What makes a pediatric or young adult patient an appropriate transplant candidate?

A 3-year-old child with chronic granulomatous disease was brought to the transplant clinic by his parents. The patient has a history of Aspergillus fumigatus pneumonia, which required mechanical ventilation, and sepsis, resulting in several intensive care stays. He has failure to thrive and developmental delay. His parents are seeking guidance whether allogeneic hematopoietic cell transplantation (HCT) is a reasonable treatment option given concerns about his upfront major health limitations.

Despite improvements in hct procedures and supportive care, nonrelapse mortality (NRM) remains a major contributor to death after allogeneic HCT, accounting for >60% of early deaths in those ≤18 years of age.¹  To assist in determining an individual patient's risk for NRM, the Hematopoietic Cell Transplantation-Comorbidity Index (HCT-CI) was created to capture incremental organ impairments and was validated to predict overall survival (OS) in both malignant²  and nonmalignant³  diseases. While its discriminative appeal has been proclaimed for its value in helping transplant physicians counsel recipients of all ages on transplant risks, the HCT-CI score has been rarely used by pediatric transplant physicians due to concern of its applicability to this age group.⁴  Children have been found to have fewer comorbidities per the HCT-CI compared to adults, resulting in relatively lower scores (0-1),⁵  despite the fact that they continue to experience substantial NRM.¹  This suggests that either the definitions used to assess HCT-CI do not comprehensively apply to children or that there are additional risk factors that are missing for this age group.⁶  This perspective also applies to adolescents and young adults (AYA)⁷ ; although they are older and therefore more likely to have testable organ dysfunction compared to infants and children, AYA also tend to have other nondisease related concerns, such as environmental and social dysfunction, that can affect OS after HCT. “Sentiment” tends to be more emotional. Finally, nonmalignant diseases represent a large percentage of indications for allogeneic HCT in children and AYA, but many of these diseases are individually unique and are predisposed to developing distinct organ toxicities that may not be fully reflected in the HCT-CI. It is also possible that these disease-defining comorbidities may be treated or cured by HCT, therefore becoming less problematic for predicting OS. Here, we discuss 1) updated applications of the HCT-CI in children and AYA populations (collectively named youth) and 2) other ways to improve pretransplant risk assessment in youth.

One method for improving applicability of HCT-CI for youth is to expand definitions and include new items that are prevalent in this age group.^8,9  For example, lack of pulmonary function testing in young children limited proper evaluation of their pulmonary comorbidities. To overcome this limitation, the definition was expanded to include history of mechanical ventilation. History of previously treated fungal infection was counted toward expanding infection comorbidity. Definitions for renal comorbidities were expanded to include the estimated glomerular filtration rate (eGFR) using either the Bedside Schwartz equation for children or the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation for young adults. Finally, failure to thrive or being underweight was added to the index to capture nutritional deficiencies more comprehensively. These additions were first tested in training sets and then validated in independent validation sets with discriminatory capacity confirmed by C-statistic (Table 1).

Despite being developed and validated in both adult and pediatric populations, adult patients and their risk factors were overly represented in the development of the HCT-CI, therefore creating the impression that it is a more adult-focused scale.²  To address this issue, comorbidities with hazard ratios <1.2 were eliminated, as these would indicate comorbidities that had minimal contribution to the predictive capacity of the model in the youth population.^8,9  Comorbidities that were eliminated in both the malignant and nonmalignant groups were arrhythmia, psychiatric disease, mild hepatic disease, moderate pulmonary disease, and peptic ulcer disease. Additional items removed for malignant diseases included inflammatory bowel disease, cardiovascular disease, being underweight, peptic ulcer disease, and rheumatologic disease, while obesity was the only additional factor removed for nonmalignant diseases (Table 1).

Applying these two strategies, two studies were performed. Youth with nonmalignant diseases (n = 2815) who received their first allogeneic HCTs between 2008 and 2017 were included in the first study to develop the youth-nonmalignant-HCT-CI (ynHCT-CI).⁹  These modifications resulted in 39% of patients having an increase in their ynHCT-CI scores with an increased hazard of mortality compared to those whose score remained the same (hazard ratio, 1.41; 95% CI, 1.01-1.98). Performance of the new model was slightly better than that of the original HCT-CI (C-statistic estimates of 65.8 versus 64.3, respectively), but the main advantage was that it represented a more youth- focused scale. Likewise, 5790 youths with malignant diseases contributed to development of the youth-malignant-HCT-CI.⁸  The youth-maliginant-HCT-CI led to an increase in comorbidity scores for 23% of youth patients and was associated with a significant risk of NRM (HR, 1.34; 95% CI, 1.02-1.74). The 3-year survival rates were 62.9%, 53.3%, and 50.1%, respectively for scores 0, 1-2 and ≥3.

Further fine-tuning of the youth HCT-CI will require thoughtful attention to data collected in future prospective studies. Psychiatric comorbidities could be expanded to include certain behaviors (eg, aggression) that have that been connected to untreated depression or anxiety in children.¹⁰  Another example is better defining comorbidities linked to specific primary nonmalignant diseases. For instance, redefining iron overload by using T2* magnetic resonance imaging or including the number of intensive care hospitalizations could better capture the burden of hemoglobinopathy-associated risk factors. It is also unknown whether elevated pre-HCT baseline biomarkers, such as urine protein to creatinine ratio or soluble C5b-9 levels, could capture individuals at higher risk for life-threatening post-HCT complications such as transplant-associated microangiopathy (Table 2).¹¹ 

Aside from comorbidities, other risk factors could further enhance our understanding of HCT risks in youths. These include categories such as neuropsychiatric conditions and socioeconomic factors.¹²  Finally, genetic variants found in recipients and their donors could adversely impact outcomes and are an emerging area of research.^13,14  A list of these potential factors is detailed in Table 2. Future prospective studies are needed to enhance risk-assessment potential for youth recipients of allogeneic HCT.

Based on the original HCT-CI, this child's risk for NRM would be negligible with a score of 0. With use of the validated ynHCT-CI, the score increases to 5, due to prior mechanical ventilation (+3), history of fungal infection (+1), and being underweight (+1), with at least a 2-fold increase in risk of NRM. The role of developmental delay is unclear and not currently validated to prognosticate survival. While HCT was ultimately recommended in this case, the family was counseled to have a more realistic sense of NRM risk.

The HCT-CI was introduced almost two decades ago and has been reliable in providing data-driven survival predictions for transplant recipients. The newly validated youth scores can further assist transplant physicians in counseling families with children. When counseling patients about transplants, additional risk factors (Table 2) need to be considered, especially those that may be discovered in future studies and added to the current models.

Based on our review of the literature, we offer the following recommendations:

1. The HCT-CI assessment should be performed in all pediatric patients undergoing HCT to assess risk of NRM. (Grade 1A).

2. Either the expanded or simplified youth HCT-CI, which provide improved definitions for certain pediatric comorbidities, should be assessed in all youth to determine risk of NRM, although real-life experience using this new scale is limited. (Grade 1B).

3. At this time, we suggest consideration of other pretransplant risk factors reported in the literature that could contribute to mortality after HCT. While they have not been validated to date in the transplant setting, future studies could broaden their applicability. (Grade 1C).

We would like to acknowledge the contributions of Drs Larisa Broglie and Brian Friend to this body of work. We thank Ms Helen Crawford for her assistance in manuscript preparation. M.L.S. was funded by research grant award R01 CA227092 from the National Cancer Institute of the National Institutes of Health. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Monica S. Thakar: on the scientific advisory boards for Proteios Technology and ImmunoVec.

Mohamed L. Sorror: reports consultancy for and receiving an honorarium from Jazz Pharmaceuticals and receiving a research grant from Massachusetts General Hospital and BlueNote.

Monica S. Thakar: Nothing to disclose.

Mohamed L. Sorror: Nothing to disclose.