Tolerability and Side-Effect Profile of rhIL-11

Introduction

Recombinant human interleukin-11 (rhIL-11, also known as oprelvekin(Drug information on oprelvekin) [Neumega]) was approved by the US Food and Drug Administration (FDA) in November 1997 to prevent severe thrombocytopenia and to reduce the need for platelet transfusions following myelosuppressive chemotherapy in patients with nonmyeloid malignancies who are at high risk of severe thrombocytopenia. The approval of a once-daily dose of 50 µg/kg administered subcutaneously was based primarily on efficacy and safety data from two randomized, placebo-controlled studies in which a total of 77 patients with solid tumors or lymphomas were enrolled.[1,2]

The tolerability of rhIL-11 has been evaluated using data from a total of 355 subjects who had been treated with rhIL-11 (vs 155 subjects treated with placebo) during the clinical development of this platelet growth factor. This data set includes healthy volunteers and patients with nonmyeloid cancer who participated in pharmacodynamic studies of rhIL-11 alone, as well as in placebo-controlled and open-label rhIL-11 studies in chemotherapy-induced thrombocytopenia.

The first clinical experience with rhIL-11 was from a phase I study that characterized the safety (and preliminary efficacy) of rhIL-11 at doses of 10, 25, 50, 75, and 100 µg/kg. Each dose of rhIL-11 was administered once daily for 14 days in 16 women receiving chemotherapy for breast cancer.[3] The rhIL-11 75-µg/kg/d dose was not technically considered to be associated with true dose-limiting toxicity, but subjectively it was not as well tolerated as doses £ 50 µg/kg/d due to the higher frequency of reversible, moderate fatigue, myalgia, and arthralgia.

Dose escalation was stopped at the 100-µg/kg dose after a patient experienced a minor thrombotic stroke after three doses of rhIL-11. No adverse sequelae resulted from this stroke and the patient’s neurologic status returned to baseline levels. Importantly, this stroke was not associated with thrombocytosis induced by rhIL-11 or changes in coagulation parameters such as prothrombin time or partial thromboplastin time. Thus, doses > 50 µg/kg/d were not evaluated in subsequent randomized, placebo-controlled studies in adults.

A placebo-controlled phase II study established the statistically significant efficacy of rhIL-11 at the 50-µg/kg/d dose, but not at the 25-µg/kg/d dose, in achieving[4] the primary efficacy end point (proportion of patients avoiding platelet transfusions).[1] Therefore, in a subsequent phase II placebo-controlled study, rhIL-11 was evaluated at only the 50-µg/kg/d dose.[2] The rhIL-11 75-µg/kg/d dose is, however, currently under investigation in pediatric patients with solid tumors or lymphomas. Pediatric patients appear to tolerate this dose well and, because of their ability to clear the drug more rapidly, a dose of 75–100 µg/kg produces plasma drug concentrations that are comparable to concentrations achieved in adults with the 50-µg/kg dose.[5]

A phase III study was initiated to further evaluate the safety and efficacy of rhIL-11 at 50 µg/kg based on the findings of the phase I and phase II studies. This study had accrued 141 patients, but was prematurely closed after the FDA granted approval of rhIL-11.

Across the two placebo-controlled, double-blind phase II studies, patients received subcutaneous daily doses of rhIL-11 50 µg/kg for 10 to 21 days. In the phase III study, rhIL-11 50 µg/kg/d was administered for 14 days. In studies that included open-label cycles, patients received up to eight sequential cycles of rhIL-11. In these studies, adverse events were assessed and graded for severity by the investigators according to the World Health Organization Toxicity Grading Scale or other prespecified definitions. In general, grades 1, 2, 3, and 4 corresponded to mild, moderate, severe, and life-threatening severities, respectively.

The results of the two randomized phase II and the one abbreviated phase III placebo-controlled, double-blind clinical studies support the acceptable tolerability and safety profile of rhIL-11 in patients undergoing chemotherapy for nonmyeloid malignancies.[1,2,5,6] Data from these studies indicate that adverse events associated with rhIL-11 therapy are generally mild, reversible, and easily managed.

Adverse-Event Profile

Phase II Randomized Placebo-Controlled Studies

Across the two phase II placebo-controlled chemotherapy studies (study I and study II), only 8 of 72 patients (11%) in the rhIL-11 50-µg/kg group prematurely discontinued the study drug because of adverse events (most commonly atrial fibrillation or flutter), compared with 4/67 patients (6%) in the placebo group.[5] A combined analysis of data from these two studies showed that the adverse events in the rhIL-11 group were similar in incidence and type to those seen in the placebo group. There was no difference between the rhIL-11 and placebo treatment groups in the respective incidence of fever (36% vs 28%; P = .4), neutropenic fever (48% vs 42%; P = .5), or flu-like symptoms (eg, chills, 25% vs 25%; P = 1.0). Thrombotic events, the number of red blood cell transfusions, and the duration of neutropenia (< 500/µL) were also similar in both groups.

The majority of adverse events reported in these two phase II studies, regardless of treatment group, were mild (grade 1) in severity, easily managed, and reversible following discontinuation of rhIL-11. Eight adverse events occurred with a statistically significantly (P < .05) higher incidence for rhIL-11 50 µg/kg compared with placebo during the double-blind study cycles (Table 1). Most of these events were mild or moderate (grade 1 or 2) in severity.

Across studies I and II, the incidence of severe or life-threatening adverse events in the rhIL-11 and placebo groups was, in general, similar. The only adverse event of grade 3 or 4 severity that occurred with a significantly higher incidence with rhIL-11 50 µg/kg compared with placebo was asthenia (grade 3, 14% [rhIL-11] vs 3% [placebo]; P = .03, Fisher’s exact test), although the overall incidence of asthenia was similar between treatment groups (65% and 64%, respectively; P = 1.0).[5]

Phase III Randomized Placebo-Controlled Double-Blind Study

The design of the phase III study differed from that of the two previously discussed phase II placebo-controlled studies by the standardized institution of a masked diuretic (hydrochlorothiazide 25 mg/triamterene 37.5 mg [Maxzide-25MG]) in patients receiving rhIL-11. This was done in an effort to ameliorate the fluid retention that was observed in prior studies.[3,5] Its use was based on a randomized crossover trial of hydrochlorothiazide(Drug information on hydrochlorothiazide) 25 mg/triamterene 37.5 mg and rhIL-11 in normal volunteers. However, the systematic use of a diuretic in this study was subsequently discontinued due to a pattern of adverse events related to diuretic-induced fluid and electrolyte imbalance, although the use of a diuretic was still permitted at the discretion of the investigator.

In this phase III trial, among 88 patients randomized to the rhIL-11 treatment group, 11 (12.5%) discontinued the study drug because of adverse events (primarily atrial fibrillation or flutter [n = 7; 8%]). These findings are similar to the two phase II controlled chemotherapy studies. In contrast to the phase II studies, metabolic adverse events led to the discontinuation of rhIL-11 in four patients (5%) in the phase III study. The adverse events included dehydration (n = 1), hypokalemia (n = 1), hypomagnesemia (n = 1), and hyponatremia (n = 1) and likely occurred secondary to diuretic therapy. Six of 88 (7%) patients in the rhIL-11 group developed hypokalemia, compared with 2 of 45 (4%) patients in the placebo group.

These occurrences, once again, were determined by the investigators to be associated with diuretic use and were not study-drug related. The only adverse events that occurred significantly more frequently in patients in the rhIL-11 group were dizziness, edema, dyspnea, atrial fibrillation/flutter, and palpitation (P < .05). Two patients in the rhIL-11 group died unexpectedly. It was concluded that these deaths were due to hypokalemia (probably related to diuretic use) and were not directly related to study-drug administration.

The incidence and type of adverse events in the phase III chemotherapy study (Table 2) were similar to those seen in the two placebo-controlled phase II studies (Table 1), with the exception of dizziness, which was attributed to concomitant diuretic use.[6] The attribution of dizziness to diuretic therapy is supported by an observed reduction in the incidence of dehydration from 8% during systematic once-daily administration of a diuretic to 0% when the diuretic was discontinued. Additionally, the incidence of dizziness in the rhIL-11 treatment group was reduced from 44% to 29% after the study protocol was changed to eliminate mandatory systematic diuretic use in the rhIL-11 group.[5]

Based on the experience in this phase III study, the routine use of diuretics during rhIL-11 therapy is not recommended, although their temporary use to alleviate edema may be indicated in some patients. If diuretic therapy is indicated, a potassium-sparing diuretic should be used and serum electrolyte levels and fluid balance should be monitored closely, as in any patient receiving a diuretic. During the phase III study, the development of severe hypokalemia (serum potassium concentrations £ 2.8 mEq/L) resulted in two fatalities.

Hypokalemia has not been shown to be attributed to treatment with rhIL-11. In healthy volunteers receiving a dose of 25 µg/kg, rhIL-11 treatment does not cause potassium depletion,[7] and rhIL-11 has not been associated with hypokalemia in patients with cancer in studies where diuretic therapy was neither encouraged nor recommended.[1,3] These findings underscore the importance of close monitoring of patients with cancer, as they often receive many therapies that complicate treatment.

As in the phase II controlled trials, the majority of adverse events in the phase III trial were mild or moderate in severity (grade 1 or 2) and resolved without sequelae after discontinuation of the study drug. Most adverse events were considered to be due to the administration of chemotherapy or the patient’s underlying disease. There were no significant differences between the rhIL-11 and placebo treatment groups in the overall incidence of grade 3 and 4 adverse events (36% [rhIL-11] and 38% [placebo]). However, the incidence of grade 3 or 4 chills was reported significantly (P = .04) more frequently in the placebo group (n = 3, 7%) than in the rhIL-11 group (n = 0). In addition, the only episodes of grade 3 or 4 gastrointestinal hemorrhage (two patients) or pulmonary infarct (one patient) occurred in the placebo group.

Fluid Retention

The majority of adverse events associated with rhIL-11 (eg, edema, dyspnea, pleural effusions, and conjunctival injection) are believed to be the result of fluid retention and increased plasma volume. Studies in volunteers have shown that treatment with rhIL-11 is associated with an expansion in plasma volume of approximately 20% as a result of sodium retention.[8] Edema is the most frequently reported adverse event; however, it is easily managed and reversible.

Edema and Dyspnea

In the two phase II placebo-controlled chemotherapy studies (studies I and II), edema and dyspnea associated with rhIL-11 therapy were most often mild or moderate in severity and were either self-limited or responded well to diuretic therapy. None of the patients withdrew from the studies because of edema. In the phase III study, mild or moderate edema and dyspnea were also two of the most common adverse events associated with rhIL-11 (Table 2); however, during the systematic use of masked diuretic, the incidence of both edema and dyspnea was lowest in patients in the rhIL-11 group.

Pleural Effusions

All of the seven events of pleural effusion that occurred during the phase II controlled chemotherapy studies represented in Table 1 occurred in a study of patients with breast cancer (study II).[2] However, patients in the rhIL-11 group may have been at higher risk overall for pleural effusions at baseline because more patients with stage IV metastatic breast cancer were randomized to the rhIL-11 treatment group (25/40, 63%) than to the placebo group (18/37; 49%). Therefore, a higher portion of patients in the rhIL-11 group were at risk of developing or having pleural effusions related to pleural metastases. All of the seven patients who had pleural effusions had stage IV metastatic breast cancer. In addition, four of these seven patients had preexisting pleural effusions that worsened during rhIL-11 treatment. No patients required thoracentesis, and most pleural effusions responded well to treatment with furosemide(Drug information on furosemide). Thus, patients with a past or current history of pleural effusions should be monitored closely for increasing fluid accumulation, as should patients with preexisting ascites.

Dilutional Anemia

Mild, reversible dilutional anemia characterized by moderate decreases in hemoglobin (15% to 19% reductions from baseline) and hematocrit levels (approximately 20% reduction from baseline) secondary to plasma volume expansion has occurred during rhIL-11 therapy in phase I studies in healthy volunteers[5,8] and nonmyelosuppressed patients with cancer.[3] In the latter population, anemia developed within 2 to 3 days of initiation of rhIL-11 therapy and reached a nadir during the second week of dosing.[3,5] However, hemoglobin levels returned to baseline levels within 2 weeks after the completion of rhIL-11 therapy.

In the two completed phase II placebo-controlled chemotherapy studies (studies I and II), rhIL-11–associated increases in plasma volume were not associated with statistically significant increases in the incidence of clinically significant anemia or mean requirements for red blood cell transfusions, compared with placebo.[1,2] Across these studies, anemia was reported in only 10% of rhIL-11–treated patients, compared with 6% of placebo-treated patients.[5] The mean number of red blood cell units transfused per patient was 1.9 in the rhIL-11 group and 1.7 in the placebo group.

A randomized, double-blind, placebo-controlled phase I study in 12 healthy, salt-restricted volunteers confirmed that the anemia associated with rhIL-11 therapy is predominantly due to an increase in plasma volume secondary to sodium retention, rather than a decrease in red blood cell mass.[8] Together, these data suggest that reductions in hemoglobin, although common, are not usually clinically significant.

rhIL-11 should be used cautiously in patients with congestive heart failure in whom fluid retention may precipitate cardiac decompensation. It may be best to avoid using rhIL-11 in some patients with severe congestive heart failure. To minimize the risk of edema and associated sequelae during therapy with rhIL-11, patients should be instructed to reduce their salt intake, monitor their weight for gradual changes, and contact their physician if they notice weight increases or tightening of rings or shoes. In some patients, edema may be alleviated by the temporary use of a potassium-sparing diuretic. It should also be noted that the occurrence of edema with rhIL-11 is reversible upon discontinuation of therapy.

Atrial Arrhythmias

Among the eight patients (12%) in the two phase II placebo-controlled chemotherapy studies (studies I and II, Table 1) who developed atrial arrhythmias during therapy with rhIL-11 50 µg/kg, only three were symptomatic (vs one placebo-treated patient). None of the patients with atrial arrhythmias developed clinical sequelae.[1,2] Among the three symptomatic patients, one patient had preexisting undiagnosed severe aortic stenosis, which was repaired surgically and maintained with calcium channel blocker (diltiazem) therapy during the study. The second patient had a prior history of atrial arrhythmia at study entry. In the third patient, atrial arrhythmia was controlled with digoxin(Drug information on digoxin) and verapamil(Drug information on verapamil). In the remaining five patients, the occurrence of atrial arrhythmias was transient (detected only by Holter monitoring), and reverted to sinus rhythm without medical intervention or electrical cardioversion. In most patients, atrial arrhythmias have not recurred during continuation of rhIL-11 therapy.

Data from all patients who received at least one dose of rhIL-11 have identified four risk factors as being most likely associated with the development of atrial arrhythmias. They are patient age, history of atrial arrhythmias, history of cardiac disorder, and history of alcohol(Drug information on alcohol) use. Although these risk factors have been established as factors associated with the occurrence of atrial arrhythmias in the general population, treatment with rhIL-11 has been shown to be significantly associated (P < .05) with the occurrence of atrial arrhythmias. This suggests that rhIL-11 may provide an additional risk in patients already at risk for atrial arrhythmias. Therefore, rhIL-11 should be used with caution in these patients, and only after consideration of the potential risks in relation to anticipated benefits.

Because of the association of rhIL-11 with atrial fibrillation/flutter in previous studies, all patients in the phase III trial were prospectively monitored with electrocardiograms and transtelephonic monitoring (TTM), which is more sensitive than Holter monitoring for detecting transient arrhythmias in symptomatic patients.[9] Fifteen patients (17%) in the rhIL-11 arm experienced atrial arrhythmias vs no patients on placebo.

Among the 15 patients who experienced atrial fibrillation/flutter in the phase III study, 10 patients were symptomatic, although 2 patients had a history of atrial arrhythmia before study entry, and 1 patient had a history of supraventricular tachycardia. Thir-teen of the 15 cases of rhIL-11–associated atrial fibrillation/flutter were uncomplicated. Of the two remaining cases, one patient developed homonymous hemianopsia concurrently with an episode of atrial fibrillation. This was considered to be due to a cerebral infarction despite a normal computed tomography scan of the brain. Another patient had electrocardiographic evidence of myocardial infarction at the follow-up visit.

The risk of the development of atrial arrhythmias during therapy with rhIL-11 may be increased in patients with advanced age or a history of more than moderate alcohol consumption, cardiac disorder, doxorubicin(Drug information on doxorubicin) therapy, diabetes, or hypertension. rhIL-11 should be used with caution in patients with a history of atrial arrhythmia. Some cases of atrial arrhythmia may have been the indirect consequence of increased plasma volume, which has been shown to induce atrial distention[10] with subsequent decreases in atrial refractory period.[11]

Analysis of TTM data from the phase III study showed no evidence of an effect of rhIL-11 on cardiac conduction intervals (PR, QRS, or QTc intervals), suggesting the absence of direct effects by rhIL-11 on cardiac function.[5]

Bleeding Complications

Retrospective analysis of the phase II placebo-controlled chemotherapy studies (studies I and II) showed a significantly (P < .01, Fisher’s exact test) lower incidence of bleeding complications—primarily ecchymosis and epistaxis—among patients in the rhIL-11 group (28%) compared with those in the placebo group (51%). None of the patients treated with rhIL-11 (n = 69, 0%) developed severe or life-threatening bleeding complications (grade 3 hemorrhage, hemorrhagic cystitis, or vaginal hemorrhage); these occurred only among placebo-treated patients (3 of 67, 4.5%). In the phase III study, bleeding events contributing to hospitalization occurred in a higher percentage of patients in the placebo group (5/45, 11.1%) than in the rhIL-11 group (3/88, 3.4%). These observations reflect the lack of adverse effects of rhIL-11 on platelet function or blood coagulation (discussed below).

Papilledema

In 6 of 355 patients treated with rhIL-11 during clinical trials, papilledema was detected following two or more cycles of rhIL-11 treatment. However, in three of these patients, papilledema may have resulted from central nervous system tumors. During the phase III study, routine ophthalmologic examinations by ophthalmologists failed to detect any signs of papilledema in 119 patients for whom safety data were available.[5]

In preclinical studies in nonhuman primates, papilledema associated with excessive doses of rhIL-11 (1,000 µg/kg/d for up to 13 weeks) has reversed completely after treatment was discontinued and has not been associated with histopathologic changes in either the eye or the central nervous system. These observations suggest that papilledema occurs rarely and is usually mild, transient, and reversible upon discontinuation of therapy. However, on the basis of clinical observations, rhIL-11 should be used cautiously and with frequent ophthalmologic monitoring in patients with preexisting papilledema or tumors involving the central nervous system.