During the period of neutropenia after stem cell transplantation (SCT), infections are a major cause of morbidity and mortality.¹ Restrictive diet, a standard of care of preventative measures for infections, has been adopted in 93% of bone marrow transplantation (BMT) centers, but its efficacy has never been examined prospectively.¹ In the 64^th American Society of Hematology (ASH) Annual Meeting and Exposition 2022, Dr. Federico Stella from the University of Verona, Italy, shared his team’s observation on the incidence of infection in a neutropenic diet, compared with a regular one and the severity of graft-versus-host disease (GVHD) due to the period of low oral feeding after SCT.¹ Hence, it is hypothesized that a restrictive and less palatable diet might contribute to low oral feeding and a greater risk of GVHD.¹

NEUTRODIET was a multicenter, randomized, prospective non-inferiority phase 3 trial that evaluated the risk of infections in patients with a non-restrictive diet (NRD), compared with a protective diet (PD) following autologous SCT (auto-SCT) or allogeneic-SCT.¹ The NEUTRODIET trial included 247 patients undergoing allogeneic- or auto-SCT and high-dose chemotherapy with >7 days of neutropenia expected.¹ In contrast, patients with active infections before neutropenia, inability to undergo oral feeding, receiving SCT for non-oncohematological diseases, undergoing second allogeneic-SCT, and transplant from umbilical cord blood or haploidentical donor were excluded.¹

Participants were randomized 1:1 to receive PD (n=111) or NRD (n=111), which consisted of foods cooked >80°C and/or thick peel fruit, along with a list of prohibited foods, including yogurt, honey, cold cuts, raw fish and meat for the former; and a diet according to hospital hygiene standards excluding raw fish and meat for the latter.¹ The primary endpoint of the study was to highlight an absence of significant differences in infections (G≥2) and mortality during the neutropenia period in both arms, while the secondary endpoints included assessment of gastrointestinal (GI) infections and fever of undetermined origin (FUO), overall survival (OS) predicted at 30 days and cumulative incidence of acute GVHD (aGVHD), as well as nutritional status considering body weight change and hospital stay duration.¹  Patient demographics were balanced, including sex (p=0.7), age (p=0.1), disease type (p≥0.5), type of transplantation (p>0.9), previous lines of therapy ≥2 (p=0.9), overall disease status at enrolment (p≥0.1), antimicrobial prophylaxis (p≥0.06), and neutropenia duration (p=0.4).¹

There was no difference in infections, including severe infection, FUO, sepsis, pneumonia, and GI infection in the 2 arms.¹ Infection (G≥2) were 72 (65%) in the PD arm and 69 (62%) in the NRD arm (RR=1.0; 95% CI: 0.8-1.3; p=0.8).¹ Although there was no difference in infection with abdominal recall, mucositis, and hospital stay duration, a greater GI infection with microbiological isolation was associated with the PD arm compared with the NRD arm (9% vs. 2.7%; RR=3.3; 95% CI: 0.1-11.0; p=0.08).¹ Concerning the microbiological isolation, Enterobacteriaceae and Clostridium difficile (C. diff.) were most commonly isolated in both arms at the blood and stool cultures.¹ Prophylactic fluoroquinolones decreased the bloodstream infection rate (3.4% vs. 15.4%; RR=0.2; 95% CI: 0.01-0.6; p=0.002), but not severe infection (G≥3) (48% vs. 45%; RR=1.1; 95% CI: 0.8-1.4; p=0.7).¹ With respect to feeding outcome, a lower body weight loss at 1 month was observed in the NRD arm (mean -2.7 vs. -3.7kg; p=0.04), while there was no difference in the use and duration of parenteral nutrition, serum albumin, and body mass index (BMI).¹  In addition, NRD was associated with greater satisfaction (16% vs. 35%; RR=0.5; 95% CI: 0.3-0.8; p=0.006), with reports of “diet prescriptions don’t negatively impact alimentation”.¹ Focusing on the allogeneic recipients, both arms were well balanced, in terms of disease, donor type, human leukocyte antigen (HLA) matching, conditioning regimens, GVHD prophylaxis, and hematopoietic cell transplantation-comorbidity index (HCT-CI).¹ The incidence of aGVHD grade ≥3 was also comparable between both arms (20% in PD vs. 9.5% in NRD; RR=2.1; 95% CI=0.5-9.1; p=0.4).¹

To conclude, the use of restrictive diet is a non-essential burden for the quality of life of patients.¹ This was demonstrated by the lack of difference in the rates of infections, death, nutritional outcomes, and aGVHD incidence between patients on PD and NRD during neutropenia.¹