Patients

Twenty-two patients with neuroblastoma were treated between January 1995 and September 1997 in six Children’s Cancer Group (CCG) institutions. All patients were treated under the auspices of a protocol approved by the institutional review board at each institution (CCG-0935).

Eligibility criteria included age younger than 22 years at initial diagnosis. Patients must have had the diagnosis of neuroblastoma based on tumor histology or bone marrow metastases and elevated urine catecholamine metabolites. Disease status could have been complete response (CR), very good partial response (VGPR), or partial response (PR). Any patient who had prior ch14.18 or 14.G2a therapy was ineligible. Any prior myeloablative chemotherapy or chemoradiotherapy regimen was acceptable. The study began as an autologous bone marrow transplantation (ABMT) protocol; however, investigators were allowed to add or substitute autologous peripheral-blood stem cells instead of ABMT 1 year into the study. Patients were required to have a Lansky performance scale score of 1, 2, or 3, and a life expectancy of more than 2 months. The following organ function criteria were required for eligibility for ch14.18 treatment: left ventricular ejection fraction of more than 47% by radionuclide angiogram or shortening fraction of more than 27% by echocardiogram, creatinine clearance or radioisotope glomerular filtration rate of more than 70 mL/min/1.73 m², AST and ALT less than 2.5 × normal and total bilirubin less than 1.5 × normal, no evidence of dyspnea at rest, no exercise intolerance, and a pulse oximetry greater than 94%. If pulmonary function tests were performed, forced expiratory volume in 1 second/forced vital capacity had to be more than 60%. Patients with veno-occlusive disease of the liver must have been either stable or improving.

Treatment Program

The protocol recommended, but did not require, that patients receive daily GM-CSF starting on day 0 of HSCT. If patients were started on another growth factor (eg, granulocyte colony-stimulating factor [G-CSF]) after HSCT, they had to be switched to GM-CSF at least 72 hours before initiation of ch14.18 treatment. Patients had to begin ch14.18 therapy within 8 weeks of HSCT after the total absolute phagocyte count (APC) was more than 1,000/μL (APC = neutrophils [segs + bands] + monocytes) after HSCT. If the increase of WBC count was slow and the APC of more than 1,000/μL had not been reached by day 25 after HSCT on GM-CSF, then the investigators were allowed to add G-CSF 10 μg/kg administered intravenously (IV) over 30 minutes daily to hasten the WBC recovery. If G-CSF was added to GM-CSF, G-CSF should have been continued until the APC was more than 1,000/μL for at least 3 consecutive days. G-CSF was required to be stopped at least 24 hours before initiation of ch14.18 treatment.

GM-CSF 250 μg/m²/d IV as a 2-hour infusion was continued during the days of ch14.18 infusion and for 3 days afterward. The same dose of GM-CSF could be administered as a subcutaneous injection on an outpatient basis. The ch14.18 infusion was started approximately 1 hour after GM-CSF when both were administered on the same day. ch14.18 was infused IV over 5 hours daily for 4 consecutive days. The initial ch14.18 dose level was 20 mg/m²/d. The subsequent dose levels were 30, 40, and 50 mg/m²/d. There was no intrapatient dose escalation.

Patients were premedicated with hydroxyzine 1 mg/kg IV over 10 minutes and acetaminophen 10 mg/kg PO to start 20 minutes before ch14.18 infusion. Also, it was recommended to begin morphine sulfate with a loading dose of 0.1 mg/kg immediately before ch14.18 administration and to continue morphine as a continuous infusion at a rate of 0.05 mg/kg/h. In addition, IV or subcutaneous epinephrine was administered to treat allergic phenomena not controlled by the above premedication regimen.

The infusion time of ch14.18 was allowed to be prolonged from 5 to 10 hours, or longer if neuropathic pain was not controlled with appropriate doses of narcotics, or if hypotension, serious allergic toxicity such as extensive urticaria or bronchospasm, or severe peripheral-sensory problems were observed.

Patients were allowed to receive up to six 4-day courses of ch14.18 therapy with GM-CSF in the absence of progressive disease. Second and subsequent treatments were planned to be administered no earlier than every 28 days. The maximum delay allowed in administering the second and subsequent courses was 120 days. Second and subsequent courses of ch14.18 were not mandatory. The HACA test results were not used to determine whether to administer the second or subsequent courses.

The following toxicities attributed to GM-CSF required decreasing the dose of GM-CSF at least 50% or discontinuing it: (1) grade 3 or 4 capillary leak syndrome/fluid retention; (2) persistent high (> 39.5°C) fevers with no obvious source of infection and negative blood cultures; (3) persistent or recurrent tachycardia and hypotension.

Measurement of HACA and ch14.

18 Levels

Definition of Dose-Limiting Toxicity (DLT)

DLT of ch14.18, when combined with GM-CSF, was defined as toxicity not related to known GM-CSF toxicity. If known GM-CSF–related toxicity was observed, then the dose of GM-CSF was to be reduced by 50% or, if necessary, discontinued while maintaining the dose of ch14.18. All toxicities used to define the DLT of GM-CSF + ch14.18 refer to those that were seen while the patient was receiving GM-CSF and ch14.18, because the primary objective of the study was to define the maximum-tolerated dose (MTD) of the combination of GM-CSF and ch14.18. The DLT of ch14.18, when combined with GM-CSF, was defined as grade 3 or 4 toxicity (CCG Toxicity Criteria and CCG Biologics Toxicity Scale), with the following exceptions: (1) grade 4 pain (requires IV narcotics), (2) grade 3 nausea and vomiting, (3) grade 3 fever, (4) grade 3 skin toxicity that improves with IV treatment (eg, with diphenhydramine) within 24 hours, (5) grade 3 electrolytes (especially hyponatremia ≤ 124 mEq/L) that improves with treatment within 24 hours, (6) grade 3 allergy (eg, moderate bronchospasm) that improves with treatment within 24 hours, (7) grade 3 hypotension and hypertension, or (8) grade 3 hepatic toxicity that returns to grade 1 before the time for the next ch14.18 treatment course.

Statistical Analysis

Patients were enrolled in cohorts of three patients each, according to the following scheme:

1. If none of the three patients demonstrated DLT, the dose was escalated one level for the next cohort.

2. If two or more of the three demonstrated DLT, the dose was considered not tolerated and reduced by one dose level. If six patients had been treated at the reduced dose level, and it was tolerated, then that level was considered the MTD. If, at most, one of six patients treated at the reduced dose level experienced DLT, then the dose was considered the MTD. Otherwise, the dose was considered to exceed the MTD and was reduced and more cohorts were enrolled, if necessary, according to the rule noted above.

3. If one of the three patients in the first cohort enrolled at a dose level experienced toxicity, then three more patients were enrolled at the same dose level. If none of the patients in the next cohort experienced DLT, then the dose was escalated and rule no. 1 noted above was applied to the next cohort. If one or more of the patients in the next cohort experienced DLT, then the dose level was considered to exceed the MTD. Subsequent cohorts were enrolled according to the procedures outlined in rule no. 2 above.

Change in Peripheral-Blood Counts

The change in each component of the peripheral-blood count was quantified as the difference between the absolute number of the first complete blood cell count reported on or after day 3 and that characteristic measured before study entry.

Course Length

For each course, the interval was calculated from the date on which the antibody was first given until the date on which the next 4-day course of antibody was given or the patient was removed from protocol therapy. If the patient was removed from protocol therapy at the end of a course other than the sixth course, then the observation on course length was considered censored. Otherwise the observation was considered complete. Median course length was calculated by the Kaplan-Meier method.

Patient Characteristics

Patient characteristics are listed in Table 1. Twenty-three patients with neuroblastoma were entered onto the study. One patient assigned to dose level 2 never received the antibody, because his parents decided against his participation in the study before the first course. On review, three of 22 patients who did receive treatment with ch14.18 and GM-CSF were subsequently deemed ineligible because of protocol treatment violations pertaining only to the timing of the GM-CSF administration before the first treatment course. Two of these three received 48 hours of GM-CSF instead of a minimum of 72 hours, as required by the protocol. In the third patient, GM-CSF was started with the antibody. This report describes the remaining 19 patients. Median age was 4.8 years (range, 2 to 15 years). Fourteen and five patients were in first and second remission before HSCT, respectively (Table 1). Median times from diagnosis to transplantation were 6 (range, 5 to 8) and 19.5 (range, 11 to 48) months for patients treated in first and second remission, respectively. As listed in Table 1, various preparatory regimens were used, but melphalan, carboplatin, and etoposide with local radiotherapy was the most common one.

Timing of the Courses

The median time between the day of transplantation and achieving an APC greater than 1,000/μL was 24 days (range, 10 to 62 days). The median time between achieving an APC greater than 1,000/μL and the first course of antibody was 11.5 days (range, 0 to 56 days). The median numbers of days between first and second, second and third, third and fourth, fourth and fifth, and fifth and sixth treatment courses were 31 (range, 21 to 70), 32 (range, 20 to 92), 30 (range, 27 to 70), 28 (range, 26 to 56), and 28 (range, 21 to 34) days, respectively. A total of 79 courses were administered. Two, one, three, six, and seven patients received one, two, three, four, and six courses of therapy, respectively.

Toxicities

There were no toxic deaths. As expected, the major toxicity was neuropathic pain (Table 2). Thirteen (68%) of 19 patients experienced severe pain during the first course of therapy. This was usually localized to the abdomen and lower extremities and was noted within 1 hour of starting the ch14.18 infusion. It usually lasted through the antibody infusion and abated 1 to 2 hours afterward. Two of these 13 episodes were classified as grade 4 (ie, severe pain despite the use of parenteral narcotic analgesics). One of these two was considered DLT at 50 mg/m²/d. In the other patient, who received 40 mg/m²/d, prolongation of the ch14.18 infusion duration from 5 to 10 hours helped to control the pain with parenteral narcotics. One of two, two of three, five of eight, and four of six had grade 4 pain at ch14.18 dose levels 1, 2, 3, and 4, respectively. There was no clear relationship of increase in pain with dose escalation; however, two episodes of severe pain despite the use of parenteral narcotic analgesics were encountered at dose levels 3 and 4. One patient treated at dose level 4 who was considered to have DLT continued to require oral morphine for 2 more weeks and refused to stand up and walk for 3 weeks after the end of ch14.18 therapy (Table 3). This reversible toxicity was the most severe toxicity attributed to ch14.18 noted in this trial. Overall, 47 (59%) of 79 courses were complicated with neuropathic pain.

Fever with no source of infection was a relatively common complication. The third most common toxicity was nausea and vomiting. Six (32%) of 19 patients experienced nausea/vomiting during the first course. One of six was considered as DLT (Table 3). The patient required readmission for vomiting and dehydration 48 hours after discharge after the first course of therapy. The next most common problem was urticarial eruption. This was observed in seven (37%) of 19 patients during the first course. Two of seven cases were classified as grade 3 (ie, generalized eruption), one of which recurred during the subsequent two courses despite treatment and was declared to be DLT (Table 3). The remaining patients had less than grade 3 skin/allergy toxicity. These patients required hydroxyzine every 4 hours and epinephrine every 6 hours on the days of ch14.18 administration. In addition, the infusion time of ch14.18 was prolonged to minimize these symptoms.

Hypotension was observed during 12 of 79 courses, of which five cases were considered to be grade 3 or 4. These patients required infusions of normal saline or albumin. The antibody infusion was completed in all patients experiencing hypotension by prolonging the infusion time. Hypotension was avoided during most of the subsequent courses by continuing with the prolonged ch14.18 infusion. Nine of 79 courses were complicated with mild capillary leak syndrome, two of which occurred during the first course. In six of nine courses, capillary leak was considered to be grade 2 (ie, face swollen, weight gain 5% to 10%, with mild peripheral edema), and the remaining three were grade 1.

Dilated pupils were noted during six of 79 courses. In four of six episodes, patients also complained of blurred or double vision. All resolved over a period of days to weeks. Transient motor weakness, mild to moderate paresthesias, and moderate to severe agitation considered to be unrelated to the neuropathic pain were infrequently observed (Table 2).

The duration of ch14.18 infusion was increased because of side effects in 10 of 19 patients. The duration was increased to 10 hours in seven patients, 12 hours in one patient, and to 20 and 23 hours in two other patients. The main reasons for prolongation of the duration of infusion were neuropathic pain, hypotension, and allergic skin reactions. The duration was increased in one of three, one of three, four of seven, and four of six patients receiving dose levels 1, 2, 3, and 4, respectively. Once the duration of ch14.18 infusion was increased, these patients received subsequent courses at the same prolonged rate.

Complete blood cell counts were obtained before the ch14.18 infusion was started and on the last day of antibody therapy. Declines of WBCs and APC were observed in 43 (56%) of 77 and 40 (56%) of 72 courses with sufficient data reported for analysis, respectively (Table 2). However, the declines in WBCs and APC were not a clinical problem. Similar to WBCs and APC, declines in platelet counts and hematocrit were also observed during therapy. The platelet count decreased in six of seven untransfused patients during the first course. Overall, platelet count declined in 44 (85%) of 52 courses given to untransfused patients during the 4-day therapy period, with a median decline of 25% (range, 2% to 91%). However, none of the courses were complicated with bleeding problems during this transient thrombocytopenic period.

Eight significant infectious problems were reported among the 79 courses. Five were herpes zoster, one was staphylococcal infection of the central venous catheter, one was an enterococcus infection of the central venous catheter and urine with bilateral lung infiltrates, and one was apparent sepsis with no organism identified. Transient and insignificant AST/ALT elevations were observed. Pulmonary toxicity was reported in four of 79 courses. Two of four were grade 3, ie, dyspnea with desaturation requiring oxygen. These were separate from the capillary leak syndrome and all resolved quickly. Metabolic problems included hyponatremia in four of 79 courses (grade 2 or less), one case of hypocalcemia (grade 4), and one case of hypokalemia (grade 3).

No GM-CSF dose modifications were required during any course, based on toxicities attributed to GM-CSF during the 4 days of ch14.18 therapy or the 3 days afterward.

Serum HACA and ch14.

18 Values

For each course of therapy, serum specimens were required before starting the ch14.18 treatment and on days 3, 7, and 13 after the initiation of ch14.18 treatment. Table 4 presents data collected and analyzed from serum specimens obtained from 18 of the 19 eligible patients. All serum samples obtained were analyzed for ch14.18 level and HACA reactivity. The values for the peak serum level of ch14.18 for the first course of antibody treatment were calculated for each patient and are presented as the mean value for all eligible patients at each dose level. Five (28%) of 18 evaluated patients showed HACA reactivity detectable after two (two patients), five (two patients), and six (one patient) courses of treatment at the dose levels listed in Table 4. No HACA reactivity was found in serum specimens from patients receiving the lowest dose of ch14.18.

Disease Progression

Ten of 19 patients experienced disease progression, with a median follow-up of 40 months (range, 25 to 50 months). Eight of 10 patients who experienced disease progression were in first remission (three in first CR, four in first VGPR, and one in first PR) and two were in second CR. Three of eight, who were in first remission, experienced disease progression while on antibody therapy after receiving one, three, and five courses of antibody therapy, respectively. The remaining five of eight patients received one, four, four, six, and four courses, and relapsed 10, 6, 8, 22, and 27 months after the last course of therapy, respectively. Two patients in second remission received one and three courses of antibody therapy and had disease progression 22 and 15 months off therapy, respectively. Progression occurred in the primary site in two patients, in distant sites in seven patients, and in the primary and distant sites in one patient. Two of 10 patients with progressive disease had developed HACA, whereas three of nine patients who did not have disease progression had HACA positivity while on study.