a pain management team or management by algorithms, in addition to consideration of IDDS. This is an important point to make, and to emphasize that the CMM group received active effective therapy, not sham or ineffective therapy. The 39% improvement in pain control for the CMM control group is strikingly similar to the 25%40% reduction in pain VAS scores observed by DuPen et al. in the randomized trial of algorithms for pain management in current oncology practice [6].
In this study, patients randomized to IDDS or who received it after failure of CMM had reduced pain scores, significant relief of most pain control drug toxicities, and possibly improved survival for the duration of this trial. Even the most refractory patients failed by CMM had a 27% reduction in pain scores, a 50% reduction in drug toxicity, and a median survival of >3 months after receiving IDDS. For patients similar to those on this trial, on at least 200 mg of oral morphine or equivalent a day without adequate pain relief or with unmanageable side-effects, IDDS is supported by sufficient evidence [4] to warrant consideration.
![]() |
Appendix |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Acknowledgements |
---|
Competing interests
Mr Pool was formerly employed by Medtronic, Inc. Authors Smith, Coyne, Staats, Rauck, Buchser, Stearns, and Boortz-Marks have received honoraria from Medtronic for lectures or consulting. Authors Smith and Staats had grant support from Medtronic to the academic institutions to perform the trial.
Received for publication September 6, 2004. Revision received November 23, 2004. Accepted for publication December 27, 2004.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
2. Smith TJ, Staats PS, Deer T et al. Randomized clinical trial of an implantable drug delivery system compared with comprehensive medical management for refractory cancer pain: impact on pain, drug-related toxicity, and survival. J Clin Oncol 2002; 20: 40404049.
3. Jacox A, Carr D, Payne R, et al. Management of Cancer Pain. Clinical Practice Guideline No. 9. Agency for Health Care Policy and Research (AHCPR publication No. 940592). Rockville, MD: AHCPR Publications; 1994.
4. Smith TJ, Coyne PJ. What is the evidence of the implantable drug delivery systems for refractory cancer pain? Support Cancer Ther 2004; 1: 185189.
5. Hillner BE, Smith TJ, Coyne PJ, Cassel B. The implantable drug delivery trial group preliminary estimates of the cost effectiveness (CE) ratio of implantable drug delivery systems (IDDS) compared to comprehensive medical management (CMM) in patients with refractory cancer pain. Proc Am Soc Clin Oncol 2003; 22: 531.
6. DuPen S, DuPen A, Polossar N et al. Implementing guidelines for cancer pain management: results of a randomized controlled clinical trial. J Clin Oncol 1999; 17: 361370.
7. Cleeland CS, Gonin R, Hatfield AK et al. Pain and its treatment in outpatients with metastatic cancer. N Engl J Med 1994; 330: 592596.
8. Vainio A, Auvinen A. Prevalence of symptoms among patients with advanced cancer: an international collaborative study. Symptom Prevalence Group. J Pain Symptom Manage 1996; 12: 310.[CrossRef][ISI][Medline]
9. Miaskowski C, Dodd MJ, West C et al. Lack of adherence with the analgesic regimen: a significant barrier to effective cancer pain management. J Clin Oncol 2001; 19: 42754279.
10. Feinstein AR, Sosin DM, Wells CK. The Will Rogers phenomenon. Stage migration and new diagnostic techniques as a source of misleading statistics for survival in cancer. N Engl J Med 2004; 312: 16041608.
11. Lewis JW, Shavit Y, Terman GW et al. Stress and morphine affect survival of rats challenged with a mammary ascites tumor (MAT 13762B). Nat Immune Cell Growth Regulation 1983; 3: 4350.
12. Shavit Y, Terman GW, Lewis JW et al. Effects of footshock stress and morphine on natural killer lymphocytes in rats: studies of tolerance and cross tolerance. Brain Res 1986; 372: 382385.[CrossRef][ISI][Medline]
13. Liebeskind JC. Pain can kill. Pain 1991; 44: 34.[CrossRef][ISI][Medline]
14. Gupta K, Kshirsagar S, Chang L et al. Morphine stimulates angiogenesis by activating proangiogenic and survival-promoting signaling and promotes breast tumor growth. Cancer Res 2002; 62: 44914498.
15. Finn JW, Pienta KJ, Parzuchowski F, Worden F. Palliative care project: bridging active treatment and hospice for terminal cancer. Proc ASCO 2002; 1452.
16. Lillemoe K, Cameron JL, Kaufman HS et al. Chemical splanchnicectomy in patients with unresectable pancreatic cancer. A prospective randomized trial. Ann Surg 1993; 217: 447455.[ISI][Medline]
17. Wong GY, Schroeder DR, Carns PE et al. The effect of neurolytic celiac plexus block on pain relief, quality of life, and survival in patients with unresectable pancreatic cancer: a prospective, double blinded, randomized clinical trial. JAMA 2004; 291: 10921099.
18. Littlewood TJ, Bajetta E, Nortier JWR et al. Effects of epoetin alfa on hematologic parameters and quality of life in cancer patients receiving non-platinum chemotherapy: results of a randomized, double-blind placebo-controlled trial. J Clin Oncol 2002; 20: 24862494.
19. Vansteenkiste J, Pirker R, Massuti B et al. Double-blind, placebo-controlled, randomized phase III trial of darbepoetin alfa in lung cancer patients receiving chemotherapy. J Nat Cancer Inst 2002; 94: 12111220.
20. Bedder MD, Burchiel K, Laron A. Cost analysis of two implantable narcotic delivery system. J Pain Symptom Manage 1991; 6: 368373.[CrossRef][ISI][Medline]
21. Hillner BE, Weeks JC, Desch CE, Smith TJ. Pamidronate in prevention of bone complications in metastatic breast cancer: a cost-effectiveness analysis. J Clin Oncol 2000; 18: 7279.
22. Earle CC, Chapman R, Baker C et al. Systematic overview of cost-utility assessments in oncology. J Clin Oncol 2000; 18: 33023317.
|