Survival and Late Effects after Allogeneic Hematopoietic Cell Transplantation for Hematologic Malignancy at less than Three Years of Age

Abstract

Very young children undergoing hematopoietic cell transplantation (HCT) are a unique and vulnerable population. We analyzed outcomes of 717 patients from 117 centers who survived relapse-free for ≥1 year following allogeneic myeloablative HCT for hematologic malignancy at <3 years-of-age, between 1987–2012. Median follow-up was 8.3 years (range 1.0–26.4 years); median age at follow-up was 9 years (range 2–29 years). Ten-year overall and relapse-free survival were 87% (95% CI 85–90%) and 84% (95% CI 81–87%). Ten-year cumulative incidence of relapse was 11% (95% CI 9–13%). Of 84 deaths, relapse was the leading cause (43%). Chronic graft-versus-host-disease 1 year after HCT was associated with increased risk of mortality (HR 2.1, 95% CI 1.3–3.3, p=0.0018). Thirty percent of patients experienced ≥1 organ toxicity/late-effect >1 year after HCT. The most frequent late-effects included growth hormone deficiency/growth disturbance (10-year cumulative incidence 23%, 95% CI 19–28%), cataracts (18%, 95% CI 15–22%), hypothyroidism (13%, 95% CI 10–16%), gonadal dysfunction/infertility requiring hormone replacement (3%, 95% CI 2–5%), and stroke/seizure (3%, 95% CI 2–5%). Subsequent malignancy was reported in 3.6%. In multivariable analysis, TBI was predictive of increased risk of cataracts (HR 17.2, 95% CI 7.4–39.8, p<0.001), growth deficiency (HR 3.5, 95% CI 2.2–5.5, p<0.001), and hypothyroidism (HR 5.3, 95% CI 3.0–9.4, p<0.001). In summary, those who survived relapse-free ≥1 year after HCT for hematologic malignancy at <3 years-of-age had favorable overall survival. Chronic graft-versus-host-disease and TBI were associated with adverse outcomes. Future efforts should focus on reducing the risk of relapse and late-effects after HCT at early age.

Introduction

Hematopoietic cell transplantation (HCT) is an important treatment modality in infants and young children with very high risk and relapsed or refractory leukemias.^1–3 Infants and very young children may be at particular risk for morbidities after HCT including from transplant-related exposures, such as total body irradiation (TBI).^{4, 5} There are, however, few reports focusing on survival and morbidities after HCT for hematologic malignancy in the first years of life.^4–8 A comprehensive characterization of outcomes following HCT for hematologic malignancy at a very young age would help to inform the care of this potentially vulnerable population.

In a retrospective, international multi-center cohort of patients reported to the to the Center for International Blood and Marrow Transplant Research (CIBMTR) between 1987 and 2012, we sought to characterize the survival and late effects of patients who underwent allogeneic, myeloablative HCT for hematologic malignancy before the age of 3 years and who were alive and relapse free for at least one year following HCT. We aimed to report overall and disease-free survival, causes of death and risk factors for mortality, and the frequency and cumulative incidence of organ toxicities and late effects.

Materials and Methods

Data Collection

We obtained data from the CIBMTR, a voluntary working group of more than 450 international transplantation centers. Centers contribute detailed pre- and post-HCT data to the Statistical Center at the Medical College of Wisconsin in Milwaukee, Wisconsin and the National Marrow Donor Program (NMDP) in Minneapolis, Minnesota. Computerized checks for discrepancies, physicians’ review of submitted data, and on-site audits of participating centers ensure data quality. Observational studies conducted by the CIBMTR are performed under guidance of the Institutional Review Board of the NMDP and are in compliance with all applicable federal regulations pertaining to the protection of human research participants.

The CIBMTR data repository includes information about patient demographics, disease type, survival, relapse, graft type, the presence of graft versus host disease (GVHD), and cause of death. A subset of CIBMTR participants is selected for more comprehensive research level data collection by weighted randomization. Late effects data were collected from this group of patients. Transplant centers report the occurrence of clinically significant organ impairment or disorders at 6-months and one year following transplant, and annually thereafter, or until death. Centers report the presence of the following specific organ toxicities and late effects: avascular necrosis, cataracts, congestive heart failure, diabetes, gonadal dysfunction/infertility requiring hormone replacement, growth hormone deficiency/growth disturbance, hemorrhagic cystitis, hypothyroidism, myocardial infarction, pancreatitis, thrombotic thrombocytopenic purpura/Hemolytic Uremic Syndrome (TTP/HUS), renal failure severe enough to warrant dialysis, stroke/seizures, bronchiolitis obliterans, pulmonary hemorrhage, cryptogenic organizing pneumonia, interstitial pneumonitis/idiopathic pneumonia syndrome, non-infectious liver toxicity, and new malignancy. This report focuses on organ impairment and disorders following one year after HCT.

Study Population

The study population consisted of patients who underwent allogeneic myeloablative HCT for hematologic malignancy at less than 3 years of age, between January 1, 1987 and December 31, 2012 (Figure 1). Patients included in this analysis survived, relapse-free, at least one year following HCT. Underlying diagnoses included acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), myelodysplastic syndrome (MDS), chronic myelomonocytic leukemia (CMML), and juvenile myelomonocytic leukemia (JMML). Twin donor and multiple donor cases were excluded. Stem cell sources included bone marrow, peripheral blood, or umbilical cord blood; patients who received multiple stem cell sources were excluded. Myeloablative conditioning regimens were defined as previously described.⁹ Patients for whom baseline or day 100 forms were unavailable were excluded. Also excluded were those for whom the CIBMTR completion team index (demonstrating the ratio of observed versus expected follow-up) was less than 80% at 5 years after HCT.

Figure 1.

Summary of patient selection and exclusions.

Statistical Analysis

Descriptive statistics for categorical variables are presented as frequencies and percentages. Median and range were used to summarize continuous variables. The primary end-point was overall survival. Secondary end-points included relapse-free survival (survival without relapse), relapse, transplant-related mortality, and occurrence of organ toxicities/late effects. Overall and relapse-free survival were estimated using Kaplan-Meier methodology. The cumulative incidence of relapse, transplant-related mortality, and of organ toxicities and late-effects were estimated using the cumulative incidence function to account for competing risks. Analyses of organ toxicities and late effects were limited to those experienced at least one year after HCT. Summaries of aggregated organ toxicities/late effects include items specified on CIBMTR reporting form fields, as noted above. Proportional hazard models were developed to explore potential risk factors for overall mortality (in the full cohort), and for the most frequently occurring late effects (limited to those undergoing HCT after the year 1994) including assessment by age at HCT, sex, underlying diagnosis, interval from diagnosis to HCT, disease status at HCT, donor type, graft type, exposure to TBI, exposure to corticosteroid for GVHD prophylaxis, history of chronic GVHD at one year after HCT, and year of HCT. Analyses were performed using SAS, version 9.3 (SAS Institute, Cary, NC).¹⁰

Results

Patient and Transplantation Characteristics

Patient selection and exclusions are summarized in Figure 1. Of 1737 patients potentially eligible based on age at HCT, diagnosis, and year of HCT, 19 patients were excluded based on donor or graft source, 22 based upon use of a non-myeloablative condition regimen, and 215 based upon missing forms or team completeness index. Of the 1481 patients potentially eligible after the exclusions above, 738 patients were excluded based on events within the first year after HCT (148 relapses, 586 deaths, and 4 lost to follow-up). Of the 586 patients who died within the first year after HCT, the majority of deaths were due to relapse (322 patients, 55%), followed by organ failure (18%), infection (12%), and GVHD (6%). Less than 1% of deaths within one year after HCT were due to new malignancy (4 cases) or graft rejection (3 cases); 7% of deaths were due to other or unknown cause. Twenty-six additional patients were excluded having undergone a second HCT within one year of initial HCT. A total of 717 patients from 117 centers were included in the analysis.

The median follow-up was 8.3 years (range 1–26.4 years). Median age at follow-up (at last contact or death) was 9 years (range 2–29 years). The CIBMTR completeness index of follow-up was 96% at 5 years, 90% at 10 years, and 82% at 15 years.

Patient and transplantation characteristics are summarized in Table I. The median age at time of HCT was 19 months (range 3–35 months), with 24% of patients less than one year of age at HCT, 40% 12–23.99 months of age, and 36% 24–35.99 months of age. The most frequent underlying diagnoses were AML (45%) and ALL (33%). Four patients had an underlying diagnosis of Down syndrome. For those with acute leukemia, 65% of patients were in a first complete remission at the time of HCT. The majority of patients (66%) underwent unrelated donor transplant. Two hundred ninety-five (41%) patients received TBI; 7% of patients received a dose of less than 1200cGy (3% having had a single dose, 4% receiving fractionated dosing), 51% of patients received a total dose of 1200 cGy in fractionated doses, 21% of patients received 1300–1375 cGy in fractionated doses, and 21% of patients received greater than or equal to 1400 cGy in fractionated doses. TBI dose was unknown one patient. Corticosteroids were included in the GVHD prophylaxis regimen in 31% of transplants.

Table 1.

Patient and Transplantation Characteristics

Variable	N (%)
Number of patients	717
Number of centers	117
Male / female	405 (56) / 312 (44)
Race
Caucasian/White	552 (77)
Asian/Hawaiian/Pacific Islander	47 (7)
Black	36 (5)
Hispanic	35 (5)
Other	34 (5)
Unknown	13 (<2)
Age at HCT, months
Median (range)	19 (3–35)
<12 months	174 (24)
12–23.99 months	286 (40)
24–35.99 months	257 (36)
Lansky score ≥90% prior to HCT (patients age ≥1yr, n=543)	457 (84)
Underlying diagnosis
AML/other acute leukemia	325 (45)
ALL	239 (33)
JMML	117 (16)
MDS	27 (4)
CMML	9 (1)
Interval from diagnosis to HCT, median (range), months	6 (<1–30)
Disease status prior to HCT (acute leukemia, n=564)
CR1	365 (65)
CR2	144 (26)
>CR2	8 (1)
Relapse	28 (5)
Primary induction failure	15 (3)
Unknown	4 (<1%)
Donor type
HLA-identical sibling	203 (28)
Other related	42 (6)
Unrelated	472 (66)
Stem cell source
Bone marrow	417 (58)
Peripheral blood	49 (7)
Umbilical cord blood	251 (35)
Conditioning regimen
TBI + cyclophosphamide ± Others	246 (34)
TBI + Others	49 (7)
Busulfan + cyclophosphamide ± Others	325 (46)
Busulfan + melphalan ± Others	80 (11)
Others	17 (2)
Corticosteroid included in GVHD prophylaxis	223 (31)
Year of HCT
1987–1989	27 (4)
1990–1994	87 (12)
1995–1999	175 (24)
2000–2004	164 (23)
2005–2009	206 (29)
2010–2012	58 (8)
Development of acute GVHD after HCT
Any grade	434 (60)
Grade 1–2	325 (45)
Grade 3–4	109 (15)
Geographical Location of HCT
United States	481 (67)
Europe	82 (11)
Australia/New Zealand	59 (8)
Canada	32 (4)
Mideast/Africa	26 (4)
Asia	23 (3)
Central/South America	14 (2)

Chronic GVHD was reported in 186 patients (26%) at one year after transplant, and was ultimately reported in a total of 226 patients (32%). Of those with chronic GVHD, 107 patients (47%) had extensive and 116 patients (51%) limited GVHD. GVHD severity was unknown in 3 patients.

Survival, Transplant-Related Mortality, and Relapse

Overall and relapse-free survival in patients who survived, relapse-free, at least one year after HCT are presented in Figure 2. The 3, 5, and 10-year overall survival estimates were 92% (95% CI 90–94%), 90% (95% CI 88–92%), and 87% (95% CI 85–90%). The 3, 5, and 10-year relapse-free survival estimates were 89% (95% CI 87–91%), 86% (95%CI 84–89%), and 84% (95%CI 81–87%). The cumulative incidence of transplant-related mortality (with relapse as a competing risk) was estimated at 3, 5, and 10 years at 3% (95% CI 2–4%), 4% (95%CI 2–5%), and 5% (95%CI 4–7%).

Figure 2.

Overall (A) and relapse-free (B) survival among patients who survived, relapse-free at least one year following HCT for hematologic malignancy (performed at less than 3 years of age.)

Of the 717 patients in this analysis, there were 84 deaths (12%). Relapse was the leading cause of death (43% of deaths), followed by infection (18%), organ failure (11%), subsequent malignancy (8%), and GVHD (7%). The cause of death was not reported in 13% of deaths. The occurrence of chronic GVHD at one year after HCT was predictive of an increased risk of mortality (hazard ratio 2.1, 95% CI 1.3–3.3, p=0.0018). In regression analysis, there were no other significant predictors of overall mortality identified. The 3, 5, and 10-year estimates of cumulative incidence of relapse were 8% (95% CI 6–10%), 10% (95% CI 8–12%), and 11% (95% CI 9–13%). Thirty patients (4%) underwent a second HCT at least one year after initial HCT.

Organ Toxicities and Late Effects

The frequencies of organ toxicities and late effects are presented in Table 2. The most frequently reported organ toxicities were growth hormone deficiency/growth disturbance (17%), cataracts (13%), hypothyroidism (11%), gonadal dysfunction / infertility requiring hormone replacement (4%) and stroke/seizure (3%). Cumulative incidences for these organ toxicities are presented in Table 3. The 10-year cumulative incidence of individual toxicities and late effects were as follows: growth hormone deficiency/growth disturbance 23% (95% CI 19–28%), cataracts 18% (95% CI 15–22%), hypothyroidism 13% (95% CI 10–16%), gonadal dysfunction / infertility requiring hormone replacement 3% (95% CI 2–5%) and stroke/seizure 3% (95% CI 2–5%). At least one organ toxicity was reported in 30% of patients, with 16% of patients reporting one toxicity, 8% two toxicities, and 6% 3 or more toxicities.

Table 2.

Frequency of Non-Malignant Organ Toxicities and Late Effects (occurring at least one year after HCT, prior to relapse or second HCT, unless otherwise specified)

Organ Toxicities and Late Effects	N (%)
Number of non-malignant late effects
0	403 (56)
1	114 (16)
2	60 (8)
≥3	41 (6)
Missing	99 (14)
Cataracts
Yes	105 (15)
No	567 (79)
Missing	45 (6)
Congestive heart failure
Yes	1 (<1)
No	505 (70)
Missing	211 (29)
Diabetes mellitus
Yes	8 (1)
No	499 (70)
Missing	210 (29)
Gonadal dysfunction/infertility requiring hormone replacement
Yes	27 (4)
No	643 (90)
Missing	47 (7)
Growth hormone deficiency/growth disturbance
Yes	122 (17)
No	539 (75)
After relapse or 2nd HCT	8 (1)
After 1 year, but relapse status unknown	1 (<1)
Missing	47 (7)
Hemorrhage cystitis
Yes	1 (<1)
No	665 (93)
Missing	51 (7)
Hypothyroidism
Yes	76 (11)
No	589 (82)
After relapse or 2nd transplant	5 (<1)
Missing	47 (7)
Pancreatitis
Yes	4 (<1)
No	503 (70)
Missing	210 (29)
TTP/HUS
Yes	3 (<1)
No	637 (89)
Missing	77 (11)
Renal failure severe enough to warrant dialysis
Yes	2 (<1)
No	672 (94)
Missing	43 (6)
Stroke/seizures
Yes	20 (3)
No	620 (86)
Missing	77 (11)
Bronchiolitis obliterans
Yes	6 (<1)
No	542 (76)
After relapse or 2nd transplant	1 (<1)
Missing	168 (23)
Interstitial pneumonitis/idiopathic pneumonia syndrome
Yes	8 (1)
No	700 (98)
After relapse or 2nd transplant	2 (<1)
Missing	7 (<1)
Non-infectious liver toxicity
Yes	1 (<1)
No	701 (98)
After relapse or 2nd transplant	4 (<1)
Missing	11 (2)

There were no events at least one year after transplant and prior to relapse or second HCT for the following organ toxicities/late effects: avascular necrosis, myocardial infarction, pulmonary hemorrhage, and cryptogenic organizing pneumonia.

Table 3.

Cumulative Incidence of Organ Toxicities and Late Effects (occurring at least one year after HCT, reported in ≥3% of patients)

Organ Toxicities and Late Effects	Cumulative incidence* % (95% CI)
Growth hormone deficiency / growth disturbance	3-years	6 (4–7)
	5-years	11 (9–14)
	10-years	23 (19–28)
Cataracts	3-years	3 (2–5)
	5-years	10 (7–12)
	10-years	18 (15–22)
Hypothyroidism	3-years	5 (3–7)
	5-years	8 (6–10)
	10-years	13 (10–16)
Gonadal dysfunction / infertility requiring hormone replacement	3- and 5-years	0 (0–1)
	10-years	3 (2–5)
Stroke / seizure	3-years	3 (1–4)
	5-years	3 (2–4)
	8-years**	3 (2–5)
New malignancy	3-years	0
	5-years	1 (0–1)
	10-years	3 (1–5)

^*Death as competing risk. Patients censored at relapse or 2^nd HCT.

^**No events reported after 8 years.

In multivariable analysis, TBI exposure was an independent predictor of increased risk for cataracts (hazard ratio 17.2, 95% CI 7.4–39.8, p<0.001), growth hormone deficiency/growth disturbance (hazard ratio 3.5, 95% CI 2.2–5.5, p<0.001), and hypothyroidism (hazard ratio 5.3, 95% CI 3.0–9.4, p<0.001) (Table 4). Donor type and stem cell source were independent predictors of growth hormone deficiency/growth disturbance, with unrelated bone marrow or stem cell donor and unrelated cord blood donor HCT associated with greater risk (hazard ratio 2.9, 95% CI 1.5–5.6, p=0.0014; hazard ratio 2.2, 95% CI 1.1–4.5, p=0.023, respectively). Age 2–3 years at HCT was associated with higher risk of cataracts than younger age at HCT (hazard ratio 2.0, 95% CI 1.0–3.7, p=0.044). Females were more likely than males to be reported with gonadal dysfunction (hazard ratio 7.0, 95% CI 2.0–24.0, p=0.0019). Neither chronic GVHD at one year after HCT nor exposure to corticosteroid for GVHD prophylaxis was identified as risk factors for these reported outcomes.

Table 4.

Multivariable Analysis of Risk for Organ Toxicities and Late Effects (occurring at least one year after HCT, reported in ≥3% of patients)

Parameter	Category	N	Hazard Ratio (95% CI)	P-value
Cataracts
Age at HCT	<1 year old	146	1.00	0.027*
	1–1.99 year old	218	1.06 (0.52–2.14)	0.88
	2–2.99 year old	204	1.95 (1.02–3.74)	0.044
TBI	No	324	1.00	<0.0001*
	Yes	244	17.20 (7.44–39.75)	<0.0001
Gonadal dysfunction/infertility
Age at HCT	<1 year old	147	1.00	0.041*
	1–1.99 year old	220	0.60 (0.15–2.41)	0.47
	2–2.99 year old	204	2.36 (0.76–7.33)	0.14
Sex	Male	326	1.00	0.0019
	Female	245	7.03 (2.06–24.01)	0.0019
GH deficiency/growth disturbance
Donor/stem cell source	HLA-identical sibling	125	1.00	0.0025*
	Other related	29	0.31 (0.04–2.41)	0.2627
	Unrelated donor (BM or PB)	174	2.91 (1.51–5.62)	0.0014
	UCB (BM or PB)	231	2.23 (1.11–4.49)	0.024
TBI	No	339	1.00	<0.0001*
	Yes	220	3.51 (2.22–5.54)	<0.0001
Hypothyroidism
TBI	No	340	1.00	<0.0001*
	Yes	223	5.28 (2..97–9.40)	<0.0001

^*Overall p-value.

GH indicates growth hormone. UCB indicates umbilical cord blood. BM indicates bone marrow. PB indicates peripheral blood.

Significant predictors are presented. There were no significant predictors of stroke/seizure identified.

A subsequent malignant neoplasm was reported in 26 patients (3.6%). One subsequent malignancy was diagnosed less than one year after HCT (spindle cell tumor of the kidney) and 3 subsequent malignancies occurred after relapse or a second HCT (including diagnoses of AML, MDS, and brain tumor). The 22 subsequent malignant neoplasms occurring after one year following HCT and before relapse or a second HCT included the following diagnoses: thyroid cancer (8 cases), non-meningioma brain tumor (6), sarcoma (4), breast cancer (1), other non-CNS solid tumor (2), and one unspecified subsequent malignancy. The 10-year cumulative incidence of subsequent malignant neoplasm was 3% (95% CI 1–5%).

Discussion

HCT is an important treatment modality in infants and young children with very high risk and relapsed or refractory leukemias. However, there is little reported that considers a range of important outcomes after HCT in the first years of life, including survival, risk factors for mortality, cumulative incidence of relapse, second malignant neoplasms, and organ toxicties.^{4, 6–8, 11} As the numbers of survivors of pediatric HCT expand over time, understanding the issues faced by survivors of HCT at early age is increasingly important.¹² In a large, multi-center, multi-national cohort, we sought to characterize the outcomes of patients who underwent allogeneic, myeloablative HCT for hematologic malignancy early in childhood, here assessed as undergoing HCT at less than 3 years of age.

We report that those who survived relapse-free for at least one year after HCT had favorable overall survival and low subsequent transplant-related mortality. These are important findings for those caring for very young children undergoing HCT for hematologic malignancy. The favorable overall survival outcomes support the use of HCT as a successful treatment modality in the youngest of patients. Nonetheless, relapse of hematologic malignancy was found to be the leading cause of death after one year of relapse-free survival (as well as within the first year after HCT). Prior reports in high risk AML, infant ALL, and JMML have also shown that relapse is a leading cause of adverse outcomes after pediatric HCT.^1–3 Our results, in the context of a relatively large cohort of patients undergoing HCT in early childhood, reinforce the importance of investigations aimed at reducing the risk of relapse in this population.

Our results also demonstrate that late effects are common in this population. Hypothyroidism, growth hormone deficiency/growth disturbance, and cataracts appeared to increase in cumulative incidence over time, whereas the occurrence of stroke/seizure appeared to plateau. Endocrine toxicities were among the most frequently reported. Indeed, endocrine late effects have been previously reported with high prevalence after HCT in childhood, including in those undergoing HCT at very young age.^{8, 13–15} Bresters and colleagues described a prevalence of thyroid dysfunction after HCT at young age of 30%.⁸ In our cohort, the cumulative incidence of hypothyroidism increased over time, suggesting an ongoing risk of developing this late effect. Survivors of HCT in childhood are considered at risk for gonadal dysfunction and infertility.^{13, 14, 16} Here, among those undergoing HCT at age less than 3 years, females were more likely than males to be reported with gonadal dysfunction/infertility requiring hormone replacement, highlighting the importance of attention to pubertal development and reproductive function in female survivors of HCT. The reporting of gonadal dysfunction and infertility is anticipated to increase with further aging of the cohort, as many in this cohort were still too young to have gonadal dysfunction fully evaluated or diagnosed, supporting careful ongoing monitoring of both male and female survivors. Some prior reports have suggested higher frequencies of short stature and growth hormone deficiency after HCT at young age than reported here, which may reflect differences in age at follow-up, means of assessment of chronic conditions, or rates and doses of TBI and other exposures.^{6, 17} We note that the mechanism underlying the relationship seen between donor type and stem cell source and growth hormone deficiency/growth disturbance is unclear, but may be related to other treatment-associated factors. Overall, we recommend long-term monitoring of those undergoing HCT at very young age, as the potential for the development of morbidities with advancing age remains a concern.

In addition, our findings reinforce the risks of TBI at very young age. TBI exposure was associated with an increased risk of endocrine toxicities, including hypothyroidism and growth hormone deficiency/growth disturbance, in keeping with prior reports.^{8, 18} Survivors of pediatric HCT, particularly those exposed to TBI, are considered at greater risk of developing insulin resistance.^{15, 16, 19} In our cohort, only 1% of patients were reported with diabetes mellitus and glucose intolerance was not systematically captured. Continued monitoring will be needed to assess late endocrine effects with increasing age, such as manifestations of gonadal dysfunction, infertility, thyroid dysfunction, and glucose intolerance. TBI exposure was also an independent predictor of increased risk of cataracts. Radiation exposure, including TBI and cranial radiation, is a known risk factor for cataract development.^{20, 21} The extent of visual impairment or need for surgical intervention associated with cataracts in this cohort is not known. Despite these risks, TBI currently remains an important modality for members of this population. Our findings highlight the risks of TBI at very young age, and reinforce the importance of ongoing clinical follow-up with attention to these risks.

Our results also underscore the impact of chronic GVHD in those undergoing HCT at very young age. Chronic GVHD has been associated with an increased risk of adverse outcomes after HCT in pediatric and adult HCT. ^{13, 22–24} Here, chronic GVHD at one year after HCT was predictive of an increased risk of mortality. It is of note, however, that we did not demonstrate an increased risk for individual organ toxicities based on the occurrence of chronic GVHD at one year after HCT. It is possible that our exploration of the potential impact of GVHD on individual late effects may have been limited by patient numbers. Armenian and colleagues, reporting on survivors of childhood HCT from the Bone Marrow Transplant Survivor Study, demonstrated that active chronic GVHD was associated with high risk of adverse health status.¹³ We note that our analysis did not account for the ongoing status of chronic GVHD in relation to the risk for organ toxicities and late effects, which could be of interest in future investigations. Nonetheless, our findings highlight the importance of improving strategies for preventing and eliminating GVHD in those undergoing HCT early in childhood.

The development of a subsequent malignant neoplasm is a particularly concerning outcome after HCT. Here, the 10-year cumulative incidence of subsequent malignant neoplasm was 3% (95% CI 1–5%), in keeping with prior reports after pediatric allogeneic HCT.²⁵ Risk factors for the development of subsequent malignant neoplasms or specific diagnoses were not explored, given the overall low number of events. Survivors of HCT at very young age should be monitored with attention to an ongoing risk for second malignant neoplasms, as the number of subsequent neoplasms is anticipated to increase with age.^25–27 There were several cases of thyroid cancer in this cohort. While this reflects a small percentage of patients, it is of note that the overall annual incidence of thyroid cancer in children less than 15 years of age has been estimated at 2.0 cases per million people per year.²⁸ Radiation exposure, including TBI is a known risk factor for the development of thyroid cancer, and younger age has been associated with increased radiation risk.^29–31 Providers caring for survivors exposed to TBI at young age should have increased awareness of the risk of thyroid cancer, an understanding of evolving screening recommendations, and should counsel survivors about this risk. Current late effects surveillance recommendations support annual palpation of the thyroid for nodules by an experienced examiner.^{16, 32} Optimal screening schedules and modalities for secondary thyroid cancers in those with TBI exposure at very young age are an important area for further investigation.

There are several limitations to this study. First, CIBMTR data forms capture a discrete listing of late complications, with other potential late effects of interest, such as hypertension, dyslipidemia, dental abnormalities, skeletal toxicities and neurocognitive impairments, not systematically captured. Therefore, our finding that approximately one third of patients experienced at least one of the CIBMTR-designated organ toxicity or late effect may underestimate the full burden of post-HCT morbidities. In considering the frequency of late effects, the extent of missing values could result in over or underestimation of late effects. Other limitations include that additional details regarding reported toxicities are not available and that the study was not powered to assess potential risk factors for less frequently occurring toxicities. In addition, these results do not assess the impact of pre-HCT treatment exposures. Finally, data from early years may have more limited applicability to current practice. These results should be considered in the context of important general changes in the care of pediatric HCT patients over the 25 years evaluated, including trends of increased use of unrelated donor and umbilical cord blood sources, as well as improvements in supportive care and a greater emphasis on monitoring for late effects. Acknowledging these limitations, we believe this study provides a comprehensive characterization of outcomes after HCT for hematologic malignancy in the first years of life.

In conclusion, those who survived relapse-free for at least one year after allogeneic myeloablative HCT at a very young age for hematologic malignancy had favorable survival. Relapse was the leading cause of death and future investigation aimed at reducing the risk of relapse should remain a priority. Organ toxicities and late effects were frequent and long-term monitoring for organ dysfunctions and toxicities is therefore a critical element in the health care of survivors. Exposure to TBI emerged as an independent predictor of multiple frequently reported toxicities. Transplant survivors treated with TBI at age less than 3 years likely warrant heightened late effects surveillance. Future efforts should not only more precisely assess the impact of HCT at very young age on long-term neurocognitive development and function, growth, endocrine function, and emerging toxicities with advancing age, but also focus on interventions aimed at mitigating these toxicities.

Highlights.

• Ten-year OS estimate for those who survived relapse-free ≥1 year after HCT was 87%

• Thirty percent of survivors reported at least one late effect

• Growth disturbance, cataracts, and hypothyroidism were the most commonly reported

• TBI and chronic graft-versus-host-disease were associated with adverse outcomes

• Efforts should focus on reducing relapse and late effects after HCT at young age