This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Halabi, S. Articles by Small, E. J. Search for Related Content PubMed Articles by Halabi, S. Articles by Small, E. J. Related Articles Related Correspondence Social Bookmarking                            What's this?

In Reply

Department of Biostatistics and Bioinformatics, Duke University; Cancer and Leukemia Group B Statistical Center, Durham, NC

Nicholas J. Vogelzang

Nevada Cancer Institute, Las Vegas, NV

Alice B. Kornblith

Dana-Farber Cancer Institute, Boston, MA

San-San Ou

Fred Hutchinson Cancer Center, Seattle, WA

Philip W. Kantoff

Dana-Farber Cancer Institute, Boston, MA

Nancy A. Dawson

Georgetown University, Washington, DC

Eric J. Small

University of California, San Francisco, San Francisco, CA

We appreciate the comments of Klepstad and Kaasa regarding our article¹ and share their view concerning the difference between pain intensity and the pain interference score. In our analysis, both pain intensity and pain interference score are prognostic factors of overall survival. Furthermore, we agree with Stenseth et al² and Klepstad and Kaasa that patients are generally not capable of discerning between loss of function caused by pain as opposed to other causes. The Brief Pain Inventory (BPI), was validated by Cleeland,³ who reported cut points of pain severity related to deterioration in the different areas of pain interference.

The entire BPI was used to assess pain in men who were randomly assigned on Cancer and Leukemia Group B trial 9480. At baseline, men were asked to rate their pain at its worst during the last 24 hours. The pain intensity question uses an 11-point numerical rating scale, with 0 representing no pain and 10 representing pain as bad as you can imagine. In the present analysis of 317 men, we examined which of the baseline factors predicted pain intensity. In addition, we explored the relationship between pain intensity and pain interference scale and investigated whether pain severity predicts worse clinical outcomes. We used a cut point of 5 or greater to represent severe pain because it has been shown that a pain severity rating of 5 or higher in patients with cancer correlates with disproportionately more functional impairment and decreased enjoyment of life.⁴ The median pain interference score at baseline was 17, and this cut point (median) was used to dichotomize the patients into two groups (low and high pain).¹

Of 317 men, 44% had a pain severity score of 5 or greater. Pain intensity was highly correlated with the BPI pain interference score (r = 0.72, P < .0001). In multiple linear regression, statistically significant predictors of pain interference score were age, race, hemoglobin, performance status, opioid analgesic use, alkaline phosphatase, and years since diagnosis. However, statistically significant predictors of pain intensity were age, performance status, opioid analgesic use, prior radiation, and years since diagnosis.

In univariate analysis, pain intensity was a statistically significant predictor of overall survival. The median survival times were 8.76 months (95% CI, 6.88 to 11.30 months) in men whose pain intensity score was 5 or greater compared with 18.15 months (95% CI, 15.47 to 20.20 months; P < .001) in men with a pain intensity score less than 5.

Adjusting for age, race, performance status, presence of visceral disease, opioid analgesic use, testosterone, hemoglobin, prostate-specific antigen (PSA), lactate dehydrogenase, alkaline phosphatase, Gleason score, years since diagnosis, prior treatment with radiation, and body mass index, pain intensity remains a statistically significant predictor of overall survival. The adjusted hazard ratio (HR) of death was 1.61 (95% CI, 1.21 to 2.13; P < .001) in men whose pain intensity score was 5 or greater compared with men with a pain intensity score less than 5.

Furthermore, when pain intensity was considered as a continuous variable, the HR for death for each unit increase in pain intensity was 1.06 (95% CI, 1.01 to 1.11; P = .026). In multivariable analysis, both pain intensity and pain interference predicted overall survival. The adjusted HR of death was 1.41 (95% CI, 1.04 to 1.91; P < .001) in men whose pain intensity score was 5 or greater compared with men whose pain intensity score was less than 5. The same variables as listed earlier were included in the multivariable proportional hazards model. Moreover, the HR of death was 1.37 (95% CI, 1.02 to 1.83; P < .001) in men whose pain interference score was 17 or greater compared with men whose pain interference score was less than 17.

We also found that pain intensity predicted time to bone progression and a 50% decline in PSA. In the proportional hazards model with the same variables as listed earlier, the adjusted HR was 1.45 (95% CI, 1.11 to 1.90; P = .006). The adjusted odds ratio for 50% decline in PSA from baseline was 1.72 (95% CI, 1.08 to 2.78; P = .023). However, pain intensity did not predict progression-free survival (adjusted HR = 1.15; 95% CI, 0.88 to 1.50; P = .307).

These results demonstrate that both pain interference score and pain intensity are significant predictors of overall survival in men with castrate-refractory prostate cancer. Furthermore, pain interference score and pain intensity are statistically significant predictors of overall survival even when adjusting for established prognostic factors. It is known that high levels of alkaline phosphatase, PSA, and lactate dehydrogenase; low levels of hemoglobin; and high Gleason scores are indicators of advanced progression and death. Thus, these findings support the hypothesis that pain intensity and pain interference score are validated measures of advanced disease in men with prostate cancer. We agree with Klepstad and Kaasa in that investigating the relationship between pain and clinical outcomes in cancer patients is a fruitful area of research.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

The author(s) indicated no potential conflicts of interest.

ACKNOWLEDGMENTS

Supported in part by Grants No. DAMD 17-03-1-0112 and W81XWH-06-1-0032 from the Department of Defense and Grant No. CA 36601 from the National Cancer Institute, Bethesda, MD.

REFERENCES

1. Halabi S, Vogelzang NJ, Kornblith AB, et al: Pain predicts overall survival in men with metastatic castration-refractory prostate cancer. J Clin Oncol 26:2544-2549, 2008[Abstract/Free Full Text]

2. Stenseth G, Bjørnnes M, Klepstad P, et al: Can cancer patients assess the influence of pain on functions? A randomised, controlled study of the pain interference items in the Brief Pain Inventory. BMC Palliat Care 6:2, 2007[CrossRef][Medline]

3. Cleeland CS: Assessment of pain in cancer: Measurement issue, in Foley KM (ed): Advances in Pain Research and Therapy. New York, NY, Raven Press, 1990, pp 47-55

4. Daut RL, Cleeland CS: The prevalence and severity of pain in cancer. Cancer 50:1913-1918, 1982[CrossRef][Medline]

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

Related Correspondence

• Predication of Cancer Patients’ Survival by Pain Interference Items Is Not the Same As Evidence That Pain Intensity Is a Survival Predictor
  Pål Klepstad and Stein Kaasa
  JCO 2008 26: 4215-4216 [Full Text]
• Predication of Cancer Patients’ Survival by Pain Interference Items Is Not the Same As Evidence That Pain Intensity Is a Survival Predictor
  Pål Klepstad and Stein Kaasa
  JCO 2008 26: 4215-4216 [Full Text]

This Article Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Halabi, S. Articles by Small, E. J. Search for Related Content PubMed Articles by Halabi, S. Articles by Small, E. J. Related Articles Related Correspondence Social Bookmarking                            What's this?