Cardiac Disturbances During the Administration of Taxol
`
`By Eric K. Rowinsky, William P. McGuire, Thomas Guarnieri, Jason S. Fisherman, Michaele C. Christian,
`and Ross C. Donehower
`
`The clinical development of taxol, a new antimicrotu-
`bule agent with a unique mechanism of cytotoxic
`action, has proceeded slowly due to serious hypersen-
`sitivity reactions (HSRs) and shortages in its supply.
`Nevertheless, large-scale phase II trials have been
`initiated as taxol has recently demonstrated impres-
`sive activity in advanced and cisplatin-refractory ovar-
`ian carcinoma. Furthermore, the incidence of HSRs has
`been reduced substantially with premedications and
`modifications in the administration schedule. How-
`ever, various manifestations of potential cardiotoxicity
`have been observed in several patients who partici-
`pated in four phase I and II studies of taxol. Asymptom-
`atic bradycardia has occurred in a high proportion of
`patients, including 29% of ovarian cancer patients who
`were treated with maximally tolerated doses of taxol in
`
`AXOL, A NOVEL antimicrotubule agent
`that induces tubulin polymerization, has en-
`tered broad phase II and III trials due to its
`unique mechanism of antineoplastic action and
`the antitumor activity that has been observed in
`advanced and refractory ovarian carcinoma, mela-
`noma, and non—small-cell lung carcinoma in phase
`I and preliminary phase II studies.1 Initially, the
`pace of taxol’s clinical development was hampered
`by both a high incidence of serious hypersensitivity
`reactions (HSRS) and a limited drug supply. Re-
`cently, Weiss et 3]2 reviewed the clinical features
`of HSRs that appear to be due to taxol’s Cre-
`mophor EL (polyoxyethylated castor oil; BASF,
`Aktiengesellshaft, Germany) vehicle.2 The admin-
`
`From the Divisions of Pharmacology and Experimental
`Therapeutics, and Medical Oncology, The Johns Hopkins
`Oncology Center, and the Division of Cardiology, The Johns
`Hopkins Medical Institutions, Baltimore; and Cancer Therapy
`Evaluation Program, Division of Cancer Treatment, National
`Cancer Institute, Bethesda, MD.
`Submitted November 7, 1990; accepted March 19, I 991.
`E.K.R. is the recipient ofanAmerican Cancer Society Career
`DevelopmentAward.
`Supported by National Cancer Institute contract N01 -CM-
`57738.
`
`Address reprint requests to Eric K. Rowinsky, MD, The Johns
`Hopkins Oncology Center, 600 N Wolfe St, Pharmacology
`1-121, Baltimore, MD 21205.
`© I 991 by American Society of Clinical Oncology.
`0732—183X/9I/0909-0021$3. 00/0
`
`a phase II study. More profound cardiac disturbances,
`including a range of atrioventricular conduction blocks,
`left bundle branch block, ventricular tachycardia (VT),
`and manifestations of cardiac ischemia, have been
`observed in seven of 140 patients (5%) who received
`taxol. Descriptions of these events are presented in this
`report to alert investigators to the potential for these
`adverse effects. Although these disturbances did not
`result in serious sequelae in most patients, investiga-
`tors should continue to maintain a high degree of
`caution until precise risk factors, frequency, and clini-
`cal significance of these adverse cardiac effects are
`determined.
`
`.1 Clin Oncol 9: l 704- 1712. © 1991 by American Society
`of Clinical Oncology.
`
`istration of taxol over longer durations with corti-
`costeroids and H1— and Hz-antihistamine premedi-
`cations resulted in a significant reduction in the
`frequency and severity of HSRs and, therefore, a
`smoother transition from phase I and limited
`phase II trials, which were concentrated at a few
`institutions,
`to broad phase II studies at many
`participating centers including community hospi-
`tals. However, various cardiac disturbances, includ-
`ing brady- and tachyarrhythmias, atrioventricular
`and bundle branch blocks, and cardiac ischemia,
`have also occurred during phase I and II studies at
`The Johns Hopkins Oncology Center (JHOC).
`The characteristics of these cardiac disturbances
`
`are presented in this report. They are of signifi-
`cance as a broader use of this important new agent
`is anticipated in the future.
`
`MATERIALS AND METHODS
`
`Study Patients and Treatment Regimens
`
`The treatment courses of patients who participated in
`four clinical trials of taxol at JHOC from 1983 to 1990 were
`
`reviewed. Details that pertain to the nature of these trials,
`patient characteristics, and drug administration schedules
`were published previously3'6 and are described briefly:
`1. A phase I study of taxol in advanced solid neoplasms
`from 1983 to 1986 accrued 30 patients who received 67
`assessable courses.3 Taxol (15 to 265 mg/mz) was initially
`infused over 1 hour (14 patients), but the infusion duration
`was subsequently increased to 6 hours (16 patients) due to
`the high incidence of HSRs in this study and concurrent
`
`1 704
`
`Journal of Clinical Oncology, Vol 9, No 9 (September), 1991: pp 1704-1712
`
`Downloaded from ascopubs.org by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`Celltrion v. Genentech
`
`IPR2017-01122
`
`Genentech Exhibit 2026
`
`

`

`CARDIOTOXICITY OF TAXOL
`
`1705
`
`trials. A premedication regimen for the prophylaxis of
`HSRs was also initiated.
`2. A phase I study of taxol in refractory leukemias from
`1987 to 1988 accrued 17 patients who received 28 assessable
`courses.4 Taxol (250 to 390 mg/mz) was infused over 24
`hours after premedication for prophylaxis of HSRs.
`3. A phase II study of taxol in advanced and refractory
`ovarian epithelial malignancies from 1987 to 1988 accrued
`47 patients who received 281 assessable courses. Taxol was
`infused over 24 hours after premedications for the prophy-
`laxis of HSRs.S Although taxol was administered over a wide
`dose range (110 to 250 mg/mz) due to hematologic toxicity in
`these heavily pretreated patients, taxol doses ranging from
`110 to 170 mg/m2 were given in 92% of courses.
`4. A phase I study of taxol and cisplatin in advanced solid
`neoplasms from 1989 to 1990 accrued 44 untreated and
`minimally pretreated patients who received 182 courses.6
`The sequence of the 24-hour taxol (110 to 200 mg/mi) and
`brief cisplatin (1 mg/min) infusions was alternated with each
`successive course. Patients received premedication for the
`prophylaxis of HSRs, as well as hydration and an antiemetic
`regimen that
`included dexamethasone, metoclopromide,
`and diphenhydramine.
`The following medications were administered before
`taxol for the prophylaxis of HSRs: dexamethasone, 20 mg
`intravenously (IV) or orally, 14 and 7 hours before taxol;
`diphenhydramine, 50 mg IV, 30 minutes before taxol; and
`one of the Hit-histamine antagonists, cimetidine 300 mg IV,
`ranitidine 50 mg IV, or famotidine 20 mg IV.
`
`Cardiac Monitoring
`
`Continuous ECG telemetry was performed during all
`courses with the Marquette 7700 ECG Telemetry System
`(Marquette Electronics, Milwaukee, WI). The system is
`equipped with an arrhythmia-processing program in which a
`patient’s baseline cardiac rhythm is initially “learned.”
`Beats are classified according to timing and width. Beats are
`then labeled as normal, premature ventricular contractions
`(PVCs), ventricular escape beats, paced beats, or artifacts
`and stored as templates. The system is also equipped with
`memory capabilities and an arrhythmia classification pro—
`gram that analyzes both individual and consecutive beats.
`Alarm defaults were typically set at 50 beats per minute
`(bpm; low heart rate), 150 bpm (high heart rate), and 6
`PVCs/min for this study, but defaults were occasionally
`reset if patients had unique intrinsic cardiac rhythms. The
`calculations of Marriot7 were used to define the normal
`limits for the QT and corrected QT (OTC) intervals.
`
`RESULTS
`
`The principal cardiac rhythm disturbances that
`were noted during the administration of taxol
`included ventricular tachycardia (VT) associated
`with ventricular ectopy in patients who were treated
`with the combination of taxol and cisplatin and a
`range of atrioventricular conduction abnormali-
`ties associated with sinus bradycardia in patients
`who received taxol as both a single agent and in
`combination with cisplatin. The majority of taxol-
`
`related cardiac disturbances were not associated
`
`with clinical symptoms and were noted inciden-
`tally during continuous cardiac monitoring. The
`clinical characteristics of these cardiac distur-
`bances are described in detail. The course of a
`
`patient who had a massive myocardial infarction
`and died during the administration of taxol is also
`described.
`
`VT
`
`Patient no. 1. A 68-year-old female with un-
`treated advanced ovarian carcinoma had a previ-
`ous history of hypertension that was controlled
`with captopril 25 mg orally every day. Her pretreat-
`ment ECG was unremarkable (QT/OTC = 376/426
`ms
`[heart
`rate, 79 bpm], PR = 136 ms, and
`QRS = 84 ms), and cardiac monitoring showed
`only rare PVCs during her first and second courses
`of taxol (170 mg/mz) and cisplatin (75 mg/mz).
`During her third course, in which taxol preceded
`cisplatin, an increased frequency of multifocal
`PVCs and bigeminy was noted early into the taxol
`infusion. She subsequently developed a six-beat
`episode of asymptomatic VT at a rate of 136 bpm
`(Fig 1) 12 hours after beginning the infusion.
`Immediately after the episode, a 12—lead ECG
`showed only a slightly prolonged QT interval
`(QT/QTc = 420/455 ms [heart rate, 72 bpm],
`PR = 156 ms, and QRS = 92 ms). Serum electro-
`lytes, including potassium, magnesium, and cal-
`cium, were within normal limits. Taxol was tempo-
`rarily discontinued, and further ectopy was not
`noted. However, an increased frequency of multi-
`focal PVCs recurred when taxol was reinstituted.
`
`Forty-eight hours of outpatient ambulatory car-
`diac monitoring, which was performed several
`days later, revealed only rare PVCs. The patient
`
`
`
`Fig I. ECG during taxol infusion (patient no. 1) showing
`brief episode of VT.
`
`Downloaded from ascopubs.org by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`

`

`l 706
`
`ROWINSKY ET AL
`
`was treated 3 weeks later with a fourth course of
`
`taxol and cisplatin that was associated with an
`increased frequency of multifocal PVCs compared
`with baseline.
`
`Patient n0. 2. A 59—year-old female with un-
`treated advanced ovarian carcinoma and no previ-
`ous cardiac history developed a four-beat episode
`of wide-complex VT (145 bpm, QRS = 150 ms) 23
`hours into the infusion of taxol during her second
`course, in which taxol (170 mg/mz) preceded cis-
`platin (75 mg/mz). Her pretreatment ECG was
`unremarkable (QT/QTC = 328/420 ms [heart rate,
`100 bpm], PR = 116 ms, and QRS = 68 ms).
`Serum electrolytes,
`including magnesium, cal-
`cium, and potassium, were normal. Cardiac moni-
`toring and relevant ECG parameters before the
`episode were also unremarkable (QT/QTc = 320/
`390 ms [heart rate, 88 bpm], PR = 120 ms, and
`QRS = 80 ms). A 12-lead ECG was unchanged
`after the event. Outpatient ambulatory cardiac
`monitoring for 72 hours showed rare and brief
`episodes of supraventricular tachycardia, prema-
`ture atrial contractions, and PVCs. VT did not
`
`occur during the patient’s first course or during six
`subsequent courses of taxol and cisplatin.
`Patient no. 3. A 48-year-old female with un-
`treated metastatic non—small-cell lung carcinoma
`and a past medical history of hypertension that
`was controlled with hydrochlorthiazide was noted
`to have rare PVCs while receiving taxol during her
`first course of taxol (170 mg/mz) and cisplatin (75
`mg/mz). Before her second course, her ECG was
`remarkable only for a mildly prolonged PR inter-
`val (QT/QTC = 380/418 ms [heart rate, 74 bpm],
`PR = 216 ms, and QRS = 88 ms). Thirteen hours
`into the taxol infusion, which followed cisplatin,
`she had a three-beat episode of VT (150 bpm).
`Taxol was continued; however, at 17 hours into
`taxol, a 72—beat episode of asymptomatic VT (167
`bpm) occurred (Fig 2). Taxol was immediately
`discontinued, and her rhythm subsequently re-
`
`
`
`Fig 2. ECG during taxol infusion (patient no. 3) showing
`segment of 72-beal episode of VT.
`
`verted to a normal sinus rhythm. Serum electro—
`lytes, including potassium, calcium, and magne-
`sium, were normal. Cardiac monitoring before the
`event and a posttreatment 12-1ead ECG showed
`essentially unchanged ECG parameters (QT/
`QTc = 360/428 ms [heart rate, 85 bpm], PR = 200
`ms, and QRS = 80 ms). Taxol was not restarted,
`and lidocaine (2 mg/min) was administered for 9
`hours. However, frequent multifocal PVCs and
`bigeminy occurred after the discontinuation of
`lidocaine. These disturbances resolved over the
`
`next 48 hours. At that time, an inpatient 24-hour
`Holter recording (Delmar Avionics, Irvine, CA)
`showed only rare PVCs. In addition, an outpatient
`ambulatory 48-hour Holter recording that was
`performed 10 days later showed rare and brief
`episodes of supraventricular tachycardia and rare
`PVCs. Three brief episodes of VT were also
`noted. These episodes consisted of 12 total beats,
`with the longest containing six beats. The patient’s
`tumor did not respond to treatment, and therapy
`was subsequently discontinued.
`
`VTAfter Left Bundle Branch Block
`
`Patient n0. 4. A 58-year-old female with un—
`treated metastatic non—small-cell lung cancer and
`no previous cardiac problems had a three-beat run
`of asymptomatic VT (160 bpm) 22 hours into the
`infusion of taxol during her second course of taxol,
`which preceded cisplatin. Immediately before the
`episode, her relevant ECG intervals were un-
`remarkable (QT/QTc = 420/428 ms [heart rate, 62
`bpm], QRS = 80 ms, PR = 180 ms) and it
`appeared that the first beat in the VT complex fell
`on a preceding T wave. Taxol was continued, and
`further arrhythmias were not noted. Serum electro-
`lytes, including potassium, magnesium, and cal-
`cium, an echocardiocardiogram, and ambulatory
`Holter recordings were unremarkable. The pa-
`tient received three additional courses of taxol and
`
`cisplatin without further cardiac disturbances un-
`til her sixth course, in which taxol preceded cis-
`platin. Her pretreatment ECG was remarkable
`only for a mildly prolonged QT interval (QT/
`QTc = 388/465 ms [heart rate, 88 bpm], PR = 172
`ms, and QRS = 80 ms; Fig 3A). She developed an
`intermittent left bundle branch block 12 hours into
`
`the infusion of taxol, which progressed to a contin—
`uous left bundle branch block (QT/QTc = 450/531
`ms [heart rate, 83 bpm], PR = 170 ms, and QRS =
`120 ms; Fig 3B). Taxol was discontinued, and the
`
`Downloaded from ascopubs.org by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`

`

`CARDIOTOXICITY OF TAXOL
`
`l 707
`
`showed nonspecific ST and T wave changes and a
`slightly prolonged QT interval (QT/QTc = 348/
`464 ms [heart rate, 109 bpm], PR = 128 ms, and
`QRS = 96 ms). Cardiac monitoring was unevent—
`ful during his initial 18 hours of taxol, with the
`exception of a mildly prolonged QT interval (QT/
`QTc = 360/450 ms [heart rate, 94 bpm], PR = 120
`ms, and QRS = 80 ms at hour 14). Fourteen hours
`into the taxol infusion and 2 hours after beginning
`hydration (300 cc total), which produced a good
`urine output in preparation for cisplatin, he com-
`plained of left arm pain that was initially attrib-
`uted to irritation from his IV catheter. However,
`
`he developed profound chest discomfort at hour
`18. A 12-lead ECG showed complete heart block
`and ST segment elevations in the inferior leads
`which were felt to be indicative of an ischemic
`
`cardiac injury. Taxol was discontinued. Soon after,
`he became hypotensive, unresponsive, and devel-
`oped bradycardia (heart rate, 40 to 50 bpm). An
`ECG demonstrated complete heart block and
`cardiopulmonary resusitation was begun, but the
`patient died 19 hours after taxol was begun.
`Postmortem examination revealed a large acute
`myocardial infarction (7 X 3 X 0.3 cm) that ex-
`tended from the atrium to the apex in the when-
`docardial region of the left ventricle. In addition,
`there was a 60% to 70% occlusion of the mid-right
`coronary artery by an atherosclerotic plaque and a
`superimposed thrombosis, a 10% to 20% occlu-
`sion of the proximal left main coronary artery, and
`30% to 40% occlusions of the left anterior descend-
`
`ing, proximal, and mid-left circumflex coronary
`arteries.
`
`Sinus Bradycardia
`
`Sinus bradycardia with heart rates ranging from
`30 to 50 bpm has been observed in a relatively high
`proportion of patients during infusions of taxol in
`both single-agent and taxol-cisplatin combination
`studies.5*6 For example, sinus bradycardia was
`noted during at least one course of taxol adminis-
`tered to 13 of 45 patients (29%) in a phase II trial
`in advanced ovarian cancer.5 Bradycardia was
`never symptomatic or associated with hemody-
`namic compromise. However, bradycardia did
`progress to higher grades of atrioventricular con-
`duction delay in two patients (patients no. 6 and
`7), including Mobitz type I block, atrioventricular
`block with 2:1 conduction, and complete heart
`block.
`
`A F
`
`(A)
`infusion (patient no. 4).
`ig 3. ECGs during taxol
`Pretreatment; (B) left bundle branch block; and (C) segment of
`12-beat episode of VT.
`
`left bundle branch resolved completely within 2
`hours. However, the patient developed an asymp-
`tomatic 12—beat episode of VT (150 bpm; Fig 3C) 6
`hours after taxol was discontinued. A slightly pro—
`longed QT interval (QT/QTc = 400/437 ms [heart
`rate, 75 bpm], PR = 160 ms, and QRS = 80 ms)
`was the only significant ECG abnormality before
`the episode. Over the next 12 hours, the patient
`had several events consistent with vasovagal epi-
`sodes that were characterized by nausea, bradycar-
`dia (heart rate, 40 to 50 bpm), and presyncopal
`symptoms. Serum electrolytes,
`including potas-
`sium, calcium, and magnesium were normal; car-
`diac enzymes (creatine phosphokinase and lactic
`dehydrogenase),
`isoenzymes, and follow-up 12-
`lead ECGs were unchanged. In addition, ambula-
`tory Holter recordings and an echocardiogram
`were unremarkable. The patient’s baseline 12-
`lead ECG remained unchanged over the next
`several months. She was subsequently treated with
`two additional courses of taxol and cisplatin and
`did not experience further cardiac conduction ab-
`normalities or increased ventricular ectopy as de—
`tected by both telemetric and Holter monitoring.
`
`Myocardial Infarction
`
`Patient no. 5. A 50-year-old male with a previ-
`ously irradiated metastatic non—small-cell
`lung
`cancer and a pertinent history of hypertension and
`hypercholesterolemia began his first course of
`taxol followed by cisplatin. A pretreatment ECG
`
`Downloaded from ascopubs.0rg by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`

`

`l 708
`
`ROWINSKY ET AL
`
`Bradycardia Associated With Atrioventricular
`Conduction Delay
`
`Patient no. 6. A 58-year—old female with ad-
`vanced ovarian carcinoma and no previous cardiac
`history was treated with taxol 135 mg/m2 after she
`did not respond to several standard chemotherapy
`regimens consisting of cisplatin, cyclophospha—
`mide, hexamethamelamine, melphalan, fluorou-
`racil, and Adriamycin (doxorubicin; Adria Labora—
`tories, Columbus, OH)
`(cumulative dose, 480
`mg/mz). Her pretaxol heart rate was 76 bpm, and
`her pretreatment 12—lead ECG was only remark-
`able for nonspecific T wave abnormalities (QT/
`QTC = 400/437 ms [heart rate, 71 bpm], PR = 150
`ms, QRS = 80 ms) (Fig 4A). Reversible asymptom-
`atic sinus bradycardia (heart rate, 40 to 50 bpm)
`was observed during her first course of taxol.
`Three weeks later, sinus bradycardia recurred
`several hours into her second taxol infusion. Sev-
`
`eral types of atrioventricular conduction abnormal-
`ities were noted hours into the infusion, including
`episodes of PR interval prolongation (200 ms),
`Mobitz type I atrioventricular block characterized
`by frequent episodes of progressive PR interval
`prolongation culminating in dropped beats, and
`atrioventricular block with 2:1 conduction (Fig 4B
`and C). These conduction disturbances were fol-
`lowed by intermittent episodes of complete heart
`block (Fig 4D). After several
`3— to 7.5—second
`episodes of asystole, taxol was discontinued, atro-
`pine was administered, and a temporary right
`ventricular pacemaker was inserted. However, a
`permanent pacemaker was later inserted when
`
`
`
`further monitoring for 36 hours showed additional
`episodes of complete heart block. Serum potas-
`sium, magnesium, and calcium levels were normal,
`and an echocardiogram showed normal left ventric-
`ular function with a small, hemodynamically insig-
`nificant pericardial effusion. Soon after, her rhythm
`reverted to normal sinus with normal atrioventric-
`ular conduction. Six additional courses of taxol
`
`were subsequently administered. Each taxol infu—
`sion was associated with pacemaker capture as the
`patient developed progressive bradycardia, and
`her heart rate fell below the demand rate of the
`
`pacemaker. Monitoring for 12-hour periods be-
`fore and after taxol revealed an unpaced rhythm
`and verified that the infusion was responsible for
`the bradycardia and subsequent demand pacing.
`To determine if premedication with ranitidine
`may have been responsible for the bradycardia
`and atrioventricular conduction delays in this
`patient, cimetidine was substituted for ranitidine
`during one course?10 Monitoring was unremark-
`able during pretreatment with cimetidine alone
`for 24 hours; however, similar cardiac effects
`occurred when taxol was administered following
`cimetidine.
`
`Patient no. 7. A 64-year—old female with ad-
`vanced ovarian cancer that progressed during
`treatment with cyclophosphamide and cisplatin
`was treated with taxol 135 mg/mz. Her pertinent
`cardiac history was remarkable for the develop-
`ment of sinus bradycardia and hypotension that
`required cardiopulmonary resuscitation during tu—
`mor debulking surgery 1 year before taxol treat-
`ment. Her pretreatment ECG was remarkable for
`poor R—wave progression in the anterior precor-
`dial leads (QT/QTc = 340/425 ms [heart rate, 94
`bpm], PR = 156 ms, QRS = 74 ms; Fig 5A). Five
`to 12 hours into each of eight courses of taxol that
`induced a partial response, she developed asymp-
`tomatic bradycardia (heart rate, 35 to 50 bpm) and
`progressive PR interval prolongation (20 to 24
`ms). This was followed by frequent episodes of
`atrioventricular block with 2:1 conduction (Fig
`SB). These conduction abnormalities resolved 3 to
`4 hours after taxol was discontinued. Outpatient
`ambulatory cardiac monitoring was unremarkable.
`
`Other Possible Cardiac Effects
`
`(A)
`infusion (patient no. 6).
`Fig 4. ECG: during taxol
`(C)
`Pretreatment;
`(B) Mobitz type I atrioventricular block;
`atrioventricular block with 2:1 conduction; and (D) complete
`heart block with segment of 7.5-second episode of asystole.
`
`Bigeminy, trigeminy, increased PVCs, and chest
`pain have also been noted during taxol infusions.2‘3
`One patient experienced chest pain associated
`
`Downloaded from ascopubs.0rg by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`

`

`CARDIOTOXICITY OF TAXOL
`
`A
`
`(A)
`infusion (patient no. 7).
`Fig 5. ECGs during taxol
`Pretreatment; (B) atrioventricular block with 2:1 conduction.
`
`with nausea and believed to be suggestive of
`angina pectoris, which was relieved by nitroglyc—
`erin. Another patient developed chest pain concur-
`rent with new onset atrial fibrillation during treat-
`ment with taxol alone.2 Atypical chest and
`abdominal pains have also been observed in associ-
`ation with symptoms indicative of taxol-related
`HSRs such as hypotension, bronchospasm, and
`urticaria}3 In addition, one patient who was treated
`with taxol alone had a synoopal episode immedi-
`ately after treatment, and two patients had one
`episode each of syncope and near-syncope within
`4 days of treatment with cisplatin and taxol. These
`events were also associated with orthostatic symp-
`toms, treatment-induced malaise, and decreased
`oral
`intake. Prolonged cardiac monitoring and
`12—lead ECGs did not reveal any abnormalities,
`and taxol was readministered uneventfully in all
`three patients.
`
`DISCUSSION
`
`A diverse spectrum of cardiac disturbances
`including ventricular arrhythmias, bradycardia, sev-
`eral degrees of atrioventricular conduction block,
`bundle branch block, and cardiac ischemia have
`been observed during the administration of taxol.
`With the exception of patient no. 5 who had a
`myocardial infarction and significant undiagnosed
`coronary artery disease, a causal relationship be-
`tween taxol and the cardiac disturbances that are
`
`described in this report is clearly supported on the
`following grounds. First, this cohort of patients
`had an excellent performance status and no rele-
`vant cardiac risk factors. In addition, there were
`no other obvious concurrent etiologic factors to
`
`1709
`
`account for these abnormalities. Second, although
`it
`is possible that
`these cardiac disturbances,
`especially increased ventricular ectopy and VT,
`occur relatively frequently in similar cohorts of
`hospitalized patients matched for age and severity
`of medical illness, these abnormalities were not
`evident during telemetric and/or Holter monitor-
`ing performed either before or after taxol. Finally,
`the temporal relationship of several of these distur-
`bances,
`including the ventricular arrhythmias,
`bradycardias, and atrioventricular conduction
`blocks, was generally consistent in that they usu—
`ally occurred several hours after taxol was begun,
`resolved shortly after taxol was discontinued, and
`often recurred when the agent was readminis-
`tered.
`
`Interestingly, all five episodes of VT and the
`myocardial infarction were observed in patients
`who were treated with the combination of taxol
`
`and cisplatin. VT has not been reported to date in
`patients receiving taxol alone. While the number
`of events is small enough that this may be coinci-
`dental, it raises the question of a synergistic effect
`of the two agents on the myocardium. Although
`the true incidence of asymptomatic VT associated
`with cisplatin is not known, clinically significant
`VT has not been associated with the widespread
`use of cisplatin. Since four of the five episodes of
`VT and the myocardial infarction occurred during
`the administration of taxol that preceded cisplatin,
`it is likely that taxol is the primary agent responsi—
`ble for the observed cardiac toxicity. However,
`potential interactions between cisplatin and taxol
`cannot be entirely dismissed. Although there was
`no evidence of concurrent electrolyte disturbances
`in these patients, one possible mechanism whereby
`cisplatin might predispose to or augment
`the
`cardiac effects of other agents is by affecting the
`renal tubules during previous courses and deplet-
`ing total body stores of largely intracellular cations
`such as magnesium, potassium, and calcium. The
`extent of the total body depletion of these cations
`may not be fully reflected in plasma electrolyte
`determinations.
`
`It is not clear whether taxol or its Cremophor
`EL formulation vehicle is responsible for these
`cardiac disturbances. In the case of HSRs, the
`Cremophor EL vehicle is known to induce hista-
`mine release, and identical reactions have been
`reported in patients treated with other similarly
`formulated drugs such as cyclosporin, didemnin B,
`
`Dovmloaded from ascopubs.org by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`

`

`'l7'lO
`
`ROWINSKY ET AL
`
`teniposide, and intravenous vitamin K311“ Al-
`though cardiovascular manifestations of excessive
`histamine release, such as profound hypotension,
`have occurred in animals and patients receiving
`other drugs formulated in Cremophor EL, primary
`ventricular arrhythmias, bradycardias, bundle
`branch blocks, and atrioventricular blocks have
`not been reported. One cannot absolutely exclude
`Cremophor EL as a contributor to the cardiotoxic-
`ity as cardiac monitoring has not been routinely
`performed during the administration of other
`similarly formulated agents. However, it is possi—
`ble that the vehicle may induce clinically inappar-
`ent cardiac disturbances in patients receiving these
`other agents that would be detected only by
`continuous ECG monitoring. Of note,
`taxol
`is
`formulated with the highest concentration of Cre-
`mophor EL per dose of all agents in clinical use.14
`If Cremophor BL is the agent responsible for
`the cardiac toxicity, the mechanism is likely due to
`its induction of histamine release. Recently, the
`roles of histamine and selective histamine recep-
`tors in cardiac tissue have been elucidated.”21
`
`Selective stimulation of cardiac H1 and H2 recep—
`tors have been demonstrated to result in increased
`
`myocardial oxygen demand in cardiac tissue, coro-
`nary vasoconstriction (H1), and chronotropic ef—
`fects (H2).15'21 There is also evidence from animal
`studies indicating that stimulation of H1 receptors
`results in prolonged atrioventricular conduction,
`possible depression of conduction in Purkinje
`tissue, myocardial cell injury, and ventricular ar-
`rhythmias.15’19'20 In addition, the absence of evi-
`dence for the systemic release of histamines does
`not rule out a pathophysiologic role for histamine
`in these cardiac effects. Histamine is released by
`the heart
`itself in response to both antigen-
`antibody reactions and various drugs.”16 Thus,
`selective activation of histamine receptors in car-
`diac tissue may explain the broad spectrum of
`cardiac manifestations that has been observed
`
`during taxol administration, including bradycar-
`dia, atriventricular conduction prolongation, bun-
`dle branch block, ventricular irritability, and car—
`diac ischemia.
`
`It is also possible that the H2 antagonists are
`partially responsible for some of the cardiac distur-
`bances described in this report. Cardiac rate and
`rhythm abnormalities, especially bradyarrhyth-
`mias, have been observed with both oral and IV
`
`administration of the H2 antagonists that are
`commonly used for the prophylaxis of taxol—
`induced HSRs.“’10 In some cases, various degrees
`of atrioventricular conduction block have also
`
`been associated with bradycardia.8 Several of the
`most severe episodes have occurred in patients
`with factors predisposing them to have a de-
`creased clearance for this class of agents, such as
`advanced age, severe systemic illness, and renal
`dysfunction.9’10 However, none of the patients
`described in this report had any of these potential
`risk factors. The most profound cardiac distur-
`bances have also occurred immediately after the
`administration of Hz-antagonists,
`in contrast to
`the disturbances described in this report.“"10 In-
`stead, the cardiac disturbances associated with the
`
`administration of taxol have been temporally re-
`lated to the infusion of taxol in its vehicle. These
`disturbances occurred in the later hours of the
`
`infusion, were not temporally associated with the
`administration of Hz—antagonists, and were usually
`self-limited or abated immediately after discontin-
`uation of the infusion.
`
`Arrhythmias and ECG changes have been ob-
`served with other antineoplastic agents, most nota-
`bly the anthracyclines and amsacrine. Although
`the incidence of anthracycline-induced acute car—
`diotoxicity as manifested by ECG changes re-
`portedly ranges from 0% to 41% with only rare
`clinically significant events, amsacrine has been re-
`ported to induce serious, life-threatening arrhyth-
`mias in approximately 1% of patients.”28 Neither
`tachyarrythmias nor bradyarrhythmias have been
`associated with the most commonly used antimicro-
`tubule agents, the vinca alkaloids and colchicine.
`However, acute cardiac ischemia, including mas-
`sive cardiac infarctions, have been associated with
`
`vincrisine, vinblastine, and vindesine.”35 Although
`the mechanisms for these adverse effects are not
`
`known, a direct antimicrotubule effect is unlikely
`as microtubules are not known to play significant
`functional roles in either the cardiac conduction
`
`system or myocardial function. Instead, the sarco-
`plasmic and sarcotubular systems play principal
`roles in both electrical impulse conduction and
`electromagnetic coupling, while myofibrils, myo-
`sin, and actin are primarily responsible for muscu—
`lar contraction.”40 However, it is conceivable that
`antimicrotubule agents including taxol may medi-
`ate cardiac muscle damage by adversely affecting
`
`Downloaded from ascopubs.0rg by Univ of Texas Health Sci Ctr At Tyler on October 27, 2016 from 132.174.254.099
`Copyright © 2016 American Society of Clinical Oncology. All rights reserved.
`
`

`

`CARDIOTOXICITY OF TAXOL
`
`1711
`
`other subcellular organelles. Both biochemical
`and ultrastructural studies des