Paediatric haematologic malignancies such as leukaemia can be treated effectively with conventional chemotherapy using combinations of anti-leukemic agents.Reference Pui, Campana and Evans1–Reference Tomizawa, Tabuchi and Kinoshita3 Some leukemic children, however, require haematopoietic stem cell transplantation, which is usually preceded by high-dose chemotherapy with or without total body irradiation. The infusion of high-dose anti-leukemic agents containing cardiotoxic drugs and the use of large amounts of infusion solutions during stem cell transfusion sometimes cause non-haematological complications that include acute heart failure and arrhythmia. However, it is difficult to predict the occurrence of such cardiac events.
Generally recognised as an index of sympathetic balance and autonomic cardiovascular control, heart rate variability has been explored in various diseases that precipitate cardiac sudden death. Recent evidence has demonstrated a strong association between decreased RR interval variability and sudden death and/or cardiac events related to cardiovascular disease.Reference Kleiger, Miller, Bigger and Moss4–Reference Ajiki, Murakawa and Yanagisawa-Miwa9 In adult patients with haematological malignancies, decreased heart rate variability was observed after chemotherapy and haematopoietic stem cell transplantation, indicating that heart rate variability monitoring could serve in predicting cardiac complications associated with chemotherapy.Reference Poręba, Poręba and Gać10,Reference Nakane, Nakamae and Koh11 Although several reports have described decreased autonomic nervous function after chemotherapy in paediatric leukaemia patients, none have addressed the changes in heart rate variability before and after haematopoietic stem cell transplantation.
The present study analysed autonomic nerve function using 24-hour Holter monitoring of heart rate variability to characterise the differences between patients undergoing chemotherapy and normal controls, and the changes in heart rate variability between before and after transplantation.
Materials and methods
Study population and protocol
We retrospectively reviewed the clinical records of two patient groups. The transplant group contained 19 patients with haematological malignancies who underwent chemotherapy followed by haematopoietic stem cell transplantation between 2006 and 2018. The non-transplant group included age- and sex-matched 19 patients with similar haematological malignancies who received chemotherapy during same period but did not require transplantation. The criteria for participation in the study were age of 5 years or older and completion of 24-hour Holter monitoring after chemotherapy in the non-transplant group and before and after transplantation in the transplant group. The patients were diagnosed as having acute lymphoblastic leukaemia, acute myeloblastic leukaemia, or malignant lymphoma. Although all patients in the transplant group had received conventional chemotherapy before transplantation, none exhibited evidence of cardiac disease. Pre-conditioning for all patients involved cyclophosphamide (120 mg/kg) and total body irradiation (12 Gy in 4 patients, 8 Gy in 14 patients, and 4 Gy in 1 patient).
The normal control group contained 32 children who were age- and sex-matched with the transplant group. The control group participants had originally undergone 24-hour Holter monitoring for the purpose of screening for heart disease due to chest pain, palpitation, and heart rate irregularities, and/or abnormal electrocardiogram findings pointed out at routine school heart examinations. All normal controls were ultimately judged as healthy.
A 24-hour Holter electrocardiogram (DSC-3300, Nihon Koden, Tokyo, Japan) with heart rate variability analysis was performed prior to pre-conditioning therapy (before transplantation) and at the time of discharge (after transplantation). During Holter monitoring, the patients were in clinically good condition, with no visible signs of heart failure, infection, severe anaemia, or abnormalities in electrolytes or C-reactive protein (Table 2). According to previously reported formula,Reference Ewer, Benjamin, Yeh, Kufe, Pollock and Weichselbaum12,Reference Sakata-Yanagimoto, Kanda and Nakagawa13 the cumulative anthracycline dose was calculated using the following ratios: doxorubicin 1.0, daunorubicin 0.5, pirarubicin 0.8, mitoxantrone 3.4, idarubicin 1.6, and epirubicin 0.6.
Any arrhythmia, heart failure, or cardiac dysfunction during preparative conditioning therapy or after transplantation was defined as a cardiac complication in patients with transplantation. Echocardiographic abnormalities included ventricular dilatation and systolic and/or diastolic dysfunction.
This study was approved by the institutional review board of Shinshu University School of Medicine (authorization number: 3718). Parental consent was obtained using an opt-out approach.
Heart rate variability analysis
Time-domain and power-spectrum analyses of heart rate variability were carried out (MemCalc/CHIRAMI ver1.1.11, GMS, Tokyo, Japan) in all examined patients following visual verification of the automatic electrocardiogram recordings. Heart rate variability was analysed for whole 24-hour Holter monitoring data and separately for awake and sleeping hours. The time-domain analysis of heart rate variability included mean heart rate, coefficient of variation of RR interval, standard deviation of all normal RR intervals, standard deviation of the average normal RR interval for all 5-minute segments, root mean square successive difference of intervals, and percentage of pairs of successive RR intervals which differed by more than 50 ms.
The power-spectrum analysis was performed in terms of total spectral power as well as for very low frequency (0.003–0.04 Hz), low frequency (0.04–0.15 Hz), and high frequency (0.15–0.50 Hz) power. Whereas the high-frequency component is mainly mediated by parasympathetic activity, low-frequency fluctuations are regulated by both sympathetic and parasympathetic function.Reference Luutonen, Antila, Neuvonen, Räihä, Rajala and Sourander14 Very low-frequency heart period rhythms are influenced by the renin–angiotensin–aldosterone system and parasympathetic mechanisms.Reference Taylor, Carr, Myers and Eckberg15 As relative value indices, low frequency/high frequency indicating sympathetic activity balance and high frequency/(low frequency+high frequency) as an indicator of parasympathetic activity balance were used. The precise definitions and abbreviations of the examined time-domain and power-spectrum parameters are listed in Table 1.
Definitions of time-domain and power-spectrum analysis parameters of heart rate variability
Statistical analysis
Statistical analysis was conducted using SPSS software ver. 24 (IBM, IL, USA). Distribution normality was assessed by the Shapiro–Wilk test. Data are expressed as the mean ± sd or the median (interquartile range). Fisher’s exact test or the chi-square test was performed to compare variables between groups stratified by category. To compare differences among the two patient groups and the control group, the Mann–Whitney U-test and the Kruskal–Wallis test followed by post hoc (Bonferroni) analysis were adopted for other variables. To determine the significance of differences between variables before and after transplantation, the Wilcoxon paired sequence test was applied. For all tests, a p-value of less than 0.05 was considered statistically significant.
Results
The clinical characteristics of the patients and age- and sex-matched controls are summarised in Table 2. The cumulative dose of anthracycline was significantly higher in the pre-transplant group than in the non-transplant group. Regarding cardiac function, the ratio of early and atrial mitral inflow velocity was higher in the control group than in the non-transplant group, while contractility and cardiothoracic ratio in chest x-rays were comparable.
Clinical characteristics and comparisons among the patient group and controls
*p < 0.01 versus pre-transplant patients.
†p < 0.05 versus non-transplant patients.
Table 3 compares the results of heart rate variability analysis among the three groups. Basic heart rhythm was sinus rhythm in all participants. During awake hours, we observed significant differences for all variables in time-domain heart rate variability and power-spectrum analysis between the pre-transplant group and the control group apart from the coefficient of variation of RR interval. The mean heart rate, power of very low frequency, and power of low frequency were significantly different between the non-transplant group and the control group. On the other hand, time-domain heart rate variability and power-spectrum analysis of sleeping hours showed significant differences for all variables between the pre-transplant group and the non-transplant group as well as the control group, apart from the coefficient of variation of RR interval and standard deviation of the average normal RR interval for all 5-minute segments.
Comparison of time-domain analysis and power spectrum analysis parameters of heart rate variability among patient groups and controls
The Kruskal–Wallis test followed by multiple comparison (Bonferroni) analysis was adopted for variables.
CVRR=coefficient of variation of RR interval; HF=high-frequency power; LF=low-frequency power; mHR=mean heart rate; pNN50=number of pairs of successive RR intervals which differed by more than 50 ms expressed as a percentage of the total number of RR intervals; rMSSD=root mean square differences; SDANN=standard deviation of the average normal RR interval for all 5-minute segments; SDNN=standard deviation of all normal RR intervals; VLF=very low-frequency power.
*p < 0.05 versus pre-transplant patients.
†p < 0.01 versus pre-transplant patients.
‡p < 0.05 versus non-transplant patients.
§p < 0.01 versus non-transplant patients.
The average interval between 24-hour Holter monitoring before and after transplantation was 95 ± 19 days. Analysis of 24-hour heart rate variability data for before and after transplantation revealed low frequency to be significantly lower after transplantation (p < 0.05). Regarding awake/sleeping periods, low frequency was significantly lower after transplantation during awake hours only (p < 0.05). No other remarkable changes were seen among the patients (Table 4).
Comparison of time-domain analysis and power spectrum analysis parameters of heart rate variability between before and after HSCT
CVRR=coefficient of variation of RR interval; HF=high-frequency power; HSCT=haematopoietic stem cell transplantation; LF=low-frequency power; mHR=mean heart rate; pNN50=number of pairs of successive RR intervals which differed by more than 50 ms expressed as a percentage of the total number of RR intervals; rMSSD=root mean square differences; SDANN=standard deviation of the average normal RR interval for all 5-minute segments; SDNN=standard deviation of all normal RR intervals; VLF=very low-frequency power.
Four of the 19 patients who required a transplant had cardiovascular complications (Table 5). All four exhibited asymptomatic non-sustained ventricular tachycardia, with one having mild cardiac dysfunction that necessitated treatment with carvedilol and a vasodilator. There were no significant differences for age, cumulative dose of anthracycline, or any variable of heart rate variability analysis (Table 6). One patient who was able to be tracked by Holter monitoring up to 2 years after transplantation had non-sustained ventricular tachycardia during preparative conditioning therapy (Case 2 in Table 5). Thereafter, the patient was asymptomatic, but the heart rate variability imbalance observed after transplantation persisted (Fig 1a and b).
Profiles of patients with cardiac complications during and after preparative conditioning of haematopoietic stem cell transplantation
ALL=acute lymphoblastic leukaemia; Allo-BMT=allogenic bone marrow transplantation; AML=acute myeloblastic leukaemia; CR=complete remission; LV=left ventricular; ML=malignant lymphoma; NSVT=non-sustained ventricular tachycardia; UR-CBT=unrelated cord blood cell transplantation.
Comparison of baseline clinical profiles and parameters of heart rate variability before HSCT between patients with and without cardiac complications
CVRR=coefficient of variation of RR interval; HF=high-frequency power; HSCT=haematopoietic stem cell transplantation; LF=low-frequency power; mHR=mean heart rate; pNN50=number of pairs of successive RR intervals which differed by more than 50 ms expressed as a percentage of the total number of RR intervals; rMSSD=root mean square differences; SDNN=standard deviation of all normal RR intervals; SDANN=standard deviation of the average normal RR interval for all 5-minute segments; VLF=very low-frequency power.
(a) Time-domain analysis of case 2 up to 2 years after haematopoietic stem cell transplantation. HR=heart rate; SDNN=standard deviation of all normal RR intervals. (b) Power spectrum analysis of case 2 up to 2 years after haematopoietic stem cell transplantation. HF=high-frequency power; LF, low-frequency power.
Discussion
In the present study, significant differences were observed in virtually all parameters between the healthy control group and the patient groups, which indicated sympathetic hyperactivity and parasympathetic hypoactivity in children with haematologic diseases. On the other hand, no differences were noted in the patient group with regard to heart rate variability after transplantation apart from improvements in low frequency.
Cardiovascular autonomic function may become impaired during the treatment of and recovery from leukaemia.Reference Nazir, AlFutaisi and Zacharia16–Reference Caru, Corbin and Périé18 Several articles have demonstrated the effect of chemotherapy on autonomic modulation assessed through heart rate variability, especially that of vincristine and anthracycline.Reference Nazir, AlFutaisi and Zacharia16–Reference Caru, Corbin and Périé18 In this study, vincristine was used in many patients, as is common in lymphoblastic leukaemia and lymphoma. There was a significant difference in the cumulative dose of anthracycline and heart rate variability during sleep between the non-transplant and pre-transplant groups. Vincristine is an anti-cancer drug member of standard leukaemia treatment protocols. Also potentially causing peripheral neuropathy, vincristine was shown to attenuate the interaction between breathing and autonomic heart rate control, which suggested parasympathetic complications in leukemic children.Reference Hirvonen, Salmi, Heinonen, Antila and Välimäki17 Anthracycline is another key drug in leukaemia treatment; however, it may additionally damage the myocardium and cardiac autonomic nerves.Reference Caru, Corbin and Périé18 Caru et al. reported that cumulative doses of doxorubicin could have a significant adverse effect on the cardiac autonomic nerve system, which supported the results of the present study. Protection strategies (i.e., dexrazoxane treatments) to counter the negative effects of the drug were considered to prevent changes observed in the cardiac autonomic nervous system.Reference Caru, Corbin and Périé18
An earlier investigation for adults which evaluated haematopoietic stem cell transplantation revealed changes in heart rate variability and heart rate turbulence, both of which were attributed to the transplantation procedure.Reference Poręba, Poręba and Gać10 The use of cyclophosphamide, total body irradiation, and cytokinesReference Straburzynska-Migaj, Ochotny and Wachowiak-Baszynska19,Reference Kunz-Ebrecht, Mohamed-Ali, Feldman, Kirschbaum and Steptoe20 should be considered as well in understanding the complex mechanisms of decreased heart rate variability. Although not investigated in the cohort, mental stress associated with treatment and diminished exercise tolerance due to long-term hospitalisation might also have affected autonomic nervous function.Reference Kunz-Ebrecht, Mohamed-Ali, Feldman, Kirschbaum and Steptoe20,Reference Bellenger, Fuller, Thomson, Davison, Robertson and Buckley21 In this paediatric study, the results in the two groups receiving chemotherapy suggested an impaired autonomic nervous system as compared with healthy controls, especially in pre-transplant patients. Further declines in heart rate variability after transplantation appear minor, unlike in adults.Reference Poręba, Poręba and Gać10 Heart rate variability decreases with ageReference Acharya, Kannathal and Krishnan22,Reference Acharya, Kannathal, Seng, Ping and Chua23 from childhood.Reference Schwartz, Gibb and Tran24 Infants have high sympathetic activity that decreases quickly between the ages of 5 and 10 years.Reference Finley, Nungent and Hellenbrand25 In adults, the attenuation of respiratory sinus arrhythmia with age is predominant.Reference Lipsitz, Mietus, Moody and Goldberger26 Age-related differences in the effects of anticancer drugs and transplants on heart rate variability can also therefore be considered. However, subclinical cardiovascular damage in children after transplantation has also been reported.Reference Borchert-Mörlins, Memaran and Sauer27,Reference Baker, Chow and Steinberger28 This impaired heart rate variability may persist for an extended time after transplantation, although the precise recovery period remains unknown. Since most related studies are cross-sectional, prospective longitudinal investigations are needed to evaluate whether such results in asymptomatic patients can be corroborated with longer follow-up results in survivors.
Although heart rate variability has been demonstrated as a strong predictor of cardiovascular mortality with high specificity, its sensitivity is insufficient to be used alone as a routine screening test.Reference Antman, Anbe and Armstrong29 In this study, there were no significant associations between mid-term cardiac complications and heart rate variability. Further trials are needed to confirm the clinical application of heart rate variability as a prognostic marker of cardiac complications in patients with haematological malignancies, especially in combination with other modalities, such as echocardiography.
This study had several limitations. First, infection, anaemia, and cardiac complications were considered as factors affecting autonomic balance, but other factors such as anxiety, mental stress, fatigue, and physical strength decline due to long-term hospitalisation were not evaluated. Second, as this is a single-centre, small-population, retrospective study, larger, prospective cohort trials at multiple facilities are warranted.
In conclusion, this study is the first to uncover heart rate variability reduction and impaired autonomic nerve system function secondary to chemotherapy in paediatric leukaemia patients, which may cause cardiovascular dysfunction. Anti-cancer drugs such as anthracycline may have an adverse effect on the cardiac autonomic nerve system. Heart rate variability is a clinically reliable and non-invasive approach to evaluate the early status of autonomic dysfunction in leukaemia patients.
Acknowledgements
The authors would like to thank all participants of this study and all individuals involved in data collection, as well as Mr. Trevor Ralph for his English editorial support.
Financial support
This research received no specific grant from any funding agency or commercial or not-for-profit sectors.
Conflicts of interest
None.
