Abstract
Background:
Treatment of infants with acute leukemia remains challenging, especially for acute lymphocytic leukemia (ALL). Infants have shown markedly higher rates of induction mortality compared with noninfants. There are limited data on presentation acuity and supportive care utilization in this age group.
Methods:
In retrospective analyses of patients treated for new onset ALL or acute myeloid leukemia (AML) at pediatric hospitals contributing to the Pediatric Health Information System, we compared presentation acuity, induction mortality, and resource utilization in infants relative to noninfants less than 10 years at diagnosis.
Results:
Analyses included 10 359 children with ALL (405 infants, 9954 noninfants) and 871 AML (189 infants, 682 noninfants). Infants were more likely to present with multisystem organ failure compared to noninfants for both ALL (12% and 1%, PR = 10.8, 95% CI: 7.4, 15.7) and AML (6% vs. 3%; PR = 2.0, 95% CI: 1.0, 3.7). Infants with ALL had higher induction mortality compared to noninfants, even after accounting for differences in anthracycline exposure and presentation acuity (2.7% vs. 0.5%, HR = 2.1, 95% CI: 1.0, 4.8). Conversely, infants and noninfants with AML had similar rates of induction mortality (3.2% vs. 2.1%, HR = 1.2, 95% CI: 0.3, 3.9), which were comparable to rates among infants with ALL. Infants with ALL and AML had greater requirements for blood products, diuretics, supplemental oxygen, and ventilation during induction relative to noninfants.
Conclusions:
Infants with leukemia present with higher acuity compared with noninfants. Induction mortality and supportive care requirements for infants with ALL were similar to all children with AML, and significantly higher than those for noninfants with ALL.
Keywords: induction mortality, infant leukemia, resource utilization
1 |. INTRODUCTION
Despite improvements in outcomes of childhood acute lymphocytic leukemia (ALL) and acute myeloid leukemia (AML) over the past few decades, treatment of infants with acute leukemia remains challenging, particularly in infants with ALL.1–8 Infants with ALL present with greater tumor burden and high-risk biological features1–9 that may require intensified chemotherapy protocols compared with noninfants with ALL. However, the historical challenge in treatment of infant ALL has been balancing the risk for early relapse with the severe toxicities associated with intensification of treatment, which could result in early mortality.8–18 Several reports have focused on targeting biological features and modifications in treatment protocols for infants with ALL. Aside from data reporting increased infectious complications in ALL, little is known about supportive care requirements in the infant ALL population.14–17,19,20
Unlike in ALL, infants with AML are treated with the same intensive regimens as older children, yet there are conflicting data on the tolerability of these regimens in infants compared to their older counterparts. Several studies have shown higher induction mortality for AML infants.21,22 Children’s Oncology Group (COG) trials and legacy Berlin Franklin Munster (BFM) studies suggest that infants with AML have higher treatment-related toxicities compared to noninfants,23–26 but their survival is similar to noninfants.23
To improve the outcomes for infants with leukemia, it is necessary to have a comprehensive understanding of the treatment-related toxicities and supportive care utilization, particularly as they relate to induction mortality. We sought to demonstrate the difference in induction mortality in infants compared to noninfants less than 10 years of age and to evaluate acuity at presentation as a potential mechanism of previously described differences in induction mortality. We also compared inpatient resource utilization (RU) by age and disease.
2 |. METHODS
2.1 |. Study design and data source
We performed a retrospective cohort study of children treated for new onset ALL and AML at children’s hospitals in the United States contributing to the Pediatric Health Information System (PHIS). PHIS is an administrative database that includes inpatient billing data from more than 40 tertiary children’s hospitals. The database consists of patient demographics, dates of service, discharge disposition, and International Classification of Diseases, Ninth Revision (ICD-9) diagnosis codes, as well as daily billing data for medications, laboratory tests, procedures, and other clinical resources. Data are anonymized at the time of submission and subject to a number of reliability and validity checks before inclusion in the database. Data quality is assured through a joint effort between the Children’s Hospital Association and participating hospitals.
2.2 |. Study population
The current study population was derived from cohorts of children who were treated at PHIS-contributing institutions for new onset ALL and AML between January 1999 and December 2015.27,28 Inclusion in these cohorts required an ICD-9 code for the given malignancy and documentation of initiation of induction chemotherapy based on manual review of the PHIS chemotherapy billing data.27,28 Patients with Down syndrome (DS) and those older than 10 years at acute leukemia diagnosis were excluded.
2.3 |. Exposures
The primary exposure of interest was patient age at diagnosis of acute leukemia. ALL and AML patients were classified based on age at diagnosis as either infants (<1 year of age) or noninfants (1–10 years of age). Secondary analyses compared infants by disease group, ALL versus AML.
2.4 |. Outcomes
The outcomes of interest included acuity at presentation, inpatient mortality during the first 35 days following the start of initial chemotherapy (induction mortality), cumulative number of inpatient days, and RU including intensive care unit (ICU) level resources. Patients were followed for the duration of their initial course of chemotherapy (i.e., induction). Specifically, follow-up started on the first day of induction chemotherapy and continued until the earliest of the following: 35 days after initiation of chemotherapy, start of the subsequent course of chemotherapy, or death.
Acuity at presentation was defined as any ICU-level resource requirement within the first 72 hours following initial leukemia admission. ICU-level resources overall and by organ system (i.e., cardiac, respiratory, renal, neurologic, and leukapheresis) were defined by procedure codes or clinical RU considered a priori as a marker of ICU-level care. Multisystem organ failure (MSOF) was defined as ICU-level care requirements involving two or more organ systems. Inpatient deaths were identified based on disposition at discharge, which is coded for each hospitalization. Daily billing data were used to determine the utilization of each of the following clinical resources: antimicrobials, blood products, analgesics, antiemetics, total parenteral nutrition (TPN), antihypertensives, diuretics, vasopressors, supplemental oxygen, ventilation, extracorporeal membrane oxygenation (ECMO), granulocyte colony-stimulating factor (GCSF), and dialysis. Rates of inpatient RU were computed as days of use per 100 inpatient days.
2.5 |. Covariates
Patient characteristics including age at diagnosis, sex, race, ethnicity, insurance status, and anthracycline exposure were ascertained from PHIS. Acuity at presentation was also considered in comparisons of induction mortality, RU, and cumulative length of stay to control for potential confounding by disease severity.
2.6 |. Statistical analyses
Patient characteristics were presented using descriptive statistics and were compared by age and disease groups using chi-square tests. Prevalence of acuity at presentation (overall and by organ system) and MSOF were compared by age and disease group using log-binomial regression. Generalized estimating equations with an exchangeable correlation matrix were used to obtain robust variance estimates to account for nonindependence of observations from the same institution. Induction mortality was compared using Cox proportional hazards regression. Cumulative numbers of inpatient days were summarized by age and disease group, and compared using linear regression models. Multivariable log-binomial regression methods were used to estimate adjusted prevalence ratios comparing the prevalence of use of each resource defined above by age group. RU rates (per 100 inpatient days) among those exposed were reported for each and were compared using Poisson regression, with inpatient days as offset and a Pearson scale adjustment to correct for possible overdispersion.
3 |. RESULTS
3.1 |. Patient characteristics
The established PHIS cohorts included 14 651 patients with newly diagnosed ALL and 1694 patients with newly diagnosed AML initiating induction chemotherapy. Of these, 10 359 ALL (71%) and 871 AML (51%) patients were 10 years of age or younger at diagnosis, with no documentation of DS and constituted the analytic study population.
Characteristics of the study population are presented in Table 1. Infants comprised 4% (n = 405) of ALL patients and 22% (n = 189) of AML patients in the study population. Infants with ALL were more likely to be female (52% vs. 46%), Black (11% vs. 7.0%), and to have received anthracycline in induction (97% vs. 28%), but had comparable distributions of ethnicity, insurance, and region compared with noninfants aged 1–10 years with ALL. Among patients with AML, there were no significant differences in the distributions of gender, race, ethnicity, insurance, or region between the infant and noninfant patients. All AML patients received anthracycline in induction. Infants with ALL were more likely to be documented as Hispanic (26% vs. 18%) and slightly less likely to be documented as either White or Black race (80% vs. 85%) than those with AML.
TABLE 1.
Demographic characteristics for infant and noninfant (1–10 years) pediatric patients with ALL and AML
| Patients with ALL (N = 10 359) | Patients with AML (N = 871) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Overall, N | <1 year, n (%) | 1–10 years, n (%) | p-Valuea | Overall, N | <1 year, n (%) | 1–10 years, n (%) | p-Valuea | p-Valueb | |
| Total number of patients | 10 359 | 405 (3.9) | 9954 (96.1) | 871 | 189 (21.7) | 682 (78.3) | |||
| Gender | .02 | .831 | .598 | ||||||
| Male | 5572 (53.8) | 194 (48.0) | 5278 (54.0) | 453 (52.0) | 97 (51.3) | 356 (52.2) | |||
| Female | 4784 (46.2) | 210 (52.0) | 4574 (46.0) | 418 (48.0) | 92 (48.7) | 326 (47.8) | |||
| Race | .01 | .251 | .029 | ||||||
| White | 7563 (75.1) | 278 (69.8) | 7285 (75.3) | 595 (70.7) | 133 (70.4) | 462 (67.7) | |||
| Black | 741 (7.4) | 44 (11.1) | 697 (7.2) | 110 (13.1) | 25 (13.2) | 85 (12.5) | |||
| Other races | 1765 (17.5) | 76 (19.1) | 1689 (17.5) | 137 (16.3) | 22 (11.6) | 115 (16.9) | |||
| Ethnicity | .20 | .106 | <.0001 | ||||||
| Hispanic/Latino | 2508 (24.2) | 105 (25.9) | 2403 (24.1) | 190 (21.8) | 33 (17.5) | 157 (23.0) | |||
| Non-Hispanic/Latino | 2999 (29) | 101 (24.9) | 2898 (29.1) | 357 (41.0) | 89 (47.1) | 268 (39.3) | |||
| Unknown | 4852 (46.8) | 199 (49.1) | 4653 (46.8) | 324 (37.2) | 67 (35.5) | 257 (37.7) | |||
| Insurance | .60 | .439 | .890 | ||||||
| Private | 4292 (41.8) | 165 (41) | 4127 (41.8) | 349 (40.8) | 76 (40.2) | 273 (40.0) | |||
| Public | 4284 (41.7) | 177 (43.9) | 4107 (41.6) | 397 (46.4) | 82 (43.4) | 315 (46.2) | |||
| Other payers | 1691 (16.5) | 61 (15.1) | 1630 (16.5) | 110 (12.9) | 29 (15.3) | 81 (11.9) | |||
| Region | .70 | .562 | .992 | ||||||
| Midwest | 2705 (26.1) | 99 (24.4) | 2606 (26.2) | 238 (27.3) | 46 (24.3) | 192 (28.1) | |||
| Northeast | 1110 (10.7) | 45 (11.1) | 1065 (10.7) | 84 (9.6) | 22 (11.6) | 62 (9.1) | |||
| South | 3468 (33.5) | 146 (36.1) | 3322 (33.4) | 289 (33.2) | 66 (34.9) | 223 (32.7) | |||
| West | 3076 (29.7) | 115 (28.4) | 2961 (29.7) | 260 (29.9) | 55 (29.1) | 205 (30.1) | |||
| Anthracycline in induction | <.001 | 1.00 | .007 | ||||||
| Yes | 3208 (31.0) | 391 (96.5) | 2817 (28.3) | 871 (100) | 189 (100) | 682 (100) | |||
| No | 7151 (69.0) | 14 (3.5) | 7137 (71.7) | 0 (0) | 0 (0) | 0 (0) | |||
Abbreviations: ALL, acute lymphoid leukemia; AML, acute myeloid leukemia.
p-Value for the comparison of demographic distributions by age group (infant vs. noninfant), separately for ALL and AML.
p-Value for the comparison of demographic distributions among infants by disease group (i.e., ALL vs. AML).
3.2 |. Acuity at presentation
Table 2 presents the comparisons of acuity at presentation. Infants with ALL (25% vs. 4%, PR: 6.7, 95% CI: 5.3, 8.6; p < .001) and AML (16% vs. 8%, PR: 2.0, 95% CI: 1.3, 3.0; p = .001) were significantly more likely to present with high acuity compared to noninfants. Trends toward greater ICU-level care requirements among infants compared with noninfants were observed for each organ system in both ALL and AML, though not all of the differences among the latter group were statistically significant. MSOF at presentation was also more prevalent in infants with ALL (12% vs. 1%, PR: 10.8, 95% CI: 7.4, 15.7; p < .001) and AML (6% vs. 3%, PR: 1.7, 95% CI: 0.95, 3.0; p = .074) compared to noninfants. Larger proportions of infants with ALL required ICU-level care (25% vs. 16%, PR: 1.9, 95% CI: 1.3, 2.7; p < .001) and had MSOF at initial presentation (12% vs. 6%, PR: 2.6, 95% CI: 1.3, 5.2; p = .007) compared to infants with AML.
TABLE 2.
Comparisons of acuity at presentation
| ALL (n = 10 359) | |||||||
|---|---|---|---|---|---|---|---|
| Overall | <1 year (n = 405) |
1–10 years (n = 9954) |
Crude PR (95% CI) |
p-Value | Adjusteda PR (95% CI) |
p-Value | |
| Acuity at presentation | |||||||
| Any ICU-level care requirement, n (%) | 469 (4.5) | 100 (24.7) | 369 (3.7) | 6.73 (5.28, 8.58) | <.001 | 6.64 (5.21, 8.57) | <.001 |
| Respiratory, n (%) | 213 (2.1) | 68 (16.8) | 145 (1.46) | 11.4 (8.18, 15.8) | <.001 | 11.4 (8.16, 15.9) | <.001 |
| Cardiac, n (%) | 184 (1.8) | 48 (11.9) | 136 (1.4) | 8.51 (6.41, 11.3) | <.001 | 8.48 (6.30, 11.4) | <.001 |
| Leukapheresis, n (%) | 156 (1.5) | 36 (8.9) | 120 (1.2) | 7.52 (4.94, 11.5) | <.001 | 7.20 (4.75, 10.9) | <.001 |
| Renal, n (%) | 105 (1.0) | 12 (3.0) | 93 (0.9) | 3.51 (1.85, 6.65) | <.001 | 3.38 (1.74, 6.56) | <.001 |
| Neurologic, n (%) | 2 (0.02) | 0 | 2 (0.02) | ne | |||
| Multisystem ICU requirement, n (%) | 156 (1.5) | 48 (11.9) | 108 (1.1) | 10.8 (7.36, 15.7) | <.001 | 10.5 (7.27, 15.3) | <.001 |
| AML (n = 871) | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Overall | <1 year (n = 189) |
1–10 years (n = 682) |
Crude PR (95% CI) |
p-Value | Adjusteda PR (95% CI) |
p-Value | Adjustedb PR (95% CI) |
p | |
| Acuity at presentation | |||||||||
| Any ICU-level care requirement, n (%) | 84 (9.6) | 31 (16.4) | 53 (7.8) | 2.07 (1.40, 3.09) | <.001 | 1.96 (1.30, 2.97) | .001 | 1.86 (1.30, 2.67) | <0.001 |
| Respiratory, n (%) | 54 (6.2) | 23 (12.2) | 31 (4.6) | 2.67 (1.69, 4.20) | <.001 | 2.46 (1.63, 3.70) | <.001 | 1.73 (1.12, 2.67) | 0.0128 |
| Cardiac, n (%) | 30 (3.4) | 10 (5.3) | 20 (2.9) | 1.81 (0.93, 3.50) | .079 | 1.73 (0.91, 3.27) | .092 | 2.81 (1.53, 5.18) | <0.001 |
| Leukapheresis, n (%) | 33 (3.8) | 9 (4.8) | 24 (3.5) | 1.34 (0.66, 2.74) | .407 | 1.07 (0.46, 2.47) | .879 | 2.53 (1.13, 5.66) | 0.0242 |
| Renal, n (%) | 6 (0.7) | 2 (1.1) | 4 (0.6) | 1.81 (0.38, 8.61) | .458 | 1.53 (0.37, 6.27) | .554 | 3.13 (0.79, 12.4) | 0.104 |
| Neurologic, n (%) | 1 (0.1) | 0 | 1 (0.2) | ne | ne | ||||
| Multisystem ICU requirement, n (%) | 31 (3.6) | 11 (5.8) | 20 (2.9) | 1.98 (1.07, 3.69) | .030 | 1.70 (0.95, 3.02) | .074 | 2.60 (1.29, 5.23) | 0.007 |
Abbreviations: ALL, acute lymphoid leukemia; AML, acute myeloid leukemia; ICU, intensive care unit.
p-Value for the comparison of demographic distributions by age group (infant vs. noninfant) separately for ALL and AML.
p-Value for the comparison of demographic distributions among infants by disease group (i.e., ALL vs. AML) adjusted for gender, race, and ethnicity.
3.3 |. Induction mortality
ALL patients exposed to anthracycline were more likely to require any ICU-level care (13% vs. 4%; adjusted RR: 2.4, 95% CI: 1.9, 3.0; p < .001) and experience MSOF (4% vs. 1%; adjusted RR: 1.8, 95% CI: 1.3, 2.4; p < .001) during induction, and had higher induction mortality (1.0% vs. 0.3%, adjusted HR: 2.1, 95% CI: 1.2, 3.9; p = .011) compared with those not exposed to anthracycline.
Comparisons of induction mortality are presented in Table 3. Infants with ALL had higher induction mortality compared to noninfants aged 1–10 years (2.7% vs. 0.5; crude HR: 5.4, 95% CI: 1.8, 10.4; p < .001). This difference persisted, but was attenuated, after accounting for differences in anthracycline exposure (adjusted HR: 3.7, 95% CI: 1.7, 7.9; p = .001) and further adjustment for acuity at presentation (adjusted HR: 2.1, 95% CI: 1.0, 4.8; p = .064). In contrast, induction mortality was similar for infants and noninfants with AML (3.2% vs. 2.1%; adjusted HR: 1.3, 95% CI: 0.4, 4.4; p = .685). Induction mortality among infants with ALL and infants with AML also did not differ significantly (2.7% vs. 3.2%; adjusted HR: 0.8, 95% CI: 0.3, 2.3; p = .646).
TABLE 3.
Comparisons of induction mortality by age category
| Total N | <1 year | 1–10 years | Crude HR (95% CI) |
p-Value | Adjusted Model 1a HR (95% CI) |
p-Value | Adjusted Model 2b HR (95% CI) |
p-Value | |
|---|---|---|---|---|---|---|---|---|---|
| ALL | 10 359 | 11 (2.7) | 46 (0.5) | 5.35 (2.76, 10.4) | <.001 | 3.66 (1.69, 7.93) | .001 | 2.15 (1.00, 4.81) | .064 |
| AML | 871 | 6 (3.2) | 14 (2.1) | 1.56 (0.59, 4.13) | .400 | 1.47 (0.64, 4.74) | .274 | 1.29 (0.38, 4.41) | .685 |
Abbreviations: ALL, acute lymphoid leukemia; AML, acute myeloid leukemia.
ALL model comparing <1 year versus 1–10-year age groups adjusted for race, ethnicity, and anthracycline exposure; AML model adjusted for race and ethnicity.
ALL model comparing <1 year versus 1–10-year age groups adjusted for race, ethnicity, anthracycline exposure, and acuity at presentation; AML model adjusted for race, ethnicity, and acuity at presentation.
3.4 |. Resource utilization
Mean cumulative number of inpatient days for infants with ALL was, as expected, longer than for noninfants with ALL (30.3 vs. 12.7 days, p < .001). In contrast, infants with AML had similar mean cumulative numbers of inpatient days to noninfant AML patients (28.6 vs. 28.7 days, p = .836). There was no difference in the cumulative number of inpatient days in infants by disease group (p = .96).
Adjusted comparisons of the prevalence of exposure to specific clinical resources and the rates of utilization among those exposed are summarized in Tables 4 and 5, respectively. Infants with ALL had greater resource requirements during induction than noninfants with ALL, even after accounting for differences in the cumulative number of inpatient days. Infants with ALL were more likely to require antimicrobial agents, antiemetics, TPN, blood products, and GCSF, which were utilized for longer durations compared to noninfants with ALL. Vasopressor, supplemental oxygen, and mechanical ventilation support were also more prevalent among infants than noninfants with ALL. Infants in both disease groups were more likely to require diuretics compared to older children. Requirements for the other evaluated resources were comparable for infants and noninfants with AML. Compared to infants with AML, infants with ALL were more likely to have a blood culture during induction, to receive GCSF, vasopressors, and supplemental oxygen, but for shorter durations. Otherwise, the prevalence of the specific resources evaluated was comparable among infants with ALL and AML.
TABLE 4.
Comparisons of the prevalence (%) of utilization of specific resources during induction
| ALL (n = 10 359) | AML (n = 871) | Infant ALL vs. Infant AMLAdjustedc PR (95% CI) |
|||||
|---|---|---|---|---|---|---|---|
| <1 year (n = 405) |
1–10 years (n = 9954) |
Adjusteda PR (95% CI) |
<1 year (n = 189) |
1–10 years (n = 682) |
Adjustedb PR (95% CI) |
||
| Complete blood count | 91.9 | 94.7 | 0.97 (0.94, 1.00) | 99.8 | 99.7 | 1.00 (0.99, 1.00) | 0.96 (0.94, 1.00) |
| Blood culture | 78.4 | 66.2 | 1.18 (1.11, 1.27) | 23.3 | 25.3 | 0.92 (0.75, 1.12) | 2.98 (2.05, 4.32) |
| Antibiotics | 95.5 | 82.5 | 1.16 (1.13, 1.19) | 98.9 | 98.6 | 1.00 (0.98, 1.02) | 0.97 (0.80, 1.17) |
| Antifungals | 79.3 | 22.5 | 3.52 (2.83, 4.39) | 93.9 | 93.2 | 1.01 (0.97, 1.05) | 0.84 (0.78, 0.89) |
| Antivirals | 10.9 | 5.8 | 1.90 (1.43, 2.52) | 4.4 | 17.1 | 0.26 (0.13, 0.50) | 2.20 (0.96, 5.03) |
| Blood products | 89.8 | 82.6 | 1.09 (1.04, 1.14) | 97.9 | 99.9 | 0.98 (0.94, 1.01) | 0.94 (0.85, 1.03) |
| GCSF | 48.9 | 1.8 | 26.7 (20.4, 35.0) | 7.40 | 9.70 | 0.98 (0.35, 2.74) | 6.35 (3.57, 11.3) |
| Non-opioid analgesics | 82.7 | 75.4 | 1.10 (105, 1.15) | 82.1 | 85.8 | 0.96 (0.90, 1.02) | 0.97 (0.89, 1.05) |
| Opioid analgesics | 97.0 | 91.5 | 1.06 (1.04, 1.08) | 92.5 | 87.9 | 1.05 (1.00, 1.11) | 1.04 (0.98, 1.11) |
| Antiemetics | 99.5 | 93.9 | 1.05 (1.04, 1.08) | 99.4 | 99.9 | 1.00 (0.98, 1.01) | 1.01 (0.93, 1.09) |
| Total parenteral nutrition | 33.5 | 9.3 | 3.61 (2.72, 4.79) | 37.9 | 39.5 | 0.96 (0.75, 1.23) | 0.83 (0.65, 1.06) |
| Antihypertensives | 27.6 | 26.3 | 1.05 (0.90, 1.22) | 29.8 | 33.6 | 0.89 (0.55, 1.43) | 0.97 (0.80, 1.17) |
| Diuretics | 58.2 | 32.7 | 1.78 (1.57, 2.03) | 68.8 | 46.2 | 1.49 (1.23, 1.80) | 0.96 (0.83, 1.12) |
| Vasopressors | 35.9 | 26.0 | 1.38 (1.18, 1.62) | 26.7 | 26.3 | 1.02 (0.73, 1.42) | 1.29 (0.96, 1.74) |
| Supplemental oxygen | 56.9 | 37.1 | 1.53 (1.37, 1.72) | 42.9 | 42.6 | 1.01 (0.80, 1.26) | 1.52 (1.19, 1.93) |
| Ventilation | 29.5 | 6.2 | 5.15 (3.50, 7.59) | 28.0 | 22.7 | 1.23 (0.88, 1.72) | 0.96 (0.78, 1.19) |
| ECMO | 0.2 | 0.04 | 4.90 (0.37, 65.1) | 0.5 | 0.6 | 0.93 (0.10, 8.49) | 0.47 (0.03, 7.42) |
| Dialysis | 2.10 | 3.10 | 0.66 (0.43, 1.03) | 2.3 | 2.2 | 1.05 (0.38, 2.89) | 0.80 (0.33, 1.93) |
Abbreviations: ALL, acute lymphoid leukemia; AML, acute myeloid leukemia; ECMO, extracorporeal membrane oxygenation; GCSF, granulocyte colony-stimulating factor.
Prevalence ratio comparing resource use among infants with ALL versus noninfants aged 1–10 years with ALL; adjusted for race, ethnicity, and acuity at presentation.
Prevalence ratio comparing resource use among infants with AML versus noninfants aged 1–10 years with AML; adjusted for race, ethnicity, and acuity at presentation.
Prevalence ratio comparing resource use among infants with ALL versus infants with AML; adjusted for race, ethnicity, and acuity at presentation.
TABLE 5.
Comparisons of the rate of utilization (use per 100 days) among users of specific resources during induction
| ALL | AML | Infant ALL vs. Infant AML Adjustedc RR (95% CI) |
|||||
|---|---|---|---|---|---|---|---|
| <1 year | 1–10 years | Adjusteda RR (95% CI) |
<1 year | 1–10 years | Adjustedb RR (95% CI) |
||
| Complete blood count | 85.9 | 87.5 | 0.98 (0.95, 1.02) | 92.5 | 93.5 | 0.99 (0.97, 1.01) | 1.01 (0.99, 1.04) |
| Blood culture | 21.0 | 22.8 | 0.92 (0.82, 1.04) | 8.3 | 10.6 | 0.79 (0.53, 1.17) | 2.59 (1.65, 4.05) |
| Antibiotics | 118.7 | 121.8 | 0.98 (0.93, 1.04) | 138.4 | 135.9 | 1.02 (0.96, 1.09) | 0.71 (0.66, 0.78) |
| Antifungals | 72.7 | 54.4 | 1.34 (1.22, 1.46) | 81.9 | 80.2 | 1.02 (0.97, 1.08) | 0.87 (0.78, 0.97) |
| Antivirals | 23.4 | 32.8 | 0.71 (0.50, 1.02) | 35.8 | 32.9 | 1.12 (0.67, 1.88) | 0.74 (0.42, 1.32) |
| Total blood products | 29.5 | 25.2 | 1.17 (1.05, 1.29) | 46.9 | 45.2 | 1.04 (0.94, 1.15) | 0.75 (0.66, 0.84) |
| GCSF | 53.7 | 26.7 | 2.01 (1.75, 2.31) | 31.2 | 29.8 | 1.05 (0.74, 1.48) | 0.72 (0.46, 1.13) |
| Non-opioid analgesics | 22.9 | 24.7 | 0.93 (0.83, 1.03) | 27.9 | 32.3 | 0.86 (0.76, 0.98) | 0.92 (0.81, 1.04) |
| Opioid analgesics | 35.5 | 34.2 | 1.04 (0.97, 1.11) | 40.8 | 35.6 | 1.15 (0.99, 1.33) | 0.90 (0.77, 1.05) |
| Antiemetics | 39.1 | 22.6 | 1.73 (1.58, 1.89) | 56.8 | 52.2 | 1.09 (0.98, 1.17) | 0.87 (0.79, 0.95) |
| Total parenteral nutrition | 37.3 | 33.2 | 1.12 (0.97, 1.30) | 43.8 | 44.8 | 0.98 (0.84, 1.14) | 0.78 (0.63, 0.95) |
| Antihypertensives | 32.0 | 33.0 | 0.97 (0.82, 1.14) | 11.0 | 12.0 | 0.91 (0.39, 2.14) | 2.05 (1.01, 4.11) |
| Diuretics | 15.5 | 17.3 | 0.90 (0.78, 1.03) | 26.9 | 22.3 | 1.21 (0.94, 1.55) | 0.56 (0.42, 0.74) |
| Vasopressors | 5.1 | 5.9 | 0.87 (0.49, 1.53) | 14.4 | 15.3 | 0.94 (0.52, 1.69) | 0.72 (0.46, 1.13) |
| Supplemental oxygen | 7.4 | 5.7 | 1.31 (0.87, 1.96) | 18.9 | 15.8 | 1.20 (0.92, 1.58) | 0.78 (0.60, 1.02) |
| Ventilation | 15.1 | 21.9 | 0.69 (0.56, 0.85) | 28.1 | 37.8 | 0.74 (0.53, 1.04) | 0.61 (0.42, 0.88) |
| ECMO | 22.9 | 23.1 | 0.99 (0.29, 3.36) | 39.30 | 32.50 | 1.59 (0.62, 4.09) | 0.36 (0.12, 1.12) |
| Dialysis | 8.40 | 15.90 | 0.56 (0.26, 1.22) | 24.20 | 25.60 | 0.94 (0.22, 4.13) | 0.58 (0.23, 1.48) |
Abbreviations: ALL, acute lymphoid leukemia; AML, acute myeloid leukemia; ECMO, extracorporeal membrane oxygenation; GCSF, granulocyte colony-stimulating factor.
Rate ratio comparing resource use among infants with ALL versus noninfants aged 1–10 years with ALL; adjusted for race, ethnicity, and acuity at presentation.
Rate ratio comparing resource use among infants with AML versus noninfants aged 1–10 years with AML; adjusted for race, ethnicity, and acuity at presentation.
Rate ratio comparing resource use among infants with ALL versus infants with AML; adjusted for race, ethnicity, and acuity at presentation.
4 |. DISCUSSION
Using a large nationally representative cohort, we demonstrate that infants with ALL present with higher acuity, have significantly higher rates of inpatient induction mortality, and have greater RU during induction compared to noninfants aged 1–10 years with ALL. In contrast, among children with AML there were no significant differences by age group after accounting for higher acuity at presentation among infants. Induction mortality and supportive care requirements were similar for infants with ALL and AML.
Though not previously described, our findings of higher acuity at presentation among infants with leukemia are not unexpected. Infants with leukemia more often present with hyperleukocytosis, hepatomegaly, splenomegaly, and higher incidence of central nervous system disease than noninfants.1–4,2,29 Infant leukemias are also known to be biologically different from those in older children with a higher prevalence of histone lysine methyltransferase 2A gene (KMT2A) gene rearrangement, negative CD10 expression, and coexpression of lymphoid and myeloid antigens.2–4,14,2,30
Our findings of greater RU among infants compared to noninfants with AML align with prior reports of higher rates of toxicities,21,22 and further suggest that higher acuity at presentation may contribute to difference in toxicities documented early in treatment. Our findings also confirmed the higher induction mortality in infants compared to noninfants <10 years with ALL, with the rates and effect estimates similar in magnitude to those previously described in the literature.13,31,32 While others have evaluated initial WBC count and leukemia subtype, none have looked at the severity of presentation as a contributor to induction mortality differences by age. In our study, higher acuity at presentation among infants with ALL partially explained the higher induction mortality compared to noninfants; however, significant differences in induction mortality persisted following adjustment for presentation acuity and anthracycline exposure. Likewise, the greater RU among infants with ALL compared to noninfants persisted following similar adjustment. Infants with ALL had resource requirements that were more comparable to infants and noninfants with AML. Among patients with AML, who are treated with the same intensive chemotherapy protocols regardless of age, induction mortality and RU were similar for infants and noninfants, despite higher acuity at presentation among infants. In contrast, infants with ALL receive more intensive chemotherapy with greater treatment-related toxicity risks than their older counterparts, which together with greater acuity at presentation, partially explain the observed residual difference in resource use and induction mortality compared to noninfants.
Previous studies of infant ALL have focused on the impact of chemotherapy modifications on overall outcomes, but have isolated differences in supportive care.16,18,20 In addition to greater utilization of antimicrobials, particularly antifungals, we found that infants with ALL were more likely to receive antihypertensives and TPN than older children. Infants overall had similar use of opioids and antiemetics, which were previously shown to be used less in vulnerable groups, such as patients with DS.33,34
Our results should be considered in the context of study limitations. First, PHIS does not capture laboratory results, thus we were not able to evaluate the contribution of biological prognostic factors in observed associations. Infant protocols use standard dose reductions based upon body weight or body surface compared to protocols for noninfants with high-risk ALL. PHIS does not include information on dosages of administered drugs; therefore, we could not perform more nuanced evaluations of treatment intensity beyond accounting for receipt of anthracyclines. PHIS is limited to inpatient billing data, which may have led to underestimation of the utilization of some resources, particularly among noninfants with ALL who are hospitalized for short durations compared to infants with ALL and AML patients. However, we limited our analysis to the induction phase, and standardized RU based on the number of inpatient days in an effort to reduce such bias, which may be exaggerated with longer follow-up periods. Additionally, the majority of the supportive care resources of interest are used in the inpatient setting, thus if unmeasured outpatient care occurred it would be a low. Death was defined by disposition documented at discharge, thus our analyses were restricted to comparisons of inpatient mortality. This may have resulted in some differential misclassification of vital status among ALL patients given the longer duration of hospitalization for infants, and a consequent upward bias of the observed hazard ratio comparing induction mortality in infants and noninfants. However, we expect the magnitude of the bias to be small, because early deaths usually occur in the inpatient setting irrespective of age. While this is one of the largest evaluations of infants with leukemia, the rarity of induction mortality prohibited the exploration of risk within more granular infant age group classifications. The generalizability of our results may be limited to institutions comparable to those included in PHIS, specifically free-standing pediatric hospitals, which may have different patient volume, levels of experience treating acute leukemias, and supportive care practices, compared with other pediatric hospitals.
In summary, our study showed that infants with acute leukemia present with significantly higher acuity compared with noninfants aged <10 years. This difference in acuity at presentation contributed to the higher rate of induction mortality among infants relative to noninfants with ALL. Supportive care requirements among infants with ALL were significantly higher than among noninfants with ALL, but were similar to rates observed for infants and noninfants with AML. The observed differences reflect the greater toxicity risk associated with intensified chemotherapy, and may account for some of the residual difference in induction mortality. Future work should quantify the contribution of treatment-related toxicity to early mortality in infants with ALL and explore supportive care strategies that would serve to equalize outcomes across age groups.
ACKNOWLEDGMENTS
Kelly D. Getz receives support for research effort from a NHLBI career development grant (5K01HL143153) and an Alex’s Lemonade Stand Foundation Young Investigator award. Lena E. Winestone, Regina M. Myers, and Caitlin W. Elgarten receive support for research effort from an Alex’s Lemonade Stand Foundation Young Investigator award. Tamara P. Miller receives support for research effort from a NCI career development grant (5K07CA211956).
Abbreviations:
- ALL
acute lymphocytic leukemia
- AML
acute myeloid leukemia
- DS
Down syndrome
- GCSF
granulocyte colony-stimulating factor
- ICU
intensive care unit
- MSOF
multisystem organ failure
- PHIS
Pediatric Health Information System
- RU
resource utilization
- TPN
total parenteral nutrition
Footnotes
CONFLICT OF INTEREST
Authors had no conflicts to declare.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the Children’s Hospital Association. Restrictions apply to the availability of these data. Data are available from the authors with the permission of Children’s Hospital Association.
This work was presented in part as a poster at the annual meeting of the American Society of Pediatric Hematology/Oncology (ASPHO), May 2018.
REERENCES
- 1.Brown P Treatment of infant leukemias: challenge and promise. Hematology Am Soc Hematol Educ Program. 2013;2013:596–600. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Biondi A, Cimino G, Pieters R, Pui CH. Biological and therapeutic aspects of infant leukemia. Blood. 2000;96(1):24–33. [PubMed] [Google Scholar]
- 3.Brown P, Pieters R, Biondi A. How I treat infant leukemia. Blood. 2019;133(3):205–214. [DOI] [PubMed] [Google Scholar]
- 4.Chessells JM, Harrison CJ, Kempski H, et al. Clinical features, cytogenetics and outcome in acute lymphoblastic and myeloid leukaemia of infancy: report from the MRC Childhood Leukaemia working party. Leukemia. 2002;16(5):2402468. [DOI] [PubMed] [Google Scholar]
- 5.Hunger SP, Loh MK, Baker SK, Schultz KR. Controversies of and unique issues in hematopoietic cell transplantation for infant leukemia. Biol Blood Marrow Transplant. 2009;15(1):79–83. [DOI] [PubMed] [Google Scholar]
- 6.Pui CH, Ribeiro RC, Campana D, et al. Prognostic factors in the acute lymphoid and myeloid leukemias of infants. Leukemia. 1996;10(6):952–956. [PubMed] [Google Scholar]
- 7.Gaynon PS, Angiolillo AL, Carroll WL, et al. Long-term results of the children’s cancer group studies for childhood acute lymphoblastic leukemia 1983–2002: a Children’s Oncology Group Report. Leukemia. 2010;24(2):320–334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Schmiegelow K, Forestier E, Hellebostad M, et al. Long-term results of NOPHO ALL-92 and ALL-2000 studies of childhood acute lymphoblastic leukemia. Leukemia. 2010;24:345. [DOI] [PubMed] [Google Scholar]
- 9.Pieters R, De Lorenzo P, Ancliffe P, et al. Outcome of infants younger than 1 year with acute lymphoblastic leukemia treated with the Interfant-06 protocol: results from an international Phase III randomized study. J Clin Oncol. 2019;37(25):2246–2256. [DOI] [PubMed] [Google Scholar]
- 10.Silverman LB, McLean TW, Gelber RD, et al. Intensified therapy for infants with acute lymphoblastic leukemia. Cancer. 1997;80(12):2285–2295. [PubMed] [Google Scholar]
- 11.Dördelmann M, Reiter A, Borkhardt A, et al. Prednisone response is the strongest predictor of treatment outcome in infant acute lymphoblastic leukemia. Blood. 1999;94(4):1209–1217. [PubMed] [Google Scholar]
- 12.Chessells JM, Harrison CJ, Watson SL, Vora AJ, Richards SM, on Leukaemia M. Treatment of infants with lymphoblastic leukaemia: results of the UK Infant Protocols 1987–1999. Br J Haematol. 2002;117(2):306–314. [DOI] [PubMed] [Google Scholar]
- 13.Rubnitz JE, Lensing S, Zhou Y, et al. Death during induction therapy and first remission of acute leukemia in childhood. Cancer. 2004;101(7):1677–1684. [DOI] [PubMed] [Google Scholar]
- 14.Hilden JM, Dinndorf PA, Meerbaum SO, et al. Analysis of prognostic factors of acute lymphoblastic leukemia in infants: report on CCG 1953 from the Children’s Oncology Group. Blood. 2006;108(2):441–451. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Pieters R, Schrappe M, Lorenzo P, et al. A treatment protocol for infants younger than 1 year with acute lymphoblastic leukaemia (Interfant-99): an observational study and a multicentre randomised trial. Lancet. 2007;370(9583):240–250. [DOI] [PubMed] [Google Scholar]
- 16.Salzer W, Dinndorf P, Dreyer Z, Hilden J, Reaman GH. Analysis of infectious complications in infants with acute lymphoblastic leukemia treated on the Children’s Cancer Group Protocol 1953: a report from the Children’s Oncology Group. J Pediatr Hematol Oncol. 2009;31(6):398–405. [DOI] [PubMed] [Google Scholar]
- 17.Salzer WL, Jones TL, Devidas M, et al. Modifications to induction therapy decrease risk of early death in infants with acute lymphoblastic leukemia treated on Children’s Oncology Group P9407. Pediatr Blood Cancer. 2012;59(5):834–839. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Dreyer ZE, Hilden JM, Jones TL, et al. Intensified chemotherapy without SCT in infant ALL: results from COG P9407 (Cohort 3). Pediatr Blood Cancer. 2015;62(3):419–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Slats AM, Egeler RM. Causes of death – other than progressive leukemia – in childhood acute lymphoblastic (ALL) and myeloid leukemia (AML): the Dutch Childhood Oncology Group experience. Leukemia. 2005;19(4):537. [DOI] [PubMed] [Google Scholar]
- 20.Salzer WL, Jones TL, Devidas M, et al. Decreased induction morbidity and mortality following modification to induction therapy in infants with acute lymphoblastic leukemia enrolled on AALL0631: a report from the Children’s Oncology Group. Pediatr Blood Cancer. 2015;62(3):414–418. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kavcic M, Fisher BT, Li Y, et al. Induction mortality and resource utilization in children treated for acute myeloid leukemia at free-standing pediatric hospitals in the United States. Cancer. 2013;119(10):1916–1923. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tomizawa D, Tawa A, Watanabe T, et al. Appropriate dose reduction in induction therapy is essential for the treatment of infants with acute myeloid leukemia: a report from the Japanese Pediatric Leukemia/Lymphoma Study Group. Int J Hematol. 2013;98(5):578–588. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Creutzig U, Zimmermann M, Bourquin JP, et al. Favorable outcome in infants with AML after intensive first- and second-line treatment: an AML-BFM study group report. Leukemia. 2012;26(4):654. [DOI] [PubMed] [Google Scholar]
- 24.Lange BJ, Smith FO, Feusner J, et al. Outcomes in CCG-2961, a Children’s Oncology Group Phase 3 Trial for untreated pediatric acute myeloid leukemia: a report from the Children’s Oncology Group. Blood. 2008;111(3):1044–1053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cooper TM, Franklin J, Gerbing RB, et al. AAML03P1, a pilot study of the safety of gemtuzumab ozogamicin in combination with chemotherapy for newly diagnosed childhood acute myeloid leukemia. Cancer. 2012;118(3):761–769. [DOI] [PubMed] [Google Scholar]
- 26.Guest EM, Aplenc R, Sung L, et al. Gemtuzumab ozogamicin in infants with AML: results from the Children’s Oncology Group trials AAML03P1 and AAML0531. Blood. 2017;130(7):943–945. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kavcic M, Fisher BT, Torp K, et al. Assembly of a cohort of children treated for acute myeloid leukemia at free-standing children’s hospitals in the United States using an administrative database. Pediatr Blood Cancer. 2013;60(3):508–511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Fisher BT, Harris T, Torp K, et al. Establishment of an 11-year cohort of 8733 pediatric patients hospitalized at United States hospitals with de novo acute lymphoblastic leukemia from health care administrative data. Med Care. 2014;52(1):e1–e6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Chen SH, Yang CP, Hung IJ, Jaing TH, Shih LY, Tsai MH. Clinical features, molecular diagnosis, and treatment outcome of infants with leukemia in Taiwan. Pediatr Blood Cancer. 2010;55(7):1264–1271. [DOI] [PubMed] [Google Scholar]
- 30.Pui CH, Raimondi SC, Srivastava DK, et al. Prognostic factors in infants with acute myeloid leukemia. Leukemia. 2000;14(4):684. [DOI] [PubMed] [Google Scholar]
- 31.Prucker C, Attarbaschi A, Peters C, et al. Induction death and treatment-related mortality in first remission of children with acute lymphoblastic leukemia: a population-based analysis of the Austrian Berlin-Frankfurt-Münster study group. Leukemia. 2009;23(7):1264. [DOI] [PubMed] [Google Scholar]
- 32.Seif AE, Fisher BT, Li Y, et al. Patient and hospital factors associated with induction mortality in acute lymphoblastic leukemia. Pediatr Blood Cancer. 2014;61(5):846–852. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Hassler A, Bochennek K, Gilfert J, et al. Infectious complications in children with acute myeloid leukemia and Down syndrome: analysis of the prospective multicenter trial AML-BFM 2004. Pediatr Blood Cancer. 2016;63(6):1070–1074. [DOI] [PubMed] [Google Scholar]
- 34.Salazar EG, Li Y, Fisher BT, et al. Supportive care utilization and treatment toxicity in children with Down syndrome and acute lymphoid leukaemia at free-standing paediatric hospitals in the United States. Br J Haematol. 2016;174(4):591–599. [DOI] [PMC free article] [PubMed] [Google Scholar]