2149

Impairment of Heart Rate Variability during Paclitaxel Therapy

Eeva M. K. Ekholm, M.D., Ph.D.1 Eeva K. Salminen, M.D., Ph.D.2 Heikki V. Huikuri, M.D., Ph.D.3 Jarmo Jalonen, B.Sc.4 Kari J. Antila, M.D., Ph.D.5 Tuula A. Salmi, M.D., Ph.D.1 Virpi T. Rantanen, M.D., Ph.D.1
1

Department of Obstetrics and Gynecology, University of Turku, Turku, Finland. Department of Oncology and Radiotherapy, University of Turku, Turku, Finland. Division of Cardiology, Department of Medicine, University of Oulu, Oulu, Finland. Cardiorespiratory Research Unit, University of Turku, Turku, Finland. Department of Clinical Physiology, University of Turku, Turku, Finland.

BACKGROUND. Paclitaxel, which has been reported to be effective in treating metastatic breast carcinoma and advanced ovarian carcinoma, has been associated with cardiac side effects. Therefore, the effect of paclitaxel on cardiovascular autonomic regulation was studied. METHODS. Twenty-four-hour ambulatory electrocardiogram measurements were recorded twice from 14 women with breast or ovarian carcinoma: once before paclitaxel treatment and once on the day after the second chemotherapy course. Heart rate variability (HRV) was assessed with spectral analysis. For the frequency domain analysis, HRV was assessed in the very low (0.005 0.040 hertz [Hz]), low (0.040 0.150 Hz), and high frequency (0.150 0.400 Hz) spectral components. RESULTS. The ratio between low frequency and high frequency HRV decreased (daytime values of 2.7% [standard deviation (SD) 1.6] vs. 1.7% [SD 0.9]; P 0.0098) after 2 courses of paclitaxel. The circadian uctuation of HRV also decreased in all studied frequency components. CONCLUSIONS. The observed changes in spectral characteristics suggest that autonomic modulation of the heart rate is impaired after paclitaxel therapy. However, from these data it is not clear whether the observed changes are permanent or whether autonomic cardiac function returns to normal some time after treatment. Further studies are needed to examine whether these indices based on HRV can be used to detect those patients at risk for cardiac side effects during chemotherapy. Cancer 2000;88:2149 53. 2000 American Cancer Society. KEYWORDS: autonomic nervous system, chemotherapy, heart rate variability, paclitaxel.

Supported by Bristol-Meyers-Squibb. Address for reprints: Eeva M. K. Ekholm, M.D., Ph.D, Department of Obstetrics and Gynecology, University of Turku, P.O. Box 52, FIN- 20521 Turku, Finland. Received July 19, 1999; revision received December 8, 1999; accepted January 3, 2000. 2000 American Cancer Society

aclitaxel is used as standard therapy for metastatic breast carcinoma and advanced ovarian carcinoma.1,2 However, treatment with this taxane compound is not recommended for patients with recent myocardial infarction, symptomatic angina, or cardiac failure,3 because cardiac side effects including ventricular tachycardia, heart block, and ischemic events have been associated with paclitaxel therapy.4 Furthermore, asymptomatic bradycardia has been reported in up to 29% of the patients receiving paclitaxel.5 Few cardiac deaths have been reported in association with paclitaxel treatment. Some patients have died of myocardial infarction,6 and other patients have shown symptoms of congestive heart failure.7,8 At present, there are no means to detect patients at risk for severe cardiac adverse events during chemotherapy. Twenty-four-hour ambulatory heart rate recordings have revealed that decreased heart rate variability (HRV) is associated with increased risk of mortality9 and susceptibility to life-threatening arrhythmias10,11 in patients with coronary artery disease without changes in the heart rate. The circadian rhythm of neural cardiac

2150

CANCER May 1, 2000 / Volume 88 / Number 9 TABLE 1 Patient Characteristics

Characteristics Patients analyzed Breast carcinoma Ovarian carcinoma Peritoneal surface carcinoma Age (yrs) (median/range) Prior radiotherapy for breast carcinoma for left-side breast carcinoma Prior anthracyclines Cumulative doxorubicin mg/m2 Cumulative epirubicin mg/m2 (median/range) Metastatic sites Visceral Liver Lung Skeletal No. of organs involved 1 2 3 Prior peripheral neuropathy (WHO grade) 0 1 Peripheral neuropathy after two paclitaxel courses (WHO grade) 0 1 2
WHO: World Health Organization.

regulation as assessed by HRV is altered in patients with heart disease,12 hypertension,13 and diabetes.14 Decreased HRV is an early sign of autonomic neuropathy,15 which can present a problem with the use of some cytotoxic drugs as shown earlier with anthracyclines and Vinca alcaloids.16 18 Twenty-four-hour recordings permit a reproducible analysis of autonomic regulation and enable estimation of circadian changes in HRV.19 We previously have reported that paclitaxel changes sympathetic control of blood pressure examined with a battery of autonomic function tests.20 To assess the efcacy of 24-hour electrocardiogram (ECG) in nding paclitaxel-induced changes in cardiac control, we studied HRV before and after treatment with paclitaxel.

No. 14 9 4 1 48 (3874) 7 4 11 450 338 (40670) 4 5 6 5 7 4 3 10 4

PATIENTS AND METHODS

Patients
Fourteen women treated for ovarian or breast carcinoma with paclitaxel were included in the study. Paclitaxel was given as single agent therapy in 10 cases, combined with cyclophosphamide in one case, and combined with cisplatin in three cases. The dose used per course was 90 200 (median 153) mg/m2. Three patients were given 90 mg/m2 paclitaxel with cisplatin 60 mg/m2. Paclitaxel was given as a 3-hour infusion with recommended antihistamine premedication. Patient characteristics are presented in Table 1. None of the patients had cardiovascular diseases or other medical history of relevance. Four of 14 patients had received prior radiotherapy for their left side breast carcinoma. Eleven patients had received prior treatment with anthracyclines (Table 1). The study was approved by the Ethics Committee of the Turku University, and all participants gave informed consent.

8 5 1

Methods
A 24-hour ambulatory ECG was recorded from all patients during normal activities with their normal sleepwake rhythm. The patients were tested once before paclitaxel treatment and on the day after the second chemotherapy course. The two-channel recordings were analyzed with a Delmar Avionics scanner (Delmar Avionics, Irvine, CA).

Analysis of HRV
The ECG data were sampled digitally and transferred from the Delmar Avionics scanner to a microcomputer. Custom-made software (Heart Signal Co., Oulu, Finland) was used for HRV analysis. The intervals between two R waves in the ECG (R-R interval) were measured. A linear detrend was applied to the R-R interval data segments of 512 samples to make them

more stationary. This was implemented by rst tting a straight line to each segment by a standard least squares method and then subtracting it from the sample value. The R-R interval series was passed through a lter that eliminates unwanted premature beats and noise and lls the resulting gaps with an average value computed in the immediate neighborhood (the average of R-R intervals including 3 cycles before and 3 cycles after the ectopic beats). An R-R interval is interpreted as a premature beat if it deviates from the previous qualied interval for more than a given tolerance level, which is a programmable parameter dependent on the prematurity index of ectopic beats for each patient. The details of this ltering technique have been described previously.11,21 Only segments with greater than 90% qualied beats were included in the analysis. To quantify the periodic components of HRV, we used spectral analysis. Very low frequency variability is associated with sympathetic vasomotor regulation. Low frequency variability relates to baroreex activity and is modulated by both sympathetic and parasympathetic control. High frequency variability is vagally mediated and reects respiratory sinus arrhythmia.22

Paclitaxel Decreases Heart Rate Variability/Ekholm et al.

2151

The computer program automatically calculates the autoregressive coefcients to dene the power spectrum density. Power spectra was quantied in four frequency bands: very low frequency power from 0.005 to 0.040 hertz (Hz), low frequency power from 0.040 to 0.150 Hz, and high frequency power from 0.150 to 0.400 Hz. The ratio between low frequency and high frequency spectra was also calculated (LF/HF ratio).23

TABLE 2 R-R Interval and Heart Rate Variability before and after Two Courses of Paclitaxel Treatment in 14 Patients
Before treatment (mean [SD]) R-R interval (ms) Daytime Sleeping hours High frequency power (ms2) Daytime Sleeping hours Low frequency power (ms2) Daytime Sleeping hours Very low frequency power (ms2) Daytime Sleeping hours LF/HF ratio Daytime Sleeping hours After treatment (mean [SD])

P value

682 [82] 823 [89] 97 [95] 209 [133] 205 [103] 428 [297]

725 [81] 793 [115] 209 [276] 209 [275] 382 [560] 325 [250]

0.013 0.319 0.032 0.568 0.354 0.135

Statistics
Study of the effect of paclitaxel on HRV was conducted during the hours of 4 9 p.m. and 12 6 a.m. First, repeated measures analysis of variance was performed using the BMDP statistical package (2V) (BMDP Statistical Software, Inc., Los Angeles, CA). If this analysis showed interaction (P 0.05), the treatment effect was calculated separately for the sleeping hours (0 6 a.m.) and awake hours (4 9 p.m.). The circadian effect on the heart rate and HRV was estimated by comparing the differences between day and night variabilities before and after paclitaxel treatment. Log transformations were performed for skewed data. The data are shown as mean (standard deviation [SD]).

520 [237] 978 [583] 2.7 [1.6] 2.4 [1.9]

737 [553] 873 [554] 1.7 [0.9] 1.9 [1.1]

0.144 0.486 0.01%

SD: standard deviation; ms: milliseconds; LF: low frequency; HF: high frequency.

RESULTS
Daytime heart rate was signicantly lower (R-R interval 725 [SD 81] milliseconds [ms]) after two courses of paclitaxel compared with pretreatment level (R-R interval 682 [SD 82] ms [P 0.013]), whereas nighttime heart rate did not change (823 [SD 89] ms vs. 793 [SD 115] ms; before vs. after, P 0.32). The LF/HF ratio decreased after paclitaxel therapy (P 0.007). This decrease was seen both in day (2.7 [SD 1.6] vs. 1.7 [SD 0.9]) and night recordings (2.4 [SD 1.9] vs. 1.7 [SD 1.1]) (Table 2). The high frequency component of HRV increased signicantly in the daytime after paclitaxel treatment (96.7 [SD 95.9] ms2 vs. 208.8 [SD 275.9] ms2, P 0.032), whereas nighttime variability showed no changes (271.1 [SD 257.4] ms2 vs. 208.6 [SD 133.4] ms2, P 0.6) (Table 2). Thus, the circadian uctuation in high frequency HRV practically disappeared after paclitaxel treatment (Table 3). Hourly values of high frequency HRV are shown in Figure 1. The nonsignicant increase in the very low and low frequency oscillations in the heart rate during daytime and a decrease during the night also resulted in decreased circadian uctuation of these oscillations of the heart rate after two courses of paclitaxel (Table 3). The recordings did not reveal a signicant number of extrasystoles in any of the patients. The results from the three patients who had not received anthracyclines did not differ from the results from patients who

TABLE 3 The Difference between Day and Nighttime Variability of the Heart Rate before Treatment and after Two Courses of Paclitaxel in 14 Patients
Before treatment (mean [SD]) R-R interval (ms) Very low frequency power (ms2) Low frequency power (ms2) High frequency power (ms2)
ms: milliseconds.

After treatment (mean [SD]) 68 [62], P 0.012 0.08 [0.15], P 0.06 0.04 [0.30], P 0.54 0.07 [0.23], P 0.29

141 [49], P 0.0001 0.27 [0.14], P 0.0001 0.31 [0.21], P 0.0001 0.42 [0.30], P 0.0002

had previously received anthracycline therapy (data not shown).

DISCUSSION
We found that paclitaxel treatment decreased the LF/HF ratio and reduced the circadian rhythm of HRV in patients with metastatic breast carcinoma or ovarian carcinoma. HRV has been used to predict mortality from 24hour ECG recordings.9 Moreover, subtle abnormalities in HRV that are related to various cardiovascular disorders can be assessed by spectral analysis.10,24 In the current study, both reduced LF/HF ratio and attenuated circadian rhythm of HRV were observed during

2152

CANCER May 1, 2000 / Volume 88 / Number 9

FIGURE 1. Mean hourly values of the high frequency component of heart rate variability before and after paclitaxel treatment in 14 patients. The bars represent standard deviation of the mean. HF: high frequency.

paclitaxel treatment. The ndings suggest that paclitaxel has an adverse effect on cardiovascular autonomic regulation, which may predispose the patients to cardiovascular complications. Previous studies have demonstrated that in healthy subjects HRV has a reproducible circadian rhythm that results from increased parasympathetic activity at night.19 This rhythm was apparent also in our patients before paclitaxel treatment but disappeared after two paclitaxel courses. Blunted circadian rhythm of spectral measures of HRV has been observed to be an early sign of cardiovascular abnormality12 and predicts mortality in patients with heart disease.25 The blunted rhythm in the patients with cardiovascular disease has been attributed both to decreased parasympathetic and increased sympathetic activity, but the etiology and exact mechanisms for this impairment are not fully understood. We have shown previously that paclitaxel does not impair HRV in short recordings obtained in the daytime.20 However, circadian variation of HRV is a more sensitive measure of damaged vagal function.12 Maximal vagal capacity that is manifest during the night cannot be attained by subjects with vagal dysfunction. However, during the daytime when vagal tone is lower, subtle impairment cannot be detected. The LF/HF ratio decreased after paclitaxel treatment. Previously, an association between reduced LF/HF ratio and mortality has been reported in patients with myocardial infarction24 and severe heart failure.26 As in our patients, the high frequency component of HRV was preserved also in the heart failure patients. The observations suggest that paclitaxel may

affect autonomic cardiac regulation in a way that predisposes some patients to cardiovascular complications. Despite the association between the reduced LF/HF ratio and cardiovascular disorders, the mechanisms responsible for this abnormality in spectral characteristics of HRV are not completely known. It has been clearly demonstrated that heart failure results in reduced LF/HF ratio.26 Overall sympathoexcitation may attenuate low frequency modulation of the heart rate and thus reduce the LF/HF ratio. Changes in central autonomic regulation also may lead to the observed decrease in LF/HF ratio. Changes in HRV may partly depend on altered blood pressure control that has been noticed after paclitaxel treatment in short recordings.20 A reduced LF/HF ratio of less than 1.3% has been reported to predict arrhythmic events and cardiac death.27 Seven of our patients had a mean LF/HF ratio of less than 1.3 during either night or daytime after two courses of paclitaxel. A reduced LF/HF ratio and circadian uctuation of HRV may reect impaired cardiac function due to incipient heart failure that predisposes to cardiac arrhythmias and conduction disorders. Most of our patients had a relapse of cancer and had received several previous chemotherapy regimens. Because it was unlikely that all the patients would tolerate, e.g., six courses, we decided to evaluate the effects of paclitaxel on autonomic cardiac function already after the second course, which leads to an interval of only 3 4 weeks between the examinations. Hence, the impairment in autonomic cardiac control is most likely not due to disease progression. Most of the serious cardiac adverse effects reported have occurred in the early phase of the paclitaxel treatment,7,8,28,29 which supports our nding on early drug-induced impairment of autonomic cardiac function. Our patients were studied on the day after the second course of paclitaxel treatment. Therefore, it is not clear whether the observed changes are permanent or whether autonomic cardiac function returns to normal some time after treatment. Because the second ECG was recorded the day after paclitaxel infusion, it is not likely that the premedication affects the results. Despite the observed changes in HRV, none of our patients suffered from any signicant cardiac side effects. Further studies are needed to examine whether these indices based on HRV can be used to identify patients at risk for cardiac side effects during chemotherapy. In conclusion, the observed decrease in the LF/HF ratio and the circadian uctuation of HRV suggest that autonomic cardiac modulation is impaired after two

Paclitaxel Decreases Heart Rate Variability/Ekholm et al.

2153

courses of paclitaxel. Whether this impairment is associated with the cardiac adverse events that have been reported with paclitaxel therapy warrants further studies.

15. 16.

REFERENCES
1. McGuire WE, Hoskins WJ, Brady MF, Kucera PR, Partridge EE, Look KY, et al. Cyclophosphamide and cisplatin compared with paclitaxel in patients with stage III and stage IV ovarian cancer. N Engl J Med 1996;334:1 6. Clemons M, Leahy M, Valle J, Jayson G, Ranson M, Howell A. Review of recent trials of chemotherapy for advanced breast cancer: the Taxanes. Eur J Cancer 1997;33:218393. Bisset D, Kaye SB. Taxol and taxotere current status and future prospects. Eur J Cancer 1993;29A:1228 31. Hochster H, Wasserheit C, Speyer J. Cardiotoxicity and cardioprotection in chemotherapy. Curr Opin Oncol 1995;7: 304 9. McGuire WP, Rowinsky EK, Rosenshein NB. Taxol: a new antineoplastic agent with signicant activity in advanced ovarian epithelial neoplasms. Ann Int Med 1989;111:2739. Rowinsky EK, McGuire WP, Guarneri T, Fisherman JS, Christian MC, Donehower RC. Cardiac disturbances during the administration of Taxol. J Clin Oncol 1991;9:1704 12. Alagaratnam TT. Sudden death 7 days after paclitaxel infusion for breast cancer. Lancet 1993;342:12323. Jekunen A, Heikkila P, Maiche A, Pyrho nen S. Paclitaxelinduced myocardial damage detected by electron microscopy. Lancet 1994;343:727 8. Kleiger RE, Miller JP, Bigger JT, Moss AJ. Multicenter postinfarction research group: decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 1987;59:256 62. Huikuri HV, Koistinen MJ, Yli-Ma yry S, Airaksinen KEJ, Seppa nen T, Ika heimo MJ, et al. Impaired low-frequency oscillations of heart rate in patients with prior acute myocardial infarction and life-threatening arrhythmias. Am J Cardiol 1995;76:56 60. Huikuri HV, Valkama JO, Airaksinen KEJ, Seppa nen T, Kessler KM, Takkunen JT, et al. Frequency domain measures of heart rate variability before the onset of nonsustained and sustained ventricular tachycardia in patients with coronary artery disease. Circulation 1993;87:1220 8. Huikuri HV, Niemela MJ, Ojala S, Rantala A, Ika heimo MJ, Airaksinen KEJ. Circadian rhythms of frequency domain measures of heart rate variability in healthy subjects and patients with coronary artery disease. Effects of arousal and upright posture. Circulation 1993;90:121 6. Chakko S, Mulingtapang RF, Huikuri HV, Kessler KM, Materson BJ, Myerburg RJ. Alterations in heart rate variability and its circadian rhythm in hypertensive patients with left ventricular hypertrophy free of coronary artery disease. Am Heart J 1993;126:1364 72. Bernardi L, Ricordi L, Lazzari P, Solda P, Calciati A, Ferrari MR, et al. Impaired circadian modulation of sympathovagal activity in diabetes. A possible explanation for altered tem17.

2.

18.

3. 4.

19.

5.

20.

6.

21.

7. 8.

22.

23.

9.

10.

24.

25.

11.

12.

26.

27.

13.

28.

14.

29.

poral onset of cardiovascular disease. Circulation 1992;86: 144352. Wheeler T, Watkins PJ. Cardiac denervation in diabetes. Br Med J 1973;4:584 6. Hirvonen H, Salmi TT, Heinonen E, Antila KJ, Va lima ki IAT. Vincristine treatment of acute lymphoblastic leukemia induces transient autonomic cardioneuropathy. Cancer 1989; 64:8015. Roca E, Bruera E, Politi PM, Barugel M, Cedaro L, Carraro S, et al. Vinca alkaloid-induced cardiovascular autonomic neuropathy. Cancer Treat Rep 1985;69:149 51. Viniegra M, Marchetti M, Losso M, Navigante A, Litovska S, Senderowicz A, et al. Cardiovascular autonomic function in anthracycline-treated breast cancer patients. Cancer Chemother Pharmacol 1990;26:22731. Huikuri HV, Kessler KM, Terracall E, Castellanos A, Linnaloto MK, Myerburg RJ. Reproducibility and circadian rhythm of heart rate variability in healthy subjects. Am J Cardiol 1990;65:3913. Ekholm E, Rantanen V, Antila K, Salminen E. Paclitaxel changes sympathetic control of blood pressure. Eur J Cancer 1997;33:1419 24. Huikuri HV, Linnaluoto MK, Seppa nen T, Airaksinen KEJ, Kessler KM, Takkunen JT, et al. Circadian rhythm of heart rate variability in survivors of cardiac arrest. Am J Cardiol 1992;70:610 5. Akselrod S, Gordon D, Madwed JB, Snidman NC, Shannon DC, Cohen RJ. Hemodynamic regulation: investigation by spectral analysis. Am J Physiol 1985;249:H86775. Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P, et al. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res 1986;59:178 93. Bigger JTJ, Fleiss JL, Steinman RC, Rolnitzky LM, Kleiger RE, Rottman JN. Frequency domain measures of heart period variability and mortality after myocardial infarction. Circulation 1992;85:164 71. Odemuyiwa O, Malik M, Farrell T, Bashir Y, Poloniecki J, Camm J. Comparison of the predictive characteristics of heart rate variability index and left ventricular ejection fraction for all-cause mortality, arrhythmic events and sudden death after acute myocardial infarction. Am J Cardiol 1991; 68:434 9. Van de Borne P, Montano N, Pagani M, Oren R, Somers VK. Absence of low-frequency variability of sympathetic nerve activity in severe heart failure. Circulation 1997;95:1449 54. Bendini MG, Intini A, Corvo P, De Martino A, Maseri A, Lanza GA. Heart rate variability as prognostic tool for cardiac events in patients with idiopathic dilated cardiomyopathy. Eur Heart J 1997;18:409. Biadi O, Mengozzi G, Gherarducci G, Strata G, Mariani M, Baldini F, et al. Evaluation of Taxol cardiotoxicity in metastatic breast cancer. Ann NY Acad Sci 1993;698:4035. Shek TWH, Luk ISC, Ma L, Cheung KL. Paclitaxel-induced cardiotoxicity. An ultrastructural study. Arch Pathol Lab Med 1996;120:89 91.