This cohort study describes the association of variation in QTc formula selection with adverse event grading and chemotherapy delivery in patients with cancer.
Key Points
Question
What association does corrected QT interval (QTc) formula selection have on the clinical management and adverse event grading for QTc prolongation for patients with cancer?
Findings
In this cohort study that included 19 955 electrocardiograms from 6881 patients, use of the Bazett formula was associated with a 3-fold increase in grade 3 QTc prolongation compared with other common formulae and likely was associated with inappropriate changes in clinical management.
Meaning
Oncologists should be aware of the discrepancy between different QTc formulae and the possible influence on clinical care.
Abstract
Importance
Monitoring of the corrected QT interval (QTc) for patients with cancer receiving chemotherapy is not standardized. Selection of QTc formula may be associated with adverse event grading and chemotherapy delivery.
Objective
To describe the association of QTc formula selection with adverse event grading and chemotherapy delivery.
Design, Setting, and Participants
This retrospective observational cohort study used data from January 2010 to April 2020 and included adult patients seen at the University of North Carolina Cancer Hospital who had an electrocardiogram (ECG) performed.
Exposures
Adjusted QTc using the Bazett, Fridericia, and Framingham formulae.
Main Outcomes and Measures
The main outcome was QTc prolongation using the Common Terminology Criteria for Adverse Events (CTCAE). Consistency between formulae was evaluated. Subsequently, appropriateness of clinical management due to prolonged QTc was assessed for a subset of patients being treated with chemotherapy agents associated with a prolonged QT interval. We hypothesized that use of the Bazett formula would be associated with higher rates of QTc prolongation and inappropriate modifications to chemotherapy.
Results
A total of 19 955 ECGs from 6881 adult patients (3055 [44.4%] women, 3826 [55.6%] men; median [IQR] age at first ECG, 60 [47-68] years) were analyzed. The percentage of ECGs with grade 3 QTc prolongation differed by formula (all patients: Framingham, 1.8%; Fridericia, 2.8%; and Bazett, 9.0%; patients receiving QT-prolonging chemotherapy [2340 ECGs]: Framingham, 2.7%; Fridericia, 4.5%; and Bazett, 12.5%). The Bazett formula resulted in a median QTc value 26.4 milliseconds higher than Fridericia and 27.8 milliseconds higher than Framingham. Of the 1786 ECGs classified as grade 3 by Bazett, 1446 (81.0%) were grade 2 or less by either Fridericia or Framingham. A total of 5 of 28 (17.9%) evaluated clinical changes associated with prolonged QTc were deemed inappropriate when using either Fridericia or Framingham formula.
Conclusions and Relevance
Findings of this cohort study suggest that the Bazett formula resulted in higher QTc values associated with a 3-fold increase in grade 3 CTCAE toxic effects compared with other common formulae. Use of the Bazett formula likely was associated with inappropriate changes in clinical management. These data support the use of a standard QTc formula (such as Fridericia or Framingham) for QTc correction in oncology.
Introduction
Prolongation of the QT interval is a common adverse event associated with many chemotherapeutic agents and can lead to life-threatening arrhythmias such as torsades de pointes (TdP). Thus, monitoring of the QT interval is critical to mitigate the risk of cardiac complications of cancer treatment. Duration of the QT interval naturally varies inversely with heart rate; hence, a corrected QT (QTc) that is adjusted for heart rate is most predictive of proarrhythmic potential.1,2,3 The QTc is widely used in clinical practice. However, key aspects of QTc monitoring lack standardization in oncology, leading to variation in practice patterns that may ultimately compromise care.
Many different formulae are available to calculate QTc. For example, the Bazett,4 Fridericia,5 and Framingham6 formulae are frequently used in clinical practice without a uniform standard. The Bazett formula predominates because many automated electrocardiogram (ECG) reporting software packages default to this formula, often without clinicians’ awareness. Unfortunately, the Bazett formula has been shown to overestimate the QTc interval at elevated heart rates.7,8,9,10 This overestimation has led many organizations to discourage the use of the Bazett formula in clinical practice.11,12 The detrimental effect of using the Bazett formula to inform treatment decisions in oncology has not been fully explored.
Along with others, our group has argued that proper selection and standardized application of a QTc formula is critical to optimize patient outcomes.11,12 The US Food and Drug Administration (FDA) prescribing information provides guidance for dose modification based on QTc for only a few chemotherapy agents and commonly does not specify which QTc formula should be used clinically. The Common Terminology Criteria for Adverse Events (CTCAE), which represents the standardized classification of adverse effects of chemotherapeutic agents published by the National Cancer Institute, also does not specify which QTc formula should be used in grading toxic effects.13 Because neither the prescribing information nor the CTCAE provide clear guidance on how to appropriately calculate QTc, clinical decisions based on prolonged QTc values may vary widely and detrimentally affect patient outcomes.
The goal of this study was to describe how variation in QTc formula selection may be associated with chemotherapy delivery in oncology. We hypothesized that there are wide variations in QTc values based on QTc formula for patients receiving chemotherapy and that clinical management decisions may inappropriately vary based on QTc formula selection. We aimed to describe the prevalence of QTc prolongation among patients with cancer and how CTCAE adverse events vary using different QTc formulae. We also sought to describe how frequently QTc prolongation influenced clinical management at our institution and whether the use of a specific QTc formula may have changed clinical decision-making.
Methods
Data Sources
We established a cohort of patients who were at least 18 years old seen at the University of North Carolina Cancer Hospital and had an ECG performed from January 2010 to April 2020. Patient age and sex and ECG parameters, including the QT, ventricular rate, atrial rate, and RR interval, were collected. These data were captured by MUSE (GE Healthcare) and are maintained for clinical use. We combined these data with medication utilization data from a central data repository to establish a subcohort of patients who were receiving chemotherapy known to prolong the QTc. The central data repository contains clinical and research data sourced from the University of North Carolina Health Care System. Data were supplemented and expanded through manual review of the electronic medical record. This study was approved by the biomedical institutional review board at the University of North Carolina. A waiver of patient informed consent was obtained owing to the minimal risk nature of the study.
ECG Measurements
For each ECG, we calculated the QTc value using the Bazett, Fridericia, and Framingham formulae (Bazett: QT interval divided by the square root of the RR interval; Fridericia: QT interval divided by the cube root of the RR interval; Framingham: QT interval + 0.154 × [1000 − RR interval]). We chose these 3 formulae because the Bazett formula is commonly reported on ECG machines, and the Fridericia and Framingham formulae have been shown to have the best rate correction in large population-based studies.10,14 The FDA recommends use of the Fridericia formula for industry in classifying QTc prolongation.15 We classified each QTc value using CTCAE guidelines, version 5.0 (grade 0: QTc <450 milliseconds; grade 1: QTc 450-480 milliseconds; grade 2: QTc 481-500 milliseconds; grade 3 QTc ≥501 milliseconds; grade 4: TdP; grade 5: death).13
Association With Clinical Management
We then evaluated treatment patterns of patients who were prescribed chemotherapy known to prolong the QTc interval. QTc-prolonging chemotherapeutic agents were identified with the FDA package insert. Patient demographics, disease state characteristics, laboratory value abnormalities, and concomitant medications were collected by individual medical record review of the electronic medical record. We captured details about the clinical management of each episode of prolongation of the QTc interval. We classified clinical changes as chemotherapeutic dose withheld, dose reduction, electrolytes supplemented, discontinuation of concomitant medication, or other. We classified each clinical change as appropriate if the QTc was prolonged by using the formula and threshold provided in the FDA package insert (if available) or if all 3 formulae resulted in grade 3 toxic effects by CTCAE grading. For patients who had multiple ECGs within 7 days, we used the highest values of QT and QTc across the 7-day period. We also captured whether the use of a specific QTc formula was documented in the medical record. Because many QTc-prolonging chemotherapeutic agents are used to treat patients with hematologic cancers, we report detailed results for those agents received by patients with hematologic cancers with a combined total of 25 or more ECGs.
Statistical Analysis
We used descriptive statistics to summarize key variables, ie, frequencies for categorical variables and mean, median, and SD for numerical variables. To examine the agreement between continuous QTc values generated by different QTc formulae, we used Bland-Altman plots and reported mean bias and 95% confidence limits of paired differences among the 3 QTc formulae. We performed post hoc analyses evaluating whether transforming data within the Bland-Altman plots (using log, log-log, or square root) or using linear regression improved reporting of confidence intervals around the mean difference.16 We evaluated the concordance of CTCAE grades among the QTc formulae using the McNemar test and κ statistic with different cutoffs for classifying patients into CTCAE grade 3 or above at 495 milliseconds and 505 milliseconds. Although there may have been multiple QTc measures from a participant, we treated clustered measures as independent when we examined concordance between different QTc formulae. This was because the intraclass correlation among the clustered paired differences (eg, QTc Bazett − QTc Framingham) was found to be very small (<1 × 10−8). This weak dependence was thus ignored to simplify the concordance analysis and interpretation. However, when we compared QTc values between the entire cohort and the subcohort receiving QTc-prolonging chemotherapy, we used a mixed effect model to account for clustering of multiple QTc values within participants. All analyses were performed using R Statistical Software, version 4.1.1 (R Foundation).
Results
Baseline Characteristics
A total of 19 955 ECGs from 6881 adult patients (3055 women [44.4%], 3826 men [55.6%]) were identified from January 2010 through April 2020. Median (IQR) age at first ECG was 60 (47-68) years (range, 18-100 years). The median (IQR) number of ECGs per patient was 2 (1-3) (range, 1-129). Median heart rate was 86 beats/min (eFigure 1 in the Supplement). Women had slightly higher heart rates compared to males (mean difference = 0.84 beats/min; 95% CI, 0.21-1.48 beats/min; P = .009). The median unadjusted QT interval was 380 milliseconds (eFigure 2 in the Supplement). The unadjusted QT interval did not significantly vary by sex.
Differences in QTc Values by Formulae
Applying CTCAE toxicity grading and using the Framingham formula, 16 371 ECGs (82.0%) were classified as normal, 2671 (13.4%) were classified as grade 1, 555 (2.8%) were classified were as grade 2, and 358 (1.8%) were classified as grade 3 (Figure 1A). Calculated QTc values using the Fridericia formula resulted in 15 684 ECGs (78.6%) classified as normal, 3038 (15.2%) classified as grade 1, 701 (3.5%) classified as grade 2, and 532 (2.7%) classified as grade 3 (Figure 1B). Calculated QTc values using Bazett formula resulted in 9958 ECGs (49.9%) classified as normal, 6263 (31.4%) classified as grade 1, 1948 (9.8%) classified as grade 2, and 1786 (9.0%) classified as grade 3 (Figure 1C).
Figure 1. Distribution of Calculated QTc Values Based on Formula for All ECGs Including Percentage of ECGs With CTCAE Grade 3 Toxicity.
Distribution of QTc values is shown for the following formulae: Framingham (A), Fridericia (B), Bazett (C), and all 3 combined (D). Grade 3 toxicity was determined by CTCAE guidelines, version 5.0. CTCAE indicates Common Terminology Criteria for Adverse Events; ECG, electrocardiogram; ms, milliseconds; QTc, corrected QT interval.
A total of 340 ECGs (1.8%) were classified as grade 3 toxicity using all 3 formulae (Figure 1D). Of 1786 total ECGs classified as CTCAE grade 3 by Bazett, 71.3% (1273 ECGs, 6.4% of all ECGs) were classified as grade 2 or less by both Fridericia and Framingham and 81.0% (1446 ECGs) as grade 2 or less by at least 1 formula. Of 358 total ECGs classified as CTCAE grade 3 by Framingham, no ECGs were classified as grade 2 or less by both Fridericia and Bazett. Of 532 total ECGs classified as CTCAE grade 3 by Fridericia, no ECGs were classified as grade 2 or less by both Bazett and Framingham.
The CTCAE grading for QTc prolongation was concordant between all 3 formulae for 56.6% (11 300) of all ECGs (Table 1). The CTCAE grading was concordant between Bazett and Fridericia for 58.3% (11 629) of ECGs (κ = 0.42); 1273 (6.4%) ECGs were classified as grade 3 by Bazett but not by Fridericia (Figure 2A). The CTCAE grading was concordant between Bazett and Framingham for 56.8% (11 329) of ECGs (κ = 0.30); 1449 (7.2%) ECGs were classified as grade 3 by Bazett but not by Framingham (Figure 2B). The CTCAE grading was concordant between Framingham and Fridericia for 94.2% (18 797) of ECGs (κ = 0.80) (Figure 2C). Bazett resulted in higher CTCAE grading than both Framingham and Fridericia in 40.9% (8153) of ECGs. Sensitivity analyses using the McNemar test demonstrated that the 3 formulae gave significantly (P < .001) inconsistent classifications regardless of the cutoff used for CTCAE grade 3 toxicity (495 milliseconds, 500 milliseconds, or 505 milliseconds).
Table 1. Concordance Among CTCAE Grading Using Bazett, Fridericia, and Framingham QTc Equations.
| Cohort | No. | CTCAE grading | Concordance | No./total No. (%) | |||
|---|---|---|---|---|---|---|---|
| Bazett, Fridericia, and Framingham | Bazett and Fridericia | Bazett and Framingham | Fridericia and Framingham | ||||
| All patients with cancer | 19 955 ECGs | All CTCAE grades | Concordant | 11 300/19 955 (56.6) | 11 629/19 955 (58.3) | 11 329/19 955 (56.8) | 18 797/19 955 (94.2) |
| Discordant | 8655/19 955 (43.4) | 8326/19 955 (41.7) | 8626/19 955 (43.2) | 1158/19 955 (5.8) | |||
| 6881 Patients | CTCAE grade 3 toxicity | Concordant | 340/19 955 (1.7) | 513/19 955 (2.6) | 340/19 955 (1.7) | 358/19 955 (1.8) | |
| Discordant | 1273/19 955 (6.4) | 1292/19 955 (6.5) | 1464/19 955 (7.3) | 174/19 955 (0.9) | |||
| Patients receiving QTc-prolonging chemotherapy | 2340 ECGs | All CTCAE grades | Concordant | 1267/2340 (54.1) | 1324/2340 (56.6) | 1271/2340 (54.3) | 2176/2340 (93.0) |
| Discordant | 1074/2340 (45.9) | 1016/2340 (43.4) | 1069/2340 (45.7) | 165/2340 (7.0) | |||
| 421 Patients | CTCAE grade 3 toxicity | Concordant | 62/2340 (2.6) | 102/2340 (4.4) | 62/2340 (2.7) | 64/2340 (2.7) | |
| Discordant | 191/2340 (8.2) | 194/2340 (8.3) | 233/2340 (10.0) | 41/2340 (1.8) | |||
Abbreviations: CTCAE, Common Terminology Criteria for Adverse Events; ECG, electrocardiogram; QTc, corrected QT interval.
Figure 2. Concordance Between the Fridericia, Bazett, and Framingham QTc Formulae.
Scatter plots of QTc values calculated with Fridericia and Bazett (A), Framingham and Bazett (B), and Fridericia and Framingham (C). Each data point represents a single ECG. Data points in the white quadrants are concordant based on CTCAE grading (dichotomized: less than grade 3 vs grade 3); shaded quadrants represent discordant CTCAE grading. Percentages represent the total percentage of discordant ECGs calculated by: number of ECGs greater than 500 by one formula but not by the other divided by the total number of ECGs (n = 19 955). CTCAE indicates Common Terminology Criteria for Adverse Events; ECG, electrocardiogram; ms, milliseconds; QTc, corrected QT interval.
Absolute QTc values differed between Bazett and Fridericia by a median of 26.4 milliseconds (Bland-Altman, mean [SD] bias = 26.3 [18.2] milliseconds; 95% CI, −9.3 to 62.0; Figure 3A), Bazett and Framingham by a median of 27.8 milliseconds (mean [SD] bias = 29.3 [21.6] milliseconds; 95% CI, −13.1 to 71.7; Figure 3B), and Fridericia and Framingham by a median of 1.5 milliseconds (mean [SD] bias = 3.0 [5.5] milliseconds; 95% CI, −7.8 to 13.8; Figure 3C) (eFigure 3 in the Supplement). Constraining values within the 1% and 99% quantiles reduced, but did not eliminate, the apparent distortion of the Bland-Altman plots (eFigure 4 in the Supplement). Data transformation using log, square root, or log-log and adjustment using linear regression of Bland-Altman plots did not substantially change the reported confidence intervals around the mean difference (eTable 1 in the Supplement).16 Subgroup analyses by sex of agreement between QTc values were similar to overall findings despite women having slightly higher heart rates than men. Excluding ECGs suspected to have atrial dysrhythmias (n = 936) or ECGs with prolonged QRS width (>120 milliseconds, eg, suspected to have bundle-branch blocks, intraventricular conduction delays) (n = 1330) also did not substantially change the overall findings.
Figure 3. Bland-Altman Plots of the Difference vs Average of Pair-Wise QTc Values.
The solid black line represents the mean difference in QTc values in milliseconds between formulae. The dotted blue lines represent 95% CIs of the mean difference. Agreement is shown between Fridericia and Bazett (A), Framingham and Bazett (B), and Fridericia and Framingham (C). ECG indicates electrocardiogram; ms, milliseconds; QTc, corrected QT interval.
Patients Receiving QT-Prolonging Chemotherapy
A total of 2340 ECGs were collected from 421 patients receiving chemotherapy known to prolong the QT interval. Electrocardiograms were collected from patients receiving 24 different chemotherapeutic agents (eTable 2 in the Supplement). Patients in this cohort had significantly higher QTc values compared with the entire cohort using the Framingham and Fridericia formulae (mean difference in milliseconds: Framingham, 3.43; P = .008; Fridericia, 3.24; P = .02; Bazett, 1.55; P = .26). A total of 62 of these ECGs (2.6%) were classified as grade 3 CTCAE toxicity using all 3 formulae (Table 1). Using the Framingham, Fridericia, or Bazett formulae, 64 (2.7%), 105 (4.5%), and 293 (12.5%) ECGs were classified as grade 3 QTc prolongation, respectively. Of the 293 ECGs classified as grade 3 by Bazett, 65.2% (191 ECGs, 8.2% of all ECGs) were classified as grade 2 or less by both Fridericia and Framingham. Of 64 total ECGs classified as CTCAE grade 3 by Framingham, none were classified as grade 2 or less by both Fridericia and Bazett. Of 105 total ECGs classified as CTCAE grade 3 by Fridericia, none were classified as grade 2 or less by both Bazett and Framingham. Differences in grade 3 CTCAE concordance by specific chemotherapeutic agent are shown in eTable 2 in the Supplement.
Clinical Changes Based on Calculated QTc
Individual encounters were evaluated for changes in clinical management for 142 patients who underwent 496 ECGs while receiving 9 QTc-prolonging chemotherapeutic agents (eTable 3 in the Supplement). No episodes of TdP were identified. Twenty-eight encounters resulted in clinical changes due to QT prolongation (Table 2). Changes included withholding, reducing, or discontinuing chemotherapy and/or concomitant QT-prolonging medications and supplementing electrolytes. A total of 56 encounters (11%) were complicated by hypokalemia (<3.5 mmol/L) and 21 (4%) by hypomagnesemia (<1.6 mg/dL). A total of 5 (17.9%) clinical management changes were deemed inappropriate due to selection of QTc formula (4 changes not in accordance with FDA labeling; 1 change with a QTc value prolonged with Bazett but not with Fridericia or Framingham). In total, 1.0% of all encounters (5 of 496) for patients receiving QTc-prolonging chemotherapy had a documented clinical change that was inappropriate based on QTc formula selection, and 6% of encounters (30 of 496) had documentation in the medical record of the QTc formula used in clinical decision-making.
Table 2. Clinical Changes for Patients Receiving QTc-Prolonging Chemotherapy Based on Calculated QTc (n = 28 Total Changes of 496 Encounters Evaluated).
| Clinical change/appropriateness | No. (%) |
|---|---|
| Clinical changes made based on prolonged QTc | |
| Total No. of changes made | 28 (100) |
| Electrolytes supplemented | 9 (32) |
| Discontinued concomitant medication | 8 (29) |
| Withheld chemotherapy | 7 (25) |
| Reduced dose of chemotherapy | 2 (7) |
| Discontinued chemotherapy agent | 1 (4) |
| Changed dosing schedule | 1 (4) |
| Appropriateness of clinical changes | |
| Appropriate | 23 (82) |
Abbreviation: QTc, corrected QT interval.
Discussion
Treating patients with chemotherapy that can prolong the QT interval presents a common challenge to oncologists: maintain dose intensity while minimizing the risk of rare but potentially fatal cardiac events. Although potentially fatal, the absolute risk of such chemotherapy-induced cardiac events such as TdP is very low, estimated at 0.001% of patients (1 out of 10 000).17,18,19 We previously advocated for monitoring the QTc interval using a standard, reliable QTc formula other than Bazett.12 The Bazett formula, which is the default formula used in many ECG software platforms, is known to overcorrect the QTc at high heart rates, and its use in patients with cancer is discouraged by the FDA.14,15 Building on our prior work, this study is, to our knowledge, the first systematic analysis of QTc monitoring in oncology to evaluate the potential outcome of QTc formula selection on chemotherapy delivery.
Analyzing data from nearly 20 000 ECGs at our cancer center, we demonstrate that the incidence of grade 3 or higher CTCAE toxicity ranges from 2.7% to 12.5% among patients receiving QT-prolonging chemotherapy depending on QTc formula selection. These incidence rates are consistent with other studies estimating QTc prolongation in oncology.20,21,22,23,24 Importantly, our data demonstrate that the use of the Bazett formula resulted in a near 3-fold increase in incidence of grade 3 or greater CTCAE toxicity for patients receiving chemotherapy compared with the Fridericia or Framingham formula. Eighty-one percent of patients classified as having grade 3 toxicity by Bazett had grade 2 or less toxicity using Fridericia or Framingham.
In absolute terms, use of the Bazett formula was associated with clinically significant differences in QTc values compared with Fridericia or Framingham: mean difference, 26.3 milliseconds and 29.3 milliseconds, respectively. These differences are substantially higher than mean differences reported in nononcology populations. For example, Vandenberk et al,14 in a similarly sized analysis of 6609 patients seen in a general hospital in Belgium, reported mean differences between the Bazett formula and Fridericia and Framingham of 8.78 milliseconds and 8.75 milliseconds, respectively. While recognizing there is no criterion standard by which to test whether the Bazett formula is “correctly” adjusting the QT for heart rate, our findings demonstrate that there is a distinctly high risk of QTc overestimation with the use of the Bazett formula compared with other formulae among patients with cancer. Furthermore, overestimation of the QTc interval is particularly harmful in oncology because chemotherapy is often modified in response to prolongation of the QTc.
We hypothesized that differences in toxicity grading based on formula selection may result in variability in practice patterns and in inappropriate withholding of chemotherapy. Here, we demonstrate that this is likely occurring in routine oncology practice. Among patients receiving QTc-prolonging chemotherapy, 5.6% of all encounters resulted in a documented change in clinical management due to prolonged QTc, including dose reductions and withholding of chemotherapy. Of these encounters, 17.9% were likely unnecessary based on inappropriate application of QTc formula. In total, 1.0% of all encounters for patients receiving QTc-prolonging chemotherapy had a documented clinical change due to prolonged QTc that was inappropriate based on QTc formula selection. Although these numbers are smaller than we anticipated, these data suggest that inappropriate formula selection influences clinical decisions. We suspect that this study underestimated the actual number of clinical changes made due to prolonged QTc because it relied on clinical documentation to identify changes. The QTc formula used to inform clinical decision-making was documented only 6% of the time, suggesting that many clinicians remain unaware that formula selection influences QTc.
Limitations
This study has several limitations. Because this was a retrospective review, it was not possible to draw causal conclusions regarding formula selection and decisions about QTc management. Causality may only be inferred from the data provided. Data from this study are also from a single institution over a period of 10 years. Practice patterns toward mitigating risk of TdP may vary at different institutions and may have changed over time. In fact, following the publication of our initial commentary in 2019,12 our oncology group began routine reporting of Fridericia results alongside Bazett in the electronic record to mitigate overinterpretation of QTc toxicity. We also defined our larger population of patients with cancer based on being treated at our cancer hospital. We did not individually verify individual cancer diagnoses for each patient. This may have resulted in patients being included in our initial estimates who were only suspected of having cancer or having a related blood disorder. We verified all diagnoses for all patients receiving QTc-prolonging chemotherapy.
Conclusions
Findings of this cohort study suggest that use of the Bazett formula was associated with substantially higher QTc values and a near 3-fold increase in grade 3 CTCAE toxicity for patients with cancer compared with other common formulae. Use of the Bazett formula likely was associated with inappropriate changes in the clinical management of patients receiving QTc-prolonging chemotherapy. These results are especially troubling given the possible lack of awareness of these issues among practicing oncologists. A critical first step to addressing this problem is to standardize QTc monitoring practices across oncology.
eTable 1. Regression models of difference versus average between formula within 1% and 99% quantiles
eTable 2. Concordance among CTCAE grade 3 toxicity using Bazett, Fredericia, and Framingham corrected QT (QTc) equations by chemotherapeutic agent
eTable 3. Demographic and clinical characteristics of patients evaluated for clinical changes due to prolonged QTc (N = 496 ECGs, 142 patients)
eFigure 1. Frequency distribution of ventricular heart rate
eFigure 2. Frequency distribution of unadjusted QT interval
eFigure 3. Distribution density of the difference between calculated QTc values using the Fridericia, Bazett, or Framingham formula
eFigure 4. Bland-Altman plots of the difference versus average of pair-wise QTc values constrained within 1% and 99% quantiles
References
- 1.Schouten EG, Dekker JM, Meppelink P, Kok FJ, Vandenbroucke JP, Pool J. QT interval prolongation predicts cardiovascular mortality in an apparently healthy population. Circulation. 1991;84(4):1516-1523. doi: 10.1161/01.CIR.84.4.1516 [DOI] [PubMed] [Google Scholar]
- 2.Noseworthy PA, Peloso GM, Hwang SJ, et al. QT interval and long-term mortality risk in the Framingham Heart Study. Ann Noninvasive Electrocardiol. 2012;17(4):340-348. doi: 10.1111/j.1542-474X.2012.00535.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Nielsen JB, Graff C, Rasmussen PV, et al. Risk prediction of cardiovascular death based on the QTc interval: evaluating age and gender differences in a large primary care population. Eur Heart J. 2014;35(20):1335-1344. doi: 10.1093/eurheartj/ehu081 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bazett HC. An analysis of the time-relations of the electrocardiograms. Heart. 1920;7:353-370. [Google Scholar]
- 5.Fridericia LS. Die systolendauer im elektrokardiogramm bei normalem menschen und bei herzkranken. Acta Med Scand. 2009;53:469-486. doi: 10.1111/j.0954-6820.1920.tb18266.x [DOI] [Google Scholar]
- 6.Sagie A, Larson MG, Goldberg RJ, Bengtson JR, Levy D. An improved method for adjusting the QT interval for heart rate (the Framingham Heart Study). Am J Cardiol. 1992;70(7):797-801. doi: 10.1016/0002-9149(92)90562-D [DOI] [PubMed] [Google Scholar]
- 7.Hodges MS, Salerno D, Erlinen D. Bazett’s QT correction reviewed: evidence that a linear QT correction for heart rate is better. J Am Coll Cardiol. 1983;1:694. [Google Scholar]
- 8.Malik M. Problems of heart rate correction in assessment of drug-induced QT interval prolongation. J Cardiovasc Electrophysiol. 2001;12(4):411-420. doi: 10.1046/j.1540-8167.2001.00411.x [DOI] [PubMed] [Google Scholar]
- 9.Al-Tweigeri T, Dent S, Al Sayed A, et al. Using the appropriate formula for QT measurement can save lives. Hematol Oncol Stem Cell Ther. Published online June 21, 2021. doi: 10.1016/j.hemonc.2021.06.001 [DOI] [PubMed] [Google Scholar]
- 10.Andršová I, Hnatkova K, Šišáková M, et al. Influence of heart rate correction formulas on QTc interval stability. Sci Rep. 2021;11(1):14269. doi: 10.1038/s41598-021-93774-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Rautaharju PM, Surawicz B, Gettes LS, et al. ; American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; American College of Cardiology Foundation; Heart Rhythm Society; Endorsed by the International Society for Computerized Electrocardiology . AHA/ACCF/HRS recommendations for the standardization and interpretation of the electrocardiogram: part IV: the ST segment, T and U waves, and the QT interval: a scientific statement from the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology; the American College of Cardiology Foundation; and the Heart Rhythm Society. J Am Coll Cardiol. 2009;53(11):982-991. doi: 10.1016/j.jacc.2008.12.014 [DOI] [PubMed] [Google Scholar]
- 12.Muluneh B, Richardson DR, Hicks C, Jensen BC, Zeidner JF. Trials and tribulations of corrected QT interval monitoring in oncology: rationale for a practice-changing standardized approach. J Clin Oncol. 2019;37(30):2719-2721. doi: 10.1200/JCO.19.00922 [DOI] [PubMed] [Google Scholar]
- 13.National Cancer Institute Cancer Therapy Evaluation Program . Common Terminology Criteria for Adverse Events (CTCAE) v5.0. Accessed February 21, 2022. https://ctep.cancer.gov/protocoldevelopment/electronic_applications/ctc.htm#ctc_50
- 14.Vandenberk B, Vandael E, Robyns T, et al. Which QT correction formulae to use for QT monitoring? J Am Heart Assoc. 2016;5(6):e003264. doi: 10.1161/JAHA.116.003264 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.US Food and Drug Administration. Guidance for Industry . E14 Clinical evaluation of QT/QTc interval prolongation and proarrhythmic potential for non-antiarrhythmic drugs. Published online 2005. Accessed March 23, 2022. https://www.fda.gov/files/drugs/published/E14-Clinical-Evaluation-of-QT-QTc-Interval-Prolongation-and-Proarrhythmic-Potential-for-Non-Antiarrhythmic-Drugs.pdf
- 16.Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res. 1999;8(2):135-160. doi: 10.1177/096228029900800204 [DOI] [PubMed] [Google Scholar]
- 17.Moss AJ, Schwartz PJ, Crampton RS, et al. The long QT syndrome: prospective longitudinal study of 328 families. Circulation. 1991;84(3):1136-1144. doi: 10.1161/01.CIR.84.3.1136 [DOI] [PubMed] [Google Scholar]
- 18.Zareba W, Moss AJ, Schwartz PJ, et al. ; International Long-QT Syndrome Registry Research Group . Influence of the genotype on the clinical course of the long-QT syndrome. N Engl J Med. 1998;339(14):960-965. doi: 10.1056/NEJM199810013391404 [DOI] [PubMed] [Google Scholar]
- 19.Essa H, Wright DJ, Dobson R, Lip GYH. Chemotherapy-induced arrhythmia—underrecognized and undertreated. Am J Med. 2021;134(10):1224-1231.e1. doi: 10.1016/j.amjmed.2021.05.026 [DOI] [PubMed] [Google Scholar]
- 20.Brell JM. Prolonged QTc interval in cancer therapeutic drug development: defining arrhythmic risk in malignancy. Prog Cardiovasc Dis. 2010;53(2):164-172. doi: 10.1016/j.pcad.2010.05.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Roden DM. Drug-induced prolongation of the QT interval. N Engl J Med. 2004;350(10):1013-1022. doi: 10.1056/NEJMra032426 [DOI] [PubMed] [Google Scholar]
- 22.Kloth JSL, Pagani A, Verboom MC, et al. Incidence and relevance of QTc-interval prolongation caused by tyrosine kinase inhibitors. Br J Cancer. 2015;112(6):1011-1016. doi: 10.1038/bjc.2015.82 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Porta-Sánchez A, Gilbert C, Spears D, et al. Incidence, diagnosis, and management of QT prolongation induced by cancer therapies: a systematic review. J Am Heart Assoc. 2017;6(12):e007724. doi: 10.1161/JAHA.117.007724 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Roden DM. A current understanding of drug-induced QT prolongation and its implications for anticancer therapy. Cardiovasc Res. 2019;115(5):895-903. doi: 10.1093/cvr/cvz013 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Regression models of difference versus average between formula within 1% and 99% quantiles
eTable 2. Concordance among CTCAE grade 3 toxicity using Bazett, Fredericia, and Framingham corrected QT (QTc) equations by chemotherapeutic agent
eTable 3. Demographic and clinical characteristics of patients evaluated for clinical changes due to prolonged QTc (N = 496 ECGs, 142 patients)
eFigure 1. Frequency distribution of ventricular heart rate
eFigure 2. Frequency distribution of unadjusted QT interval
eFigure 3. Distribution density of the difference between calculated QTc values using the Fridericia, Bazett, or Framingham formula
eFigure 4. Bland-Altman plots of the difference versus average of pair-wise QTc values constrained within 1% and 99% quantiles