How a 100-Year-Old Formula Could Stop Lifesaving Cancer Therapy

Alexander M. Castellino, PhD

August 30, 2019

While leading his team on a hospital round, Joshua F. Zeidner, MD, saw a note recommending that arsenic trioxide (Trisenox, Teva) should be temporarily discontinued in a 44-year-old woman with acute promyelocytic leukemia (APL).

The recommendation was made according to the product's prescribing information, after a routine ECG reported a QT interval of 400 ms, a heart rate of 95 beats per minute, and a corrected QT (QTc) interval of 503 ms.

But should the drug really be stopped because of this finding?

The question is pertinent since this treatment is a lifesaver, Zeidner noted.

In addition, he was aware that the QT prolongation he was seeing was calculated from a 100-year-old formula known as the Bazett correction, the accuracy of which has been called into question.

Although the lengthened QT interval may increase the risk of cardiac arrhythmia, that potential (and theoretical) risk needed to be balanced against the very real risk of an  aggressive leukemia progressing if the drug was withheld.  

"This representative patient is not an isolated phenomenon," Zeidner, from the University of North Carolina's Lineberger Comprehensive Cancer Center, told Medscape Medical News.

"It happens all the time," Zeidner noted, "and frequently in the outpatient setting, where some of these medications are pills that can potentially lead to QT prolongation."

Many of the oral tyrosine kinase inhibitors (TKIs) approved for use in cancer also have recommendations for monitoring the QT interval in their prescribing information, and frequently contain a strong recommendation on the QTc thresholds that require reduction of the dose or stopping the drug altogether.

However, there is no guidance in the prescribing information on which formula to use when calculating the QTc interval. But this is crucially important, Zeidner said, as the use of different corrections provide different values.  

Four different methods for measuring QTc are available for monitoring patients who may be at risk for arrhythmias: Bazett, Fridericia, Framingham, and Hodges.

All the formulae correct for heart rates, but "Bazett's formula overcorrects for high heart rate and provides a higher QTc value compared with other corrections," Zeidner said.

"That is why using Bazett's formula, it is easy to see how lifesaving treatments may be withheld for some patients," he said. But he questioned whether they need to be.

With several colleagues from UNC, Zeidner recently highlighted the ambiguities that exist in using QTc, and suggested ways in which this could be rectified for optimal patient management in a Comments and Controversies article published online July 19 in the Journal of Clinical Oncology.

The authors present a compelling argument against how we measure QTc today, said Richard M. Steingart, MD, a cardiologist at Memorial Sloan Kettering Cancer Center in New York City, who was approached for comment. "It makes sense that physicians should reexamine the method used for calculating mean QTc," he told Medscape Medical News.  

Joerg Herrmann, MD, from the cardio-oncology clinic at Mayo Clinic, Rochester, Minnesota, agreed. He argues that this is true even for most cardiologists. "This is a topic of high interest to oncologists and very pertinent for TKI therapies among others. The nuances need to be highlighted," he told Medscape Medical News.

How QTc Is Measured Makes a Difference

In a quest for determining which formula provided the best correlations between QTc and clinical parameters and treatment, Zeidner and colleagues uncovered a report that analyzed over 3000 electrocardiograms from 113 patients who were treated in a clinical trial (J Clin Oncol. 2014;32:3723-3728).

"The report concluded that using a specific QTc threshold has little value for reducing and holding the dose of arsenic trioxide in these patients," Zeidner said.

In another study where QTc was assessed in over 6000 patients using five different calculations, Bazett provided the worst correction due to over- and underestimation of QTc during fast and low heart rate, while the Fridericia and Framingham methods provided the best corrections (J Am Heart Assoc. 2016;5:5)

Knowing how QTc is measured is important, Zeidner said.

Ambiguity Lies in Not Knowing Which Formula to Use

Zeidner and his colleagues looked at the protocols of the drugs that were approved and that required monitoring patients for QTc, and noted that the prescribing information of the drugs did not provide any guidance on which formula to use.

Arsenic trioxide is a case in point. Trisenox was approved after a landmark study showed its benefit in combination with all-trans retinoic acid for patients with newly diagnosed, low-risk APL. The protocol of this study used the Framingham formula.

The product prescribing information recommends QT interval monitoring, and also recommends that the drug should be stopped on finding QTc >450 ms in males and >460 ms in females. However, it does not mention that QTc measurement specified in the protocol used the Framingham formula.

In the case of the APL leukemia patient whom Zeidner encountered during rounds, the QTc that raised the alarm had been calculated using Bazett formula, which is likely to be an overestimate, and hence Zeidner decided not to follow the recommendation to temporarily stop the drug.

The patient continued on arsenic trioxide without any cardiac complications.     

This is an issue for several other oncology drugs, the authors note. Examples include gilteritinib (Xospata, Astellas), approved for acute myeloid leukemia (AML); ivosidenib (Tibsovo, Agios), also approved for AML; and encorafenib (Braftovi, Array Biopharma/Pfizer), which is approved for use in melanoma in combination with binimetinib (Mektovi, Array Biopharma/Pfizer). The approved prescribing information stipulates that the drug should be stopped at certain QTc measurements, but does not specify how the QT interval was measured.

Only in the case of ribociclib (Kisqali, Novartis), approved for breast cancer, does the product prescribing information stipulate that QTc should be measured by the Fridericia formula.

Medscape Medical News reached out to the manufacturers of these drugs. All spokespersons indicated that the Fridericia correction (QTcF) was used in the pivotal clinical trials associated with the drugs' approvals and their prescribing information specified it as QTcF.

However, there was no guidance on which formula to use in clinical practice.

"Clinical decisions are based on physician discretion," Astellas' Marjorie Moeling said in an email response.

Jessica Smith from Pfizer's Global Media Relations, said in an email: "The FDA approved label is for use by physicians," and pointed out that because of the debate associated with which QTc correction to use, the label and FDA leave it to the physician to interpret the ECG.

That is precisely the ambiguity we do not need in clinical practice, Zeidner commented.

The QTcF in the prescribing information could well stand for Fridericia or Framingham. Why not specify what it was, he asked. And why saddle the physician with the decision as to which formula to use, when this can ultimately impact patient care and change management?

All the clinicians to whom Medscape Medical News spoke for this article agreed that most ECG machines in clinics and hospitals use the Bazett correction as a default.

That means the QTc measurement that is usually recorded is not in alignment with what was used in the clinical trial protocol for each drug, Zeidner pointed out.

"This discrepancy leads to a significant difference in how patients are managed and how lifesaving therapy can be withheld for the wrong reason, leading to suboptimal care for patients," he said.

After reading the JCO article, Memorial Sloan Kettering's Steingart said he also personally went through the prescribing information of the drugs mentioned. He was surprised to see that there was detailed guidance on QTc values that would lead to dose reductions or stopping the drugs, but that there was no guidance on which formula to use for QTc.

The choice of the formula does make a difference, he emphasized.

At MSK, physicians use Bazett in clinical practice and the Fridericia for research purposes. But he told Medscape Medical News that MSK physicians are aware of the pitfalls in using the Bazett correction and will manually calculate QTc when necessary.

However, not all clinicians are aware that Bazett may not be the optimal formula to use for QTc, or that their ECG machines are using Bazett as the default.

Brian Jensen, MD, a coauthor on the article and a cardiologist colleague of Zeidner's at UNC, said that he was familiar with QT corrections. He noted that although most ECG machines use Bazett as the default, some major software makers are incorporating algorithms for using Framingham or Hodges. He also indicated that UNC is making these changes for general in-patient machines.

The JCO commentary should serve as a call to action, Jensen said. He added that even some cardiologists may not be aware of the issues in using Bazett's correction.

"Cardiology fellows are often taught only about Bazett's formula and how to calculate it," Herrmann said, but studies have shown that Fridericia and Framingham do better.

"It makes no sense to have multiple ways to calculate QTc, especially when there is an inconsistency in their usage," Zeidner added. He calls for transparency, and urges pharmaceutical companies to make it clear which formula was used in the study protocol and regulatory authorities to mandate the information in the product information.

"Opting for options other than Bazett's is important," said Mayo Clinic's Herrmann. Calculators are available online, but an automated way would be very valuable, he added.

The American Society for Clinical Oncology (ASCO) has not issued guidelines on the best method for QTc calculation.

In 2004, the US Food and Drug Administration issued a guidance document for measuring QTc, in which it discouraged physicians from using Bazett. Again in 2015, the agency reasserted its recommendation and stipulated that the Fridericia formula is likely appropriate in most instances, but also encouraged using others, Zeidner told Medscape Medical News.

The Heart Rhythm Society, the American Heart Association, and the American College of Cardiology Guidelines provide guidance on using Framingham or Hodges in preference to either Fridericia or Bazett, Zeidner added.

Zeidner's hope is that the century-old Bazett formula can be put to rest once and for all. He is also hopeful that the JCO commentary raises awareness and that regulatory agencies and national organizations can come together and foster a unified approach.

Herrmann noted that oncologists often make their own decisions and at times cardio-oncology gets asked for input, but not often.

"We serve patients best by working in collaboration," commented Jensen.

Zeidner reports receiving honoraria from, and is a consultant for, Agios, Celgene, Daiichi Sankyo, Tolero Pharmaceuticals, Pfizer, AsystBio Laboratories, and Covance, and has received expenses from Takeda, Celgene, Agios, Daiichi Sankyo, and Dava Oncology. Steingart is a consultant for Pfizer. The other study coauthors have disclosed no relevant financial relationships.

J Clin Oncol. Published online July 16, 2019. Full text

For more from Medscape Oncology, join us on Twitter and Facebook


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.