How to Maintain a Healthy Gut Microbiome in Children with Cancer? Gut Microbiome Association with Diet in Children with Solid Tumors Postchemotherapy

Abstract

Malnutrition is a common complication in children with cancer. Cancer treatment and malnutrition can disrupt gut microbiome diversity and composition. The gut microbiome is of broad interest to better understand the mechanisms of malnutrition in cancer therapy. This study aimed to compare the gut microbiome between children with solid tumors postchemotherapy and healthy controls, and investigated the association of the putative microbiome differences with diet. Study participants were 27 children (7–18 years) with solid tumors within the first year after the completion of chemotherapy and 22 healthy controls. The study groups did not have a statistically significant difference in age, race, sex, and body mass index. At study intake, the participants completed the Block Kids Food Screener for dietary intakes in the past week. Fecal specimens were collected and analyzed for the gut microbiome. The cancer and control groups differed in gut microbial β-diversity and abundance analyses. The macronutrient intakes such as carbohydrates, fiber, beta-carotene, and vitamin B6 were positively associated with α-diversity. Children with adequate vitamin B6 had a higher Chao1 diversity index than children with inadequate or excessive intake (p = 0.0004). Children with excessive selenium intake had a trend for higher Pielou's_e index than children with inadequate intake (p = 0.091). Maintaining a healthy gut microbiome is critical among children with cancer. This study provides new insights on the linkages between dietary intakes and the gut microbiome in children with solid tumors postchemotherapy. These findings, if replicated in future independent studies, may help anticipate malnutrition and plan for personalized nutrition approaches during chemotherapy in pediatric cancers.

Introduction

Approximately 16,000 children and adolescents are diagnosed with cancer in the United States each year (Snaman et al., 2018). Among children with cancer, about 30% of them are diagnosed with extracranial solid tumors (Sharma et al., 2017). Malnutrition is a common complication in children with cancer (Begum et al., 2012). Current literature has reported that both malnutrition and undernutrition are highly prevalent from diagnosis until the completion of therapy, particularly among children with solid tumors (Gaynor and Sullivan, 2015; Iniesta et al., 2015).

The gut microbiome, defined as the collection of microbes and their genomes in the gastrointestinal (GI) tract (Gilbert et al., 2018), plays a critical role in human health and disease (Lynch and Pedersen, 2016). Accumulating evidence has demonstrated that long-term diet is a primary driver of the diversity and composition of the gut microbiome (Johnson et al., 2019; Xu and Knight, 2015), accounting for 44% of the total variation in average microbiome composition. A previous study showed that there was significant longitudinal pairing of diet with the microbiome for 78% of the subjects (Singh et al., 2017). Intake of specific dietary components further indicated how certain bacteria respond to specific nutrients (Leshem et al., 2020). Nutrients such as protein, fat, digestible and nondigestible carbohydrate, prebiotics, and polyphenols could individually induce shifts in the gut microbiome with secondary effects on host immunologic and metabolic markers (Leshem et al., 2020; Singh et al., 2017). Thus, it is important to build a healthy gut microbiome through modulating diet (Johnson et al., 2019).

Maintaining a healthy gut microbiome is critical among children with cancer as dysbiosis in the gut microbial composition has been widely reported across the continuum of cancer treatment (Bai et al., 2018; Bhatt et al., 2017; Chua et al., 2020; Rajagopala et al., 2016) and even survivorship (Cozen et al., 2013). Dysbiotic gut microbiome (i.e., loss of keystone taxa, loss of diversity, shifts in metabolic capacity, or blooms of pathogens) (Levy et al., 2017; Vangay et al., 2015) not only interferes with cancer chemotherapeutic metabolism, but also serves as a potential biomarker of GI toxicity in children with cancer, including mucositis, diarrhea, constipation, and infections (Stringer et al., 2013).

Based on the microbiome-gut-brain axis (Bajic et al., 2018; Song and Bai, 2021), disrupted gut microbiome was associated with psychoneurological toxicities such as inflammatory pain, fatigue, anxiety, depression, and cognitive dysfunction (Bai et al., 2018, 2020; Song and Bai, 2021).

Currently, dysbiotic gut microbiome profiles have been reported in children with cancer receiving treatments (e.g., chemotherapy) and cancer survivors (Rajagopala et al., 2016; Touchefeu et al., 2014). Specifically, children and adolescents with acute lymphoblastic leukemia (ALL) reported a lower diversity of the gut microbiome than healthy controls (Liu et al., 2020; Rajagopala et al., 2016); compared with the day before chemotherapy, the number of bacteria dramatically decreased after chemotherapy started (Huang et al., 2012). Additionally, Cozen et al. (2013) found that cancer survivors of adolescent and young adult Hodgkin lymphoma showed a significantly lower value of unique operational taxonomic units (OTUs) of the gut microbiome than healthy controls.

There is a lack of research focusing on the gut microbiome in children with solid tumors (Bai et al., 2018) and relationships between diet and changes in the gut microbiome have yet to be studied (Xu and Knight, 2015). As children with cancer experience both malnutrition and alterations in the gut microbiome across cancer treatments, understanding associations between diet and the gut microbiome could provide new insights into biological mechanisms of cancer treatment-related symptoms and toxicities. Finding out the relationship between the gut microbiome and diet in children with solid tumors could help clinicians better understand how to use diet to modulate the gut microbiome, therefore relieving treatment-related GI toxicities (e.g., stomatitis, constipation, and diarrhea) and central nervous system-related toxicities (e.g., anxiety and cognitive dysfunction).

The objectives of this study were to (1) compare the intake of macronutrients and antioxidant nutrients between children with solid tumors postchemotherapy (within 1 year) and those of healthy controls; and (2) examine the association between macronutrients, antioxidant nutrients and the gut microbiome in this population.

Materials and Methods

Design and setting

This study used a cross-sectional design. We enrolled 7- to 18-year-old children with solid tumors after they were consented to participate in this study from the Aflac Cancer and Blood Disorder Center in Children's Healthcare of Atlanta in Atlanta, Georgia. Age, sex, and race-matched healthy controls were recruited via flyers, online e-news blast, and ResearchMatch (a disease-neutral web-based recruitment registry).

Participants

This study included two groups of children: one group with solid tumors (case group) and one group of healthy children (control group). For the case group, eligible children had to meet the following criteria: (1) were diagnosed with solid tumors (e.g., sarcomas, germ-cell tumors, and neuroblastoma); (2) received at least one cycle of chemotherapy; (3) completed chemotherapy within 1 year; and (4) agreed to participate. Children were excluded if they did not receive any chemotherapy, could not understand, and answer the questionnaires, or had a cognitive impairment, such as Down's syndrome.

Regarding the control group, healthy children who had not received antibiotics within the past 4 weeks and who had not been diagnosed with chronic or autoimmune diseases or conditions that can influence the gut microbiome profiles were included. These two groups were matched by age, sex, and race during recruitment.

Study endpoints and measures

Gut microbiome

The gut microbiome was assessed using fecal specimens. According to the Human Microbiome Project (HMP) protocol (Human Microbiome Project Consortium, 2012), parents and children were taught to collect fecal specimens using the stool collection kit and store them in freezer at home before shipping to the laboratory. During the hospital visit, the trained research staff provided the parent with the fecal sample collection kit. The samples were frozen before being shipped to the Biobehavioral Laboratory at School of Nursing, Emory University. Once received by the research staff, the stool samples were stored in a −80°C freezer until DNA extraction and assaying.

Dietary intakes

The dietary intake of nutrients was measured using the Block Kids Food Screener (BKFS), which includes 41 items developed by NutritionQuest (Berkeley, CA, USA). This instrument has been validated to evaluate dietary intake of nutrients and food groups among children aged 2–18 years. Parents, together with their child, completed the BKFS to estimate the child's intake of fruit, vegetables, dairy, whole grains, protein sources, saturated fat, and sources of added sugars. The frequency of food and beverage consumption ranges from “none” to “every day.”

Studies have proved that BKFS has good relative validity to examine the nutrients and food groups in children and adolescents. Overall correlations with 24-h dietary recalls and with the Food Frequency Questionnaires (FFQ) were high and Bland–Altman plots showed strong agreements between BKFS and FFQ (Hunsberger et al., 2015). In this study, macronutrients and antioxidant nutrients were analyzed (Goñi and Hernández-Galiot, 2019; Lobo et al., 2010).

Demographic and clinical variables

Child's demographic data (e.g., age, gender, race/ethnicity, height, weight, and body mass index [BMI] percentile), health history (e.g., use of antibiotics and disease history), cancer diagnosis, and treatment data (e.g., diagnosis and cycles of chemotherapy) were obtained from the electronic medical record.

Procedures

The study was reviewed and approved by the authors' institutional review board (Emory University IRB00091722). Participants in the case group were recruited during their routine outpatient clinic visits. Clinical collaborators from Children's Healthcare of Atlanta in Atlanta identified eligible patients and asked them whether they were willing to discuss the study with our research team. After they agreed, one trained research staff described the study, consented parents, and assented age-eligible patients. Questionnaires were distributed to the children to complete during clinic visits, and parents were instructed on the stool specimen collection at home. The electronic medical records of the pediatric patients with solid tumor were used to collect the demographic information, health history, cancer diagnosis, and treatment-related information. For the control group, all the procedures were the same, excluding the use of the electronic medical records.

DNA extraction and sequencing

Based on the HMP standard operating protocol, the microbial DNA was extracted from fecal samples using the Power Soil isolation kit (Mo Bio Laboratories, Carlsbad, CA, USA) at Environmental Microbial Genomics Laboratory in Georgia Institute of Technology. The 16S rRNA amplicon libraries were prepared for the 16S rRNA V4 region (Gevers et al., 2012; Poretsky et al., 2014). These 16S rRNA amplicons were generated using KAPA HiFi HotStart ReadyMix (KK2600; KAPA Biosystems) and primers specific to 16S V4 region of Bacteria and indices were attached using the Nextera XT Index kit (FC-131-1001; Illumina).

Clean-up was performed on the indexed libraries using AMPure XP beads. The 16S libraries were pooled in equal amounts based on fluorescence quantification. Each run included a control template to test for the polymerase chain reaction (PCR) accuracy and possible contamination. Final library pools were quantitated via quantitative PCR (catalog KK4824; Kapa Biosystems). The pooled library was sequenced on an Illumina miSeq using miSeq v3 600 cycle chemistry (catalog MS-102-3003; Illumina) at a loading density of 8 pM with 20% PhiX, at PE300 reads. The microbial sequencing led to paired-end sequences.

Bioinformatics and statistical analysis

The macronutrients and micronutrients intake from the BKFS were calculated by the NutritionQuest. Based on the recommended daily nutritional intake from 2015 to 2020 dietary guideline, the child's nutritional intake was categorized into three levels: inadequate level (less than 90% of the recommended dietary allowance [RDA]); adequate level (within the range of 90–110% of the RDA); and excessive level (more than 110% of the RDA).

Quantitative Insight into Microbial Ecology 2 (QIIME 2) was used to analyze the taxonomic composition and diversity of the gut microbiome (Bai et al., 2019; Bolyen et al., 2019). QIIME 2 default parameters were used for sequencing data, and sequence quality was filtered with DADA2 (Callahan et al., 2016) to infer exact sequence variants. By removing the length of primers, the raw sequences were trimmed at 0 and 0 base pairs, and then truncated at 250 and 215 base pairs based on the Phred Quality score >30.

Taxonomies were assigned by a Naive Bayes classifier trained on the Greengenes database with sequences adapted to the 16S rRNA V4 gene region. The α-diversity (richness and evenness of the gut microbiome within samples) was calculated using four different parameters: Shannon's index, Chao1, Faith's phylogenetic diversity (Faith's_PD), and Pielou's evenness (Pielou's_e). UniFrac distance metrics were used to describe dissimilarities between bacterial communities based on their phylogenetic information. Both the weighted and unweighted UniFrac distance metrics as measures of β-diversity and principal coordinates analysis were used to visualize diversity patterns.

Pairwise permutational multivariate analysis of variance (PERMANOVA) (Kelly et al., 2015) was used to test taxa dissimilarities between cancer and healthy control groups. Mann–Whitney U tests were used to compare dietary status between children with cancer and healthy controls. Spearman's correlation was used to explore correlations between the relative abundance of gut microbiome taxa and the α-diversity index with diet. PERMANOVA (Kelly et al., 2015) was used to test taxa dissimilarities between nutritional intake levels. The analysis of composition of microbiomes (ANCOM) (Mandal et al., 2015) was used to analyze associations between diet and the abundance of the gut microbiome.

Results

Characteristics of participants

Forty-nine children were enrolled including 27 cancer cases and 22 healthy controls (Table 1). The cancer group included 13 boys and 14 girls, with a mean age of 14.4 years. The control group consisted of 9 boys and 13 girls, with a mean age of 12.1 years. No significant differences were found between the two study groups in age (p = 0.053), gender (p = 0.774), race (p = 0.172), and BMI (p = 0.346).

Table 1.

Description of Demographic and Clinical Variables

Variable	Cancer (n = 27)	Control (n = 22)	Total (n = 49)	p
Age in year, mean (SD)	14.4 (4.0)	12.1 (4.2)	13.3 (4.2)	0.053
Race, n (%)
Black/biracial	11 (30.6)	11 (50.0)	23 (38.3)	0.17
White	20 (55.6)	7 (31.8)	28 (46.7)
Other	5 (13.9)	4 (18.2)	9 (15.0)
Gender, n (%)
Male	13 (48.1)	9 (40.9)	22 (44.9)	0.77
Female	14 (51.9)	13 (59.1)	27 (55.1)	0.77
BMI, mean (SD)	21.0 (5.4)	19.9 (4.0)	20.5 (4.8)	0.35
Level of BMI, n (%)
Underweight	3 (11.1)	1 (4.5)	4 (8.2)
Normal	15 (55.6)	15 (68.2)	30 (61.2)
Overweight	6 (22.2)	2 (9.1)	8 (16.3)
Obesity	3 (11.1)	4 (18.2)	7 (14.3)
Post-treatment in month, mean (SD)	5.2 (4.3)	NA
Chemotherapy, n (%)
Yes	27 (100)	NA	27 (55.1)	NA
No	0	NA	22 (44.9)	NA
Cancer type, n (%)
Sarcomas	16 (59.3)	NA
Germ cell tumor	3 (11.1)
Neuroblastoma	3 (11.1)
Wilms' tumor	2 (7.4)
Others	3 (11.1)

BMI, body mass index; NA, not applicable; SD, standard deviation.

Dietary intakes in children with cancer and healthy controls

Table 2 shows comparisons of dietary intake between the cancer and healthy control groups. Macronutrients were analyzed both on the net weight of intake and the percentage they took up in total intake based on calories. The cancer group had a significantly higher total daily calories intake (p = 0.047) while there were no significant differences of percentage intake of macronutrients between the two groups. As for micronutrients and trace elements, the cancer group had a significantly higher intake in vitamin E (p = 0.026), vitamin C (p = 0.022), and selenium (p = 0.027).

Table 2.

Comparison of Dietary Intakes Between Cancer Children and Healthy Controls

Variables	Control (n = 22)	Cancer (n = 27)	All (n = 49)	p
Macronutrients
Daily calories, kCal	1055.04 (355.1)	1339.55 (569.2)	1211.81 (501.1)	0.047
Protein
Total protein, g	45.58 (16.3)	56.38 (24.0)	51.53 (21.4)	0.078
Protein%	17.12	17.46	17.32	0.61
Fat
Total fat, g	44.11 (15.9)	56.74 (27.5)	51.07 (23.7)	0.062
Fat%	37.80	37.62	37.69	0.55
Carbohydrate
Total carbohydrate, g	122.39 (44.9)	155.76 (68.4)	140.77 (60.8)	0.055
Carbohydrate%	46.41	46.38	46.39	0.43
Fiber
Total fiber, g	8.59 (4.0)	11.23 (5.9)	10.04 (5.3)	0.08
FiberWt%	1.79	1.77	1.78	0.38
Antioxidant micronutrients
Vitamin E, mg	3.13 (1.1)	4.19 (1.8)	3.63 (1.6)	0.026
Vitamin A, mcg	402.26 (191.5)	502.94 (186.2)	450.20 (193.5)	0.092
Beta-carotene, mcg	1302.22 (1037.6)	2056.00 (1676.8)	1661.16 (1414.1)	0.084
Lycopene, mcg	3865.14 (3152.0)	5511.93 (3940.2)	4649.32 (3602.2)	0.14
Lutein and zeaxanthin, mcg	950.57 (829.7)	1501.81 (1080.1)	1213.07 (985.3)	0.07
Vitamin C, mcg	56.04 (26.6)	87.97 (56.3)	71.25 (45.7)	0.022
Selenium, mcg	51.55 (20.1)	68.9 (28.5)	59.81 (25.7)	0.027

Bold font indicates a p-value that is or shows trend of being statistically significant.

Carbohydrate%, percentage of carbohydrate in total intake based on calories; Fat%, percentage of fat in total intake based on calories; FiberWt%, percentage of fiber in total intake based on weight; Protein%, percentage of protein in total intake based on calories.

Profiles of the gut microbiome in children with solid tumors and healthy children

The raw sequence count per sample ranged from 3456 to 234,172 for the gut microbiome samples, with an average sequence count of 58,613 per sample. After the DADA2 process, 1138 features were reported, with a total frequency of 1,926,036. Frequencies per feature ranged from 2 to 249,452, with a median frequency of 72; feature frequencies per sample ranged from 1857 to 165,900, with a median frequency of 33,888. By using the trained classifiers based on Greengenes 13_8 99% OTUs (taxonomic assignment based on a 99% similarity), the bacterial taxonomy of the fecal specimens included 11 bacterial phyla and 133 genera. The top dominant bacterial phyla (Fig. 1A) were Bacteroidetes, Firmicutes, Proteobacteria, Verrucomicrobia, and Actinobacteria, whereas the dominant bacterial genera (Fig. 1B) included Bacteroides, Faecalibacterium, Prevotella, Roseburia, and Ruminococcus.

FIG. 1.

The gut microbial taxa in 27 children with solid tumors and 22 healthy children. (A) Presents all the phyla among children with cancer and healthy children; (B) presents top 30 genera among children with cancer and healthy children.

All α-diversity estimates, including Shannon diversity index, Observed OTUs, Faith_PD, and Pielou's_e, showed no significant associations between children with cancer and healthy controls. Significant dissimilarities between the cancer and control groups were found in β-diversity based on Jaccard distance (p = 0.009), while unweighted UniFrac distance suggested a trend of dissimilarities in gut microbiome between the two groups (p = 0.074).

Abundance analysis was conducted using ANCOM. At phylum level, children with cancer showed lower abundance of Firmicutes, Bacteroidetes, and Proteobacteria. At genera level, compared with healthy controls, children with solid tumors had lower abundance in taxa such as Clostridium, Oscillospira, Coprococcus, Bacteroides, Faecalibacterium, Ruminococcus, and Dialister, but higher abundance in taxa Anaerostipes, Dorea, and Sutterella.

Associations between dietary intakes and the gut microbiome diversity

Table 3 shows correlations between diet and the α-diversity indices of the gut microbiome. Total protein intake showed negative associations with Shannon's index (p = 0.02) and Pielou's_e index (p = 0.03); based on the calorie's percentage, intake of carbohydrate (p = 0.07) and fiber (p = 0.07) showed trends of positive associations with Chao1. Regarding the micronutrients, the amount of beta-carotene intake had a positive correlation with Faith's_PD (p = 0.02) and a trend of positive association with Chao1 (p = 0.08); however, the amount of selenium intake was negatively correlated with Shannon's index (p = 0.05) and Pielou's_e (p = 0.03), and vitamin A showed a trend of negative association with Pielou's_e (p = 0.06).

Table 3.

Correlation Between Continuous Dietary Status and Microbial α-Diversity

Variable	Faith's_PD			Chao1			Shannon			Pielou_e
Variable	r	p	n	r	p	n	r	p	n	r	p	n
Daily calories	−0.06	0.67	49	−0.08	0.59	49	−0.19	0.19	49	−0.18	0.21	49
Total protein	−0.12	0.41	49	−0.15	0.32	49	−0.33	0.02	49	−0.30	0.03	49
Protein%	−0.08	0.63	49	−0.18	0.27	42	−0.23	0.15	42	−0.11	0.50	42
Total fat	−0.10	0.51	49	−0.15	0.31	49	−0.16	0.27	49	−0.12	0.43	49
Fat%	−0.16	0.31	49	−0.23	0.14	42	−0.01	0.97	42	0.13	0.40	42
Total carbohydrate	0.05	0.71	49	0.06	0.70	49	−0.08	0.60	49	−0.13	0.37	49
Carbohydrate%	0.16	0.31	49	0.28	0.07	42	0.13	0.41	42	−0.04	0.79	42
Total fiber	0.16	0.28	49	0.15	0.31	49	0.06	0.66	49	−0.03	0.86	49
FiberWt%	0.22	0.16	49	0.28	0.07	42	0.26	0.10	42	0.15	0.35	42
Vitamin E	0.08	0.60	42	0.04	0.81	42	0.04	0.82	42	0.01	0.93	42
Vitamin A	0.20	0.20	42	0.07	0.64	42	−0.06	0.73	42	−0.29	0.06	42
Beta-carotene	0.35	0.02	42	0.27	0.08	42	0.17	0.28	42	−0.07	0.67	42
Lycopene	0.20	0.21	42	0.23	0.14	42	0.07	0.67	42	−0.04	0.81	42
Lutein and zeaxanthin	0.25	0.11	42	0.17	0.28	42	0.12	0.44	42	−0.06	0.69	42
Vitamin C	0.19	0.23	42	0.22	0.16	42	0.22	0.16	42	0.14	0.38	42
Selenium	−0.06	0.69	42	−0.13	0.41	42	−0.30	0.05	42	−0.33	0.03	42

Bold font indicates a p-value that is or shows trend of being statistically significant.

Spearman correlation was used.

The α-diversity was compared between three nutritional intake levels (inadequate, adequate, and excessive) (Table 4). Compared with the group with inadequate carbohydrates intake, the adequate intake group had a significantly higher Chao1 (p = 0.005) and Faith's_PD (p = 0.008), and a trend of higher Shannon's index (p = 0.083). Children with adequate vitamin B6 had a higher Chao1 diversity index than children with inadequate or excessive vitamin B6 (p = 0.0004). Children with excessive selenium intake had a trend of higher Pielou's_e index than children with inadequate selenium intake (p = 0.091).

Table 4.

Different α-Diversity Values of the Gut Microbiome Compared Between Groups with Different Levels of Nutrient Intake

α-Diversity	Nutrient	Group A	Group B	H	p
Chao1	Vitamin B6	Adequate	Excessive	4.18	0.041
Chao1	Carbohydrate%	Inadequate	Adequate	7.99	0.005
Faith's_PD	Vitamin B6	Inadequate	Excessive	3.42	0.065
	Vitamin B6	Adequate	Excessive	3.75	0.053
	Carbohydrate%	Inadequate	Adequate	7.03	0.008
Shannon	Vitamin B6	Adequate	Excessive	3.75	0.053
Shannon	Carbohydrate%	Inadequate	Adequate	3.00	0.083
Pielou's_e Evenness	Selenium	Inadequate	Excessive	2.86	0.091

Bold font indicates a p-value that is or shows trend of being statistically significant.

Associations between levels of dietary intakes and the gut microbial abundance

ANCOM was used to analyze associations between gut microbiome abundance and levels of nutritional intake. At the phylum level, children with adequate fiber showed a higher abundance of Cyanobacteria. At the genus level, children with excessive total calories intake had a higher abundance of Catenibacterium. Children with inadequate fat intake based on calories percentage had higher abundances in family S24-7 and in genus Megasphaera while children with adequate fat intake had higher abundances in bacterial family Erysipelotrichaceae and Peptostreptococcaceae. Children with adequate fiber intake had a higher abundance of bacterial order YS2.

Discussion

This study compared dietary intakes between children with solid tumors and healthy children and examined associations between dietary intakes and the gut microbiome in these children. We found that children with cancer reported significantly higher intakes of macronutrients and antioxidant nutrients than healthy children, but no differences in major energy ratios. These two groups showed no significant differences in α-diversity measures while the β-diversity suggested that the two groups have different compositions of the gut microbiome. Additionally, we found significant associations between macronutrients (e.g., carbohydrates and fiber) and micronutrients (e.g., selenium intake and vitamin A) and the gut microbiome α-diversity.

The lower abundance of microbial taxa in children with solid tumors after receiving chemotherapy is consistent with the results from previous studies focusing on gut microbiome composition in children with cancer, although the specific microbial genera that differ between cancer and control groups in this study are at variance from those in previous studies. Studies focusing on the gut microbiome composition in children with ALL found that the children with cancer had significantly lower amounts of Clostridium, Roseburia, Bifidobacteria, Lactobacillus, and Escherichia coli (Huang et al., 2012; Rajagopala et al., 2016). In this study, the patient group had lower abundance in Clostridium, Oscillospira, Coprococcus, Bacteroides, Faecalibacterium, Ruminococcus, and Dialister. Differences in the microbial taxa listed between this study and previous studies might be due to the disparate cancer development and pathogenesis in between cancer types.

Clostridium, Coprococcus, Bacteroides, and Faecalibacterium are bacterial taxa that are associated with immunity and anti-inflammation functions in the bowel. Specifically, Clostridium is believed to have a role in directing the responses of Treg cells (Forbes et al., 2018), a type of immune cells, and a decrease in Clostridium might be related with disturbances in the immunity in the gut and may cause GI toxicities. As reported in previous study, Coprococcus plays a role in protecting the liver from inflammation (Pedersen et al., 2019). Lower levels of Bacteroides and Faecalibacterium are both found to be associated with intestinal disorders, such as inflammatory bowel disease (Martín et al., 2017; Zhou and Zhi, 2016). All altered levels of these bacteria in the cancer group probably suggested the role of the gut microbiome in the cancer treatment-related GI symptoms such as stomatitis and diarrhea.

Oscillospira is another genus that is lower in abundance in the cancer patient group. Multiple studies have observed that Oscillospira is enriched in lean subjects and is associated with leanness or lower BMI levels in both adults and children (Konikoff and Gophna, 2016; Petrov et al., 2017). Given these observations of the strong association between higher levels of Oscillospira and lower BMI, it is reasonable to suggest that the decrease of Oscillospira in the cancer group might be associated with the trend of higher BMI levels in these subjects.

Ruminococcus is found to serve to degrade and convert complex polysaccharides into a variety of nutrients for the hosts (La Reau and Suen, 2018), and there might be a relationship between the decreased level of Ruminococcus and poorer absorption of nutrients or even malnutrition. The cancer group had higher abundance in taxa Anaerostipes, Dorea, and Sutterella, which are found to associate with inflammation. Dorea is positively correlated with inflammatory markers (Brahe et al., 2015) and its increase is associated with inflammatory bowel disease (Guinane and Cotter, 2013), and Sutterella has mild pro-inflammatory capacity in the human GI tract (Hiippala et al., 2016).

In this study, children from the cancer group had higher intake of macronutrients and micronutrients. Specifically, they showed significantly higher intakes of daily calories, and a trend of higher intakes of total protein, fat, carbohydrates, and fiber than the control group. However, there were no significant differences between the two study groups if the amount of macronutrients intake is viewed based on intake percentage. The compromised GI functions and manifestations of cancer treatment-related GI symptoms affect the absorption of nutrients among children with cancer, and therefore they need to compensate by increasing intakes to meet the required energy for daily activities and cancer recovery.

Due to the importance of nutrients in cancer recovery, the European Society for Clinical Nutrition and Metabolism (ESPEN) guideline strongly recommends the energy intake of the patients ranging between 25 and 30 kcal/(kg·day) to meet the energy expenditure, and the protein intake above 1 g/(kg·day) and even up to 1.5 g/(kg·day). Clinically, this point is also emphasized to the parents, possibly explaining the reason behind higher nutrition intake in children with cancer. The ESPEN guideline strongly recommends against any dietary provisions that restrict energy intake in patients with or at risk of malnutrition (Arends et al., 2017). Thus, more attention should be paid to adequate dietary intakes which may be associated with cancer-related toxicities, such as fatigue and comorbidities such as obesity.

A positive correlation was found between beta-carotene intake and α-diversity index Faith's_PD. A high diversity of the gut microbiome had more healthy effects and a low gut microbiome diversity was associated with a higher weight gain in the long term (Menni et al., 2017). When the intake of beta-carotene increases, there is a higher microbial richness in our sample. Beta-carotene is the most abundant vitamin A carotenoid precursor in the human diet and can only be acquired through food or supplements (Wassef et al., 2014). Both beta-carotene and vitamin A function as antioxidants and participate in the regulation of host immune responses by activating immune cells such as macrophages and natural killer cells (Lyu et al., 2018).

Studies have shown that retinoic acid, converted from vitamin A, is a critical regulator for the intestinal immune response. In mice, a lack of carotenoids and vitamin A in the diet reduces commensal microbes, and thus suppresses pro-inflammatory Th17 cell generation in the gut (Cha et al., 2010). Mechanisms of the association between the beta-carotene intake and microbial richness might be due to the fact that the supplementation of beta-carotene increases IgA production and regulates the immune responses in the GI system (Lyu et al., 2018), which in turn protect the commensal microbes in the gut and help maintain a high microbial α-diversity.

Through the analyses of associations between the gut microbiome and the diet and nutritional intake levels, positive correlations were reported between α-diversity and carbohydrates and vitamin B6 intakes. These findings were consistent with previous studies. Vitamin B6 functions as an essential cofactor for enzymes involved in various metabolic activities and an increase of vitamin B6 aids in polyunsaturated fatty acid metabolism, and biosynthesis of arachidonic acid and hepatic cholesterol (Li et al., 2016). An increase of vitamin B6 also reduces the production of lithocholate (Li et al., 2016), a toxic bile acid, and promotes the homeostasis of microbial communities in the distal gut (Rodionov et al., 2019), therefore leading to a higher diversity in the gut microbiome.

Adequate intake of carbohydrates and fiber is positively correlated with α-diversity, probably because children are therefore less likely to have excessive intake of fat, and the energy density of diet is reduced. Fiber plays a critical role in the diversity of healthy gut microbiome. As reported, an increased fiber intake produces more short-chain fatty acids, which in turn promote intestinal gluconeogenesis and liponeogenesis (Menni et al., 2017).

Further work suggested that transitions to the refined diet that lacks soluble fiber is the primary driver of gut microbiota alterations (Morrison et al., 2020). Therefore, adequate intake of carbohydrates and fiber promotes GI tract health and prevents infections and colonization of the gut by pathogenic microbes (Ngalavu et al., 2020). Until now, the exact mechanisms behind the relationship of carbohydrate intake level and gut microbial diversity are still not well studied and should be explored in future investigations.

This study has several limitations. We have a small sample size, and all children were recruited from Children's Healthcare of Atlanta, Georgia. Our findings may not be generalized into other clinical settings. In addition, we only analyzed the correlations between diet and α-diversity and microbiome abundance. Lastly, the microbiome differences observed between children with solid tumors and controls can be potentially attributed to numerous factors other than the dietary intakes, such as chemotherapy and the type of cancer. Our analyses were conducted without controlling these primary confounders, which should be considered in future work.

Conclusions

Compared with healthy children, children with cancer exhibited a higher intake of daily calories, and thus had a higher intake of both macro- and micronutrients, probably attempting to compensate their compromised GI function affected by cancer and treatment. No significant differences in alpha diversity were found between the cancer and control groups, but these two groups distinguished from each other in gut microbial β-diversity and abundance analyses. Positive associations were found between the intake of certain macronutrients (carbohydrates and fiber), micronutrients (beta-carotene and vitamin B6) and α-diversity. Future studies should confirm our findings in larger independent samples with an eye to further decipher the impact of gut microbial alterations on cancer treatment symptoms and toxicities.

Maintaining a healthy gut microbiome is critical among children with cancer. In this context, this study provides new insights regarding the linkages between dietary intakes and the gut microbiome postchemotherapy in children with solid tumors. As the gut microbiome and its clinical relevance continue to be unraveled, these findings, if replicated in future independent studies, may help anticipate malnutrition and plan for personalized nutrition approaches during chemotherapy in pediatric cancers in the future.

Footnotes

Authors' Contributions

S.Z., J.B., and D.W.B. contributed to the design and conceptualization; J.B., M.M., C.P., K.S.S., B.G., and T.O. contributed to the implementation and data collection; J.B. and K.T.K. for data analysis; S.Z. contributed to writing the article. All authors made a significant intellectual contribution and agreed to submit the article for consideration for publication.

Acknowledgments

We would like to acknowledge all the participants for their participation in this study.

Author Disclosure Statement

The authors declare they have no conflicting financial interests.

Funding Information

This study was supported by the National Institutes of Health/National Institute of Nursing Research (1K99NR017897-01 and 4R00NR017897-03), the Southern Nursing Research Society/American Nurses Foundation, and Oncology Nursing Foundation Grant. The article was submitted to a preprint server Authorea (DOI: 10.22541/au.162526069.90949263/v1). The article was not previously published.

Abbreviations Used

References

Arends

, Bachmann

, Baracos

, et al. (2017). ESPEN guidelines on nutrition in cancer patients. Clin Nutr, 36, 11–48.

Bai

, Behera

, and Bruner

. (2018). The gut microbiome, symptoms, and targeted interventions in children with cancer: A systematic review. Support Care Cancer, 26, 427–439.

Bai

, Bruner

, Fedirko

, et al. (2020). Gut microbiome associated with the psychoneurological symptom cluster in patients with head and neck cancers. Cancers (Basel),, 12, 2531.

Bai

, Jhaney

, Daniel

, and Watkins Bruner

. (2019). Pilot study of vaginal microbiome using QIIME 2 in women with gynecologic cancer before and after radiation therapy. Oncol Nurs Forum, 46, E48–E59.

Bajic

, Johnston

, Howarth

, and Hutchinson

. (2018). From the bottom-up: chemotherapy and gut-brain axis dysregulation. Front Behav Neurosci, 12, 104.

Begum

, Jahan

, Tawfique

, and Mannan

. (2012). Out come of induction of remission in undernourished children with acute lymphoblastic leukaemia. Mymensingh Med J,, 21, 691–695.

Bhatt

, Redinbo

, and Bultman

. (2017). The role of the microbiome in cancer development and therapy. CA Cancer J Clin, 67, 326–344.

Bolyen

, Rideout

, Dillon

, et al. (2019). Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol, 37, 852–857.

Brahe

, Le Chatelier

, Prifti

, et al. (2015). Specific gut microbiota features and metabolic markers in postmenopausal women with obesity. Nutr Diabetes, 5, e159.

10.

Callahan

, McMurdie

, Rosen

, Han

, Johnson

, and Holmes

. (2016). DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods, 13, 581–583.

11.

Cha

, Chang

, et al. (2010). Downregulation of Th17 cells in the small intestine by disruption of gut flora in the absence of retinoic acid. J Immunol, 184, 6799–6806.

12.

Chua

, Rajasuriar

, Lim

YAL

, Woo

, Loke

, and Ariffin

. (2020). Temporal changes in gut microbiota profile in children with acute lymphoblastic leukemia prior to commencement-, during-, and post-cessation of chemotherapy. BMC Cancer, 20, 151.

13.

Cozen

, Yu

, Gail

, et al. (2013). Fecal microbiota diversity in survivors of adolescent/young adult Hodgkin lymphoma: A study of twins. Br J Cancer, 108, 1163–1167.

14.

Forbes

, Chen

C-y

, Knox

, et al. (2018). A comparative study of the gut microbiota in immune-mediated inflammatory diseases—Does a common dysbiosis exist?. Microbiome, 6, 221.

15.

Gaynor

, and Sullivan

. (2015). Nutritional status and nutritional management in children with cancer. Arch Dis Child, 100, 1169–1172.

16.

Gevers

, Knight

, Petrosino

, et al. (2012). The Human Microbiome Project: A community resource for the healthy human microbiome. PLoS Biol, 10, e1001377.

17.

Gilbert

, Blaser

, Caporaso

, Jansson

, Lynch

, and Knight

. (2018). Current understanding of the human microbiome. Nat Med, 24, 392–400.

18.

Goñi

, and Hernández-Galiot

. (2019). Intake of nutrient and non-nutrient dietary antioxidants. contribution of macromolecular antioxidant polyphenols in an elderly Mediterranean population. Nutrients, 11, 2165.

19.

Guinane

, and Cotter

. (2013). Role of the gut microbiota in health and chronic gastrointestinal disease: Understanding a hidden metabolic organ. Therap Adv Gastroenterol, 6, 295–308.

20.

Hiippala

, Kainulainen

, Kalliomäki

, Arkkila

, and Satokari

. (2016). Mucosal prevalence and interactions with the epithelium indicate commensalism of Sutterella spp. Front Microbiol, 7, 1706.

21.

Huang

, Yang

, Liu

, et al. (2012). Effect of high-dose methotrexate chemotherapy on intestinal Bifidobacteria, Lactobacillus and Escherichia coli in children with acute lymphoblastic leukemia. Exp Biol Med, 237, 305–311.

22.

Human Microbiome Project Consortium. (2012). Structure, function and diversity of the healthy human microbiome. Nature, 486, 207–214.

23.

Hunsberger

, O'Malley

, Block

, and Norris

. (2015). Relative validation of Block Kids Food Screener for dietary assessment in children and adolescents. Matern Child Nutr, 11, 260–270.

24.

Iniesta

, Paciarotti

, Brougham

, McKenzie

, and Wilson

. (2015). Effects of pediatric cancer and its treatment on nutritional status: A systematic review. Nutr Rev, 73, 276–295.

25.

Johnson

, Vangay

, Al-Ghalith

, et al. (2019). Daily sampling reveals personalized diet-microbiome associations in humans. Cell Host Microbe, 25, 789–802.e785.

26.

Kelly

, Gross

, Bittinger

, et al. (2015). Power and sample-size estimation for microbiome studies using pairwise distances and PERMANOVA. Bioinformatics, 31, 2461–2468.

27.

Konikoff

, and Gophna

. (2016). Oscillospira: A central, enigmatic component of the human gut microbiota. Trends Microbiol, 24, 523–524.

28.

La Reau

, and Suen

. (2018). The Ruminococci: Key symbionts of the gut ecosystem. J Microbiol, 56, 199–208.

29.

Leshem

, Segal

, and Elinav

. (2020). The gut microbiome and individual-specific responses to diet. mSystems, 5, e00665-20.

30.

Levy

, Kolodziejczyk

, Thaiss

, and Elinav

. (2017). Dysbiosis and the immune system. Nat Rev Immunol, 17, 219–232.

31.

, Shu

, Xu

, et al. (2016). Integrative analysis of metabolome and gut microbiota in diet-induced hyperlipidemic rats treated with berberine compounds. J Transl Med, 14, 237.

32.

Liu

, Zou

, Ruan

, et al. (2020). Pediatric acute lymphoblastic leukemia patients exhibit distinctive alterations in the gut microbiota. Front Cell Infect Microbiol, 10, 558799.

33.

Lobo

, Patil

, Phatak

, and Chandra

. (2010). Free radicals, antioxidants and functional foods: Impact on human health. Pharmacogn Rev, 4, 118.

34.

Lynch

, and Pedersen

. (2016). The human intestinal microbiome in health and disease. N Engl J Med, 375, 2369–2379.

35.

Lyu

, Wu

, Wang

, Shen

, and Lin

. (2018). Carotenoid supplementation and retinoic acid in immunoglobulin A regulation of the gut microbiota dysbiosis. Exp Biol Med (Maywood), 243, 613–620.

36.

Mandal

, Van Treuren

, White

, Eggesbo

, Knight

, and Peddada

. (2015). Analysis of composition of microbiomes: A novel method for studying microbial composition. Microb Ecol Health Dis, 26, 27663.

37.

Martín

, Miquel

, Benevides

, et al. (2017). Functional characterization of novel Faecalibacterium prausnitzii strains isolated from healthy volunteers: A step forward in the use of F. prausnitzii as a next-generation probiotic. Front Microbiol, 8, 1226.

38.

Menni

, Jackson

, Pallister

, Steves

, Spector

, and Valdes

. (2017). Gut microbiome diversity and high-fibre intake are related to lower long-term weight gain. Int J Obes (Lond), 41, 1099–1105.

39.

Morrison

, Jašarević

, Howard

, and Bale

. (2020). It's the fiber, not the fat: Significant effects of dietary challenge on the gut microbiome. Microbiome, 8, 15.

40.

Ngalavu

, Jiang

, El-Ashram

, et al. (2020). Effect of dietary fiber sources on in-vitro fermentation and microbiota in monogastrics. Animals (Basel), 10, 674.

41.

Pedersen

, Pulliam

, Patel

, et al. (2019). Liver tumorigenesis is promoted by a high saturated fat diet specifically in male mice and is associated with hepatic expression of the proto-oncogene Agap2 and enrichment of the intestinal microbiome with Coprococcus. Carcinogenesis, 40, 349–359.

42.

Petrov

, Saltykova

, Zhukova

, et al. (2017). Analysis of gut microbiota in patients with Parkinson's disease. Bull Exp Biol Med, 162, 734–737.

43.

Poretsky

, Rodriguez

, Luo

, Tsementzi

, and Konstantinidis

. (2014). Strengths and limitations of 16S rRNA gene amplicon sequencing in revealing temporal microbial community dynamics. PLoS One, 9, e93827.

44.

Rajagopala

, Yooseph

, Harkins

, et al. (2016). Gastrointestinal microbial populations can distinguish pediatric and adolescent Acute Lymphoblastic Leukemia (ALL) at the time of disease diagnosis. BMC Genomics, 17, 635.

45.

Rodionov

, Arzamasov

, Khoroshkin

, et al. (2019). Micronutrient requirements and sharing capabilities of the human gut microbiome. Front Microbiol, 10, 1316.

46.

Sharma

, Ahmad

, Bhat

, Aziz

, Lone

, and Bhat

. (2017). A profile of pediatric solid tumors: A single institution experience in Kashmir. Indian J Med Paediatr Oncol, 38, 471–477.

47.

Singh

, Chang

, Yan

, et al. (2017). Influence of diet on the gut microbiome and implications for human health. J Transl Med, 15, 73.

48.

Snaman

, Kaye

, Baker

, and Wolfe

. (2018). Pediatric palliative oncology: The state of the science and art of caring for children with cancer. Curr Opin Pediatr, 30, 40–48.

49.

Song

, and Bai

. (2021). Microbiome-gut-brain axis in cancer treatment-related psychoneurological toxicities and symptoms: A systematic review. Support Care Cancer, 29, 605–617.

50.

Stringer

, Al-Dasooqi

, Bowen

, et al. (2013). Biomarkers of chemotherapy-induced diarrhoea: A clinical study of intestinal microbiome alterations, inflammation and circulating matrix metalloproteinases. Support Care Cancer, 21, 1843–1852.

51.

Touchefeu

, Montassier

, Nieman

, et al. (2014). Systematic review: The role of the gut microbiota in chemotherapy- or radiation-induced gastrointestinal mucositis—Current evidence and potential clinical applications. Aliment Pharmacol Ther, 40, 409–421.

52.

Vangay

, Ward

, Gerber

, and Knights

. (2015). Antibiotics, pediatric dysbiosis, and disease. Cell Host Microbe, 17, 553–564.

53.

Wassef

, Wirawan

, Chikindas

, Breslin

, Hoffman

, and Quadro

. (2014). β-carotene-producing bacteria residing in the intestine provide vitamin A to mouse tissues in vivo. J Nutr, 144, 608–613.

54.

, and Knight

. (2015). Dietary effects on human gut microbiome diversity. Br J Nutr 113, Suppl, S1–S5.

55.

Zhou

, and Zhi

. (2016). Lower level of Bacteroides in the gut microbiota is associated with inflammatory bowel disease: A meta-analysis. Biomed Res Int, 2016, 5828959.