Abstract
Arsenic trioxide (ATO) and all-trans retinoic acid (ATRA) combination therapy yields high complete remission and disease-free survival rates in acute promyelocytic leukemia (APL). ATO is dosed on actual body weight and high ATO doses in overweight patients may contribute to increased toxicity. We performed a retrospective, two-center study comparing toxicities in patients who received the Lo-Coco et al ATRA/ATO regimen with capped ATO, ≤10 mg/dose, and non-capped ATO, >10 mg/dose. A total of 44 patients were included; 15 received doses ≤10 mg and 29 received >10 mg. During induction, there was no difference in the incidence of grade ≥3 hepatotoxicity, grade ≥3 QTc prolongation, neurotoxicity, and cardiac toxicity between groups. In consolidation, patients receiving >10 mg/dose experienced a greater incidence of neurotoxicity (66.7% vs 22.2%; p = 0.046). Capping doses saved $24634.37/patient and reduced waste of partially-used vials. At a median follow-up of 27 months, no disease relapses occurred in either group. This represents an opportunity to improve the safety profile of this highly effective regimen.
Keywords: Acute promyelocytic leukemia, arsenic trioxide, obesity, toxicity, capped dose
Introduction
Arsenic trioxide (ATO) and all-trans retinoic acid (ATRA) in combination is the standard of care treatment for low- and intermediate-risk acute promyelocytic leukemia (APL).1 This regimen represented a turning point APL treatment, yielding complete remission (CR) and overall survival (OS) rates over 90% at two years, while offering patients freedom from long-term complications of anthracyclines (i.e. cardiac dysfunction and secondary malignancies).2,3 However, patients receiving ATO may still experience clinically significant complications during treatment, including hepatotoxicity, QTc prolongation, neurotoxicity, and myelosuppression. The long-term implications of ATO treatment are only now being realized and studied.4
In this highly curable disease, adopting a dosing strategy that maintains anti-leukemic efficacy while reducing toxicity is critical. The ATO labeling recommends 0.15 mg/kg/dose for induction and consolidation in low/intermediate risk APL with no maximum dose.5 This is consistent with published dosing of ATO from Lo-Coco et al.,2 and reflects the American Society of Clinical Oncology’s (ASCO) clinical practice guideline recommendations to dose chemotherapy based on actual body weight (ABW), with few well-studied exceptions.6 These recommendations excluded patients with leukemia, but are still commonly extrapolated to leukemia patients. As a result, ATO dosing practices for low- and intermediate-risk APL vary among institutions, particularly in obese patients. However, clinical experience and growing evidence suggests obese patients with APL may experience increased toxicity compared to their non-obese counterparts. In a Phase I study of ATO in adults with APL, sudden deaths occurred in 3 of 10 patients, all of whom were obese, prompting dose capping at 150% of ideal body weight (IBW).7 In pediatric patients, enhanced toxicity was seen in patients with body mass indexes (BMI) >30 kg/m2, even with using an IBW-based dose.8 Furthermore, a retrospective, cohort analysis of low- and intermediate-risk APL patients found that when ATO was dosed according to ABW with no maximum, obese patients experienced significantly more doses held due to toxicity than non-obese patients.9 Pharmacokinetic data may provide a rationale for increased toxicity in obese patients. ATO has a volume of distribution of 5.6 L/kg, suggesting that larger individuals have wider drug distribution.10 Pharmacokinetic studies have confirmed a slower terminal elimination phase, and accumulation after multiple doses.5
A large proportion of the APL population has comorbid obesity. In a case control study of 469 patients with AML, 81% of APL patients were obese, while only 41.7% of non-APL AML patients were obese.11 A pooled analysis of 446 APL patients in four CALGB clinical trials supported an increased incidence of obesity in patients with APL and found worse OS in obese patients.12
Given the high incidence of obesity in APL and possible inferior outcomes, identifying an optimal ATO dosing regimen is an important clinical question. To date, comparative studies of toxicity and efficacy outcomes with ATO dosed on adjusted body weight or with dose maximums vs. ABW have not been published. The Johns Hopkins Hospital (JHH) Sidney Kimmel Comprehensive Cancer Center initiated dose-capping of ATO to the vial size in June 2017. The University of Maryland Medical Center (UMMC) Greenebaum Comprehensive Cancer Center serves a comparable regional population, and uses ABW with no dose cap. The primary objective of this study is to compare the toxicity profile of ATO between patients who did and did not receive doses capped at the 10 mg vial size.
Methods
Patients
This was an Institutional Review Board-approved, two-center, retrospective review of patients with low/intermediate risk APL receiving ATO and ATRA per the Lo-Coco et al regimen.3 Adult patients with APL who received ATO and ATRA per the Lo-Coco regimen were identified in the JHH and UMMC medical records. Patients without complete medication administration records, who received alternative induction regimens, were high-risk per the Sanz score (initial white blood cell count [WBC] >10,000/microliter), or received an anthracycline or gemtuzumab ozogamicin were excluded.13 Patients who received a starting ATO dose capped at or whose weight-based dose was ≤10 mg/dose were compared to patients who received >10 mg/dose.
Study design
Retrospective chart review was utilized to evaluate for inclusion and exclusion criteria. Baseline demographics and laboratory values including sex, race, age, weight, height, aspartate transaminase (AST), alanine transaminase (ALT), total bilirubin (TBILI), and QTc (Framingham correction) were collected. Maximum values of AST, ALT, TBILI, and QTc were obtained during induction. Grade ≥3 hepatotoxicity was defined according to the Common Terminology Criteria of Adverse Events (CTCAE) version 5. Grade ≥3 QTc prolongation was defined as a QTc ≥500 milliseconds (msec) per the CTCAE. Patients with grade ≥3 QTc prolongation were assessed for receipt of concomitant QTc prolonging agents in the previous 24 hours prior to the electrocardiogram. Cardiac toxicity and neurotoxicity were defined as the incidence of new-onset CTCAE-defined cardiac disorders and nervous system disorders, as determined through chart review. Stroke, intracranial hemorrhage, and headache were excluded as neurotoxicity due to the probability of underlying disease biology leading to stroke and intracranial hemorrhage, while headache is a common toxicity with ATRA. Medication administration data were utilized to confirm all doses and timing of ATO received. Absolute neutrophil count (ANC), platelet count (PLT), and hemoglobin (HGB) values were collected at time of induction evaluation bone marrow biopsy. Time to last packed red blood cell (PRBC) and platelet transfusion were obtained using blood bank administration data. In the absence of diffuse intravascular coagulopathy, active bleeding, or other patient-specific factors, packed red blood cell and platelet transfusion thresholds were 7 g/dL and 10000/microL, respectively. Response to induction therapy was assessed through documentation of morphologic CR by bone marrow biopsy or documented early death (ED) during induction. In consolidation, ATO starting dose, dose reductions/omissions, and clinical toxicities were assessed. Patients who did not have consolidation data available for evaluation were excluded for analysis of ATO toxicity in consolidation.
Endpoints
The primary endpoint was the composite incidence of grade ≥3 hepatotoxicity, grade ≥3 QTc prolongation, cardiac toxicity, and neurotoxicity during induction. Secondary endpoints included the evaluation of grade ≥3 neutropenia at time of induction evaluation bone marrow biopsy, time to last PRBC and PLT transfusions during induction, and evaluation of the ATO toxicity profile with respect to non-obese (BMI < 30 kg/m2) vs obese (BMI ≥30 kg/m2) patients and ATO dose < 0.15 mg/kg (defined as < 0.145 mg/kg) and ATO dose ≥0.15 mg/kg (defined as ≥0.145 mg/kg). The sec ondary analysis of ATO toxicity profile by obesity status was performed due to previous studies that found an increased incidence of toxicity in obese patients; the analysis by ABW-based dose (<or ≥0.15 mg/kg) was performed as patients ≤66.7 kg would receive doses of ATO of 0.15 mg/kg, but still be included in the ≤10 mg/dose cohort. Additional secondary endpoints included the evaluation of the induction cycle mean cumulative ATO dose in patients who did or did not experience toxicity. In consolidation cycles, a composite endpoint of neurotoxicity events was used to compare the incidence of neurotoxicity between the two groups. A cost analysis was performed by comparing the cost of treatment based on the mean weight for the study population vs capped 10 mg dose using the average wholesale price (AWP) reported from March 2020.
Statistical analysis
Baseline characteristics and endpoints were compared using the chi-squared and Fisher’s exact test for dichotomous variables. Continuous variables were compared using the Student’s t-test and Mann-Whitney U-tests. Multivariate logistic regression was performed to assess the effect of age, treatment center, obesity status, and dose cohort on the composite ATO toxicity endpoint. Statistical analysis was performed using IBM SPSS version 26.
Results
From August 2015 to December 2019, a total of 44 patients, 24 from JHH and 20 from UMMC, were eligible for inclusion. All patients prior to these time periods did not have sufficient data for analysis. Baseline characteristics are reported in Table 1. There were 15 patients in the ≤10 mg/dose cohort and 29 in the >10 mg/dose cohort. Seven patients in the ≤10 mg/dose cohort received doses ≤0.15 mg/kg. Patients in each cohort were similar, with the exception of median weight 66.4 kg vs 88.6 kg (p = 0.005), BMI 25.8 kg/m2 vs 31.6 kg/m2 (p = 0.008), and median ATO dose 10 mg vs 12 mg (p < 0.001). Additional baseline characteristics are compared between institutions in Table 2. Age, weight, and BMI were similar between institutions; median starting ATO dose, weight-based dose, and baseline AST and ALT were higher in UMMC patients.
Table 1.
Baseline characteristics by initial ATO dose received.
| Total (n = 44) | Dose < 10 mg (n = 15) | Dose > 10 mg (n = 29) | P-value* | |
|---|---|---|---|---|
| Institution, n | 24 JHH; 20 UMMC | 12 JHH; 3 UMMC | 12 JHH; 17 UMMC | 0.025 |
| Sex, % female | 50 | 40 | 55 | 0.526 |
| Age, years, mean (range; SD) | 50.6 (20–87; 16.7) | 47.9 (20–87; 15.7) | 52.0 (23–83; 17.3) | 0.447 |
| Weight, kg, median (range) | 87.7 (52.2–228.5) | 66.4 (52.2–149.5) | 88.6 (71.6–228.5) | 0.005 |
| BMI, kg/m2, median (range) | 30.2 (18.1–62) | 25.8 (18.1–54.8) | 31.6 (22.5–62.0) | 0.008 |
| WHO Obesity Class | Underweight: 1 | Underweight: 1 | Underweight: 0 | |
| Normal: 7 | Normal: 5 | Normal: 2 | ||
| Overweight: 13 | Overweight: 6 | Overweight: 7 | ||
| Obese: 12 | Obese: 0 | Obese: 12 | ||
| Obese II: 3 | Obese II: 0 | Obese II: 3 | ||
| Obese III: 8 | Obese III: 3 | Obese III: 5 | ||
| Starting ATO dose, mg, median (range) | 12 (7.8–20) | 10 (7.8–10) | 12 (11–20) | <0.001 |
| Dose, mg/kg, median (range) | 0.15 (0.067–0.157) | 0.149 (0.067–0.157) | 0.147 (0.082–0.157) | 0.586 |
| Baseline AST, IU/L, median (range) | 25 (11–155) | 25 (12–43) | 25 (11–155) | 0.951 |
| Baseline ALT, IU/L, median (range) | 26.5 (8–173) | 22 (8–48) | 33 (8–173) | 0.113 |
| Baseline TBILI, mg/dL, median (range) | 0.6 (0.2–3.2) | 0.6 (0.2–3.2) | 0.7 (0.2–1.5) | 0.363 |
| Baseline QTc, msec, median (range) | 415 (296–479) | 420 (381–479) | 415 (296–466) | 0.647 |
| Duration of induction, days, median (range) | 28 (3–67) | 29 (14–34) | 28 (3–67) | 0.154 |
| Cumulative ATO dose during induction mg, mean (range; SD) | 333.5 (140–540; 91.0) | 261.9 (140–320.1; 51.9) | 374.8 (240–540; 83.0) | <0.001 |
| Dose omissions for any reason, n | 6 | 4 | 2 | |
| Dose reductions for any reason, n | 3 | 2 | 1 | |
| CR after induction, n (%) | 41 (93.2) | 15 (100) | 26 (89.7) | 0.540 |
| Relapse, n | 0 | 0 | 0 | |
| Death, n | 3 | 0 | 3 |
JHH: The Johns Hopkins Hospital; UMMC: University of Maryland Medical Center; WHO: World Health Organization.
P-values generated with Mann Whitney U test (non-parametric) and t-test (normally distributed).
Table 2.
Baseline characteristics by treatment center.
| Total (n = 44) | Johns Hopkins Hospital (n = 24) | University of Maryland Medical Center (n = 20) | P-value* | |
|---|---|---|---|---|
| Sex, % female | 50 | 37.5 | 65 | 0.129 |
| Age, years, mean (range; SD) | 50.6 (20–87; 16.7) | 50.9 (20–83; 15.2) | 50.3 (23–87; 18.8) | 0.905 |
| Weight, kg, median (range) | 87.7 (52.2–228.5) | 85.2 (52.2–149.5) | 88.3 (61.7–228.5) | 0.962 |
| BMI, kg/m2, median (range) | 30.2 (18.1–62) | 27.5 (18.1–55.7) | 31.4 (22.5–62.0) | 0.311 |
| WHO Obesity Class | Underweight: 1 | Underweight: 1 | Underweight: 0 | |
| Normal: 7 | Normal: 4 | Normal: 3 | ||
| Overweight: 13 | Overweight: 10 | Overweight: 3 | ||
| Obese: 12 | Obese: 2 | Obese: 10 | ||
| Obese II: 3 | Obese II: 1 | Obese II: 2 | ||
| Obese III: 8 | Obese III: 6 | Obese III: 2 | ||
| Starting ATO dose, mg, median (range) | 12 (7.8–20) | 10.5 (7.8–17.0) | 13.0 (9.3–20.0) | 0.006 |
| Starting ATO dose > 10 mg, n (%) | 29 (65.9) | 12 (50.0) | 17 (85.0) | 0.025 |
| Dose, mg/kg, median (range) | 0.15 (0.067–0.157) | 0.144 (0.067–0.157) | 0.149 (0.088–0.157) | 0.015 |
| Baseline AST, IU/L, median (range) | 25 (11–155) | 21.5 (11–39) | 32.5 (15–155) | 0.003 |
| Baseline ALT, IU/L, median (range) | 26.5 (8–173) | 22.5 (8–47) | 35.5 (17–173) | 0.001 |
| Baseline TBILI, mg/dL, median (range) | 0.6 (0.2–3.2) | 0.6 (0.2–3.2) | 0.7 (0.2–1.5) | 0.222 |
| Baseline QTc, msec, median (range) | 415 (296–479) | 415 (381–453) | 416 (296–479) | 0.888 |
| Duration of induction, days, median (range) | 28 (3–67) | 28 (14–67) | 28 (3–34) | 0.953 |
| Cumulative ATO dose during induction mg, mean (range; SD) | 333.5 (140–540; 91.0) | 301.1 (140–514.4; 80.6) | 375.0 (240–540; 88.6) | 0.008 |
| Dose omissions for any reason, n | 6 | 4 | 2 | |
| Dose reductions for any reason, n | 3 | 2 | 1 | |
| CR after induction, n (%) | 41 (93.2) | 23 (95.8) | 18 (90) | 0.583 |
| Relapse, n | 0 | 0 | 0 | |
| Death, n | 3 | 1 | 2 |
WHO: World Health Organization.
P-values generated with Mann Whitney U test (non-parametric) and t-test (normally distributed).
The primary endpoint variables are reported in Table 3. There were no differences in hepatotoxicity outcomes. Cardiac toxicity occurred in one (6.7%) ≤10 mg/dose (atrial fibrillation) and three (10.3%) >10 mg/dose patients (all grade ≥3: pericarditis, STEMI, and arrhythmia with cardiac arrest) (p = 1.00). Grade ≥3 QTc prolongation occurred in one patient in the ≤10 mg/dose cohort and in two patients in the >10 mg/dose cohort; in all three patients, no concomitant QTc prolonging medications were administered within 24 hours of the documented grade ≥QTc prolongation. Neurotoxicity occurred in two (13.3%) ≤10 mg/dose and six (20.7%) >10 mg/dose patients (p = 0.695). Grade ≥3 neurotoxicity occurred in three patients (severe neuropathy in one and encephalopathy in two patients), one in the ≤10 mg/dose cohort and two in the >10 mg/dose cohort. There were no differences myelosuppression outcomes. The composite endpoint occurred in eight (53.3%) ≤10 mg/dose and 13 (44.8%) >10 mg/dose patients (p = 0.752). The multivariate logistic regression found no effect of age, treatment center, obesity status, or dose cohort on the composite ATO toxicity endpoint. There were three deaths during induction, all patients received >10 mg/dose of ATO. Two patients died of complications of intracerebral hemorrhage. One patient died of complications of diffuse alveolar hemorrhage. At a median follow-up of 27 months, there were no relapses in either cohort.
Table 3.
Induction toxicity outcomes by initial ATO dose received.
| Total (n = 44) | Dose < 10 mg (n = 15) | Dose > 10 mg (n = 29) | P-value | |
|---|---|---|---|---|
| Hepatotoxicity | ||||
| Maximum AST, IU/L, median (range) | 120.5 (35–1273) | 163 (59–381) | 103 (35–1273) | 0.131 |
| Maximum ALT, IU/L, median (range) | 129.5 (24–1067) | 170 (36–481) | 114 (24–1067) | 0.151 |
| Maximum TBILI, mg/dL, median (range) | 0.95 (0.4–5.9) | 0.9 (0.4–3.5) | 1.0 (0.4–5.9) | 0.442 |
| Grade ≥3 hepatotoxicity | 15 (34.1) | 7 (46.7) | 8 (27.6) | 0.315 |
| Grade 1 hepatotoxicity | 15 (34.1) | 4 (26.7) | 11 (37.9) | |
| Grade 2 hepatotoxicity | 12 (27.3) | 4 (26.7) | 8 (27.6) | |
| Grade 3 hepatotoxicity | 13 (29.5) | 7 (46.7) | 6 (20.7) | |
| Grade 4 hepatotoxicity | 2 (4.5) | 0 | 2 (6.9) | |
| Time to hepatotoxicity, days, mean (range; SD) | 9.4 (0–21; 5.5) | 9.3 (3–21; 5.3) | 9.5 (0–21; 5.7) | 0.932 |
| Cardiac toxicity | ||||
| Maximum QTc, msec, median (range) | 445 (391–613) | 447 (418–613) | 442 (391–506) | 0.795 |
| Max change in QTc, msec, median (range) | 29.5 (‒16.2–195.2) | 29.8 (‒16.2–192.6) | 29.1 (0–125.5) | 0.701 |
| Grade ≥3 QTc prolongation | 3 (6.8) | 1 (6.7) | 2 (6.9) | 0.736 |
| Time to max QTc, days, median (range) | 13 (0–26) | 16 (7–26) | 12 (0–26) | 0.143 |
| CV toxicity during induction | 4 (9.1) | 1 (6.7) | 3 (10.3) | 1.00 |
| Grade >3 CV toxicity | 3 | 0 | 3 (10.3) | 0.540 |
| Time to CV toxicity, days, median (range) | 12.5 (3–21) | 21 | 4 (3–21) | 0.346 |
| Neurotoxicity | ||||
| Neurotoxicity | 8 (18.2) | 2 (13.3) | 6 (20.7) | 0.695 |
| Grade ≥3 neurotoxicity | 3 (6.8) | 1 (6.7) | 2 (6.9) | 1.00 |
| Time to neurotoxicity, days, mean (range; SD) | 21.6 (3–36; 11.8) | 12 (8–16; 5.7) | 24.8 (3–36; 11.8) | 0.106 |
| Myelosuppression | ||||
| ANC at bone marrow biopsy, cells/microL, median (range) | 800 (20–4000) | 800 (50–3328) | 885 (20–4000) | 0.425 |
| Grade ≥3 neutropenia at bone marrow biopsy | n = 41* 23 (56.1) |
n = 15 9 (60.0) |
n = 26* 14 (53.8) |
0.754 |
| Time from induction to last platelet transfusion, days, median (range) | 16 (0–30) | 19 (0–30) | 17.5 (0–28) | 0.289 |
| Time from induction to last RBC transfusion, days, median (range) | 23 (1–40) | 22 (1–31) | 23 (1–40) | 0.750 |
| Composite toxicity | ||||
| Grade >3 hepatotoxicity or QTc prolongation, neurotoxicity, or cardiac toxicity | 21 (47.7) | 8 (53.3) | 13 (44.8) | 0.752 |
Values are n, (%) unless otherwise noted.
Three patients with mortality during induction were not eligible for analysis.
Characteristics based on non-obese (n = 21) obese (n = 23) and status are reported in Table 4. There were no differences in toxicity outcomes in these groups. Patients categorized by ABW-based dose (<0.15 mg/kg vs ≥0.15 mg/kg) did not have differences in the toxicity outcomes evaluated (Table 4). Maximum changes in QTc were 19.4 msec and 33.5 msec (p = 0.082); grade ≥3 QTc prolongation occurred in zero and three (10.0%) patients respectively (p = 0.540). During induction, patients who experienced neurotoxicity and grade ≥3 QTc prolongation received more cumulative ATO than patients without these toxicities. There was no difference in cumulative ATO dose in cardiac toxicity, hepatotoxicity, or myelosuppression outcomes.
Table 4.
Induction toxicity outcomes by obesity status and actual body weight initial ATO dose received.
| Non-obese (n = 21) | Obese (n = 23) | P-value | <0.l5mg/kg (n = 14) | >0.15 mg/kg (n = 30) | P-value | |
|---|---|---|---|---|---|---|
| Hepatotoxicity | ||||||
| Maximum AST, IU/L, median (range) | 123.0 (48–672) | 108.0 (35–1273) | 0.630 | 95.5 (35–1207) | 127.5 (48–1273) | 0.147 |
| Maximum ALT, IU/L, median (range) | 140.0 (51–442) | 112.0 (24–1067) | 0.312 | 113.0 (24–481) | 169.5 (44–1067) | 0.162 |
| Maximum TBILI, mg/dL, median (range) | 0.9 (0.4–5.9) | 1.0 (0.4–2.6) | 0.613 | 1.0 (0.4–3.5) | 1.0 (0.4–5.9) | 0.801 |
| Time to hepatotoxicity, days, mean (range; SD) | 10.1 (1–21; 6.3) | 8.8 (0–19; 4.7) | 0.426 | 9.9 (3–19; 4.7) | 9.2 (0–21; 5.9) | 0.663 |
| Grade ≥3 hepatotoxicity | 8 (38.1) | 7 (30.4) | 0.597 | 4 (28.6) | 11 (36.7) | 0.738 |
| Grade 1 hepatotoxicity | 7 (33.3) | 8 (34.8) | 6 (42.9) | 9 (30.0) | ||
| Grade 2 hepatotoxicity | 5 (23.8) | 7 (30.4) | 3 (21.4) | 9 (30.0) | ||
| Grade 3 hepatotoxicity | 8 (38.1) | 5 (21.7) | 3 (21.4) | 10 (33.3) | ||
| Grade 4 hepatotoxicity | 0 | 2 (8.7) | 1 (7.1) | 1 (3.3) | ||
| Cardiac toxicity | ||||||
| Maximum QTc, msec, median (range) | 447.2 (391–613) | 442.0 (402–506) | 0.842 | 438.6 (402–472) | 451.5 (391–613) | 0.124 |
| Time to max QTc, days, median (range) | 14 (0–26) | 12 (0–26) | 0.663 | 11 (0–26) | 13 (0–26) | 0.791 |
| Max change in QTc, msec, median (range) | 34.6 (0–192.6) | 23.9 (‒16.2–125.5) | 0.605 | 19.4 (‒16.2–69.2) | 33.5 (0–192.6) | 0.082 |
| Grade ≥3 QTc prolongation | 1 (4.8) | 2 (8.7) | 0.609 | 0 | 3 (10.0) | 0.540 |
| CV toxicity during induction | 2 (9.5) | 2 (8.7) | 0.924 | 1 (7.1) | 3 (10.0) | 1.00 |
| Time to CV toxicity, days, median (range) | 12 (3–21) | 12.5 (4–21) | 0.683 | 21 | 4 (3–21) | 0.346 |
| Neurotoxicity | ||||||
| Time to neurotoxicity, days, mean (range; SD) | 12 (8–16; 5.7) | 24.8 (3–36; 11.8) | 0.106 | 28 | 20.7 (3–36; 12.4) | 0.603 |
| Neurotoxicity | 2 (9.5) | 6 (26.1) | 0.245 | 1 (7.1) | 7 (23.3) | 0.402 |
| Grade ≥3 neurotoxicity | 1 (4.8) | 2 (8.7) | 1.00 | 0 | 3 (10.0) | 0.540 |
| Myelosuppression | ||||||
| ANC at bone marrow biopsy, cells/microL, median (range) | 740 (20–3328) | 970 (110–4000) | 0.575 | 970 (200–2570) | 770 (20–4000) | 0.377 |
| Grade ≥3 neutropenia at bone marrow biopsy | n = 20* l2 (60.0) |
n = 21* 11 (47.8) |
0.756 | n = 13* 7 (53.8) |
n = 28* 16 (57.1) |
1.00 |
| Time from induction to last platelet transfusion, days, median (range) | 7 (35.0) | 7 (33.3) | 1.00 | 18.5 (6–30) | 17 (1–28) | 0.360 |
| Time from induction to last RBC transfusion, days, median (range) | 16 (0–30) | 18 (0–28) | 0.856 | 23 (12–31) | 23 (1–40) | 0.696 |
| Composite toxicity | ||||||
| Grade > 3 hepatotoxicity or QTc prolongation, neurotoxicity, or cardiac toxicity | 9 (42.9) | 12 (52.2) | 0.563 | 5 (35.7) | 16 (53.3) | 0.342 |
Values are n, (%) unless otherwise noted.
Three patients with mortality during induction were not eligible for analysis.
A total of 29, 26, 25, and 24 patients had data available for consolidation cycles one, two, three, and four, respectively (Table 5). However, there were 30 patients eligible for evaluation of composite neurotoxicity (one patient in the >10 mg dose cohort did not have data from cycle one, but subsequently had data for assessment in future cycles). The incidence of neurotoxicity increased from cycles one through four in the >10 mg dose cohort. The composite incidence of neurotoxicity among patients receiving ≤10 mg/dose was 2 (22.2%) compared to 14 (66.7%) in the >10 mg/dose cohort (p = 0.046). No patients had reported hepatotoxicity or cardiac toxicity leading to ATO dose reductions/omissions. Dose reductions and omissions in consolidation cycles by dosing group are characterized further in Table 6.
Table 5.
Neurotoxicity incidence in consolidation.
| Consolidation cycle | Dose < 10 mg | Dose > 10 mg | P-value |
|---|---|---|---|
| Cycle One (n = 29) | 2 (22.2) | 7 (35.0) | |
| Cycle Two (n = 26) | 2 (25) | 8 (44.4) | |
| Cycle Three (n = 25) | 1 (12.5) | 9 (52.9) | |
| Cycle Four (n = 24) | 1 (12.5) | 11 (68.8) | |
| All Cycles (n = 30) | 2 (22.2) | 14 (66.7) | 0.046 |
Values are n, (%).
Table 6.
Dose reductions/omissions in consolidation.
| ATO consolidation | Dose reductions/omissions | |
|---|---|---|
| Dose < 10 mg | Dose > 10 mg | |
| Cycle One (n = 29) | 0 | 1 (admitted for neutropenic fever) |
| Cycle Two (n = 26) | 0 | 2 (neutropenia) |
| Cycle Three (n = 25) | 1 (neutropenia) | 2 (neutropenia, neuropathy) |
| Cycle Four (n = 24) | 1 (neutropenia, previous tolerability) | 3 (neutropenia, severe headache/vision changes) |
The estimated cost for a 10 mg/10 mL vial of ATO, based on AWP, is $712.80. With a mean initial weight of 87.7 kg in the study population, the recommended weight-based dose for ATO is 13.2 mg. The cumulative dose received for a 28-day induction and four cycles of consolidation is 1425.6 mg ($101616.77) with this approach, compared to 1080 mg ($76982.40) for dose-capped patients. The relative cost difference is $24634.37 per patient. This does not account for the cost of the partially used second vial, which is not stable for later use, and may double the cost relative to capped-dose patients.
Discussion
This study describes the association of ATO dose capping to 10 mg with ATO toxicity in a Mid-Atlantic United States population. The composite outcome of grade ≥3 hepatotoxicity, grade ≥3 QTc prolongation, cardiac toxicity and neurotoxicity during induction was not different between patients capped at or receiving ≤10 mg/dose and those receiving >10 mg/dose. However, in consolidation, patients who received ≤10 mg of ATO per dose had significantly less neurotoxicity than those receiving >10 mg (12.5% vs 68.8%; p = 0.027). Importantly, the cumulative incidence of relapse was zero in both groups after a median follow-up of 27 months, thus capping doses to 10 mg does not appear to compromise the effectiveness of this regimen. The study by Lo-Coco et al reported rates of Grade ≥3 hepatotoxicity, QTc prolongation, cardiac toxicity numerically comparable to this population, with the exception of neurotoxicity, which was not characterized in the Lo-Coco et al study.3 While the incidence of composite toxicity during induction was not different between the dosing cohorts in this study population, the current study may be underpowered to detect differences in outcomes due to the low incidence and many confounding variables that contribute to these complications during treatment of newly-diagnosed APL. This is evidenced by several secondary endpoint findings. When toxicity profile was analyzed by weight-based dose received (<0.15 mg/kg or 0.15 mg/kg) there was a nominally higher frequency of grade ≥3 hepatotoxicity, grade ≥3 QTc prolongation, and greater maximum change in QTc from baseline in patients receiving ≥0.15 mg/kg. Furthermore, the mean cumulative ATO dose received during induction was higher among patients who experienced neurotoxicity (395.1 mg vs. 318.6 mg; p = 0.031) and grade ≥3 QTc prolongation, (437.3 mg vs 325.3 mg p = 0.039).
Patients in the >10 mg/dose cohort experienced more neurotoxicity by consolidation cycle four compared to the patients receiving ≤10 mg ATO per dose (68.8% vs 12.5%; p = 0.027). There was also an increase in the incidence of neurotoxicity in the >10 mg/dose patients with each subsequent consolidation cycles. These findings suggest cumulative ATO doses may have an association with neurotoxicity. For example in the >10 mg/dose cohort, the incidence of neurotoxicity was 20.7% during induction, 35.0% in consolidation one, and reached 68.8% by consolidation four. This observation is similar to known properties of other neurotoxic chemotherapy agents (e.g., platinum agents, taxanes, vinca alkaloids, proteasome inhibitors).14–18 This is also supported by a study of relapsed APL patients who received ATO at 0.08 mg/kg vs 0.16 mg/kg, with similar rates of efficacy but reduced toxicity in the lower dose cohort.19 As such, limiting cumulative ATO received by capping doses may significantly reduce the incidence of, and potentially severity of, neurologic complications; these complications may compromise the ability to complete this curative therapy and are also known to significantly impact survivor quality of life.15
There were no statistically significant differences in toxicity seen between patients stratified by obesity status, as compared to previous literature which identified that obese patients were more likely to have ATO doses held due to adverse events than non-obese patients.9 The previous study also utilized the WHO class II obesity (BMI ≥35 kg/m2) and class III (BMI ≥40 kg/m2) for analysis, as compared to the class I obesity (BMI ≥30 kg/m2) definition used in this analysis. In the current study, three obese patients received capped doses to 10 mg, which may have balanced the increased risk of toxicity in obese patients that has been previously reported.7–9
As ATO is priced per vial, the non-capped dose group had higher cost of treatment than patients capped to the vial size. The cost savings for an entire treatment course was calculated as $24634.37/patient using the mean weight of the population. This largely underestimates the overall cost savings since unused ATO must be discarded due to stability.5 Capping ATO doses to the vial prevents wasting unused drug left in a partially used second vial, potentially saving twice the treatment cost ($153964.80 vs. $76982.40 per patient for the entire regimen). Though not analyzed in this study, further cost benefits may include a reduction in the cost of treatment-related complications.
The findings of this study cannot be applied to the ATO dosing utilized in the AML17 trial, in which induction starts with ATO 0.3 mg/kg days 1–5 in week 1, then 0.25 mg/kg twice weekly for weeks 2–8.20 Subsequent consolidation cycles include similar dosing of 0.3 mg/kg day 1–5 in week 1, followed by 0.25 mg/kg twice weekly for weeks 2–4. This regimen offers significantly less cumulative ATO than the Lo-Coco et al regimen during induction (5 mg/kg vs 9 mg/kg for a full 60-day induction). The cumulative ATO dose in consolidation is the same (12 mg/kg vs 12 mg/kg for four cycles of consolidation). The AML17 trial reported a lower incidence of QTc prolongation and transaminitis than the Lo-Coco et al trial, with similar efficacy outcomes. Unfortunately, neither trial consistently reported neurologic toxicity outcomes. Using the presented capped dosing strategy for the Lo-Coco regimen, patients up to a weight of 120 kg would still receive 5 mg/kg of ATO for a full 60-day induction. These comparisons suggest there may be multiple opportunities to optimize ATO dosing and reduce toxicity. Further dosing strategies should also harness the pharmacokinetics/pharmacodynamics of ATO, such as in the AML17 trial, which administered ATO twice weekly after an initial loading period. This reduces chemotherapy infusion chair time and promotes patient adherence and quality of life.
Limitations of this study include its retrospective nature, which may underreport the incidence of non-laboratory toxicities documented in the electronic medical record, such as cardiac and neurological toxicities. Moreover, while this study did not find relapses in either dosing cohort, a larger sample with longer follow-up would be needed to investigate this question in a patient population with very high rates of long-term disease-free survival. Additionally, this study did not analyze long-term ATO toxicities, which might include hepatic disorders, cardiovascular events, cognitive impairment, and/or secondary malignancies.4 It is unclear if patients who receive a higher cumulative dose of ATO during their treatment are more likely to have long-term ATO toxicities.
Conclusions
In this study, which compared the toxicity profile of ATO between patients receiving ≤10 mg/dose and patients receiving >10 mg/dose, patients with dose capping to 10 mg had a lower incidence of neurotoxicity during consolidation. The ATO toxicity profile during induction was not statistically significantly different between groups. Secondary analysis showed a potential relationship with higher cumulative ATO doses in patients who experienced neurotoxicity and grade ≥3 QTc prolongation during induction, suggesting that more significant differences in toxicity may be realized with a larger sample size of patients receiving doses capped at 10 mg. The cost of treatment was reduced by $24634.37 per patient on average by dose capping, but this underestimates the cost savings from avoidance of partially-used vials. Additionally, all patients who received doses capped to 10 mg achieved complete remission after induction, and at a median follow-up of 27 months, no patients have relapsed using this dosing strategy. These findings suggest that capped ATO doses may reduce both ATO and financial toxicity, and represents an opportunity to further optimize this highly curative regimen through larger studies.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: G.G. is supported by the National Institutes of Health, National Cancer Institute (P01-CA225618, P30-CA00793).
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References
- 1.National Comprehensive Cancer Network. Acute myeloid leukemia (Version 3.2019), https://www.nccn.org/professionals/physician_gls/pdf/aml.pdf (2019, accessed 29 January 2020).
- 2.Lo-Coco F, Avvisati G, Vignetti M, et al. Retinoic acid and arsenic trioxide for acute promyelocytic leukemia. N Engl J Med 2013; 369: 111–121. [DOI] [PubMed] [Google Scholar]
- 3.Platzbecker U, Avvisati G, Cicconi L, et al. Improved outcomes with retinoic acid and arsenic trioxide compared with retinoic acid and chemotherapy in non–high-risk acute promyelocytic leukemia: final results of the randomized Italian-German APL0406 trial. J Clin Oncol 2017; 35: 605–612. [DOI] [PubMed] [Google Scholar]
- 4.Zhu H, Hu J, Chen L, et al. The 12-year follow-up of survival, chronic adverse effects, and retention of arsenic in patients with acute promyelocytic leukemia. Blood 2016; 128: 1525–1528. [DOI] [PubMed] [Google Scholar]
- 5.Cell Therapeutics, Inc. Trisenox (package insert). Seattle, WA: Cell Therapeutics, Inc., 2000. [Google Scholar]
- 6.Griggs JJ, Mangu PB, Anderson H, et al. Appropriate chemotherapy dosing for obese adult patients with cancer: American Society of Clinical Oncology Clinical Practice guideline. J Clin Oncol 2012; 30: 1553–1561. [DOI] [PubMed] [Google Scholar]
- 7.Westervelt P, Brown RA, Adkins DR, et al. Sudden death among patients with acute promyelocytic leukemia treated with arsenic trioxide. Blood 2001; 98: 266–271. [DOI] [PubMed] [Google Scholar]
- 8.Fox E, Razzouk BI, Widemann BC, et al. Phase 1 trial and pharmacokinetic study of arsenic trioxide in children and adolescents with refractory or relapsed acute leukemia, including acute promyelocytic leukemia or lymphoma. Blood 2008; 111: 566–573. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hickey E, Clemons B, Griffin S, et al. Multicenter evaluation of arsenic trioxide dosing in obese patients with low–intermediate risk acute promyelocytic leukemia. Leuk Lymphoma 2019; 60: 3557–3560. [DOI] [PubMed] [Google Scholar]
- 10.Hosseini R, Mandegary A, Alimoghadd K, et al. Pharmacokinetic of arsenic trioxide in newly diagnosed acute promyelocytic leukemia patients. J Appl Sci 2008; 8: 4617–4623. [Google Scholar]
- 11.Tedesco J, Qualtieri J, Head D, et al. High prevalence of obesity in acute promyelocytic leukemia (APL): implications for differentiating agents in APL and metabolic syndrome. Ther Adv Hematol 2011; 2: 141–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Castillo JJ, Mulkey F, Geyer S, et al. Relationship between obesity and clinical outcome in adults with acute myeloid leukemia: a pooled analysis from four CALGB (alliance) clinical trials: obesity APL AML survival. Am J Hematol 2016; 91: 199–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Sanz MA, Lo-Coco F, Martin G, et al. Definition of relapse risk and role of nonanthracycline drugs for consolidation in patients with acute promyelocytic leukemia: a joint study of the PETHEMA and GIMEMA cooperative groups. Blood 2000; 96: 1247–1253. [PubMed] [Google Scholar]
- 14.Saif MW and Reardon J. Management of oxaliplatin-induced peripheral neuropathy. Ther Clin Risk Manag 2005; 1: 249–258. [PMC free article] [PubMed] [Google Scholar]
- 15.Tay CG, Mun Lee VW, Ong LC, et al. Vincristine-induced peripheral neuropathy in survivors of childhood acute lymphoblastic leukaemia. Pediatr Blood Cancer 2017; 64: e26471. [DOI] [PubMed] [Google Scholar]
- 16.Lavoie Smith EM, Lang L, Chiang C, et al. Patterns and severity of vincristine-induced peripheral neuropathy in children with acute lymphoblastic leukemia. J Peripher Nerv Syst 2015; 20: 37–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.van Gerven JM, Moll JW, van den Bent MJ, et al. Paclitaxel (taxol) induces cumulative mild neurotoxicity. Eur J Cancer 1994; 30A: 1074–1077. [DOI] [PubMed] [Google Scholar]
- 18.Zhao W, Wang W, Li X, et al. Peripheral neuropathy following bortezomib therapy in multiple myeloma patients: association with cumulative dose, heparanase, and TNF-a. Ann Hematol 2019; 98: 2793–2803. [DOI] [PubMed] [Google Scholar]
- 19.Shen Y, Shen Z-X, Yan H, et al. Studies on the clinical efficacy and pharmacokinetics of low-dose arsenic trioxide in the treatment of relapsed acute promyelocytic leukemia: a comparison with conventional dosage. Leukemia 2001; 15: 735–741. [DOI] [PubMed] [Google Scholar]
- 20.Burnett AK, Russel NH, Hills RK, et al. Arsenic trioxide and all-trans retinoic acid treatment for acute promyelocytic leukaemia in all risk groups (AML17): results of a randomised, controlled, phase 3 trial. Lancet Oncol 2015; 16: 1295–1305. [DOI] [PubMed] [Google Scholar]