The chemo-gut study: investigating the long-term effects of chemotherapy on gut microbiota, metabolic, immune, psychological and cognitive parameters in young adult Cancer survivors; study protocol

The incidence of cancer among young adults has dramatically increased in recent decades [10], with over 8000 young Canadians diagnosed annually [12]. Between 1992 and 2013 the incidence of cancer has increased among young Canadian’s aged 15 to 29 by 18.2, and 11.9% among 30 to 39 year olds [12]. Considering their age at diagnosis, young adults experience unique personal, physical, and socioeconomic challenges throughout their cancer experience [12, 13]. As most young adult survivors are exposed to toxic therapies (e.g. chemotherapy) many will face years of chronicity living with treatment-related physical and psychosocial side-effects. For instance, Asher [14] reports that executive functioning, attention, concentration, processing speed, reaction time, motor speed and dexterity were the cognitive domains most frequently associated with chemotherapy induced cognitive impairment in all cancer populations. Additionally, Seitz et al. [15] examined the prevalence of posttraumatic stress, depression and anxiety among young adult survivors (n = 820) of adolescent cancer with a mean current age of 30.4 years [15]. They found that 90.5% cancer survivors had received chemotherapy, and 22.4% reported clinical symptoms of posttraumatic stress, anxiety and/or depression compared to only 14.0% of controls [15]. Posttraumatic stress symptoms in both male and female survivors were found to be more than three times that of controls [15]. Furthermore, female survivors reported symptoms of depression and anxiety significantly more often than controls, and 24.3% of the survivors met DSM-IV criteria for at least one mental disorder, relative to only 15.3% of the controls [15]. Indeed, the increased prevalence of young adult cancers in conjunction with the improved survival rates has resulted in a large group of survivors facing years of chronicity, unique challenges, and specific needs emerging from their cancer experience.

While the detrimental effects of chemotherapy on psychosocial and cognitive function are established, evidence for the underlying physiological mechanisms and system interactions that drive the psychological and cognitive impairments remains limited. The gut-brain axis refers to a bidirectional communication system within an organism that enables gut microbes to communicate with the brain and vice versa via neural, endocrine, immune and metabolic pathways [4, 16]. The role of the gut-brain axis in the context of the young adult is a critical consideration, since environmental toxins can interfere with normal development, potentially increasing individuals’ risk for developing psychiatric disorders such as depression and anxiety, which typically develop during adolescence and young adulthood.

Chemotherapy has been shown to adversely affect the gut microbiota in both pediatric and adult cancer cohorts. For example, Huang et al. [17] investigated changes in gut microbiota among 36 children with acute lymphoblastic leukemia (ALL) before and after receiving high-dose methotrexate chemotherapy, finding that the total gut microbial content in fecal samples decreased by 29.6%, relative to healthy controls, with significant reductions in Bifidobacteria, Lactobacillus and E. coli [17]. Furthermore, using a pre-post-measure design, Montassier et al. [18] examined 28 adult non-Hodgkin’s lymphoma patients treated with chemotherapy to assess the functional mechanisms of the gut microbiota in mediating the pathophysiology of mucositis. Mucositis is a condition characterized by painful inflammation and ulceration of the mucous membranes lining the digestive tract, affecting 70% of patients following chemotherapy [18, 19]. Significant decreases in Firmicutes and Actinobacteria, and increases in the abundance of Proteobacteria were observed alongside a reduced capacity for nucleotide, energy and vitamin metabolism, suggesting that chemotherapy induces severe compositional and functional imbalance in the gut microbiome that is associated with gastrointestinal mucositis [18]. While recent research demonstrates that chemotherapy is acutely associated with dysbiosis, research has not yet examined whether these effects persist in the longer-term (i.e. over months) following the cessation of chemotherapy. Therefore, as part of this study we will conduct a novel examination of the longer-term effects of chemotherapy on gut microbiota up to 6-months post-chemotherapy.

The gut-brain axis exerts regulatory effects on immune function and behaviour. Indeed, previous research has found that under certain conditions the protective epithelial layer in the gut can become compromised leading to increased intestinal permeability and consequently increased systemic circulation of the bacterial endotoxin, lipopolysaccharide (LPS), a component of the cell wall of Gram-negative bacteria [11, 20]. This “leaky gut” results in a peripheral inflammatory immune response, such that systemic levels of pro-inflammatory cytokines are increased [20], inducing sickness behaviours [21].

The potential for gut microbiota to affect the immune system and inflammatory mechanisms is a critical consideration in the context of anxiety and depression, as inflammation has been implicated as a driving factor in depressive illness [3, 22]. Behaviours indicative of sickness include lethargy, depression, anxiety, anorexia, social withdrawal and isolation, and have consistently been associated with augmented levels of pro-inflammatory cytokines, especially tumour necrosis factor alpha (TNF-α), interleukin 6 (IL-6), and interleukin 1 beta (IL-1ß) [21]. Additionally, studies involving Interferon alpha (INF-α) treatment for melanoma patients often report the development of cytokine induced depressive symptoms and sickness behaviours following INF-α treatment [23].

The HPA-axis may also play a role in mediating gut-brain axis functions. Previous studies have shown that gut bacteria can activate stress circuits via the vagus nerve [2], which modulates gut-brain axis function, subsequently affecting behaviour [5]. Corticotrophin releasing factor (CRF), a stress-related neuropeptide, is highly associated with gut-brain axis functioning [24]. Furthermore, results from human and animal studies have found that increased systemic inflammation (e.g. augmented pro-inflammatory cytokines from endotoxemia) is associated with increased levels of adrenocorticotropic hormone (ACTH) and cortisol, further implicating the role of immune function in moderating HPA-axis function [25]. As evidence continues to accumulate demonstrating that the gut-brain axis contributes to physiological function and affective behaviour related to the HPA-axis and immune system, it is important to consider the potential bi-directional communication between these systems, particularly in the context of disease.

Chemotherapy has been associated with increased intestinal wall permeability, mucositis, and the onset of a systemic inflammatory immune response, dysregulation of the HPA-axis, and subsequent manifestation of sickness behaviours [11, 20, 21]. Gut microbiota may play a role in mediating such effects [7]. Moreover, in addition to adversely affecting the gut microbiota, chemotherapy-induced mucositis is also associated with significant elevations in systemic pro-inflammatory cytokines, especially IL-1ß and IL-6 [26], which are accompanied by increased symptoms of pain, anxiety, and depression [27]. As such, the potential disruption of normal physiological and psychological development among young cancer survivors and the role of gut microbiota in moderating such changes is of particular concern, as many survivors face lifelong health challenges following a cancer experience, which is why these immune and metabolic components have been included within our study.

Gut microbiota dysbiosis has been implicated in cognitive and psychological deficits. Previous studies demonstrate that alterations in gut microbiota can affect both anxiety and depressive behaviour in animals and humans [2, 28‐31]. Neurotransmitter dysregulation is widely implicated in depressive illness, and certain neurotransmitters including serotonin (5-HT), dopamine, GABA (gamma-Aminobutyric acid), and noradrenalin, are affected by the gut microbiota [4, 32]. This is an important consideration given that neurotransmitters, especially serotonin as it participates in mood regulation, are implicated in the neurological dysregulation observed in depressed patients [33] and cancer survivors frequently experience comorbid depression following a cancer experience [15]. Furthermore, a small but growing body of pre-clinical evidence suggests that the gut-brain axis also plays a role in cognitive function, including spatial learning and working memory [34], and alterations in brain derived neurotrophic factor (BDNF) expression in regions of the brain that are associated with specific cognitive functions (e.g. the hippocampus, which is involved in learning and memory) [34, 35].

While the majority of evidence stems from animal research, some human studies have also shown associations between the gut microbiota, and psychological and cognitive function. For instance, Prehn-Kristensen et al. [36] investigated the gut microbiota composition of 14 male attention deficit hyperactivity disorder (ADHD) patients compared with healthy male controls, and found that alpha diversity of ADHD patients was significantly decreased, and augmented levels of Bacteroidaceae, and both Neisseriaceae and Neisseria spec. Were identified as potential biomarkers for the ontogeny of juvenile ADHD [36]. Additionally, Naseribafrouei et al. [37] examined associations between human fecal microbiota and depression in 55 adults (37 patients versus 18 controls), finding that Bacteroidales were overrepresented, while Lachnospiraceae species were underrepresented among depressed patients. Additionally, the Oscillibacter genus, of which valeric acid is the main metabolic end product and a homolog of the neurotransmitter GABA, a neurotransmitter implicated in major depressive disorder, was also correlated with depression [37].

Furthermore, in a triple-blind, placebo-controlled, randomized, pre- and post-intervention study with 40 healthy participants, Steenbergen et al. [29] investigated the effects of a multispecies probiotic versus a placebo control on cognitive reactivity to sad mood. Participants who received the 4-week probiotic intervention showed a significant attenuation in overall cognitive reactivity to sad mood, which appeared to be a function of reduced rumination and aggressive thoughts [29]. Finally, Hilimire et al. [28] conducted a cross-sectional study with 710 young adults to examine whether consumption of fermented foods, which contain probiotics, interacts with neuroticism to predict social anxiety symptoms, using self-reported measures of fermented food consumption, neuroticism, and social anxiety [28]. For individuals reporting high traits of neuroticism, greater consumption of fermented food was associated with fewer symptoms of social anxiety, suggesting that fermented foods containing beneficial probiotic bacteria may exert protective effects against social anxiety symptoms for young adults with higher trait neuroticism [28]. Although the underlying mechanisms mediating these cognitive-behavioural effects remain to be elucidated, it is plausible that alterations in gut microbiota resulting from dietary supplementation with health promoting bacteria play a role.

Chemotherapy has been associated with cognitive and psychological impairments [14, 15]. For example, Prasad et al. [38] examined associations between neurocognitive dysfunction and psychosocial outcomes on the achievement of social milestones among adult survivors of adolescent and early young adult (AeYA) cancer who had been diagnosed between 11 to 21 years of age. Relative to healthy siblings, AeYA survivors’ were more likely to experience neurocognitive dysfunction in domains related to task efficiency, emotion regulation, and memory, were less likely to be married or employed, and reported greater levels of emotional distress characterized by anxiety, somatization, and depression [38]. Given this, the potential role of gut microbiota in mediating or moderating chemotherapy’s effects is an important direction for novel investigation. Such knowledge will help us to better understand the ontogeny of cognitive and psychological deficits in young cancer survivors, while also presenting new opportunities for targeted interventions perhaps via prebiotic and/or probiotic supplementation of key health promoting bacteria that may be depleted following chemotherapy.

Findings regarding associations between gut microbiota and obesity also hold implications related to body composition among young adult cancer survivors who tend to be more overweight and obese relative to healthy peers. For example, recent data from a sample of 49 young adult survivors of adolescent cancer found that over 57% of these survivors were overweight or obese, which is higher than the national average [39], and consistent with previous research on obesity in adolescent and young adult (AYA) cancer survivors [40]. In human and animal studies, diet and/or antibiotic induced gut microbiota dysbiosis has been shown to play a strong role in the development of obesity [41‐44]. Given the established link between obesity and gut microbiota dysbiosis [45], it is possible that chemotherapy may adversely affect metabolic processes mediated by the gut microbiota, resulting in an increased risk for young adult cancer survivors to become overweight or obese. However, additional research is needed to examine the potential relationship between chemotherapy, gut microbiota, and fat mass among young cancer survivors; a rationale for our inclusion of these elements in our study.

Clinical and demographic factors including current age, age at diagnosis, income, occupation, ethnicity, education, living arrangements, cancer diagnosis and treatments received, history of treatment related mucositis, current diet and exercise regimes, bowel habits, glucocorticoid medication, antibiotic, and probiotic use within the last 2 years, smoking, birth delivery mode (i.e. caesarian section vs. vaginal), and breastfeeding during infancy will be collected as these variables are known to impact the gut microbiota.

Investigators will provide participants with a stool collection kit and instructions for collecting samples. Stool samples will be collected into a sterile conical tube and placed in a biohazard bag that will be stored in the participant’s freezer until the date of their appointment (up to 3 days). Samples will be brought to the University of Calgary on ice and stored at − 80 °C until processed. Microbial profiling will be carried out according to our previously published protocol [42]. Bacterial DNA will be extracted from stool samples using a FastDNA spin kit for feces (MP Biomedicals, Santa Anna, CA, USA) and the V3-V4 region sequenced by the Centre for Health Genomics and Informatics (University of Calgary) using the Illumina 16S rRNA sequencing platform. Sequencing data from MiSeq will be demultiplexed and converted to fastq format using Illumina’s bcl2fastq software. Primers will be removed from reads using CutAdapt and initial quality trimming performed. Following primer removal, length sequences of shorter than 10 base pairs will be removed, and the quality trimming threshold set at Q20. R package DADA2 will be used for sequence/OTU clustering. Chimeric sequences will then be removed, and taxonomic assignment performed using the RDP classifier and RDP database as reference. Further downstream analysis will be performed using Phyloseq R package. The Shannon and Simpson index will be used to measure alpha diversity. Non-metric multidimensional scaling (NMDS) on Bray-Curtis dissimilarity matrix will be used to evaluate beta diversity. Differential abundance analysis between groups will be carried out using the LEfSe algorithm [47], with an alpha = 0.05 to determine significance of differentially abundant features.

Body composition will be measured using dual-energy-x-ray absorptiometry (DXA) [48]. Specifically, fat mass, lean mass, body fat percentage, and bone mineral concentration will be quantified using whole-body dual-energy x-ray absorptiometry (DXA; Hologic QDR 4500; Hologic, Inc., Bedford, MA), which is a valid and reliable measure of body composition in healthy adolescents and adults, as well as cancer patients [48, 49].

Pro-inflammatory cytokines interleukin 6 (IL-6), interleukin 1 beta (IL-1ß), and tumour necrosis factor alpha (TNF-α), anti-inflammatory cytokine interleukin 10 (IL-10), and C-reactive protein (CRP) will be measured in serum using MesoScale multiplexing [50], in consideration of the robust plate to plate reproducibility over time and the dynamic range with a lower limit of detection in the sub picograms per milliliter (i.e. pg/mL) range.

BDNF will be assayed from blood serum using MesoScale multiplexing [50], in consideration of the robust plate to plate reproducibility over time and the dynamic range with a lower limit of detection in the sub picograms per milliliter (i.e. pg/mL) range.

Serotonin will be measured using Serotonin ELISA (enzyme-linked immunosorbent assay) kits from Enzo Life Sciences [51] according to manufactures instructions. The ELISA is a plate-based assay technique designed for detecting and quantifying substances, including neurotransmitters and proteins.

Lipopolysacchide (LPS), a gram-negative bacterial endotoxin associated with intestinal permeability, systemic inflammation, and sickness behaviours, will be measured from serum using MesoScale multiplexing [50].

Cortisol will be analyzed using hair samples, which provide a reliable and valid measure of long-term cortisol activity [52, 53]. Hair cortisol will be quantified using liquid chromatography tandem mass spectrometry, and is reported to be more representative of total cortisol production relative to single saliva or serum measures, allowing us to quantify cortisol production over an extended period of time.

Post-Traumatic Stress symptoms will be measured using the Impact of Events Scale. Depression, anxiety, pain behaviour, fatigue, cognitive function, and social isolation will be measured via the NIH Patient-Reported Outcomes Measurement Information System (PROMIS). PROMIS is a set of person-centered measures that evaluates physical, mental, and social health, and is a valid measure of psychosocial outcomes among young adult cancer patients [54]. Details of these measures can be found in Table 1. An objective measure of executive function will also be collected using the Sustained Attention to Response Task (SART) is a computer-based go/no-go task that is designed to measure working memory, sustained attention, and impulse/inhibitory control [55], cognitive deficits that are frequently experienced by survivors who have received chemotherapy. The SART evaluates participants’ ability to withhold behavioral response to a single, infrequent target presented against a background of frequent non-targets [55]. Participants are asked to respond by pressing a button or the spacebar key after seeing the non-target on the screen and to inhibit their response to seeing the target. This task requires participants to be attentive to their responses, such that, at the appearance of a target, they can override the dominant motor response and substitute the directly antagonistic response (i.e. withhold button press). This study will use the online version of the SART, delivered through the Inquisit web software provider Millisecond.​com. In this version of the task, a practice block (~ 30 s) is followed by two blocks (~ 3 min each), totaling approximately 6 min to complete the SART task.

Once identified, participants will be contacted by telephone or email to further verify inclusion and exclusion criteria, to obtain written informed consent, and schedule their first appointment. We will aim to assess participants at 3 different time points: < 2 months (T1), 3/4 (T2) and 5/6 (T3) months post chemotherapy, with healthy controls ideally being tested on the same day or within 5 days from when the patient is tested.

Participants will then receive a package via mail or in person, with instructions and materials for stool sample collection, and a link to fill out the self-report questionnaires, and demographic and clinical data online using REDCap. REDCap is a free, secure, browser-based application designed to support Electronic Data Capture for research studies provided through the Clinical Research Unit in the Cumming School of Medicine at the University of Calgary. Medical charts will also be accessed to confirm information regarding cancer diagnosis, treatment, glucocorticoid and antibiotic medication use, and mucositis.

Participants will be instructed to collect stool samples at home which can be stored in a freezer up to 3 days prior to their appointment, and to complete their online questionnaires prior to the appointment. On the day of their appointment, we will collect their stool sample from them. Participants will then be taken to have a hair sample collected for cortisol analysis (approximately 10 min), complete the cognitive task (i.e. the SART) on a computer (approximately 6 min), and complete their DXA scan (approximately 10 min), and have their blood drawn for serum biomarker analyses (approximately 10 min). This whole process should take approximately 45 min. Participants’ T2 and T3 month follow-up appointments will be scheduled at each subsequent visit and the procedure will be the same with respect to psychological and cognitive measures completed online, followed by their in person testing and biological sample collection. Participants will be provided a voucher for parking, and may choose from compensation in the form of either $30 at the end of the study or a pass to a fitness class after each appointment (they will have options to choose from in advance). If desired, participants will also receive a package with their individual results summarized and interpreted at the end of the study.