|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
Journal of Clinical Oncology, Vol 24, No 32 (November 10), 2006: pp. 5171-5172 © 2006 American Society of Clinical Oncology. DOI: 10.1200/JCO.2006.08.6157
In ReplyPrincess Margaret Hospital, University of Toronto, Toronto, Ontario, Canada We thank Schagen and van Dam for their comments about our recent article in the Journal of Clinical Oncology.1 We acknowledged in the discussion of our article that part of the apparent improvement in cognitive function of women at 1 and 2 years after chemotherapy might be accounted for by practice effect; although, we were remiss in not giving this reservation more prominence by including it in the abstract of the article, for example. The longitudinal study that we reported1,2 was designed in 1999 to evaluate the extent and time course of cognitive dysfunction of women receiving chemotherapy for breast cancer. At the time of its design there was very little information about cognitive effects in such a population, and very little guidance as to how to best measure them. While a battery of neuropsychological tests has been the gold standard for people with other medical conditions, prolonged testing is quite stressful for women undergoing chemotherapy, and the multiplicity of tests makes it difficult to identify a single primary end point of cognitive dysfunction that facilitates the design and analysis of a clinical trial. We elected to use the High Sensitivity Cognitive Screen (HSCS) because it contains a brief selection of cognitive tests (it takes about 20 minutes to complete), appeared to be reasonably sensitive both in prior evaluation3 and in our pilot study that evaluated cognitive dysfunction among patients with breast cancer,4 and provided a single end point of cognitive function that prevented problems associated with multiple significance testing. However, as pointed out by Schagen and van Dam, our subsequent work has shown that the HSCS is subject to practice effect, at least when readministered to the same subjects after a short interval of time (median, 17 days5). While such practice effect is likely to decline with increasing intervals between administrations, there is evidence for such an effect from the data for the control women in our recent study, where the apparent incidence of cognitive dysfunction improved from 5% to 3.6% to 0% after 1- and 2-year intervals.1 We have now estimated changes in test results due to practice effect by assuming that it is of the same magnitude in women who received chemotherapy and in controls. We computed the practice effect as the mean difference between the follow-up and the baseline score for the control group. As reported in the literature,6 we found the practice effect greatest between the first and second assessments. We then calculated the reliable change index (RCI) for each individual using the practice effect from the control group (RCI = [Assessment 2 score – Assessment 1 score] – [practice effect estimated from controls]/[SE of the difference in scores]). The RCI score can then be interpreted as a Z score or a CI with changes greater than a designated cut point (generally the 90th or 95th percentile with a Z score of 1.64 or 1.96, respectively) being considered a statistically significant change.7 After adjusting for practice effect from baseline to 1 year, 81% of patients who received chemotherapy showed no change in cognitive function, 10% had decline, and 8% had improvement using the 90th percentile; 92% were stable and 8% improved using the 95th percentile. From baseline to 2 years, 96% of women had no change, 2% showed decline, and 2% showed improvement using the 90th percentile; 98% remained stable and 2% declined using the 95th percentile. These results suggest that cognitive function did not return to baseline levels at 2 years. The other point raised by Schagen and van Dam addresses the variability in incidence of cognitive dysfunction among women who have received chemotherapy for breast cancer, and in particular the variability reported in our own studies.1,2,4,5 However, as these authors have themselves shown,8 the incidence of such dysfunction depends on the chemotherapy received, being more common in women who have received cyclophosphamide, methotrexate, and fluorouracil as in our earlier study4 than where anthracycline-based chemotherapy was used more commonly, as in our subsequent study.1,2 Moreover, our recent pilot study which investigated practice effect5 recruited patients who self-reported that their memory and concentration was not as good as before they commenced chemotherapy. Consistent with the literature,8-10 we showed no association between our objective test results and perception of cognitive impairment. However it is still possible that these inclusion criteria, as well as self selection of patients who were motivated to take part in a detailed study of cognitive dysfunction, rather than an unselected group during chemotherapy,1,2 resulted in a higher incidence of cognitive dysfunction. Cognitive dysfunction is a subtle but important adverse effect of cancer and its treatment, and there are many problems with its characterization. In current studies we are using a more detailed neuropsychological battery for assessment of patients, together with computer-based testing that is less dependent on language skills, and are careful to compare with control groups that allow correction for practice effect. It appears that the impairment seen on the HSCS at baseline in the patient group remains a problem for many women, with only 8% showing improvement after adjusting for practice effect at 1 and 2 years. The effect of practice in improved performance may reflect the adaptation that patients make in dealing with cognitive deficits in their everyday life. Authors' Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest. REFERENCES
1. Mar Fan H, Houédé-Tchen N, Yi Q-L, et al: Fatigue, menopausal symptoms and cognitive function in women following adjuvant chemotherapy for breast cancer: One and two year follow-up of a prospective controlled study. J Clin Oncol 23
: 8025
-8032, 2005 2. Tchen N, Juffs HG, Downie FP, et al: Cognitive function, fatigue, and menopausal symptoms in women receiving adjuvant chemotherapy for breast cancer. J Clin Oncol 21
: 4175
-4183, 2003 3. Fogel BS: The high sensitivity cognitive screen. Int Psychogeriatr 3 : 273 -288, 1991[CrossRef][Medline] 4. Brezden CB, Phillips KA, Abdolell M, et al: Cognitive function in breast cancer patients receiving adjuvant chemotherapy. J Clin Oncol 18
: 2695
-2701, 2000 5. Vardy J, Wong K, Yi QL, et al: Assessing cognitive function in cancer patients. Support Care Cancer doi:10.1007/s00520-006-0037-6 6. McCaffrey RJ, Cousins JP, Westervelt HJ, et al: Practice effects with the NIMH AIDS abbreviated neuropsychological battery. Arch Clin Neuropsychol 10241 -10250, 1995 7. Temkin NR, Heaton RK, Grant I, et al: Detecting significant change in neuropsychological test performance: A comparison of four models. J Int Neuropsychol Soc 5 : 357 -369, 1999[CrossRef][Medline] 8. Schagen SB, Muller MJ, Boogerd W, et al: Late effects of adjuvant chemotherapy on cognitive function: A follow-up study in breast cancer patients. Ann Oncol 13
: 1387
-1397, 2002 9. van Dam FS, Schagen SB, Muller MJ, et al: Impairment of cognitive function in women receiving adjuvant treatment for high-risk breast cancer: High-dose versus standard-dose chemotherapy. J Natl Cancer Inst 90
: 210
-218, 1998 10. Ahles TA, Saykin AJ, Furstenberg CT, et al: Neuropsychologic impact of standard-dose systemic chemotherapy in long-term survivors of breast cancer and lymphoma. J Clin Oncol 20
: 485
-493, 2002
Related Correspondence
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||
|
|||||||||||
|
|
Copyright © 2006 by the American Society of Clinical Oncology, Online ISSN: 1527-7755. Print ISSN: 0732-183X
|
