2Pharmacy, and 3Nursing, Fox Chase Cancer Center, Philadelphia, PA, USA. Summary. Purpose: To estimate the incidence of hypersensitivity reac- tions using ...
Annals of Oncology 8: 611-614, 1997. © 1997 Khmer Academic Publishers. Printed in the Netherlands.
Short report Short-course intravenous prophylaxis for paclitaxel-related hypersensitivity reactions M. A. Bookman,1 D. D. Kloth, 2 P. E. Kover,3 S. Smolinski3 & R. F. Ozols' Departments of Medical Oncology. 2Pharmacy, and 3Nursing, Fox Chase Cancer Center, Philadelphia, PA, USA
Summary
sensitivity reactions were documented in 13 patients (4.6%) during the first or second cycle with a 95% confidence interval Purpose: To estimate the incidence of hypersensitivity reac- (CI) of 2.2% to 7.0%. Reactions resolved rapidly without tions using a short-course intravenous prophylactic regimen in sequelae and did not require hospitalization. Only two reacpatients receiving outpatient therapy with paclitaxel. tions (0.7%) were graded as serious with a 95% CI of 0.2% to Patients and methods: Patients were identified from a retro- 1.2%, based on the use of bronchodilators and presence of spective search of a computerized pharmacy database covering angioedema. Therapy was continued with modification in 10 a two-year period from January 1994 through December 1995. patients without recurrent hypersensitivity reaction. Therapy Eligible outpatients received paclitaxel as a one- to three-hour was discontinued in two patients without rechallenge and infusion 30 minutes after intravenous dexamethasone (10 or 20 discontinued in one patient after rechallenge with a recurrent mg), diphenhydramine (50 mg), and cimetidine (300 mg) or hypersensitivity reaction. ranitidine (50 mg). Charts from all patients were then manuConclusion: A short-course single-dose regimen of intraally reviewed to verify drug administration and to record any venous dexamethasone, diphenhydramine, and cimetidine (or evidence of hypersensitivity reactions during the first two ranitidine) offers a safe and convenient alternative for prevencycles of therapy. tion of hypersensitivity reactions associated with outpatient Results: A total of 283 outpatients were identified from the paclitaxel administration. pharmacy database and all charts reviewed. All patients received intravenous dexamethasone (5 to 20 mg) 30 minutes Key words: Cremophor, dexamethasone, hypersensitivity, prior to paclitaxel without prior oral dexamethasone. Hyper- paclitaxel
Introduction Paclitaxel, originally isolated from the bark of the western yew, Taxus brevifolia, has rapidly emerged as an important drug in the treatment of ovarian, breast, lung, head and neck, bladder, and other epithelial cancers [1]. Poor solubility hampered early drug development, and paclitaxel is formulated with a polyoxethylated castor oil vehicle (Cremophor EL). Early clinical trials encountered a high frequency of serious hypersensitivity reactions which resulted in the empiric adoption of a prolonged infusion (24 hours) and addition of prophylactic measures including corticosteroids and histamine Hi/H 2 antagonists [2]. In general, these reactions occur rapidly, and are associated with symptoms attributable to direct release of histamine or other vasoactive compounds, similar to reactions following administration of radiologic contrast agents. The majority of reactions appear secondary to the polyoxethylated castor oil vehicle, as similar reactions have been described with other drugs using related formulations. In addition, the vehicle itself has been found to induce histamine release in dogs [3]. With the use of oral premedication, a randomized
clinical trial documented a 1.3% incidence of severe hypersensitivity reactions and a 42% incidence of minor reactions among patients receiving either a three-hour or 24-hour infusion of paclitaxel [4]. Due to convenience, reduced myelosuppression, and equivalent efficacy (as second-line therapy), the three-hour outpatient regimen has become more commonly utilized than the 24hour infusion. Standard premedication has included 20 mg oral dexamethasone administered 12 and 6 hours prior to treatment with paclitaxel. However, this regimen requires increased patient vigilance in order to avoid treatment delay in the event of missed dosing and also exposes the patient to multiple high doses of corticosteroid. Anecdotal experience, initially from patients who omitted doses of oral dexamethasone, suggested that the use of intravenous dexamethasone 30 minutes prior to paclitaxel might also be effective. With early safety experience, this became the preferred regimen at our institution, unless the oral regimen was required by clinical protocol design or physician preference. Sufficient numbers of outpatients were treated in a consistent manner over two years to provide the basis for a retrospective review to estimate the frequency and severity of reactions.
612 Patients and methods A centralized pharmacy database (NDC Datastat) is maintained at Fox Chase Cancer Center for recording of drugs that are dispensed to the outpatient infusion area and other patient treatment areas. This database was queried to generate a report spanning from 1 January 1994 through 31 December 1995 for all patients that received intravenous paclitaxel preceded by dexamethasone, diphenhydramine, and cimetidine (or ranitidine), without other selection criteria. Sequential records were selected blindly, without regard to clinical outcomes or toxicity. Medical charts were then manually reviewed for all patients to verify drug delivery and the occurrence of HSRs during the first and second cycles. Third and subsequent cycles were not reviewed, as the majority of HSRs have been reported during the first two cycles in other studies. The use of additional chemotherapy drugs in combination with paclitaxel was not analyzed, as these drugs would generally be administered after completion of the paclitaxel infusion, and would not be expected to influence the rapid occurrence of hypersensitivity reactions during the paclitaxel infusion. All hypersensitivity reactions were reviewed in detail, based on materials from the written medical record. Reactions were scored as 'serious' if there was angioedema, hypotension requiring pressors, respiratory distress requiring bronchodilators, or generalized urticaria, which were the same criteria prospectively employed in the EuropeanCanadian randomized trial [4]. Although records were carefully reviewed, we anticipated that our retrospective analysis would underestimate the true incidence of minor reactions (i.e., flushing) that were not associated with interruption of treatment or administration of other medications, as these events would not be uniformly recorded in the patient chart To reduce the risk of under reporting, a separate hospital-wide computerized registry of adverse drug reactions was also searched for patients that received paclitaxel during the study period. Reports to this registry were voluntarily contributed by hospital staff, and would therefore tend to bias our sample toward the occurrence of reactions, as patients without adverse events would not be entered into the registry. These data were cross-referenced with the larger pharmacy database for a more conservative estimate of adverse events.
Results A total of 283 outpatients were identified that received at least one cycle of therapy during the two years selected from the pharmacy database. Three of these patients were initially identified through cross-reference with the hospital-wide adverse drug reaction database. The remainder of events were accurately recorded in both database systems. Medical charts and treatment records were available for review on all patients and study population characteristics are summarized in Table 1. The ratio of women to men was 1.8, which reflects the early utilization of paclitaxel in the treatment of breast and ovarian cancer. However, the most common primary site was lung, reflecting recent experience in nonsmall-cell lung cancer. Intravenous prophylaxis for hypersensitivity reactions was administered 30 minutes prior to paclitaxel and consisted of dexamethasone (10 or 20 mg), diphenhydramine (50 mg), and cimetidine (300 mg) or ranitidine (50 mg). Among the treated population, a total of 13 patients (nine women and four men) experienced a hypersensitivity reaction during the first or second cycle of therapy equivalent to an overall frequency of 4.6%, with a 95%
CI of 2.2% to 7.0%. The median age of patients with hypersensitivity reactions was 61 years (range 44 to 83 years). Reactions occurred throughout the two-year period of this analysis without evidence of clustering or association with other phenomena, such as paclitaxel lot number. Reactions were generally mild and rapidly resolved with discontinuation of paclitaxel and other measures as summarized in Table 2. None of the patients required admission for acute management although two patients were electively hospitalized within the same day for additional corticosteroids followed by a 24-hour paclitaxel infusion. Only two reactions were graded as serious, one due to the requirement for bronchodilators in one and development of angioedema in the other. Thus, the incidence of serious hypersensitivity reactions was 0.7% with a 95% CI of 0.2% to 1.2%. There were no treatment-related deaths. Therapy with paclitaxel continued in 10 of the 13 patients with various modifications, such as lengthening of the infusion from 3 to 24 hours, or addition of oral dexamethasone premedication without recurrent hypersensitivity reactions on subsequent cycles. Only one patient discontinued therapy due to recurrent hypersensitivity reaction at rechallenge and two patients discontinued therapy without attempt at rechallenge. All three of these patients went on to receive alternative chemotherapy without paclitaxel. Discussion In view of the increased utilization of outpatient paclitaxel infusions and the inconvenience of prolonged oral prophylaxis with dexamethasone, we wished to evaluate the safety of short-course intravenous dexamethasone administered 30 minutes prior to paclitaxel. A sequential cohort of outpatients spanning two years was identified retrospectively from a pharmacy database that unbiased with regard to adverse events. A small number of
Table 1. Patient characteristics. Total number of patients Male Female Total number of treatment cycles Median age (range) years Primary Tumor Lung Ovary Breast Esophageal or gastric Head and neck Endometrial Miscellaneous* 1
283 101 (35.7%) 182(64.3%) 534 60 (27 to 86) 97 (34.3%) 58 (20.5%) 57(20.1%) 25 (8.8%) 16(5.7%) 7 (2.5%) 23(8.1%)
Miscellaneous includes: 3 bladder, 3 lymphoma, 3 sarcoma, 2 prostate, 1 colon, 1 fallopian, 1 glioblastoma, 1 hepatocellular carcinoma, 2 melanoma, 1 mesothelioma, 1 parotid, 1 peritoneal, 1 renal, 1 testicular, 1 unknown.
613 Table 2. Patients with hypersensitivity reactions." Patient characteristics
Treatment data
Age
Sex
Primary tumor
Date
49
F
Melanoma
75
F
56
Cycle
Dose (mg/
Sched (hrs)
Vol (ml)
2/94
1
175
5
Ovary
3/94
1
135
6
Lung
8/94
2
175
150
9/94
2
175
1
1
10/94
1
175
3
6
45
M
62
M
Esophageal Lung
62
F
Ovary
11/94
1
175
3
< 1
69
F
Breast
11/94
1
175
3
10
Lung
11/94
1
175
3
6
72 56
M
Lung
7/95
2
175
1
10
61
F
7/95
1
200
1
50
83
F
Head and neck Bladder
9/95
1
175
3
9
54
M
9/95
1
200
3
9
44
F
12/95
1
175
3
7
Head and neck Ovary
Description of reaction and management
Flushing, angioedema Cl (serious). Infusion aborted. Resolved with DPH, DEX. Admitted for additional DEX followed by 24-hour infusion without HSR. No HSR C2. Flushing, dyspnea, chest pain, diarrhea Cl. Infusion aborted. Resolved with DPH, DEX. Admitted for additional DEX followed by 24-hour infusion without HSR NoHSRC2. N o H S R C l . Hypertension and fever during C2 Resolved with AAP, infusion completed. HSR occurred while taking oral steroids for brain metastases. No HSR with C3. No HSR Cl. Flushing, dyspnea, hypotension during C2. Resolved with hydration. Completed dose next day after oral DEX. No HSR C3 Palpitations, dyspnea Cl. Resolved spontaneously. Completed dose next day without HSR. No HSR C2 with oral DEX. Flushing, hypotension Cl. Infusion aborted. Resolved with DPH and DEX Subsequently admitted for 24-hour infusion without HSR. No HSR C2. Flushing, chest tightness Cl Infusion aborted. Resolved with hydration. Rechallenged after three weeks with recurrent HSR and paclitaxel discontinued. Flushing, wheezing, dyspnea Cl. Resolved spontaneously. Paclitaxel discontinued. No HSR Cl. Dyspnea, chest discomfort C2. Resolved spontaneously, infusion completed same day. No HSR C3. Flank pain, parasthesia, dyspnea Cl. Infusion aborted. Paclitaxel discontinued. Dyspnea Cl (serious). Resolved with bronchodilator, DEX. No HSR C2 with 24-hour infusion. Coughing, facial flushing C1. Resolved with HC and DPH. Retreated five days later after oral DEX, no HSR. No HSR C2 with oral DEX. Dyspnea, chest pain, flushing Cl. Infusion aborted. Resolved without further therapy Retreated after oral DEX. No HSR C2 with oral DEX.
* Abbreviations: Sched - scheduled infusion duration; Vol - volume infused when reaction occurred; HSR - hypersensitivity reaction; AAP acetaminophen; DPH - diphenhydramine; DEX - dexamethasone; HC - hydrocortisone; C - treatment cycle.
additional outpatients were identified from a hospitalwide computerized registry of adverse drug reactions. Records from all patients were then manually reviewed to determine the incidence and nature of all reactions occurring during the first two cycles of paclitaxel. The overall incidence of all hypersensitivity reactions was 4.6% and serious reactions limited to 0.7%, which compares favorably to published data with conventional oral prophylaxis prior to either a 3-hour or 24-hour paclitaxel infusion [4]. In a retrospective series of 43 patients that received short-course intravenous dexamethasone (16 mg) there were no reactions, but patient selection criteria and the duration of paclitaxel infusion were not specified [5]. Although we attempted to collect data on hypersensitivity reactions from multiple sources, we assume that our written and computerized hospital records underestimate the true incidence of minor reactions that were not associated with any intervention or change in treatment program. Careful prospective monitoring in the context of a randomized clinical trial using oral dexamethasone prophylaxis documented a 42% incidence of
minor reactions, including transient flushing, at some point during treatment [4]. This is much higher than our reported overall incidence of 4.6%. In contrast, our 95% Cl for serious events (0.2% to 1.2%) overlap with the prospective randomized study, and we have no reason to believe that our data for serious reactions would be subject to the same negative reporting bias as minor reactions. Therefore, on the basis of our retrospective data, we conclude that short-course intravenous prophylaxis is at least equivalent to oral prophylaxis. All 13 reactions occurred quickly and resolved rapidly without sequelae and none of the patients required hospitalization for acute management. Ten patients were rechallenged with paclitaxel after modification in premedication or infusion duration and were able to continue therapy without recurrence of hypersensitivity reactions. Paclitaxel was discontinued in one patient due to recurrent hypersensitivity. Two patients were not rechallenged and received alternative chemotherapy. Of interest, one patient experienced an atypical reaction (fever and hypertension) during the second cycle after beginning therapy with oral dexamethasone for man-
614 agement of cerebral edema associated with brain metastases. This confirms the observation that hypersensitivity reactions can occur in spite of both oral and intravenous prophylaxis. In summary, we recommend that patients receive dexamethasone 10 or 20 mg, diphenhydramine 50 mg, and cimetidine 300 mg (or ranitidine 50 mg) intravenously 30 minutes prior to outpatient paclitaxel infusions of one to three hours. This regimen appears safe, effective, and convenient when retrospectively evaluated in a diverse patient population. We are not aware of any specific contraindications to this approach or situations when the oral regimen would be preferred. If the patient develops a reaction, paclitaxel should be interrupted and supportive care administered until the reaction resolves. The majority of hypersensitivity reactions resolve without sequelae and do not require hospitalization. Following additional intravenous or oral premedication, paclitaxel can be cautiously resumed with monitoring of vital signs within 24 hours. In view of the small number of reactions and variability in patient management, it is not possible to recommend a uniform approach for rechallange based on our retrospective data. In conclusion, this single-dose intravenous regimen is a reasonable choice for outpatients receiving paclitaxel by a one- to three-hour infusion, reducing the need for monitoring compliance during multidose oral regimens, and reducing the total amount of steroids administered.
Acknowledgement
The authors gratefully acknowledge support for data collection provided to Fox Chase Cancer Center as an unrestricted grant from Bristol-Myers Squibb, Princeton, NJ.
References 1. Rowinsky EK, Donehower RC. Paclitaxel (Taxol). N Engl J Med 1995; 332: 1004-14. 2. Weiss RB, Donehower RC, Weirnik PH et al. Hypersensitivity reactions from taxol. J Clin Oncol 1990; 8: 1263-8. 3. Lorenz W, Reimann HJ, Schmal A et al. Histamine release in dogs by cremophor EL and its derivatives: Oxethylated oleic acid is the most effective constituent. Agents Actions 1977; 7: 63-7. 4. Eisenhauer EA, ten Bokkel Huimnk WW, Swenerton KD et al. European-Canadian randomized trial of paclitaxel in relapsed ovarian cancer. High-dose versus low-dose and long versus short infusion. J Clin Oncol 1994; 12: 2654-66. 5. Pankh B, Khanolkar S, Advani SH et al. Safety profile of singledose dexamethasone premedication for paclitaxel. J Clin Oncol 1996; 14: 2189-90. Received 13 February 1997; accepted 13 May 1997. Correspondence to. Michael A. Bookman, MD Department of Medical Oncology Fox Chase Cancer Center 7701 Burholme Avenue Philadelphia, PA 19111, USA
