Analysis of ventricular late potentials in signal-averaged ECG of people with epilepsy

Authors


Address correspondence to Dr. Konrad Rejdak, Department of Neurology, Medical University of Lublin, 8 Jaczewskiego Str., 20-954 Lublin, Poland. E-mail: krejdak@europe.com

Summary

Purpose:  There has been growing interest in cardiac disturbances in epilepsy patients and their etiologic role in the context of sudden death. Ventricular late potentials (VLPs) recorded on signal-averaged electrocardiography (SAECG) reflects delayed ventricular depolarization and identifies the structural or functional substrate for the ventricular tachycardia in the reentry mechanism. Therefore, abnormal SAECG poses the potential of identifying patients at increased risk of malignant ventricular arrhythmias and sudden cardiac death. The aim of this exploratory study was to screen epilepsy patients who were treated with established doses of antiepileptic drugs (AEDs) on the presence of VLPs.

Methods:  Forty-five consecutive patients with the diagnosis of epilepsy and 19 healthy volunteers, aged younger than 46 years, participated in the study. Exclusion criteria included symptoms or signs of diseases other than epilepsy, in particular relating to heart disease or medication influencing the cardiovascular system, as well as seizure reported by patients that occurred <3 days before the ECG examination. The electrocardiogram was recorded according to the standard protocol. The seizure frequency was calculated based on the available data of epileptic events within the preceding 3 months. Disease duration was estimated by determining the time from the first reported seizure to the present.

Key Findings:  There were 22 patients (48%) in the epilepsy group and only one patient (5%) in the control group fulfilling the criteria for VLP (p = 0.0005). Subsequently, epilepsy patients were divided into two subgroups according to VLP presence. Patients with VLP had longer disease duration (p = 0.03) compared to those without VLP. Similarly, patients with VLP more frequently had refractory epilepsy (p = 0.03) and had higher monthly seizure frequency (p = 0.02). Analysis of the proportions of generalized seizures (GS) and focal seizures (FS) showed a tendency for higher number of generalized tonic–clonic seizures in the VLP group, but this did not reach statistical significance (p = 0.06). VLP patients tended to be more often on polytherapy (defined as more than one AED per patient) (p = 0.07) as compared to epilepsy patients without VLP. However, if the numbers of AEDs per patients among the subgroups were compared, patients with VLP were treated with more AEDs than patients without VLP (p = 0.01). The study was not sufficiently powered to pinpoint any particular drug or AED combination to influence the appearance of VLP in epileptic patients. In particular, there was no difference in valproate or carbamazepine exposure, considering the percentage of patients exposed or the total daily dose administered.

Significance:  Epilepsy patients more frequently display abnormal SAECGs with VLPs as compared to the control population, and their presence correlates with the disease duration, uncontrolled seizures, and polytherapy. Further longitudinal studies are needed in order to stratify the risk of life-threatening ventricular events in epilepsy patients with VLPs.

There has been growing interest in cardiac disturbances in epilepsy patients and their etiologic role in the context of sudden death (Nei, 2009; Surges et al., 2010a). Indeed, epilepsy is associated with an increased mortality rate compared with the general population, and sudden unexpected death in epilepsy (SUDEP) represents about half of the seizure-related deaths in this group. The main risk factors for SUDEP are associated with poorly controlled seizures. Among cardiac factors, bradyarrhythmias and asystole, as well as tachyarrhythmias and alterations to cardiac repolarization, should be considered (Surges et al., 2009, 2010c). The difficulty in identifying the patients at risk stems from the fact that the prevalence of interictal cardiac arrhythmias in patients with epilepsy is similar to that in the healthy population (Blumhardt et al., 1986; Massetani et al., 1997). However, decreased heart rate variability among people with epilepsy, particularly when the epilepsy is refractory, suggests an alteration of the autonomic balance in such patients (Sevcencu & Struijk, 2010). Therefore, there is a need for more sensitive methods to detect cardiac dysfunction in order to prevent the life-threatening events.

Signal-averaged electrocardiography (SAECG) is a noninvasive method for the detection of high frequency, microvolt-level cardiac electrical activity in the terminal portion of QRS complexes and has been widely used to record ventricular late potentials (Batchvarov et al., 2004; Iravanian et al., 2005; Jaroszyński et al., 2005). Ventricular late potentials (VLPs) are thought to result from fragmentation of electromotive forces in abnormal areas of ventricular myocardium, where activation is delayed by slow conduction. Many reports have demonstrated that the presence of abnormal SAECG reflects delayed ventricular depolarization and identifies the structural substrate for the ventricular tachycardia in the reentry mechanism (Gomes et al., 1993; Santangeli et al., 2008). Therefore, abnormal SAECG poses the potential of identifying patients at increased risk of malignant ventricular arrhythmias, and sudden cardiac death (Das & El Masry, 2010).

The current study was undertaken to screen epilepsy patients treated with established doses of antiepileptic drugs (AEDs) for the presence of VLPs.

Methods

The study was approved by the local ethics committee and all subjects provided informed consent to be enrolled in the study groups.

The study was carried out at the Outpatient Clinic, Department of Neurology, Medical University of Lublin, Poland. Forty-five consecutive patients aged younger than 46 years participated in the study. First, patients had to have the diagnosis of epilepsy confirmed, according to the recommendations of the International League Against Epilepsy (ILAE) (Berg et al., 2010). Secondly, the exclusion criteria included symptoms or signs of diseases other than epilepsy, in particular relating to heart disease such as chest pain, exercise intolerance, fainting, dyspnea, palpitation, and so on, or medication influencing the cardiovascular system. In addition, patients underwent standard blood workup (liver and renal function, serum electrolytes, basic hematologic indices) performed not longer than 1 week prior to enrollment. Serum levels of cardiac enzymes were not measured for the purpose of that study. All study subjects had physical examination before ECG recording, and any significant abnormalities such as substantial valvular heart murmur, third heart sound, and sings of pulmonary congestion or edema were excluded. In addition, patients were asked to provide recent chest x-ray done no longer than 6 months prior to study enrollment in order to exclude apparent pathologic changes.

All patients were carefully interviewed and clinically examined, and the epilepsy type was classified according to the recommendations (Berg et al., 2010). Patients were categorized according to the disease activity and treatment response into two subgroups: refractory and controlled according to the recent guidelines (Kwan et al., 2010). The seizure frequency was calculated based on the available data of epileptic events within the preceding 3 months. Seizure reported by patients occurring <3 days before the ECG examination were within the exclusion criteria. Disease duration was estimated by determining the time from the first reported seizure to the present.

Control group consisted of 19 healthy volunteers (aged <46 years), who underwent ECG protocol similar to that of patients with epilepsy.

ECG recording

The ECG was recorded according to the standard protocol in a quiet room between 10:00 a.m. and noon, (to avoid circadian rhythm bias), during spontaneous quiet breathing, and following 10 min of adjustment in a supine position. Firs, the standard ECG recording was performed as a screening procedure and any abnormalities including cardiac chamber enlargement or hypertrophy, ST elevation and ST depression, pathological Q waves, or arrhythmia were excluded. Based on the recording, the heart rate was measured using the RR interval (RRI, in milliseconds; ms), and P wave (ms), PQ intervals (ms), and QRS (ms) were calculated. QT intervals were manually measured from the start of the QRS complex to the end of the T wave (defined by the intersection with the isoelectric line). QT and preceding RR intervals were determined from three to five successive ECG complexes, and corrected QT intervals (QTc) were calculated using the Bazett’s formula [QTc = QT/√(RR) in ms] (Bazett, 1920).

Secondly, SAECG was performed using equipment constructed in the National Institute of Cardiology (Warsaw, Poland) and at least 300 QRS complexes were recorded and averaged for further analysis (Fig. 1). Signals received from three bipolar orthogonal Frank leads were amplified and filtered with the use of Butterworth filters between 40 and 250 Hz. Recordings were accepted for analysis if two of the following criteria were met: number of beats averaged >250 and mean noise levels <1.0 μV. We considered SAECG parameters abnormal if signal-averaged QRS duration (SA-QRSd) >114 ms, duration of low amplitude signals lower than 40 μV (LAS40) >38 ms, and root mean square voltage of signals in the last 40 ms of the high frequency QRS intervals (RMS40) <20 μV. An SAECG study was considered positive for VLPs if two of these three parameters were abnormal according to previous recommendations (Breithardt et al., 1991).

Figure 1.

 Graph demonstrating the method of calculation of SAECG parameters: QRSd, duration of the filtered QRS complex (ms); RMS40, root mean square voltage of the last 40 ms of the high frequency QRS intervals (μV); and LAS40, duration of low-amplitude signals lower than 40 μV (ms) for the VLP criteria assessment. We considered SAECG parameters abnormal if SA-QRSd >114 ms, LAS40 > 38 ms, or RMS40 < 20 μV. SAECG was considered positive for VLP if two of these three parameters were abnormal.

Echocardiography

Those patients, who had VLP detected on SAECG were offered an additional examination with standard transthoracic echocardiography using HP Sonos 5,500 ultrasound machine (Hewlett Packard, Duluth GA, U.S.A.), in order to exclude asymptomatic structural heart disease according to current guidelines (Feigenbaum et al., 2009).

Statistical analysis

The baseline data (presented in Tables 1 and 2) were compared using the Mann-Whitney-Wilcoxon test or the Fisher’s exact test on the distribution of respective characteristics between the groups. The Spearman rank correlation coefficient was used for correlation analysis between the duration of ECG parameters and the clinical characteristics. Values were expressed as median with 95% percentile range or mean ± standard deviation (SD) throughout the study. Two-sided tests were used throughout and a p-value of <0.05 was considered statistically significant. Because the sample size in this exploratory study was relatively small, the correction for multiple comparisons has not been performed. The GraphPad InStat 3.05 (GraphPad Software Inc., San Diego, CA, U.S.A.) statistical program was used.

Table 1.   Patients’ characteristics
CharacteristicEpilepsyControlp-value
  1. n, no. of patients; pGE, primary generalized epilepsy; FE, focal epilepsy; F, female; M, male; QTc, corrected QT interval; SA-QRS, signal-averaged QRS duration (ms); LPD, late potential duration (ms); RMS40, root mean-square voltage of the terminal 40 ms of the QRS complex in microvolts (μV); LAS40, duration of low-amplitude signals lower than 40 μV (ms).

N4519
Gender (F/M)27/1812/7ns
Age (years)30 (21–42)29 (20–45)ns
Seizure type (%)   
 pGE10 (22)na
 FE35 (78)na
Etiology (%)   
 Structural/metabolic9 (20)na
 Unknown36 (80)na
Disease duration (years)8 (1–34)na
Seizure control (%)   
 Refractory23 (51)na
 Remission22 (49)na
Therapy (%)   
 AED monotherapy19 (42)na
 AED polytherapy26 (58)na
ECG characteristics   
 P wave (ms)82 (74–89)80 (69–90)ns
 PQ interval (ms)184 (177–190)181 (169–192)ns
 QRS (ms)90 (86–93)85 (70–90)ns
 QTc (ms)378 (332–423)368 (307–428)ns
 SA-QRS (ms)113 (78.5–142.9)105 (91.1–136.6)ns
 RMS40 (μV)16.9 (4.1–63.2)19.2 (9.3–38.2)ns
 LAS40 (ms)26.7 (12.6–55.0)23.6 (15.7–42.4)ns
VLP present/VLP absent22/23 (48%)1/19 (5%)0.0005
Table 2.   Patients’ characteristics
CharacteristicVLP absentVLP presentp-value
  1. n, no. of patients; pGE, primary generalized epilepsy; FE, focal epilepsy; GS, generalized seizures; F, female; M, male; RRI (ms), interbeat interval in milliseconds; QTc, corrected QT interval; QTd, QT dispersion; SA-QRS, signal-averaged QRS duration (ms); LPD, late potential duration (ms); RMS40, root mean-square voltage of the terminal 40 ms of the QRS complex in microvolts (μV); LAS40, duration of low-amplitude signals lower than 40 μV (ms).

N2322 
Gender (F/M)14/913/9ns
Age30 (28–33)30 (27–31)ns
Seizure type (%)   
 pGE7 (30)5 (23)ns
 FE16 (70)17 (77)
Etiology (%)   
 Structural/Metabolic5 (22)4 (18)ns
 Unknown18 (78)18 (82)
Disease duration (years)6 (5–10)11 (8–14)0.03
Seizure control (%)   
 Refractory8 (34)15 (69)0.03
 Remission15 (66)7 (31)
Total seizures/month0 (1–5)4 (4–8)0.02
GS/month0 (1–3)2 (2–3)0.06
ECG characteristics   
 P wave (ms)81 (71–90)83 (73–92)ns
 PQ interval (ms)183 (176–189)181 (169–192)ns
 QRS (ms)86 (81–90)92 (87–96)ns
 QTc (ms)385 (330–420)377 (327–426)ns
 SA-QRS (ms)100.5 (96–104)117.8 (116–123)<0.0001
 RMS40 (μV)21.1 (18–27)14.4 (12–17)0.001
 LAS40 (ms)23.6 (20–25)27.5 (27–36)0.001

Results

Demographic and clinical characteristics of patients are presented in Table 1. There were no significant differences regarding gender or age between the epilepsy patients and controls. There was no significant difference between the epileptic group and control group in the number of patients with the cardiovascular risk factors: smoking, hypertension, or hypercholesterolemia, or in the other diseases present (data not shown). Among the epileptic patients, the majority (35 patients) had localization related epilepsy (LRE), whereas 10 patients presented with primarily generalized seizures (GS). Regarding the etiology of the epilepsy syndrome, there were nine patients with structural changes on brain magnetic resonance imaging (MRI): seven with remote posttraumatic lesions and two had records of benign tumor resected prior to the onset of epilepsy (detailed information not available but demonstrated nonprogressive character of the neoplasm). The remainder of the patients had epilepsy with “unknown” etiology. In the total group of epilepsy patients, 23 were with uncontrolled epilepsy (refractory) and 22 reached sustained remission according to a recently proposed definition (Kwan et al., 2010). Nineteen patients received monotherapy and the remaining 26 were kept on polytherapy with two or more AEDs. Among the AEDs, valproate (77%) and carbamazepine (43%) were the most frequently used, either as monotherapy or polytherapy. None of the study participants received any cardiologic treatment.

Standard diagnostic blood workup excluded any other significant abnormalities including serum electrolytes both in epileptic patients and controls (data not shown).

None of the included control or epilepsy patients had underlying ECG abnormalities that would have warranted further cardiologic evaluation. There were no significant differences in QTc intervals between the controls and epileptic patients (Table 1). Similarly, the median values for RMS40, LAS40, and SA-QRS did not differ between the epilepsy group and controls. However, in the epilepsy group there were 22 patients with abnormal SA-QRS values (49%), 17 patients with abnormal RMS40 values (38%), and 7 with LAS40 values outside the normal range (17%), whereas in the control group five patients had abnormal SA-QRS values (26%), none had abnormal RMS40, and one had abnormal LAS40 values (5%). In the epilepsy group, 22 patients (48%) had at least two abnormal parameters and thus fulfilled the criteria for late potentials (an example of a recording with VLP is demonstrated in Fig. 2 and without VLP in Fig. 3). Among them there were 5 (11%) with all three parameters abnormal. In the control group, there was only one patient (5%) fulfilling the criteria for late potentials with two abnormal parameters (p = 0.0005).

Figure 2.

 SAECG with filtered QRS complex meeting positive VLP criteria: QRS duration = 121 ms, root mean square voltage of the last 40 ms of the high frequency QRS intervals (RMS40) = 14.4 μV, and duration of low amplitude signals lower than 40 μV (LAS40) = 40 ms.

Figure 3.

 SAECG not meeting the criteria for positive VLPs: QRS duration, 113 ms; root mean square voltage of the last 40 ms of the high frequency QRS intervals (RMS40), 44 μV; and duration of low amplitude signals lower than 40 μV (LAS40), 23 ms.

Subsequently, patients were divided into two subgroups according to VLP presence, and their demographic and clinical characteristics are described in Table 2.

No statistically significant differences were found in either age or gender or in etiology or seizure phenotype between the two subgroups. Patients with VLPs had longer disease duration (p = 0.03) compared to those without VLPs. Similarly, patients with VLPs more frequently had refractory epilepsy (p = 0.03) and had more seizures per month (p = 0.02) based on frequency calculated for the 3-month period preceding the ECG examination. When analyzing the proportions of generalized (GS) and focal seizures (FS), there was a tendency for a higher number of generalized tonic–clonic seizures in the VLP group, but it did not reach statistical significance (p = 0.06) (Table 2). VLP patients tended to be more often on polytherapy (p = 0.07) as compared to epilepsy patients without VLP (Table 3). However, if the numbers of AEDs per subject among the subgroups were compared, subjects with VLP were treated with more AEDs than patients without VLP, which reached statistical significance (p = 0.01) (Table 3). The study was not sufficiently powered to pinpoint any particular drug or AED combination to influence the appearance of VLP in epileptic patients.

Table 3.   Patients’ characteristics on drug treatment
CharacteristicsVLP absentVLP presentp-value
  1. n, no. of patients; AED, antiepileptic drugs; VLP, ventricular late potentials.

N2322 
Therapy (%)   
 AED monotherapy13 (56)6 (27)0.07
 AED polytherapy10 (44)16 (73)
No of AEDs/patient (median; 95% CI)1 (1.2–1.8)2 (1.7–2.5)0.02
 (median daily dose; 95% CI, mg)   
Drugs, n (%)   
 Carbamazepine8 (34)11 (50)ns
 (median daily dose; 95% CI, mg)9001,200ns
 Valproate16 (69)19 (86)ns
 (median daily dose; 95% CI, mg)1,3001,000ns
 Lamotrigine5 (22)6 (27)ns
 Levetiracetam0 (0)2 (9)ns
 Topiramate1 (4)2 (9)ns
 Gabapentin1 (4)1 (5)ns
 Vigabatrin0 (0)2 (9)ns

In particular, there was no difference in valproate or carbamazepine exposure, considering the proportion of patients exposed or the total daily dose administered (Table 3).

Regarding the ECG characteristics, there was no significant difference in P wave, QT intervals, QRS duration, or QTc intervals between the epileptic patients with or without VLP (Table 2). As expected, the median values for SA-QRS and LAS40 were significantly higher in VLP patients (p < 0.0001; p = 0.001, respectively) as compared to the VLP absent group. Conversely, median RMS40 value was significantly lower in the VLP subgroup than in epileptic patients without VLP (p = 0.001).

Twenty-two subjects from epilepsy group and one subject from the control group were examined by standard transthoracic echocardiography, and none was found to have detected structural cardiac abnormalities. All studied cardiac dimensions including left ventricular internal dimensions in diastole, left ventricular internal dimensions in systole, interventricular septal thickness at end diastole, left ventricular posterior wall thickness, right ventricular diastolic diameter, left atrial diameter, left ventricular ejection fraction, shortening fraction, peak velocity of mitral E wave, peak velocity of mitral A wave, the E/A ratio, isovolumic relaxation time, and deceleration time of mitral E wave were within the normal range.

Discussion

The novel finding of this exploratory study is to demonstrate that a large proportion of epilepsy patients display VLPs in the SAECG as assessed interictally. This seems important and interesting, given that a cohort of subjects studied was in the 18–45 year age range and did not report any subjective or objective complaints from cardiovascular system. The occurrence of VLPs in our epilepsy patients was significantly higher compared to healthy controls from our study, but also of importance is the general population data, in which VLPs were reported in 3–6% in otherwise asymptomatic subjects (Graham & Handelsman, 1998; Yakubo et al., 2000).

The occurrence of VLPs was significantly correlated with refractory epilepsy, longer disease duration, and higher number of AEDs used in therapy compared to epilepsy patients with normal SAECG parameters. Other factors such as seizure type or epilepsy syndrome did not seem to play the role in the emergence of VLPs. Another approach to assess the disturbances in cardiac conductance is to measure QT interval duration and interval dispersion (QTD). Recent retrospective study reported higher values for heart rate corrected QT interval duration and QTd in refractory epilepsy patients as compared to controls (Neufeld et al., 2009). In our patients there was no significant increase of the above parameters comparing to control group; this can be explained by the fact that our epilepsy patients were younger and represented different outcomes with regard to seizure control. In addition, we aimed to exclude the direct effect of seizures on ECG parameters (Surges et al., 2010c) by performing the analysis in the interictal period.

The exact mechanism by which VLPs appear in that population of patients remains unclear. The VLPs constitute low-amplitude electrical signals at the end of the QRS complex that are generated by diseased myocardium, activated slower and later than its usual timing in the cardiac cycle (Lander et al., 1993). Because of their very low amplitude, VLPs cannot be usually recognized on standard 12-lead ECG, and amplified high-resolution ECG recording is required for their identification. VLPs represent delayed conduction through a diseased myocardium and, therefore, indicate the presence of a potential anatomic substrate for reentrant ventricular arrhythmias (Santangeli et al., 2008). Most of the anatomic substrates for VLPs consist of myocardial areas within and surrounding myocardial infarct regions but a number of other pathologic conditions are associated with such anatomic substrates including dilated cardiomyopathy (Mancini et al., 1993) and arrhythmogenic right ventricular cardiomyopathy (Nava et al., 2000), hypertrophic cardiomyopathy (Cripps et al., 1990), and other. Arrhythmia triggers, such as acute ischemia, imbalance in autonomic tone, or the onset of clinical heart failure, may provide the link between presence of VLPs and occurrence of spontaneous ventricular tachyarrhythmias (Santangeli et al., 2008). None of our patients had diagnosed cardiac disease, and any organic heart abnormalities were within exclusion criteria; therefore, the above mechanisms based on anatomic substrate can be excluded in our patients. Another point to consider is that the slowing down and fragmentation of the electrical signal, resulting in VLPs, are not necessarily related only to an anatomic substrate, but can also be related to slow and discontinuous conduction caused by abnormalities in gap junction distribution and function, thus forming a functional, rather than anatomic, substrate for reentry (Antzelevitch, 2005; Santangeli et al., 2008). In that respect, the effects of drugs influencing the cellular conduction such as AEDs should be seriously considered. Our study has not been sufficiently powered to demonstrate that a particular drug influenced the cardiac conductance in the form of delayed potentials, although polytherapy seemed to be a risk factor in that matter. Alternatively, centrally mediated mechanisms with reduced activity in amygdale–prefrontal circuits might influence brainstem neurons and the sympathovagal balance in the cardiac efferent peripheral autonomic nervous system. Both neurologic pathology and alterations of psychic states can produce autonomic imbalance resulting, at their most serious intensity, in cerebrogenic sudden death. Indeed, different alterations in ECG recordings, such as QT-interval prolongation, premature ventricular beats, late ventricular potentials, nonsustained ventricular tachycardia (VT), and the R on T phenomenon, which may be predictive of sudden death, were reported in different neurologic and psychiatric disorders (Baranchuk et al., 2009). These include the full range of conditions characterized by clear organic brain lesions such as subarachnoid hemorrhage, intracerebral hemorrhage, and ischemic stroke, but also diseases without any organic changes found, as happens in psychiatric patients. Interestingly, recent study by Nahshoni et al. (2010) reported a high frequency of VLPs in schizophrenia patients chronically treated with neuroleptics. Autonomic imbalance in epilepsy patients, especially those presenting drug refractory disease, is a well-known phenomenon (Sevcencu & Struijk, 2010). It has been reported that the autonomic alterations in epilepsy may be minor in early disease stages and increase proportionally with the duration of the disease (Persson et al., 2007). In addition, the magnitude of heart rate variablilty changes seems similar in patients with drug-controlled and refractory TLE (Ansakorpi et al., 2002), whereas the interictal baroreflex alterations may depend on the degree of seizure control (El-Sayed et al., 2007). Although, we did not study the indices of autonomic balance in our patients, it is plausible to assume that the appearance of VLPs in that cohort of patients might result from some complex brain regulated, central mechanisms.

Further longitudinal studies are needed in order to stratify the risk of life-threatening ventricular events in epileptic patients with VLPs. This could be done by long-term ECG monitoring using Holter technique, but it has not performed in our patients yet. It is also of interest whether the presence of interictally detected VLP predisposes to other conductance cardiac abnormalities during the ictal activity. In particular, it is difficult to speculate at present if VLPs might be related to QT-interval prolongation or shortening described after different seizure types (Brotherstone et al., 2010; Surges et al., 2010b), but it would be an interesting issue for further investigation. On the basis of previous experience in patients after myocardial infarction, the VLPs have rather high negative predictive value. However, when present, VLPs help better stratify the arrhythmic risk of patients under several disease conditions (Santangeli et al., 2008). This would also apply to epilepsy patients with many different clinical factors involved, including seizure activity, drug therapy, and comorbidities.

Acknowledgment

We express our thanks for valuable help of Dr. Jacek Gawłowicz.

Disclosure

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Ancillary