Medical and Nutritional Outcomes Are Similar Among Autologous Transplant Patients on Enteral Nutrition When Compared to Parenteral Nutrition. A Randomized Pilot Study

ABSTRACT

Autologous hematopoietic stem cell transplantation (AHSCT) is the treatment for myeloma and lymphoma. posttreatment, significant nutritional and medical issues and malnutrition assessed by Subjective Global Assessment (SGA) arise. No established effective treatment for using either parenteral (PN) or enteral routes (EN) to improve nutritional status, reduce medical complications, and be cost‐effective is available. We investigated the effectiveness of EN versus PN in terms of nutritional path of supplementation. AHSCT patients were randomized to either EN or PN and were followed at baseline, 15 and 30 days posttransplant. Age, body mass index, SGA, length of stay (LOS), medical complications, severity of complications, infections, overall survival (Day 100), albumin, random blood glucose, and C‐reactive protein were evaluated. Descriptive statistics, Spearman's, chi square, correlations, and uni‐ and multivariate by type of feed, using SPSS v 29. Thirty‐six patients with complete medical and laboratory data were followed. No significance in any of the medical or nutritional parameters between the two groups was found. No correlations between SGA at any time point and type of feeding were identified. No relationship between SGA, LOS, complications, albumin, CRP, or random blood glucose at all three time points was seen. EN is a safe, convenient, and cost‐effective option for AHSCT patients since medical and nutritional outcomes were similar between those receiving EN compared to PN.

1. Introduction and Background

Autologous (which involve the use of the patient's own stem cells) hematopoietic stem cell transplant (AHSCT) is used as standard of treatment for patients with multiple myeloma, and as part of salvage consolidation for patients with Hodgkin's and non‐Hodgkin's lymphoma [1, 2]. Since AHSCT requires intensive chemotherapy for the procedure, it is associated with several complications, including systemic as well as local toxicities that involve the gastrointestinal (GI) tract. Patients may experience severe mucositis, esophagitis, oral ulcerations, esophageal or gastric dysmotility, nausea, and vomiting, all of which can disrupt normal oral intake, digestion, and absorption [3]. The consequences of this nutritional deprivation can lead to a significant decline in nutritional status, including weight losses greater than 2%–5% of actual body weight, hypercatabolism, electrolyte disturbances, and malnutrition [4]. In both pediatric and adult transplant patients, malnutrition is a negative prognostic indicator for survival [5].

The American Society of Parenteral and Enteral Nutrition (ASPEN) recommends the use of nutritional support during AHSCT for patients who are malnourished, have decreased oral intake, or decreased intestinal absorption for any prolonged period [6]. Nutritional support can be delivered via enteral or parenteral routes. For practical reasons, parenteral nutrition (PN) is the default nutrition support used in some centers for AHSCT patients as they already have a central venous catheter inserted for medical treatment. PN has been proven to be safe and effective in this patient population and has been shown to preserve body mass and decrease catabolism [7]. Conversely, it is also associated with an increased number of complications, including primary infections related to the central line as well as increased costs [8, 9]. The complications are varied with a major risk of hyperglycemia; however, these altered blood sugars can be managed with insulin and strict clinical protocols [10].

Enteral Nutrition (EN) is the most appropriate form of nutritional support in those with a functional GI tract and is feasible in a wide variety of cancer treatments [9, 11]. However, clinical and individual perceptions of enteral feeds are that they can be traumatic and uncomfortable for the patient, especially when there is the possibility of GI toxicities [11]. In a nonrandomized study by Seguy et al., in allogeneic HSCT (donor stem cell transplantation) patients reported both a decreased incidence of graft versus host disease (GVHD) and reduced infection‐related mortality at Day 100 with the use of EN feeds [9]. This same group conducted another study confirming benefits of EN versus PN in allogenic transplant patients [12]. Further, EN nutrition support is associated with lower risks of infectious complications and hyperglycemia, as well as decreased intestinal permeability [9, 13]. In contrast, the use of EN can increase the prevalence of aspiration pneumonia in gastric and/or post‐pyloric feeds [14]. Adequate use of EN feeds during the early transplantation course (pre‐ and postoperatively) is associated with reduced non‐relapse mortality and improved survival rate at 5 years posttransplant in those who had undergone allogenic hematopoietic cell transplantation [15]. Furthermore, use of EN nutrition support is associated with a lower incidence of both overall and gut acute graft versus host than PN nutrition support in part due to the ability of enteral feeds to maintain mucosal integrity, modulate the immune response to intensive chemo and radiotherapy, and support the GI tract environment, including gut microflora [15].

In 1987, Detsky published a landmark article on Subjective Global Assessment (SGA) which has become the gold standard to assess patients' nutritional status [16]. In 2018, the Global Leadership Initiative on Malnutrition (GLIM) reached a consensus on the diagnosis of malnutrition, recommending that all patients should be screened and assessed to determine if they are at risk for malnutrition and then provided with the appropriate nutrition intervention [17]. ASPEN does recommend perioperative nutritional support (any route) for those who are moderately to significantly malnourished [9]. Most patients who undergo AHSCT are started on PN feeds when their intake drops to < 50% of requirements. A randomized study of 50 well‐nourished AHSCT patients was conducted to determine the effect of early nutrition support (commenced when oral intake was < 80% of estimated requirements) compared with usual care (commenced when oral intake < 50% of estimated requirements) [18]. Both groups demonstrated significant weight loss over time (p = 0.0005) [18]. Thus, it is vitally important to initiate early nutritional support for patients soon after transplant even in those who were previously well‐nourished. AHSCT often negatively impacts nutritional status with side effects like oral mucositis and GI disturbances which can result in inadequate oral intake and treatment‐induced malnutrition. The amelioration of taste buds further leads to a reduction of appetite and decreased oral intake. There is reduced caloric intake, decreased protein uptake, and increased protein GI loss compounded by increased stress metabolism; effectively causing a negative catabolic state. The American Society of Enteral and Parenteral Nutrition (ASPEN) recommends alternative nutrition support using EN or PN in this patient population.

Previous research has not examined the impact of EN support compared to PN support in AHSCT patients. Moreover, research has not fully considered the medical complications in this patient population. Much of the current evidence is extrapolated from patients undergoing allogenic hematopoietic stem cell transplants, which have complications unique to that population. There are no established effective treatments for AHSCT patients that will maintain or improve nutritional status, reduce medical complications, and be cost‐effective.

Thus, we sought to investigate the effectiveness of EN as compared to PN in terms of the nutritional path of supplementation in a pilot randomized study. The objectives of our study were to: (1) explore the differences in the medical outcomes between EN versus PN nutrition support in AHSCT patients; and (2) examine differences in transplant‐related, biochemical, and nutrition‐related outcomes in AHSCT patients randomized to enteral versus PN arms.

2. Materials and Methods

A pilot prospective randomized study was conducted in a large academic teaching hospital in Southwestern Ontario between 2019 and 2022 and was interrupted due to emergence of COVID outbreaks in the hospital during the early period of the study. Consent was obtained prior to admission at the time of pretransplant visit. All patients did undergo myeloablative chemotherapy which consists of either single agent melphalan for myeloma patients or combination chemotherapy consisting of carmustine, etoposide, cytarabine, and melphalan (BEAM) starting 2 days before or 6 days prior to stem cell reinfusions (considered Day 0), respectively. Patients do develop side effects from intensive chemotherapy such as oral and GI mucositis, nausea, vomiting, diarrhea, anorexia post chemotherapy with side effects peaking 3–5 days post chemotherapy (Day +3, +5). However, most patients continued to maintain their oral intake and nutritional status even after chemotherapy, but their nutritional intake does decrease by Day 5 posttransplant. Most patients start their nutritional intake orally by Day 12–15 posttransplant, after progressive healing of the oral and GI mucosa occurs. Hence, nasogastric tubes were inserted on Day 5 ± 1 post‐AHSCT treatment if oral intake was less than 80% of recommendations. Nutrition support using peptamen 1.2 AF, continued until oral intake was > 50% of recommendations or until the patient was ready for discharge. Central venous catheters were placed for AHSCT treatment and also used for PN administration with a two‐in‐one premixed formula with 20% SMOF Kabiven lipid, as per hospital protocol.

Patients were assessed using validated standardized tools at three time points: baseline (at time of transplant) and then Day 15 and Day 30 post‐AHSCT. The assessments included anthropometric measures: body mass index: BMI (weight in kg/height m²) and percent weight change [(UBW − CBW)/UBW] × 100. Nutritional status was evaluated using SGA, which is easy to perform, inexpensive, and requires minimal training [19] and is used to assess nutrition status. Research has shown that SGA may be a predictor of poor clinical outcomes in critically ill patients [20, 21]. Laboratory data including albumin, pre‐albumin, CRP, and random blood glucose, utilizing standardized hospital laboratory assays. Quadricep muscle layer thickness (QMLT) was measured using Sonosite ultrasound to assess thickness as a measure of lean muscle mass. Treatment‐related variables, including complications, severity of complications, and total length of stay (LOS), were also collected. Additionally, body composition (BC) using bioelectrical impedance analysis (BIA) to assess percent body fat, lean muscle mass, and phase angle, along with hand grip strength (HGS) measured by Jamar dynamometer and dietary measures (3‐day food records), which were also completed at all three time points.

2.1. Inclusion Criteria

The following criteria were used for inclusion in the study: all adult patients aged 18–75 years; patients admitted to our center undergoing autologous HSCT; patients diagnosed with the following conditions: non‐Hodgkin's lymphoma (all types), Hodgkin's lymphoma (all subtypes), and multiple myeloma; patients receiving any of the following: conditioning chemotherapy: melphalan, etoposide/melphalan, or BEAM (carmustine, etoposide, cytarabine, melphalan) and with a functional GI tract.

2.2. Exclusion Criteria

The following were excluded from the study: intestinal obstruction; patients with nasal deformities, tumors of nasal tracts or upper nare obstruction; patients with active bacteremia while proceeding with transplant; patients with active malignancy of the upper GI tract, not in remission as evidenced by recent imaging studies (< 4 weeks); patients with any GI bleeding, paralytic ileus, obstruction, or any other GI condition which excludes use of the GI system for nutritional support, as these patients will require PN feeding only and cannot be randomized.

2.3. Consent Process

Patients were consented by the study coordinator following the Western University, Research Ethics and Lawson Health Science Research board parameters, and the informed consent procedure conformed to the ICH guidelines on Good Clinical Practice. Ethical approval was obtained from the Hospital and University research boards (WREB # 113204).

2.4. Baseline Evaluation

After patients were consented, the following were collected or measured prior to infusion of the stem cells for AHSCT: serum albumin; serum pre‐albumin; C‐reactive protein (CRP); fasting blood glucose; anthropometrics (weight [kg] and height [cm]); calculation of BMI; QMLT using ultrasound; SGA; and estimated oral intake of calories and protein from 3‐day diet history recall.

2.5. Day 15 Post Treatment

After completion of treatment on Day 15 of either EN or PN, a post‐trial evaluation was completed including all the above parameters as well as the number of days to transition to oral intake and the percentage of calories and protein needed to meet 100% via EN or PN.

2.6. One Month Follow‐Up

Post discharge, the above parameters were repeated at Day 30 with a more comprehensive estimation of macro and micronutrient intakes using a 3‐day food record (3DFR) at their regularly scheduled clinic visit was completed.

The data related to transplant associated complications were collected prospectively and entered in the database, as complications can occur any time during the posttransplant period from Day 0 until Day 30 of the transplant.

2.7. Sample Size

The hematology unit where patients undergo AHSCT has approximately 70 patients per year. Given the limited evidence with AHSCT patients and nutritional support, this pilot study was conducted to obtain preliminary data to determine the appropriate size based on the variable of interest [22]. Therefore, a decision was made that 40 patients would be a reasonable number to recruit over a 1‐year period. Thus, 40 patients were randomized to either PN or EN, resulting in 20 patients per treatment arm. The study period continued until 40 patients were enrolled.

2.8. Randomization Process

Randomization was a 1:1, in permutated blocks independently by a statistician into both the EN and PN arms.

2.9. Statistical Analysis

Descriptive statistics were conducted to compare baseline measures and differences between groups, using two‐sample t‐tests for continuous variables and Spearman's, chi square, and Fisher's exact test for categorical variables. We ran correlations and uni‐ and multivariate analysis to determine the magnitude and direction of the relationships between variables by type of feed. Analyses were completed using SPSS v29; p < 0.05 was considered significant.

3. Results

The study design was a prospective randomized study with 1:1 randomization in permuted blocks. Forty participants were initially enrolled and consented, two patients were discharged early, and two refused their randomized nutrition therapy and were removed from the study. Here we report data on 36 patients. Table 1 describes the baseline patient characteristics between the EN and PN groups. Both groups had equal myeloma and lymphoma patients (Hodgkin's and non‐Hodgkin's). All myeloma patients received melphalan, and lymphoma patients received BEAM conditioning chemotherapy. All patients had KPS (Karnofsky Performance Scale) of > 80% at the time of transplant. Table 2 provides results on medical complications and bloodwork at baseline, Day 15, and Day 30 posttreatment. Nutritional parameters are presented in Table 3 for all three time points between the two groups. No significant changes in posttransplant weight gain or loss between the groups were seen (data not shown). We were not able to demonstrate statistical significance in any of our medical or nutritional parameters between the two groups. Table 4 demonstrates the cost savings in the EN group compared to the PN group. Figures 1 and 2 depict SGA results from baseline and Day 15; unfortunately, SGA at Day 30 was not completed due to COVID dietetic staffing shortages.

TABLE 1.

Baseline characteristics EN compared to PN.

Variable	EN, N = 19	PN, N = 17
Gender F/M %	37/63	70/30
Age (mean ± SD)	61.1 ± 7.9	60.8 ± 11.6
BMI (mean ± SD)	28.4 ± 7.4	25.9 ± 4.7
SGA categories (%)
SGA‐A	15	35
SGA‐B	45	30
SGA‐C	40	35
BMI categories (%)
Underweight	0	6
Normal weight	26	53
Overweight	47	12
Obese	26	29

Note: SGA categories: A: well‐nourished, B: mild to moderate malnutrition, C: severely malnourished. BMI categories: underweight: < 18.5 kg/m²; normal weight: 18.5–24.9 kg/m²; overweight: 25.0–29 kg/m²; obese: ≥ 30 kg/m². p = NS for all variables.

TABLE 2.

Medical complications and laboratory investigations in patients on EN compared to PN at baseline, Day 15, and Day 30.

Variable	EN, N = 19	PN, N = 17
LOS (days)	16.7 ± 2.5	18.5 ± 5.6
Complications (%) yes/no	89/11	100
Severity (%)
Mild	31	12
Moderate	44	35
Severe	25	53
Infections (%)
Yes/no	31/69	29/71
OS (%)
Still alive Day 100	89	88
ALB (mean ± SD)
Baseline	38.0 ± 4.3	37.9 ± 3.9
Day 15	32.6 ± 3.6	32.9 ± 2.4
Day 30	39.3 ± 5.3	38.9 ± 4.5
RBG (mean ± SD)
Baseline	6.0 ± 1.4	5.8 ± 0.8
Day 15	6.1 ± 1.2	6.4 ± 2.3
Day 30	6.7 ± 1.5	6.0 ± 1.0
CRP (mean ± SD)
Baseline	7.6 ± 9.0	3.0 ± 3.7
Day 15	16.8 ± 13.0	14.8 ± 10.2
Day 30*	4.2 ± 3.3	6.1 ± 6.3
Day 30**	4.2 ± 3.3	23.4 ± 54.9

Abbreviations: ALB, albumin; CRP, C‐reactive protein; LOS, length of stay; OS, overall survival (Day 100); RBG, random blood glucose.

* with the outlier removed, ** with outlier; p = NS for all variables.

TABLE 3.

Nutritional parameters at baseline, Day 15, and Day 30 between the two groups.

Variable	EN, N = 19	PN, N = 17
BMI (mean ± SD)
BMI‐baseline	28.4 ± 7.4	25.9 ± 4.7
BMI‐Day 15	28.8 ± 8.2	24.8 ± 4.4
BMI‐Day 30	27.9 ± 7.3	24.0 ± 3.9
SGA‐baseline (%)
SGA‐A	15	35
SGA‐B	45	30
SGA‐C	40	35
SGA‐Day 15 (%)
SGA‐A	9	17
SGA‐B	91	83
SGA‐C	0	0
BMI categories (%) baseline
Underweight	0	6
Normal weight	26	53
Overweight	47	12
Obese	26	29
BMI categories (%) Day 15
Underweight	0	6
Normal weight	44	62
Overweight	19	6
Obese	37	25
BMI categories (%) Day 30
Underweight	0	0
Normal weight	47	69
Overweight	26	25
Obese	26	6

Note: SGA‐A, well‐nourished; SGA‐B, mild‐moderately malnourished; SGA‐C, severely malnourished. SGA was not completed at Day 30 due to dietetic staffing shortages from COVID. BMI categories: underweight: < 18.5 kg/m²; normal weight: 18.5–24.9 kg/m²; overweight: 25.0–29 kg/m² obese: ≥ 30 kg/m². p = NS for all variables.

Abbreviations: BMI, body mass index; SGA, Subjective Global Assessment.

TABLE 4.

Cost analysis EN compared to PN.

		EN	PN
A	Formula Cost/1 L	$20.00	$53.00
B	LOS (days) a	16.7	18.5
C	Cost formula per average LOS (assumption 1 L/d) [a*b]	$334.00	980.50
D	Total number of patients 70/year	35	35
E	Total cost per formula for 35 patients [d*c]	$11 690.00	$34 317.50
F	Cost/day for hospital stay b	$2500.00	$2500.00
G	Cost for average hospital stay per feed type [b*f]	$41 750.00	$46 250.00
H	Total cost per stay for 35 patients [g*35]	$1 461 250.00	$1 618 750.00
	Total cost for feed + stay for 35 patients for 1 year [h + c]	$1 472 940.00	$1 653 067.50
	Difference in total cost		$180 127.50

^a From Table 2, LOS on EN was 16.7 ± 2.5 days compared to LOS on PN was 18.5 ± 5.6.

^b Based on CIHI hospital stay costs in Ontario [23], costing from Nestle Canada for EN; costing from Fresenius‐Kabi for PN.

FIGURE 1.

Weight loss over time EN versus PN. p = NS for all variables. SGA, Subjective Global Assessment; SGA‐A, well‐nourished; SGA‐B, mild‐moderately malnourished; SGA‐C, severely malnourished. p = NS.

FIGURE 2.

SGA at Day 15, EN compared to PN. SGA, Subjective Global Assessment; SGA‐A, well‐nourished; SGA‐B, mild‐moderately malnourished; SGA‐C: severely malnourished. None of the patients at Day 15 were classified as severely malnourished. No SGA was completed at Day 30. p = NS for all variables.

There were significant correlations between SGA at baseline and infection (r = −0.407, p = 0.01). As well, significant correlations were seen with SGA at baseline and overall survival (OS) Day 100 (r = −0.493, p = 0.003) and again at Day 15 (r = −0.681, p = 0.005), respectively. There was borderline significance between SGA at baseline and severity of infection (r = −0.337, p = 0.055). Further, BMI at baseline was significantly correlated with OS (r = −0.344, p = 0.040); and BMI‐Day 30 was significantly correlated to severity of infection (r = −0.429, p = 0.013). However, there was no correlation between SGA at baseline or Day 15 and type of feeding. As well, no relationship was seen between SGA at either time point, LOS, complications, and blood work including albumin, CRP, or random blood glucose at all three time points. Results from multivariate analysis indicated no relationship with nutritional or medical complications and type of feeding path.

4. Discussion

To our knowledge, this is the first study to scientifically weigh nutritional intake and its relation to complications in AHSCT patients prospectively randomized to either EN or PN. While we report no statistical significance between the two groups, these data are clinically relevant from both a safety and cost benefit perspective. This is vitally important to both our patients and the health care institution based on our current Canadian health care crisis. These results allow us to say with some confidence that EN should not be seen as an inferior method of nutrition support for AHSCT patients moving forward. EN remains a safe and acceptable method of nutrition to AHSCT patients, though with nonsignificant benefit as compared to the PN arm. There is less risk of GI mucositis and less duration of nutritional supplementation required in the EN method as compared to the PN method.

We report that 85% of AHSCT patients on EN were classified as SGA‐B or ‐C (moderately to severely malnourished) compared to 65% of PN patients at baseline. By Day 15, 81% of EN patients and 83% of PN patients were classified as moderately malnourished, respectively. None of the patients were classified as severely malnourished at Day 15. To our knowledge, no studies have examined SGA between AHSCT patients on EN and PN. In a retrospective study by Brotelle et al. examining malnutrition in allo‐HSCT patients, 26% of patients were malnourished upon admission compared to 57% at discharge [24]. The same study also found no significant difference between undernourished or well‐nourished groups regarding the type of nutritional support (EN or PN) [24]. Furthermore, in the multivariate analysis, Brotelle et al. found that malnutrition at hospital admission for allo‐HSCT increased the risk of malnutrition after allo‐HSCT (OR = 3.60 [0.95, 13.67], p = 0.06) [24]. Notably, in a prospective cohort study in 123 autologous and allogenic hematopoietic stem cell transplant patients, 59.7% of participants were SGA‐B or ‐C at discharge [19]. Additionally, an observational prospective study by Pereira et al. in 76 critically ill patients reported that 60.5% of patients were classified as SGA‐B–C, and that these patients had a longer LOS (p < 0.003) and a higher mortality rate compared to well‐nourished patients [21].

Malnutrition, including underweight patients, appears to be a significant ongoing issue in hospitalized patients. Unfortunately, most of the research on malnutrition in stem cell transplant patients has focused on allogeneic stem cell transplant patients. However, Navarro 2010 examined both allogeneic and AHSCT in a retrospective observational study of 4213 with ~9% AHSCT patients [25]. They reported that underweight patients experienced poor clinical outcomes, including lower odds of survival (OS) (relative risk [RR]: 1.92, 95% confidence interval [CI]: 1.28–2.89, p = 0.002); significantly higher transplant‐related mortality (TRM) (RR: 2.23, 95% CI: 1.17–4.25, p = 0.014) and higher relapse rates (RR: 2.06, 95% CI: 1.20–3.54, p = 0.009) [25]. Similarly, in a large retrospective observational study examining 12 050 allogenic patients, it was reported that underweight patients had lower OS (HR: 1.10, 95% CI: 1.02–1.19, p = 0.018) and a higher relapse rate when compared to normal weight patients (HR: 1.16, 95% CI: 1.06–1.28, p = 0.002) [26]. In contrast, several other studies involving allogenic patients reported no adverse outcomes in those patients with low BMI and/or underweight patients [27]. Interestingly, a prospective observational study of 77 allogenic patients reported no correlation between BMI and neutrophil engraftment or GVHD [28]. However, there was significantly better survival in obese patients [28]. This is not surprising, as it has been reported in interstitial lung disease patients that increased BMI resulted in less morbidity and mortality posttransplant [29, 30].

Poor clinical outcomes have also been reported when examining the effect of weight loss posttransplant. In a large retrospective observational study, examining 12 050 allogenic patients reported that underweight patients had lower odds of survival (HR: 1.10, 95% CI: 1.02–1.19, p = 0.018) and a higher relapse rate when compared to normal weight patients (HR: 1.16, 95% CI: 1.06–1.28, p = 0.002) [26]. Furthermore, the percentage weight loss may also play a significant role in morbidity and mortality posttreatment [31]. A prospective observational study involving 180 allogenic recipients reported a significantly longer LOS in those patients with > 10% weight loss [31].

While we report no difference in LOS between EN and PN groups, very little has been published on LOS in autologous transplant patients. In a review of 49 studies, including 27 RCT's, Cadena et al. reported no difference in hospital LOS between EN and PN among critically ill patients [32]. Similarly, in a multicenter randomized controlled trial by Reignier et al. examining EN and PN in critically ill adults, they also reported no significant difference in ICU LOS or hospital LOS [33]. In contrast, a systematic review and meta‐analysis of randomized controlled trials by Elke et al. reported that the use of EN significantly decreased the length of ICU LOS compared to PN in critically ill patients (95% CI –1.23, −0.37, p = 0.0003) [34]. However, no significant effect was seen with subgroup analysis regarding hospital LOS, likely due to patients not meeting their caloric requirements while on EN compared to PN [34].

Previous and current systematic reviews have been able to lend support for the concept that feeding the gut may be the best approach in critically ill patients. A 2017 systematic review by Baumgartner et al. reporting on 13 allo‐HSCT studies with 18 167 patients reported that EN was associated with better survival, lower complication rates, less acute GVHD, and faster neutrophil recovery compared to those on PN [23]. Furthermore, these authors reported negative associations of malnutrition and clinical outcomes, based on moderate quality studies, indicating that EN may be a superior nutrition modality versus PN. Similarly, a more recent systematic review by Zama et al. examining eight studies on allo‐HSCT, the incidence of acute GVHD in a total of 495 patients (RR: 0.69; 95% CI: 0.56–0.86; p = 0.0007) and gut GVHD (RR: 0.44; 95% CI: 0.30–0.66; p < 0.0001) in 396 patients were lower in the EN group than in the PN group [35]. However, no differences were found in the two groups in terms of OS at day 100 (RR: 1.07; 95% CI: 0.95–1.21; p = 0.29) [35].

Based on a recent consensus statement (2022), Bechtold et al. reported that EN may be effective in improving clinical outcomes and reducing the risk of nutrition issues in HSCT patients [36]. In support of this, a retrospective cohort study by Beckerson et al. examining 484 allo‐HSCT patients reported increased non‐relapse mortality with adequate PN compared to adequate EN (HR: 2.9; 95% CI: 1.6–5.4, p < 0.001) [15]. There were also increased incidences of all GVHD in patients who received PN compared to EN (OR: 1.8; 95% CI: 1.1–3.0; p = 0.018) [15]. These findings are consistent with a prospective study of 121 allo‐HSCT patients by Seguy et al. who reported improved clinical outcomes for those patients on EN, such as better OS (HR: 0.2, 95% CI: 0.05–0.77, p = 0.019); shorter time to engraftment (p = 0.004); fewer transfusions (p = 0.004); less GVHD (p = 0.009), and less severe GVHD (p = 0.0007) [12]. Similarly, in a retrospective cohort study by Guièze et al., they reported that patients on EN experienced minimal morbidity issues, including less days with fever (2 vs. 5 days; p < 0.01) and less use of antifungal therapy (7 vs. 17 patients; p < 0.01) [37]. As well, no differences were seen in OS, and they reported a 14% mortality in both groups at day 100 [37]. Taken together, these data suggest that EN may be a suitable alternative to PN in allo‐HSCT patients. Whether this can be extrapolated to AHSCT patients warrants further research. In contrast, in a recent observational, nonrandomized, retrospective cohort of AHSCT patients receiving 11 days of PN nutrition support compared to those not receiving PN, the authors reported no improved OS between the two groups [38].

Our study has some limitations that need to be addressed. This pilot study is limited by the sample size, which impedes the ability to determine significance. Using the results in a post hoc analysis to better predict the needed sample size, we determined we would need between 63 and 93 participants per arm to achieve a power of 80% and a level of significance of 5% (two sided) based on using LOS, CRP, and/or BMI as variables. The means for each group are comparable, and the p values were not approaching significance, suggesting that even with a larger sample size, cost will likely remain the only significant difference between these two modalities.

However, we are currently organizing a multicenter trial to improve our sample size, and our collective findings should be able to provide more robust conclusions. Furthermore, this study was interrupted by the COVID‐19 pandemic, which prolonged the study and limited accessibility to participants. As well, with hospital and research resources diverted to some of the more pressing needs, we were unable to complete all data collection, as some patients presented with COVID‐19 related complications.

In conclusion, our results support the concept that EN is a safe, convenient, and cost‐effective nutritional support modality for AHSCT patients since medical and nutritional outcomes were similar between those receiving EN compared to PN. The increasing use of EN in AHSCT patients is the need of the hour and should alert clinicians working with this patient population to consider implementing EN. However, a larger multicenter trial is needed to further validate these results.

Author Contributions

J.M. and B.H. provided expert consultation, analyzed data, critical review, drafted the manuscript. H.R., M.M., A.A., and C.S. collected the data, drafted, and critically revised the manuscript. A.F., M.A., and A.X. were responsible for clinical care, data collection, and manuscript review. U.D. and B.H. designed the protocol, critically revised the manuscript, and supervised the project.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Research data are not shared.