Symptoms of Survivors of Pediatric Hematopoietic Stem Cell Transplant by Age, Sex, and Transplant Type

Abstract

Purpose: The purpose of this study was to describe symptoms experienced by survivors of pediatric hematopoietic stem cell transplant (HSCT), and demographic and treatment-factors associated with ongoing symptomology. Methods: Fifty pediatric survivors completed a cross-sectional pilot study. Questionnaires were administered online via REDCap to assess symptoms experienced in the last week. Survivors also consented to a medical record chart review. Results: Survivors were on average 5.4 years post-HSCT (range 1.1 to 9 years), male (58%), and Caucasian (80%) who received an allogeneic HSCT (92%). The most commonly reported symptoms were difficulty concentrating (42.5%), pain (38%), worry (38%), nervousness (37.5%), and lack of energy/fatigue (34%). Survivors reported up to 14 symptoms, with 90% of the sample experiencing at least one symptom in the previous week. Average number of symptoms varied by age group between 2.1 (8–9 years) and 6.8 (18 and older). Age and female gender were associated with higher levels of fatigue. Conclusions: The majority of survivors experienced at least one symptom in the previous week. Neuropsychological symptoms and pain endure well into survivorship that can influence outcomes such as function and health-related quality of life (HRQOL). Research is needed on biological mechanisms of ongoing symptomology, effective interventions to prevent or mitigate symptoms, and the impact of symptoms on patient outcomes including daily functioning and HRQOL. Implications Survivors of pediatric HSCT continued to experience symptoms for up to nine years. Survivors should be frequently screened for symptoms, as symptoms may affect function, learning/employment outcomes, and HRQOL.

Introduction

In the United States, more than 2,800 children, adolescents, and young adults under the age of twenty-one receive hematopoietic stem cell transplants (HSCT) each year (Health Resource Services Administration, 2018). Pediatric HSCT is often the only available curative treatment option for patients with a variety of conditions, including oncologic, hematologic, immunologic, and metabolic disorders. Patients undergoing HSCT receive chemotherapy, radiation, immunosuppressants, as well as other treatment and medications, which can have acute as well as long-term health consequences. As the number of HSCT survivors continues to increase (Majhail et al., 2013), a better understanding of treatment-related effects, risk factors, and effective interventions are needed.

Symptoms result from treatment-related effects that can have debilitating consequences on survivors of HSCT. During the acute phase of HSCT treatment, pediatric patients report a variety of psychological and somatic symptoms such as anxiety, depression, behavioral and social problems, post-traumatic stress reactions, pain, nausea, appetite loss, sleep disturbances, drowsiness, fatigue, and anorexia (Packman et al., 2010; Phipps et al., 2002; Ullrich et al., 2018; Vasquenza et al., 2015). Although many symptoms tend to decline over the course of treatment (Phipps et al., 2002), survivors of pediatric HSCT report persisting difficulties with a variety of treatment-related late-effects up to 12 years post-HSCT, such as post-traumatic stress disorder, somatic complaints (e.g., pain, fatigue, sleepiness), and impaired social competence which encompasses social skills, social communication, and interpersonal communication (Graef et al., 2016; Phipps et al., 1995; Taskiran et al., 2016; Vannatta et al., 1998; Zanato et al., 2017). In addition, symptoms, such as physical and mental fatigue, pain, and emotional distress have been found to affect survivors of pediatric HSCT well into adulthood (Berbis et al., 2013; Hacker et al., 2017; Kenzik et al., 2015). Thus, the long-term symptoms experienced by survivors of pediatric HSCT may contribute to the impediment of quality of life during both childhood (Packman et al., 2010; Vasquenza et al., 2015) and adulthood (Cohen et al., 2012; Kenzik et al., 2015; Sanders et al., 2010; Schultz et al., 2014).

The National Institutes of Health (NIH) Hematopoietic Cell Transplant Late Effects Initiative (Bevans et al., 2017) was a working group tasked to report on the state of the science of the health-related quality of life (HRQOL), including patient-reported outcomes, and literature for pediatric and adult HSCT survivors. The group reported that many late effects and patient-centered outcomes are not well understood in HSCT survivors, particularly in pediatric HSCT survivors, due to lack of consistent measurement and timing of administration (Bevans et al., 2017). Late effects noted to be poorly understood included symptoms such as pain, fatigue, sleep dysfunction, distress, depression, and cognitive dysfunction, and mechanisms by which symptoms occur. It was also noted that global HRQOL was not well defined for pediatric survivors of HSCT (Bevans et al., 2017).

Previous research examining the symptoms of adult survivors of pediatric cancer demonstrated that pain, fatigue, sleep disturbance, psychological distress, and difficulty concentrating are common (Finnegan et al., 2009). Utilizing these findings, Finnegan and colleagues (Finnegan et al., 2010) developed the Symptom Cluster Experience Profile Framework (SCEP). The SCEP Framework depicts the process through which risk and protective factors combine with stress status to act upon subgroups of symptom profiles, either mitigating or exacerbating symptoms, and the resulting impact on one's quality of life. Risk and protective factors include cancer and treatment exposures, comorbidities, lifestyle behaviors, affect, social support, physical environment, and stressful life events. These factors combined with the resulting stress response influence symptom experience, which in turn influence quality of life and functional status (Finnegan et al., 2010).

As the life expectancy of HSCT recipients continues to increase, it is necessary for clinicians to better understand the symptoms patients are experiencing, the risk and protective factors for ongoing symptomology, the impact of symptoms on HRQOL, and patient-centered outcomes in order to inform the development of appropriate assessment and treatment options. A research study was undertaken to test a conceptual model adapted from SCEP (Finnegan et al., 2010), in order to better understand symptoms experienced by survivors of pediatric HSCT, and potential demographic and treatment-related variables that are associated with symptom experience and HRQOL. The purpose of this study was to describe symptoms experienced by survivors of pediatric HSCT, and demographic and treatment-factors associated with ongoing symptomology. It was hypothesized that reduced-intensity conditioning would be associated with an endorsement of fewer symptoms. It was also hypothesized that allogeneic transplant would predict higher symptom endorsement than autologous or cord blood transplant when controlling for comorbid conditions (i.e., graft-versus-host-disease), age, and gender.

Methods

Design

We performed an observational cross-sectional descriptive pilot study of pediatric survivors and their caregivers to better understand the symptoms reported by survivors of pediatric HSCT as well as possible demographic and treatment variable associated with symptom types. Questionnaires were administered using Research Electronic Data Capture (REDCap) at a single time point. In addition to questionnaire data, an electronic medical record (EMR) chart review was completed to collect additional demographic and HSCT data.

Sample

The sample for this study was comprised of families of children who were at least one year post-HSCT. Dyads of survivors of pediatric HSCT and a caregiver were recruited, however, if a survivor wished to participate but a caregiver declined, the survivor was still eligible to participate with either consent if the survivor was 18 years of age or older, or with assent and parental permission. Participants were recruited until the goal sample size of 50 dyads of survivors and caregivers was met. The study was comprised of a convenience sample conducted within a cohort of previous research subjects. All participants had agreed to be re-contacted for additional research during a previously completed research study conducted by one of the authors (AP). Survivors of pediatric HSCT were eligible if they were ages 8 to 27 years old, at least one-year post-HSCT, and able to read and speak English. The upper age limit was determined by the oldest participant from the cohort of previous research subjects and the lower age limit was related to the lower age limit for questionnaire self-report. Survivors were eligible if they were inpatient, however, none of the patients approached for recruitment were inpatient at the time. Caregivers were excluded if they were under the age of 18 or unable to speak or read English, and survivors were excluded if they were too ill to participate (i.e., intubated, actively dying).

Instruments

Various questionnaires were used in this study to measure symptoms of survivors of pediatric HSCT including Patient-Reported Outcomes Measurement Information System (PROMIS) measures and the Memorial Symptom Assessment Scale (MSAS). There are three versions of the MSAS that vary the number of symptoms assessed based on the age of the participant: the MSAS measures 32 symptoms in adults 18 years of age or older, the MSAS (10–18) measures 30 symptoms in 10- to 18-year-olds, and the MSAS (7–12) measures 8 symptoms in 7- to 12-year-olds.

The MSAS is validated for use with adults 18 years of age and older, and was administered to study participants 18 years of age and older. The MSAS includes 32 physical and psychological symptoms with questions on prevalence/frequency, severity of the symptom, and symptom distress. Symptom scores are calculated by taking the average of the frequency, severity and distress for each individual symptom. The MSAS physical symptom subscale ([MSAS-PHYS], α  =  .87) is the average distress associated with the 12 prevalent physical symptoms (lack of appetite, lack of energy/fatigue, pain, feeling drowsy, constipation, dry mouth, nausea, vomiting, change in taste, weight loss, feeling bloated, and dizziness). The physical subscale is evaluated in increments of 0.8 with zero being no symptoms present, 0.8 being symptoms present but no distress, and increasing levels indicating increased distress. The psychological symptom subscale ([MSAS-PSYCH], α  =  .83) is the average frequency of six prevalent psychological symptoms (worry, feeling sad, feeling nervous, difficulty sleeping, feeling irritable, and difficulty concentrating). The psychological symptom subscale is scored in increments of one, with zero for no symptom and four for “almost constantly”. The Global Distress Index (GDI) is a measure of the overall symptom distress for 10 prevalent symptoms, and is calculated by taking the average frequency of four psychological symptoms (feeling sad, worry, feeling irritable, and feeling nervous) and the average distress of six physical symptoms (lack of appetite, lack of energy/fatigue, pain, drowsy, constipated, and dry mouth).

The Total MSAS score is the average symptom scores of all 32 symptoms, α  =  .76-.89 (Lynch Kelly et al., 2015; Portenoy et al., 1994).

The MSAS (10–18) is validated for use in children ages 10 to 18 and was administered to study participants ages 10 to 17. The MSAS (10–18) is similar to the MSAS with two items removed (sexual function and bloating) for a total of 30 symptoms. The MSAS (10–18) has been used in pediatric cancer populations (α  =  .85) and the scoring is the same as the full 32-symptom MSAS questionnaire (Yeh et al., 2009).

The MSAS (7–12) is validated for use in children ages seven to twelve, and includes 8 physical and psychological symptoms (fatigue, sadness, itching, pain, worry, lack of appetite, vomiting, insomnia). The MSAS (7–12) was created as a shorter version when children under the age of 10 had difficulty completing MSAS (10–18) (Collins et al., 2002). Although validated in children eight to twelve years old, the MSAS (8–12) was only administered to 8- and 9- year olds in this study in order to maximize the number of possible symptoms available for assessment (Collins et al., 2000; Collins et al., 2002).

The PROMIS-29 v2.0 and the Pediatric PROMIS-25 v2.0 were also administered to measure symptoms of survivors. The PROMIS-29 v2.0 was administered to participants 18 years of age and older and is comprised of 29-items divided into subscales measuring fatigue, anxiety, depression, sleep disturbance, pain interference, and pain intensity. There are also subscales of physical function and social function. Participants rate items experienced over the past seven days on a 5-point Likert scale ranging from 1, not at all, to 5, very much. All raw scores are converted to T scores. Scoring interpreted according to the American Psychological Association level 2 guidelines where less than 55 is interpreted as none to slight, 55–59.9 as mild, 60–69.9 as moderate, and ≥70 as severe. Hays et al., also validated physical (α = .98) and mental health (α = .97) summary scores (Craig et al., 2014; Hays et al., 2018). Participants 18 years of age or older completed the PROMIS-29. The PROMIS-25 v2.0 was administered to participants 17 years and younger and has 25-items divided into subscales of anxiety, depression, fatigue, pain interference, and one question on pain intensity. There are also subscales for physical function mobility and peer relationships. Participants rate items experienced over the past seven days on a 5-point Likert scale ranging from 1, not at all, to 5, very much. All raw scores are converted to T scores where greater than 50 are considered higher than the general population with a chronic condition (including cancer).

Survivors also completed two instruments on pain quality: the PROMIS Pain Quality-Sensory SF 8a and PROMIS Pain Quality-Affective SF 8a. The PROMIS Pediatric Pain Quality-Sensory short form 8a asks participants to rate how their pain felt over the last seven days and consists of 8-items on a 5-point Likert scale ranging from 1, not at all, to 5, very much. The PROMIS Pediatric Pain Quality-Affective short form 8a asks each participant how their pain felt over the last seven days and consists of 8-items which are rated in a yes or no format. All raw scores were converted to T scores where greater than 50 is considered higher than the general population with a chronic condition.

Procedures

Following Institutional Review Board approval, patients and caregivers were consented, or assented with parental permission if they were under 18 years of age. Patients were recruited from the HSCT Survivorship Clinic of a Midwestern Children's Hospital. Patients were recruited in person at a clinic visit or over the phone.

Data sources included questionnaires collected at a single time point and an electronic medical record (EMR) chart review. Questionnaires were administered using REDCap (Harris et al., 2009) either in person on an iPad, or via a secure link sent by email. Several participants requested paper questionnaires, which were then double entered into the database by members of the research team (SD, NB) and validated to ensure data were entered correctly. Questionnaires took participants between 15 to 45 min to complete and they were compensated for their time with a $50 debit card. An EMR chart review was completed to collect additional demographic and treatment-related data (i.e., transplant type, donor source, conditioning regimen).

Analysis

The analysis was completed using SPSS version 25 (IBM, Armonk, NY) statistical software. Descriptive statistics (i.e., frequency, percentage, mean) were used to describe the sample and symptoms. Continuous variables were analyzed using Pearson's correlation coefficient. ANOVA was used to examine the relationships between nominal demographic and treatment variables and symptoms. Significance was set at .05. Cohen; (1988) methods were used for determining correlational effects sizes (.1 small effect, .3 moderate effect, and .5 or greater large effect).

Results

Seventy eligible patients and caregiver dyads were contacted for participation in the study. Of the 70 dyads, 15 did not respond, 4 declined, and 51 enrolled for a recruitment rate of 93%. One 19-year-old male did not complete the questionnaires and was excluded from the analysis, for a total sample size was 50 pediatric survivors of HSCT and 51 family caregivers. Pediatric HSCT survivors included in the analyses (n  =  50), were on average 14.2 years of age (range 8 to 23) at the time of their study visit. The mean age at HSCT was 9.3 (range 2.2 to 21.9) years of age, and the average time between HSCT and their study visit was 5.4 years (range 1.1 to 9). More than half the sample were male (58%). The majority were Caucasian (80%) race and non-Hispanic ethnicity (90%).

The majority (92%) of participants received allogeneic transplants with ablative conditioning (74%) and had matched unrelated donors (66%). The majority of matched related donors (n  =  13, 26%) were siblings with one mother serving as a donor. Participants were transplanted for a variety of conditions including bone marrow failures, immune disorders, hematologic disorders, and cancer diagnoses. See Table 1 for more detail.

Table 1.

Sample Demographics.

Characteristic	Range	Average
Age at transplant (years)	2.2-21.9	9.3
Time since transplant (years)	1.1–9	5.4
	Number (N = 50)	Percentage
Age at study visit (years)
8–12	19	38
13–17	20	40
18 & older	11	22
Male	29	58
Race
Caucasian	40	80
African–American	6	12
Mixed	3	6
None selected	1	2
Ethnicity
Non-hispanic	45	90
Hispanic	5	5
Diagnosis
Bone marrow failure	18	36
Aplastic anemia	3	6
Fanconi anemia	12	24
Dyskeratosis congenita	1	2
Blackfan diamond anemia	1	2
Immune disorder	10	20
HLH	5	10
SCID	2	4
CGD	2	4
XLP	1	2
IPEX	1	2
Hematologic disorder	2	4
Sickle cell anemia	2	4
Cancer	20	40
AML	7	14
ALL	6	12
MDS	3	6
Neuroblastoma	1	2
Ewing's sarcoma	1	2
PNET	2	4
Transplant type
Allogeneic	46	92
Autologous	4	8
Donor type
Matched related	13	26
Matched unrelated	33	66
Self	4	8
Graft type
Bone marrow	26	52
Peripheral blood stem cells	19	38
Cord blood	3	6
Cord  +  bone marrow	2	4
Conditioning
Reduced intensity	13	26
Ablative	37	74
Graft versus host disease (GVHD)
Acute GVHD	16	32
Chronic GVHD	18	36

Note. HLH  =  hemophagocytic lymphohistiocytosis; SCID  =  severe combined immune disorder; CGD  =  chronic granulomatosis disease; XLP  =  x-linked lymphoproliferative disorder; IPEX  =  immunodysregulation polyendocrinopathy enteropathy x-linked; AML  =  acute myeloid leukemia ALL  =  acute lymphoblastic leukemia; MDS  =  myelodysplastic syndrome; PNET  =  primitive neuroectodermal tumors.

Based on the MSAS, the top five symptoms experienced by survivors of pediatric HSCT were difficulty concentrating (42.5%), pain (38%), worry (38%), nervousness, (37.5%) and lack of energy or fatigue (34%). The average number of symptoms varied by age group with 8 to 9 year olds reporting they experienced 2.1 (SD ± 1.4) out of 8 symptoms, 10 to 17 year olds experienced 4.7 (SD ± 3.6) out of 30 symptoms, and those 18 years of age and older experienced 6.8 (SD ± 5) of 32 symptoms in the previous week. Please see Table 2 for additional detail.

Table 2.

MSAS Symptom and Scale Descriptive Statistics.

	MSAS 8-9yo (n  =  10)		MSAS 10-17yo (n  =  29)		MSAS 18 ± (n  =  11)
	n	Mean (SD)	n	Mean (SD)	n	Mean (SD)
Symptom
Lack of energy	4	1.00 (0)	8	1.96 (0.81)	5	1.73 (0.72)
Sad	0	0	7	1.86 (1.03)	6	1.50 (0.55)
Itchy	4	1.33 (0.38)	4	1.42 (0.42)	2	1.50 (0.71)
Pain	2	1.67 (0.47)	11	2.00 (0.80)	6	2.06 (0.77)
Worried	1	1.33	11	1.61 (0.49)	7	1.52 (0.5)
Lack of appetite	1	0.50	6	1.28 (0.77)	1	1.33
Vomiting	1	1.00	0	0	0	0
Insomnia	1	2.00	8	2.50 (0.69)	2	2.33 (1.41)
Difficulty concentrating	-	-	12	1.64 (0.56)	5	1.73 (0.72)
Cough	-	-	6	1.50 (0.46)	4	1.33 (0.38)
Feeling nervous	-	-	10	1.60 (0.56)	5	1.73 (0.64)
Dry mouth	-	-	3	1.22 (0.51)	0	0
Nausea	-	-	2	1.33 (0.94)	3	1.00 (0.33)
Feeling drowsy	-	-	8	1.33 (0.56)	5	1.80 (0.77)
Numbness/tingle in hands/feet	-	-	3	1.89 (0.19)	5	1.53 (0.51)
Problems urinating	-	-	0	0	0	0
Short of breath	-	-	7	1.33 (0.61)	1	0.67
Diarrhea	-	-	2	1.33 (0.94)	2	1.33 (0.94)
Sweats	-	-	2	0.67 (0)	1	3.33
Dizziness	-	-	2	1.50 (0.71)	0	0
Difficulty swallowing	-	-	0	0	0	0
Feeling irritable	-	-	7	1.52 (0.54)	3	1.78 (0.69)
Mouth sores	-	-	4	1.13 (0.48)	2	2.00 (0)
Change in taste	-	-	2	0.50 (0)	0	0
Weight loss	-	-	3	0.50 (0)	1	0.50
Hair loss	-	-	2	2.00 (0.71)	3	2.33 (0.29)
Constipation	-	-	3	1.67 (1.26)	0	0
Swelling arms/legs	-	-	0	0	1	3.00
“I don’t look like myself”	-	-	0	0	1	2.50
Changes to skin	-	-	1	4.00	1	0.50
Feeling bloated	-	-	-	-	1	3.67
Problems with sexual activity or interest	-	-	-	-	1	1.67
Total number of symptoms	10	2.1 (1.37)	29	4.69 (3.60)	11	6.8 (5)
Global distress index				1.39 (0.83)		1.56 (0.48)
Psychological symptom scalea				1.99 (0.73)		1.74 (0.55)
Physical symptom scaleb						1.30 (0.71)
Total MSAS score						1.51 (0.46)

Note. Mean symptom score calculated by averaging frequency, severity and distress of each symptom. MSAS = Memorial Symptom Assessment Scale. Dash (-) indicates symptom not assessed on questionnaire.

^a Psychological symptom scale calculated by averaging symptom frequency for 6 symptoms.

^b Physical symptom scale calculated by averaging symptom distress for 12 symptoms.

According to the MSAS (7–12), completed by participants between the ages of eight and nine (n  =  10), nine out of the ten participants experienced at least one of the eight symptoms on the questionnaire. Three participants in this age group experienced one symptom, and two participants experienced two, three, and four symptoms over the last week. On average, participants in this age group experienced two symptoms in the previous week that occurred for a short frequency and were mildly distressing. The two most reported symptoms for this group were “feeling tired” and itching which were both endorsed by four participants. Tiredness occurred at a short to medium frequency with little to no distress. Itching occurred for one of the four at a medium frequency and for two with moderate severity, causing little to moderate distress. Worry and insomnia were both endorsed by one participant each in this age group with moderate distress.

The MSAS (10–18), which measures 30 symptoms, was completed by 10- to 17-year-olds (n  =  29). Of the 29 participants in this age group, four reported not experiencing any symptoms during the previous week. The mean number of symptoms was 4.7 (SD ± 3.6) symptoms, with a maximum of 14 symptoms reported by a single participant for this age group. The top four symptoms reported by this age group were difficulty concentrating (41.4%), pain (38%), worry (38%), and “feeling nervous” (34.5%). While most symptoms endorsed as being present over the past week were rated as mildly to moderately severe and distressing, there were a few symptoms a portion of participants rated as highly severe and distressing. Seven participants (24.1%) endorsed “feeling irritated”, 11 endorsed “feeling worried” (38%), and eight endorsed “lack of energy”/fatigue (27.6%) with moderate severity and distress. The majority of participants who endorsed “feeling sad” (n  =  7, 24.1%) rated mild to moderate severity and distress, however, there was a participant who rated high severity and distress. The majority of participants rated difficulty concentrating (n  =  12, 41.4%) as mildly to moderately severe and distressing, however, several participants rated it highly severe and distressing. Insomnia was rated by the majority (n  =  8, 27.6%) as moderate to high in severity and distress.

The MSAS measures 32 symptoms and was completed by 11 participants 18 years of age and older. The top three symptoms endorsed by this age group were feeling worried (63.6%), feeling sad (54.4%), and pain (54.4%). All participants in this age group endorsed at least one symptom with a range of one to fourteen symptoms and a mean of 6.8 (SD ± 5) symptoms experienced in the previous week. One participant experienced 14 symptoms, one 12 symptoms, and three experienced 11 symptoms. Although there were several symptoms that were only endorsed by one or two participants in this age group, they were rated as moderate to high in severity and distress including insomnia, sweats, hair loss, swelling in arms and legs, bloating, and “I don’t look like myself”. Please see Table 2.

The MSAS Global Distress Index (GDI) is an approximation of overall symptom distress based on ten prominent symptoms. The GDI for 10- to 17-year-olds was 1.39 (SD ± 0.83), and for those 18 years of age and older it was 1.56 (SD ± 0.48), meaning participants experienced mild to moderate overall symptom distress. The Psychological Symptom Scale (MSAS-PSYCH) is an indicator of the amount or frequency of symptoms participants are experiencing. The MSAS-PSYCH score for 10- to 17-year-olds was 1.99 (SD ± 0.73), and for those 18 years of age and older it was 1.74 (SD ± 0.55). The MSAS-PSYCH is evaluated in increments of one, indicating the frequency that participants in this study were experiencing symptoms occasionally to frequently.

The Physical Symptoms Scale (MSAS-PHYS) and Total MSAS score were only able to be calculated in the full version of the MSAS administered to participants 18 years of age and older (n  =  11) since the MSAS-PHYS includes bloating, which is only on the full scale, and the Total MSAS score is calculated with the symptom scores of all 32 symptoms. The MSAS-PHYS is an indicator of the amount of symptom distress participants are experiencing and is evaluated in increments of 0.8. Participants who completed the MSAS full version had an overall MSAS-PHYS score of 1.30 (SD ± 0.71) indicating symptoms are present and mildly distressing. This also coincides with the Total MSAS score which was 1.51 (SD ± 0.46), indicating symptoms are present in low to moderate amounts and mildly distressing overall.

The PROMIS-25 v2.0 was completed by participants 17 years of age and younger (n  =  39). Means for all subscale T scores fall at or below national average for chronic conditions, however, the majority of subscales had maximum T scores within a moderate range (60–69.9). Similar results were seen on the PROMIS-29 v2.0 completed by participants 18 years of age and older (n  =  11). The PROMIS Pain Quality-Affective SF 8a was completed by 22 participants. The T scores ranged from 34.6 (none/slight) to 63 (moderate), with a mean of 44.06 (SD ± 6.59). Similarly, the PROMIS Pain Quality-Sensory SF 8a (n  =  22) had a range from 36.1 (none/slight) to 54.7 (mild) with a mean of 44.57 (SD ± 4.32). See Table 3 for more detail.

Table 3.

PROMIS Symptom T-Score Descriptive Statistics.

	PROMIS 25 (N  =  39)			PROMIS 29 (N  =  11)
Construct	Mean	SD	Range	Mean	SD	Range
Physical function	52.02	7.62	35–57.1	48.49	8.24	41–56.9
Anxiety	45.82	8.57	35.6–59.7	48.90	8.23	40.3–63.5
Depression	51.15	7.59	37.7–68.8	50.69	7.23	41–62.2
Fatigue	43.62	8.92	35.4–61.7	53.82	8.49	33.7–57.1
Peer relationships/social roles	44.91	7.97	34.5–61.1	49.31	9.67	33.7–57.1
Pain	44.41	9.18	36.7–65.3	46.23	8.71	41.6–61.3
Sleep				45.20	8.49	32–66.4

Note. PROMIS = Patient-Reported Outcomes Measurement Information System. Sleep not present on the PROMIS 25.

Relationships Between Variables

The Total number of symptoms endorsed on the MSAS was also compared by time since HSCT. Participants were placed in groups by time since HSCT: one to three years post HSCT (n  =  12, 24%), four to six years post HSCT (n  =  20, 40%), and seven to nine years post HSCT (n  =  18, 36%). The mean age per group was 13, 13.95, and 15.39 years respectively. The average number of symptoms increased as time since HSCT increased, however, this was not significant (p  =  .089). Participants one to three years post HSCT reported an average of 2.59 (SD ± 2.5, CI [0.99–4.17]) symptoms, four to six years post HSCT reported an average of 4.9 (SD ± 3.68, CI [3.18–6.62]) symptoms, and those seven to nine years post HSCT reported an average of 5.72 (SD ± 4.54, CI [3.47–7.98]) symptoms.

Pearson's correlations between MSAS and demographic variables are displayed in table 4. Correlations were run between MSAS symptoms and total chemo and radiation doses without significance. Correlations were also calculated with demographics and MSAS symptoms of itching, vomiting, cough, dry mouth, problems urinating, diarrhea, sweats, dizziness, difficulty swallowing, feeling irritable, mouth sores, change in taste, weight loss, constipation, swelling in arms and legs, change in looks, changes to skin, feeling bloated, and problems with sexual activity or interest without significance.

Table 4.

Pearson Correlation Table of Demographics and MSAS.

	Patient age	Age at HSCT	Time since HSCT	Fatigue	Sad	Pain	Worry	Lack of appetite	Insomnia	Difficulty concentrating	Nervous	Nausea	Drowsy	Numbness/ tingling	Short of breath	Hair loss	Total number of symptoms	Psych subscale	Physical subscale	Global distress index
Patient age	1
Age at HSCT	.876**	1
Time since HSCT	.259	-.234	1
Fatigue	.370	.408	.024	1
Sad	-.083	.122	-.330	.781	1
Pain	.035	.072	-.127	.235	.485	1
Worry	.140	.043	.116	.744	.715*	.739*	1
Lack of appetite	.546	.609	.074	.901	1.00**	.855	.817*	1
Insomnia	.088	-.023	.189	.730	.160	.473	.319	.826	1
Difficulty concentrating	.152	.142	.135	.410	-.133	.088	.428	.400	-.075	1
Nervous	.112	.205	-.142	.054	.541	.520	.412	.000	-.152	.109	1
Nausea	-.190	-.495	-.088	-.381	-.125	.747	-.500	c	1.00**	.452	-.870	1
Drowsy	.476	.495	.134	.657	.510	.069	.488	1.00**	.163	.669	.300	.191	1
Numbness/ tingling	-.280	-.217	-.522	.690	.441	-.631	.839	c	c	-.135	.404	-.500	.309	1
Short of breath	-.372	-.229	-.165	.884**	.844	.000	.982	.866	.866	.500	.186	−1.0**	.978**	c
Hair loss	.618	.889*	-.610	.591	c	.686	.577	1.00**	c	1.00**	.868	.500	.867	.189	1.00**	1
Total Number of symptoms	.424**	.259	.325*	.662**	.379	.411	.553*	.643	.272	.297	.612*	-.891*	.630*	-.249	.491	.825	1
Psych subscale	-.083	-.111	.025	.768**	.806**	.697**	.846**	.861*	.672*	.673**	.657**	.074	.609*	-.011	.776*	.834	.326	1
Physical subscale	.438	.213	.104	.981**	.549	.411	.774	c	1.00**	.851	.731	.974	.953*	.597	c	.881	.688	.826*	1
Global distress index	.083	.119	-.096	.673*	.843**	.819**	.615**	.925**	.406	.145	.441	.485	.673*	-.125	.501	.976**	.262	.667**	.888**	1

Note. C  =  Cannot be computed because at least one of the variables is constant due to small sample size.

**Correlation is significant at the 0.01 level (2-tailed).

*Correlation is significant at the 0.05 level (2-tailed).

A significant strong correlation was observed between patient's age at HSCT and hair loss (r  =  .889, p < .05). Although only five participants reported experiencing hair loss as a symptom, those who endorsed it reported high levels of severity and distress. Patient age was also moderately correlated with total number of symptoms (r  =  .424, p < .01), however, younger patients were surveyed on fewer symptoms. Multiple symptoms were strongly correlated with each other including worry with sadness (r  =  .715, p < .05) and pain (r  =  .739, p < .05), and short of breath with fatigue (r  =  .884, p < .01) and drowsiness (r  =  .978, p < .01). Several symptoms were also correlated with the psychological subscale, physical subscale, and global distress index on the MSAS.

Pearson's two-tailed correlations between PROMIS symptom inventories and demographic variables are displayed in Table 5. Patient age at the time of the study visit was moderately correlated with anxiety (r  =  .337, p < .05). Age at HSCT was not significantly correlated with any symptoms. Depression T scores were significantly correlated with moderate to large effect with all other symptom T scores at p < .01, except sleep which was significant at p < .05. Anxiety T scores were significantly correlated with moderate to large effect with all T scores except peer support and physical function. Fatigue T scores were significantly correlated with moderate to large effect with all symptom T scores except peer support and sleep at p < .01. Physical functioning T scores were significantly correlated with a large effect to fatigue (r = -.721, p < .01) and with moderate effect to all other symptom T scores except anxiety and sleep. Pain T scores were significantly correlated with small to moderate effect with all symptom T scores except peer support and sleep. Peer/social support T scores were only significantly correlated with depression (r = -.567, p < .01) and physical function (r  =  .311, p < .05) T scores with moderate to large effect. Sleep T scores were only correlated with anxiety (r  =  .618, p < .05) and depression (r  =  .658, p < .05) with large effect.

Table 5.

Correlations Between Demographic and PROMIS Symptom Scale T Scores.

	Patient age	Age at HSCT	Time since HSCT	Anxiety	Depression	Fatigue	Physical function	Pain	Peer/ social	Sleep
Patient age	1
Age at HSCT	.876**	1
Time since HSCT	.259	-.234	1
Anxiety	.337*	.247	.166	1
Depression	.247	.126	.232	.621**	1
Fatigue	.128	.037	.177	.475**	.515**	1
Physical function	-.044	-.031	-.009	-.232	-.398**	-.721**	1
Pain	.242	.098	.286*	.349*	.444**	.501**	-.387**	1
Peer/social	-.055	-.001	-.106	-.258	-.567**	-.266	.311*	-.182	1
Sleep	.266	-.164	.374	.618*	.658*	.384	-.291	.255	-.428	1

**. Correlation is significant at the 0.01 level (2-tailed).

*. Correlation is significant at the 0.05 level (2-tailed).

One-way ANOVA were completed to examine differences between demographic and treatment variable groupings and symptom scores. There were no significant differences by graft types, race, ethnicity, previous HSCT, previous chemotherapy, posterior reversible encephalopathy syndrome (PRES), thrombotic microangiopathy (TMA), or primary disease relapse on any symptom. Several significant group differences in symptoms by demographic and treatment variables were present. Females reported significantly more fatigue than males on the MSAS (p  =  .032). Patients who had an autologous transplant (n  =  4) reported significantly more sadness and lack of appetite than those who had allogeneic transplants. There was also a significant difference in pain reported on the MSAS between patients with a hematologic diagnosis, who had a higher mean pain score, and those with a diagnosis of an immune disorder (p  =  .043). Having acute graft versus host disease (GVHD) was significantly associated with nervousness (p  =  .016), mouth sores (p  =  .02), and difficulty sleeping T scores (p  =  .013). Having chronic GVHD was significantly associated with itching (p  =  .002) and nervousness (p  =  .037).

Discussion

Survivors of pediatric HSCT reported multiple symptoms that are significant and persist for years following HSCT. Survivors of acute lymphoblastic leukemia who had undergone HSCT were five times more likely to suffer late-effects than survivors who received conventional chemotherapy, with 71.5% of survivors of HSCT reporting at least one late-effect (Berbis et al., 2013). Participants (N  =  256) were on average 11.9 (SD ± 6.4) years post-HSCT. Harrington et al. (2010) found similar results in adult survivors of a variety of cancers who experienced treatment-related symptoms more than 10 years following treatment completion. In their review of the literature, Wu and Harden (2015) reported that up to a third of cancer survivors experienced symptoms after cessation of treatment. Of the 50 survivors of pediatric HSCT in this study, 90% experienced at least one symptom, only five did not experience any symptoms during the past week, with two participants experiencing the maximum reported 14 symptoms.

The most frequently reported symptoms on the MSAS varied by age group. For 8- to 9-year-olds, being tired and itching were the most reported symptoms of the 8 symptoms surveyed. For 10- to 17-year-olds, difficulty concentrating, pain, worry, and nervousness were the most reported symptoms of the 30 symptoms surveyed. Wu et al. (2018) found that 71% of HSCT survivors experienced cognitive impairment, with 42% reporting difficulty in attention and concentration. Difficulty concentrating has particular impact on educational and work outcomes. For those 18 years of age and older, the most reported symptoms were worry, sadness, and pain of the 32 symptoms surveyed. Parsons et al. (2012) found that more than 50% of adolescent and young adult (15–39 years of age, N  =  388) cancer survivors who were attending work or school full-time prior to diagnosis experienced problems returning to work or school, and more than 30% in full-time work or school reported difficulties with attention and concentration. Mental health symptoms such as anxiety, sadness, nervousness, and worry, along with pain were prominent symptoms experienced by multiple age groups that impeded their ability to complete work or school tasks (Parsons et al., 2012). Jim et al. (2018) found a prevalence of 42% of adult survivors of HSCT suffered from insomnia in the six months following treatment, including 22% reporting insomnia at baseline prior to treatment that did not improve over time. Anxiety and depression have been reported to be difficult for health care providers to accurately diagnose in pediatric cancer populations due to the overlap in characteristics of anxiety and depression and treatment-related effects such as difficulty concentrating, difficulty sleeping, and changes in appetite (Lauer, 2015). The NIH HSCT Late Effects Workgroup reported psychological outcomes as undefined for survivors of pediatric HSCT (Bevans et al., 2017), suggesting this is a knowledge gap. In adults, the prevalence of depression is between 12 to 30% and emotional distress is between 22 and 43% of HSCT survivors (Bevans et al., 2017).

More than a quarter of the study sample (27%) endorsed fatigue as a symptom they had experienced in the last week. Tomlinson et al. (2019) also reported older aged adolescents and young adults (11–18 years) survivors of cancer or HSCT were significantly more likely to experience extremely bothersome fatigue than children 10 years of age or younger. Fatigue was also experienced significantly more by females than males in this sample, which is frequently reported in the literature (Bevans et al., 2017; Cheon et al., 2021).

The Symptom Cluster Experience Profile Framework (Finnegan et al., 2010) models subgroup-specific symptom profiles that are driven by risk and protective factors. In the framework risk and protective factors are moderated by stress response to impact symptoms and symptom cluster. Examples of risk and protective factors include demographics, treatment exposures, environment, life strain, health behaviors, social support, and outlook on life. Symptom profiles directly impact quality of life dimensions and functional status. Finnegan et al. (2010) found that in adult survivors of pediatric cancer, pain, fatigue, sleep disturbance, psychological distress, and difficulty concentrating were the symptoms present in the high-risk profile. In this study, multiple symptoms correlated with each other which may indicate clusters of symptoms exist, however, there was insufficient power to run the cluster analyses. Of particular interest, depression was significantly correlated with moderate to large effect with all symptoms on the PROMIS instruments. Finnegan et al. (2009) also found overlap between the symptoms in the high-risk profile for pediatric cancer survivors, and the most commonly reported symptoms for survivors of pediatric HSCT including difficulty concentrating, lack of energy/fatigue, and pain. Survivors of pediatric HSCT also reported numerous psychological symptoms including nervousness, worry, anxiety, and sadness. Further research with larger samples is needed to determine if these symptoms form a cluster or high-risk profile and if they are generalizable to the wider HSCT survivor community.

Several demographic and treatment variables were associated with symptoms. Patients with autologous transplants were statistically different than patients who received allogeneic transplants for increased sadness and lack of appetite, however, only four patients in this sample received an autologous transplant so further research with larger samples is needed to test this relationship. Initial hypotheses that 1) reduced intensity conditioning would be associated with fewer symptoms than ablative conditioning, and that 2) allogeneic HSCT would be associated with more total symptoms than cord or autologous symptoms were rejected. However, small sample size could have impacted analyses since only a quarter of the sample had reduced intensity conditioning, four participants had autologous HSCT, and three participants had cord blood HSCT. Of all the demographic and treatment variables, age was the only significant variable related to hair loss on the MSAS and anxiety on the PROMIS instruments. Nuss and Wilson (2007) found that in survivors of pediatric HSCT (N  =  31), older age at time of transplant was associated with higher procedural anxiety (r  =  .498, p  =  .005).

Acute and chronic GVHD were associated with nervousness. Acute GVHD was also associated with mouth sores and difficulty sleeping. Chronic GVHD was associated with itching, which is logical since the majority of chronic GVHD in this sample was located in the skin. Cheon et al. (2021) also found chronic GVHD to be associated with fatigue in adult survivors of HSCT, however that was not found in this study.

Limitations

This was a pilot study which limited group analyses due to the small sample. This study was also a single time point which prevents analysis of trends in symptoms over time. Survivors in this sample were on average 5.4 years out from their transplant date with a wide range from one to nine years post-treatment. Although time from HSCT was not statistically significant, it is possible symptoms experienced by survivors farther out from HSCT are unrelated to HSCT treatment and are related to co-morbidities, however, across a variety of cancers and treatment regimens it is common for adult cancer survivors to experience symptoms related to treatment for more than 10 years after treatment (Harrington et al., 2010). There was also a wide range in age of participants which affected the total number of symptoms they were asked to rate on the MSAS and limited some of the statistical age group comparisons.

Implications for Practice

Survivors who participated in this study continued to experience symptoms up to nine years post-HSCT. Mental health symptoms and pain were endorsed across all age groups, however, mental health is often overlooked in survivorship. Survivors should be frequently screened and referred to psychological or pain services when appropriate. Survivorship care plans can help facilitate successful transitions in care and increase awareness of late effects (Denzen et al., 2019). Mental health should also be included in survivorship care plans for all survivors of HSCT (Denzen et al., 2019). Difficulty concentrating was the most common symptom reported by survivors. Difficulty concentrating and other neuropsychological symptoms can have a profound effect on school and work outcomes (Parsons et al., 2012). Educators and caregivers should be aware of the potential impact of HSCT treatment on learning outcomes, provide assessment, and work to develop learning plans that fit the individual needs of survivors. Information on evidence-based methods to manage symptoms should also be available for survivors, for example, the Oncology Nursing Society has evidence-based resources available (https://www.ons.org/the-pep-topics) on symptoms management including interventions for commonly reported cancer treatment-related symptoms. The American Society of Transplantation and Cellular Therapy (ASTCT) also has recommendations for survivors of HSCT including survivorship care plans for all survivors of HSCT (Denzen et al., 2019) with information included on mental health, sexual health, and the immune system; and screenings for symptoms such as sleep disruption (Jim et al., 2018), and fatigue which is known to be associated with adverse adjustment outcomes (Park et al., 2019).

Conclusions

Survivors of pediatric HSCT continue to experience symptoms years into survivorship. The most commonly reported symptoms for this sample were difficulty concentrating, pain, worry, nervousness, and lack of energy/fatigue, with 90% of the sample experiencing at least one symptom in the previous week. Survivors of HSCT should be screened regularly for symptoms in the clinical setting and symptom screening should be incorporated into survivorship care plans. Future research is needed to examine the biological mechanisms that contribute to the ongoing symptomology described in this study and to develop effective interventions to minimize or prevent symptoms throughout survivorship.