Diets for patients undergoing haematopoietic stem cell transplant—a review

Diet recommendations for patients undergoing haematopoietic stem cell transplant (HSCT) have evolved over the years, but no standard of practice currently exists and there is broad variability between centres in terms of nutrition-related practices and guidelines on all levels—local, national, and international (Lipkin et al. 2005; Fox and Friefeld 2012; Vicenski et al. 2012; Lassiter and Schneider 2015; Baumgartner et al. 2017a, 2017b; Moody et al. 2018; Wolfe et al. 2018; Barnett 2019; Moody 2019; Taggart et al. 2019). Despite a lack of evidence linking intake of specific foods to incidence of infection, post-HSCT complications such as Graft vs. Host Disease (GvHD), overall survival rates, or indeed conferral of a protective effect in any way secondary to avoidance of particular foods, many centres continue to recommend restrictive diets for their HSCT patients. These recommendations are based largely on historical practice related to theoretical risk, adherence to outdated policies, and a lack of consensus across centres or in the literature (Tramsen et al. 2016; Baumgartner et al. 2017b, 2018; Barnett 2019; Moody 2019; Taggart et al. 2019).

Restrictive diets for HSCT patients appeared in the 1960s, in combination with other practices thought to be protective against infection. These practices included patient isolation, sterilization of linens, clothing, and equipment, and prophylactic antibiotic use to “decontaminate” and “cleanse” the gut in addition to a heavily restricted diet, meant to be followed for at least 6 months after allogeneic HSCT (Fox and Friefeld 2012; Lassiter and Schneider 2015; Moody et al. 2018; Wolfe et al. 2018; Moody 2019; Taggart et al. 2019). The diet has been called many things over the years, primarily the “neutropenic diet” (ND), used for both oncology and HSCT patients, and more recently the “low bacteria diet” (LBD). The rationale for the ND/LBD diet came from a belief that bacteria commonly found in food and introduced into a gut weakened by harsh conditioning regimens and post-HSCT side effects, such as mucositis, would increase the risk and incidence of bloodstream infections as a result of translocation from the gut into the bloodstream (Fox and Friefeld 2012; Lassiter and Schneider 2015; Wolfe et al. 2018; Taggart et al. 2019). Particular emphasis was placed on bacteria commonly found on raw fruit and vegetables, including enterobacter and pseudomonas bacteria, as these were considered to be particularly dangerous pathogens in a significantly immunocompromised patient (Wolfe et al. 2018). As a result, fresh fruits and vegetables in addition to other foods considered high risk (such as raw nuts, cream (fresh or shelf stable) and custards, yoghurt with live cultures, some pickled foods, undercooked meat and eggs, and unpasteurized dairy) were restricted. The restrictions imposed by the ND/LBD led to diets high in processed foods, and limited in sources of fibre and vitamin C due to the complete removal of fresh fruits and vegetables (Wolfe et al. 2018; Taggart et al. 2019), potentially predisposing post-HSCT patients to other diet-related complications including vitamin and mineral deficiencies, hypertension due to increased sodium intake (from processed foods), and decreased overall bowel health due to minimal fibre intake (Taggart et al. 2019). Despite literature showing that incidence of food-borne illness post-HSCT is extremely rare (an 0.3% incidence over an 11 year period, as shown by Boyle et al. (2014)), and a better understanding of the function of the gut, variations of the ND/LBD persist, and are still widely recommended across HSCT centres, even as other total-protective-environment practices such as sterilization of equipment and linen have been abandoned (Lipkin et al. 2005; Fox and Friefeld 2012; Baumgartner et al. 2017b; Wolfe et al. 2018).

With an increasing body of evidence supporting use of a more liberal diet, many North American centres are moving towards a less restricted, food-safety based (FSB) approach, backed further by guidelines from food and health agencies including the CDC, USDA, FDA, and Health Canada (Tramsen et al. 2016; Baumgartner et al. 2017a; Barnett 2019). Most paediatric centres in Canada now follow the FSB approach, which patients and families find far more palatable than the restrictive diets of the past, and as a result, much easier to adhere to (Neumann et al. 2016; Tramsen et al. 2016; Baumgartner et al. 2018; Barnett 2019). At present, SickKids Hospital in Toronto, Ontario, follows the FSB approach, recommending a significantly liberalized diet compared to past practice (Stuhler 2017, 2019). Out of familiarity, the diet is still called the “low bacteria diet” (Table 1, 2) but is entirely food-safety based, and adapted from Health Canada’s diet recommendations for immunocompromised individuals (Health Canada 2012, 2015). The food-safety approach followed by SickKids and many other centres focuses on safe food handling and preparation (Table 3), and avoidance of foods that are known to have the potential to cause food-borne illness, including undercooked eggs, meat, poultry, and fish, and unpasteurized dairy products. Fresh fruits and vegetables, heavily restricted in the past, are allowed, enhancing the overall nutritional composition of the diet with the reintroduction of vitamins, minerals, and fibre. Yoghurt and kefir are allowed, as are pickled and some fermented products such as kimchi and sauerkraut. Food from restaurants, if safely prepared, is permitted. Municipal tap water is allowed, and sterile or boiled water not recommended except in the case of untested well water supply or for preparation of infant or therapeutic enteral formulas (Health Canada 2015; Stuhler 2017, 2019). Palatability and access to food (particularly for families far from home and without good kitchen facilities) is increased with the FSB approach. There may also be institutional cost-saving advantages to adoption of the FSB diet, as less work, staff, and special products are required at the hospital-level to support the diet (Moody et al. 2018). Despite this trend, enormous amounts of variability still exists in terms of centre-to-centre practices around diet recommendations. In a survey of 17 Swiss centres performing HSCTs, Baumgartner et al. (2017b) found no consensus in terms of existence of an HSCT-specific diet, particularly for autologous-HSCT patients (as some centres did not require any type of special diet for this group), recommended length of time a diet should be followed if it was in place, and which foods were allowed while following the diet. While there were similarities noted in terms of food preparation and recommended avoidance of undercooked meat/fish/eggs and unpasteurized dairy, some centres continued to advise against eating raw fruits and vegetables while others allowed it. Vicenski et al. (2012) describe similar inconsistencies across 17 centres in Brazil, and other studies report the same variation in practice in other parts of the world (Lipkin et al. 2005; Fox and Friefeld 2012; Baumgartner et al. 2017a; Wolfe et al. 2018; Barnett 2019; Moody 2019). A body of literature is emerging showing minimal benefit of the ND/LBD approach in comparison to a FSB diet, which may help standardize approaches in the future.

There is still so little known about the role nutrition plays in those undergoing HSCT. However, based on the findings of the groups above, choosing a food safety based diet for HSCT patients is not only safe, but superior given the positive impact the diet can have on quality of life and overall nutritional status. Discarding the old neutropenic or low bacteria diets in favour of the FSB diet is a safe and responsible choice for the future. Ensuring that all staff involved in provision of care for HSCT patients are educated about the diet, and resources available to patients and families are easily accessible and up-to-date is key.