Application of the National Comprehensive Cancer Network-distress thermometer in pediatric patients during autologous and allogeneic hematopoietic stem cell transplantation and relationship to blood parameters of the stress axis

In this prospective monocentric study, 40 pediatric HSCT patients between 10 and 18 years of age, and one parent of each patient, were prospectively and consecutively enrolled at in-patient admission for scheduled HSCT between January 2019 and December 2020.

Patients were excluded if they were under 10 years of age, or if the patients and/or their parents did not have sufficient German or English language skills to complete the questionnaires used in this study. Furthermore, the patients were excluded from the study if they had already been diagnosed with a mental illness prior to HSCT.

The observation period was defined as the period from one day before start of the conditioning regimen until the date of discharge after HSCT.

The NCCN-DT is a screening tool used to assess distress in patients with hemato-oncological malignancies, developed by the United States National Comprehensive Cancer Network (NCCN). It is a single item screening instrument with an eleven-point visual analogue scale on what the subjectively experienced distress level can be rated from 0 (no distress) to 10 (extreme distress). A cutoff score ≥ 5 is internationally recommended as an indicator for a distressed patient. In addition, a problem list identifies potential sources of distress. We have made slight changes to make it easier for children and young people to see the categories: the thermometer has been colored from green (no stress) to red (maximum stress). The questions for the problem scales were not changed.

The children and one parent (always the same parent; the parent who was admitted to the in-patient stay) per child were given a distress diary with the DT included on every page to self-assess their distress level. The problem lists were handed out at the following time points to analyze the possible cause of distress: day of admission, five days before HSCT (day − 5), the day of HSCT (day 0), day + 7, day + 14, day + 21, day + 28, day + 35, day + 42, day + 50, day + 60 after HSCT and the day of in-patient discharge. Using these stress diaries with the DT and the problem lists, the reported stress levels were recorded and evaluated during the time period on the ward on the day of in-patient admission, day − 5, day 0, day + 7, day + 14, day + 21, day + 28, day + 35, day + 42, day + 50, day + 60 after HSCT and the day of in-patient discharge.

The stress parameters cortisol, the thyroid stimulating hormone (TSH), free triiodothyronine (fT3) and thyroxine (fT4) were determined in the blood of the pediatric patients at the same time points for analysis of the DTs to compare the results. TSH, as well as fT3 and fT4 were quantified in heparinized peripheral blood using enzyme immunoassays, i.e. the ADVIA Centaur^® XPT Immunoassay-System (Siemens-Healthcare GmbH, Erlangen, Germany).

Cortisol is a hormone of the stress response and recruits energetic compounds and molecules through activating catabolic processes. Cortisol was determined in the blood plasma using the corticosterone (Rat/Mouse 9 ELISA Kit from DRG Instruments GmbH, Marburg, Germany). The sampling and preparation of the peripheral blood taken from routine blood sampling were performed according to the manufacturer information, including the determination of negative and positive control samples.

All blood parameters were determined from blood samples taken as part of routine diagnostics, which took place daily in the morning between 6 and 7 am. An influence of the circadian rhythm on these parameters could thus be minimized. No cortisol parameters were recorded during steroid substitution. Results were determined in the blood of the pediatric patients at the same time points for analysis of the DTs in order to compare the results.

Laboratory data were analyzed descriptively (mean, median, range, 25th percentile, 75th percentile, minimum, and maximum) and tested for normal distribution using the Shapiro–Wilk normality test. Multiple comparisons between baseline values and values at different observation days were analyzed with a 2-way ANOVA analysis of variance. Significant deviations above or below the upper or lower normal limits were detected using the Wilcoxon-signed rank test. The results are displayed as mean ± standard error of the means (SEM) in the graphs.

The maximum problem awareness of mental and somatic problem percentages was displayed as mean ± 95% CI. Comparisons between results of parents and patients were performed using an unweighted Cohen’s Kappa. The agreement measures were defined as < 0.00 (poor), 0.00–0.20 (slight), 0.21–0.40 (fair), 0.41–0.60 (moderate), 0.61–0.80 (substantial), 0.81–1.00 (almost perfect).

All p values were adjusted for multiple testing. p values of < 0.05 (*), p < 0.01 (**), p < 0.001 (***), and p < 0.0001 (****) were defined as statistically significant.

The collection of laboratory data was performed using the laboratory information system Swisslab LAURIS (version 16.10.05.5, Roche Diagnostics IT Solutions GmbH, Berlin, Germany). All medical reports derived from the SAP system of the University Children’s Hospital Tübingen (SAP Netweaver SAP GUI for Windows, Version 7300.3.10.1084, SAP SE, Walldorf, Germany). The data were collected using Microsoft Excel (Microsoft Excel 2010 Version 14.0 for Windows, Microsoft Germany GmbH, Munich, Germany). Graphs and statistical tests were created with GraphPad Prism for Windows, version 9.4, 2022 (GraphPad Software Inc., La Jolla, CA, USA).

Patients characteristics are summarized in Table 1. The patient group consisted of 40 pediatric patients who received an HSCT because of a hematologic–oncologic malignant or nonmalignant underlying disease. From these patients, 22 patients (55.0%) were male, 18 patients (45.0%) were female. The median age was 14.6 years (range 10–18 years). Patients received grafts from matched sibling donors (MSD; n = 13; 32.5%), matched unrelated donors (MUD; n = 15; 37.5%), mismatched family donors (MMFD; n = 7; 17.5%), and themselves (autologous; n = 5; 12.5%). The conditioning regimen was in 37 out of 40 patients myeloablative (92.5%) and in three cases intensity-reduced (7.5%), composite in one case of cyclophosphamide (4 × 50 mg per kilogram (mg/kg)) and antithymocyte globulin (ATG; 3 × 10 mg/kg), in one case of fludarabine (3 × 40 mg per square meter (mg/m²)), thiotepa (15 mg per kilogram (mg/kg)), total lymphoid irradiation (TLI, 1 × 7 Gray (Gy)) and ATG (2 × 15 mg/kg) and in the third case of fludarabine (5 × 30 mg/m²), total body irradiation (TBI; 4 Gy), Thiotepa (2 × 5 mg/kg), ATG (6 mg/kg) and post-transplant cyclophosphamide (2 × 50 mg/kg). Out of the 40 patients, 17 (42.5%) suffered from acute leukemia or myelodysplastic syndromes and four patients (10.0%) had a lymphoma. Furthermore, seven patients (17.5%) were diagnosed with a nonhematologic solid tumor and twelve patients (30.0%) had a nonmalignant underlying disease. During the clinical course 17 patients (42.5%) suffered from sepsis or systemic inflammatory response syndrome (SIRS) and viremia. An acute graft-versus-host disease (aGvHD) grade 1 or 2 occurred in eleven patients (27.5%). There was no case of higher grade aGvHD during the observation period. A veno-occlusive disease (VOD) was diagnosed in seven patients (17.5%) within the clinical course and three patients (7.5%) had a graft failure. The median observation period was 51.5 days (range 24–155 days).

The median baseline DT value of the patients was 5 (range 3–9). On day − 5 before HSCT, the patients reported a peak median DT value of 6 (range 2–9, DT day − 5 vs. base: p = 0.76), followed by a downward trend with a median of five 1 week after HSCT (range 0–9, DT day + 7 vs. base: p > 0.99), 5 on day + 14 (range 0–10, DT day + 14 vs. base: p > 0.99), 4 on day + 21 (range 0–7, DT day + 21 vs. base: p = 0.17), 3 on day + 28 (range 0–8, DT day + 28 vs. base: p = 0.0224), 4 on day + 35 (range 1–9, DT day + 35 vs. base: p = 0.22), 3 on day + 42 (range 1–10, p = 0.0142), 2 on day + 50 (range 2–6, day + 50 vs. base: p = 0.0057), 2 on day + 60 (range 1—5, day + 60 vs. base: p = 0.0056) and 2 on the day of discharge (range 0—4, DT day of discharge vs. base: p < 0.0001). The median values of the DT at the time points of 4 weeks, 6 weeks, day + 50 and day + 60 after HSCT and on the day of discharge were significantly below the baseline value on the day of in-patient admission (Fig. 1).

The median baseline DT value of the parent group was 7 (range 2–9). As in the pediatric patients, the parents showed an increase of DT scores from the day of admission to the day of HSCT, with a peak on the day of HSCT (median 8.5, range 2–10), followed by a decreasing trend in the DT value over the further course of time to day + 7 (median 6.5, range 1–9, DT day + 7 vs. base: p > 0.99), day + 42 (median 3, range 1–10, DT day + 42 vs. base: p = 0.0015), day + 60 (median 4, range 2–6, DT day + 60 vs. base: p = 0.0363) and on the day of discharge (median 3, range 0–4, DT day of discharge vs. base: p < 0.0001). The median value of the DT on the day of HSCT was significantly compared to baseline (Fig. 1).

Comparing the two groups, the parents of the patients reported higher DT values than the children themselves at all time points. On the day of HSCT, parents scored significantly higher DT values as compared to their children (p = 0.0157) (Fig. 1).

The mean cortisol level of the patients at baseline was 280 ± 21.2 nmol/l and thus within the normal range (125 to 420 nmol/l). The cortisol levels exceeded the upper limit value on the day of HSCT and remained above the upper limit until discharge with a significant difference from the upper normal value on day + 7 after HSCT and on the discharge day (mean cortisol value day + 7 (525 ± 46 nmol/l) vs. upper normal value (420 nmol/l): p = 0.0293; mean cortisol value discharge day (534 ± 37 nmol/l) vs. upper normal value: p = 0.0046; Fig. 2a).

The mean TSH level of the pediatric patients at in-patient admission was 3.7 ± 0.3 mU/l. The mean TSH values of the patients were never outside the normal range during the in-patient stay (0.8 to 5.4 mU/l). They dropped after admission and reached their lowest point on day + 7 after HSCT. Subsequently, they showed an increasing trend over time (Fig. 2b).

The mean values of fT3 and fT4 at in-patient admission were 4.6 ± 0.2 pmol/l and 12.7 ± 0.4 pmol/l, respectively. The mean values of fT3 and fT4 were never outside the normal range (normal range fT3: 3.5 to 6.5 pmol/l, normal range fT4: 12 to 23 pmol/l) and also showed no upward or downward trend over time (Fig. 2c, d).

The results of the problem lists are presented in Table 2 and in Fig. 3. In summary, it can be stated that more than half of the children reported problems in the areas of exhaustion, worry, nausea, and inflammation. Among the parents, more than 50% reported problems in the areas of worry and anxiety.

Regarding the problem area of exhaustion over time, it is noticeable that more children reported problems in this area than their parents. This difference reaches significance at the time points of day − 5 (p < 0.0001), day 0 (p < 0.0001), day + 7 (p < 0.0001), day + 14 (p < 0.0001), day + 21 (p < 0.0001), day + 28 (p = 0.0021), day + 35 (p < 0.0001), day + 42 (p = 0.0172) and the day of discharge (p = 0.0037). The rate of exhaustion in the pediatric patients increased rapidly until the day of HSCT, after which the curve tended to decline slowly (Fig. 4a).

Concerning the problem area of worries, it is noticeable that the parents initially indicated more problems in this area than the pediatric patients with significant differences at the time points of day of admission (p = 0.0036), day − 5 (p = 0.0001), day 0 (p < 0.0001), day + 7 (p = 0.0001) and day + 14 (p < 0.0001). However, while the parents' awareness of worries reaches its maximum on the day of HSCT (67.5%) and subsequently drops from day + 21 after HSCT onwards, the rate of worries among patients increases over time and is significantly higher than that of the parents at day + 42 (p = 0.0172) (Fig. 4b).

The awareness of problems in the area of fears was higher among the parents than among the patients over the entire time course with significant differences at the time points of day of admission (p = 0.0039), day − 5 (p = 0.0003), day 0 (p < 0.0001), day + 7 (p < 0.0001), day + 14 (p = 0.0003), day + 28 (p = 0.0427), and the day of discharge (p = 0.0028). In both groups, the maximum percentage of problem perception was the highest on the day of HSCT and subsequently decreased except for the day of discharge, at which time the number of parents perceiving fears increased again (Fig. 4c).

The stress diaries were given to only one parent—either the mother or the father. It is possible that the perception of the stress level and also the different problem lists differ between the parents. This may result from gender differences or the individual parent’s own coping strategies and levels of resilience. These aspects should be investigated in more detail in future studies.

The in-patient stay during HSCT is quite heterogeneous depending on the course of therapy and occurring complications. Most of the patients are discharged between day + 35 and day + 42 after HSCT, but the range with the discharge day occurring between day + 17 and day + 148 after HSCT is quite wide. Of course, this has an influence on the patients' stress perception, whether they are still in in-patient treatment or already at home. For this reason, data were collected exclusively during the in-patient stay in order to maintain as consistent a setting as possible. As a result, the sample size decreases towards later time points. Differences in stress perception during the in-patient stay or during the time afterwards should be further investigated.

The DT has already been used successfully for a long time in adult oncology to assess stress levels, is therefore very well investigated according to the comprehensive studies, and is widely used due to its quick and uncomplicated application. Few studies have proven its feasibility, validity and specificity in pediatric cancer patients, however not in German and not in the HSCT context (Patel et al. 2010, 2011, 2014). Therefore, our aim was to investigate the implementation and the benefit of the NCCN-distress thermometer for our patients.