Dear Editor
Compared with other immunocompromised populations (including renal transplant recipients), Nontyphoidal salmonella (NTS) bacteremia is relatively uncommon in allogeneic hematopoietic stem cell transplant (HSCT) recipients, with only a few reports in the literature. In addition to illustrating an unusual outcome for a rare infection in the posttransplant setting, the present letter provides relevant clinical reminders on the epidemiology, clinical course, and management of the condition. We also highlight the need to routinely evaluate for NTS infection in HSCT recipients who develop a febrile diarrheal illness, especially among those with chronic graft-versus-host disease (GVHD) and/or those who pursued oncologic treatment in densely populated, lower resource areas.
Case Report
The patient’s family provided both verbal and written informed consent for the use and disclosure of his de-identified clinical data in the present case report. A previously healthy 32-year-old male from Fortaleza, Brazil, was diagnosed with high-risk, 17p-deleted acute myelogenous leukemia. In early 2020, the patient underwent “7+3” remission induction (cytarabine/daunorubicin) followed by nonmyeloablative conditioning (busulfan/fluda-rabine) and haploidentical HSCT (performed in March 2020). Complete remission was achieved. Despite triple prophylaxis with posttransplant cyclophosphamide, cyclosporin A, and mycophe-nolate mofetil, he developed acute GVHD of the upper gastrointestinal tract at posttransplant day 30. Symptoms improved with prednisone and did not recur after steroid discontinuation. His posttransplant course was also complicated by cytomegalovirus viremia (treated with ganciclovir for 14 days) and Epstein-Barr virus reactivation (treated with rituximab for 4 weeks). By November 2020, all immunosuppressive agents had been discontinued (including GVHD prophylaxis, prednisone, and rituximab) with no new complications and with persistent complete remission. In May 2021, the patient arrived at the local bone marrow transplant (BMT) clinic reporting fever, severe abdominal pain, and watery diarrhea since the previous night. He was found to have a blood pressure of 60/30 mm Hg, temperature of 100.5 °F, and diffuse abdominal tenderness/guarding, which prompted hospital admission and sepsis protocol initiation. Laboratory studies suggested stage 2 acute kidney injury and moderate transaminitis, and computed tomography imaging showed mild ascites and distension of the small bowel. Initial workup was otherwise unremarkable. Despite intravenous meropenem/vancomycin and aggressive fluid resuscitation, the patient remained hypotensive and anuric, leading to vasopressor initiation and transfer to intensive care unit. Thereafter, on account of increased vasopressor requirements, severe metabolic acidosis, and worsening renal dysfunction, the patient was started on mechanical ventilation and hemodialysis support. Serum fungal markers, real-time polymerase chain reaction (PCR) for Epstein-Barr virus and cytomegalovirus, HIV screening, and hepatitis panel were negative. However, stool multiplex PCR panel was positive for NTS. Additionally, blood cultures collected on admission grew multisensitive NTS (unspecified serovar on final laboratory report). Transthoracic echocardiogram showed left ventricular ejection fraction reduction to 40% but no vegetations, thrombi, or wall motion abnormalities. Given his persistent fevers, continued hemodynamic instability, and progressive leukocytosis (>40?000 cells/?L), antimicrobial therapy was broadened by adding amikacin and polymyxin B. Despite intensive supportive measures, the patient died 2 weeks after presentation.
Discussion
Transmission of NTS (Gram-negative bacilli of the Enterobacteriaceae family) usually occurs via ingestion of contaminated food/water.1,2 A wide clinical spectrum can result, including asymptomatic carriage, self-limited gastroenteritis, focal suppurative/metastatic infection, and invasive bacteremic illness.2,3 For actively infected patients, NTS identification in the bacteriology laboratory is not difficult. Stool and blood cultures are often preferred, as PCR-based methods may have limited sensitivity.4 Although bacteremia occurs in <5% of patients with nontyphoidal salmonellosis, some populations (eg, malnourished children, individuals with sickle cell disease or recent malaria, and HIV-infected adults) have a particularly higher risk for invasive disease.4,5 Compared with healthy adults, renal transplant recipients taking immunosuppressants have a 20-fold greater risk of NTS infection.6 Conversely, HSCT recipients rarely develop NTS bacteremia, with only a few reports in the literature.3,7,8 In a prospective study to examine causes of acute diarrhea in HSCT recipients (n = 126), GVHD was more frequent than enteric infections (48% vs 13% of cases) and no patients had NTS isolated (Clostridioides difficile, astrovirus, and adenovirus were the main infectious etiologies).9 In a retrospective study to analyze bacterial foodborne illness incidence within the first year post-BMT, only 2 of 4069 patients (0.05%) developed an NTS infection, and neither progressed to invasive disease.10 In a 15-year retrospective cohort from a single center with nearly 500 HSCTs per year, NTS bacteremia occurred in 8 allogeneic and 4 autologous BMT patients, with only 1 death due to NTS-related sepsis, which was in an individual with advanced cirrhosis and extensive chronic GVHD.3 Although the above-cited studies were conducted in the United States, our patient was an inhabitant of Fortaleza, which is one of the most densely populated cities in northeastern Brazil. Data on NTS bacteremia remain scarce, with very few studies reporting incidence from Asia and no such studies from Latin America. The limited resources in those areas may also hinder the determination of specific serovars, as illustrated by this case. Nevertheless, disease burden is likely higher in low to middle income countries, as illustrated in (Figure 1).5 A prospective study following 65 individuals who underwent BMT in South India showed that, after 20 days of the procedure, diarrhea was more frequently related to enteric infections than GVHD (60% vs 44%).11 Meanwhile, a Chinese cohort study (n = 423) found a 2.13% incidence of NTS bacteremia in HSCT recipients, with full recovery after antibiotics in all cases and no fatal outcomes. Affected individuals were mainly men from 30 to 40 years old, and median infection onset was at posttransplant day 315 (range, 207-629 days); these findings matched our patient’s characteristics. Multivariate analysis supported nonmyeloablative conditioning (odds ratio, 4.604; P = .037) and extensive GVHD (odds ratio, 8.054; P = .003) as risk factors for NTS bacteremia.8 Ultimately, our BMT patient died from NTS-related sepsis, which is an unusual outcome for a rare post-HSCT complication. The severity of the infection was somewhat unexpected, especially considering the >1-year interval from his transplant date (a time at which T cells and memory B cells have typically recovered),12 lack of additional risk factors for NTS dissemination (eg, HIV infection, sickle cell disease, recent malaria, anatomic/functional asplenia, inherited immunodeficiency, current immunosup-pressant use, cirrhosis, or chronic GVHD),2-5 and timely initiation of antibiotic therapy. We hypothesize that delayed or abnormal reconstitution of innate/adaptive immunity could have contributed to his dismal clinical course. Despite the continued emergence of increasingly sophisticated strategies to improve posttransplant outcomes, immune reconstitution after haploidentical HSCT is a complex multipathway process, with great variations in the depth of immunosuppression experienced by individual patients.13,14 For instance, the source of stem cells (eg, marrow, peripheral blood, or cord blood) and the degree of donor-recipient HLA compatibility both affect the pace of immune reconstitution.12,14 Risk factors associated with late infections include persistently low B-cell counts, inverted CD4/CD8 ratio, and decreased immunog-lobulin synthesis.14 Although nontyphoidal salmonellosis rates are likely higher in developing areas (such as northe-astern Brazil, where our patient lived), recurrent outbreaks in wealthy countries highlight the risk that NTS poses to immunocompromised populations worldwide.3,5 For instance, 2447 NTS outbreaks (65?916 individual cases) occurred during the period from 1998 to 2015 in the United States, with a food vehicle being identified in 49% of these.15 Given current resistance trends, life-threatening NTS disease should be treated with both a third-generation cephalosporin and a fluoroquinolone until antimicrobial susceptibility is known.4,16 In the setting of worsening septic shock, the patient’s regimen was escalated to meropenem/vanco-mycin/amikacin/polymyxin B, which may not be as active intracellularly against NTS.16-18
Conclusions
Despite its rarity in the post-HSCT setting, NTS bacteremia should be considered as a differential diagnosis for BMT patients presenting with acute diarrhea and fever/sepsis. This is particularly relevant for individuals with chronic GVHD who have received nonmyeloablative conditioning and/or pursued oncologic treatment in densely populated, lower resource areas. In order to prevent NTS infection, progression to invasive disease, and sepsis-related death, clinicians should direct attention, respectively, to the following 3 factors: food safety practices in allogeneic HSCT recipients, prompt evaluation for NTS infection in those who develop a febrile diarrheal illness, and empiric antibiotic therapy that accounts for local resistance patterns while maintaining a high level of intracellular activity against the bacteria.
References:
DOI : 10.6002/ect.2022.0317
From the 1Department of Medicine, Icahn School of Medicine at Mount Sinai (Morningside/West), New York, New York; the 2Walter Cantídio University Hospital, Universidade Federal do Ceará; and the 3Department of Medicine, Centro Universitário Christus, Fortaleza, Brazil
Acknowledgements: The authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author: Bruno Almeida Costa, Department of Medicine, Icahn School of Medicine at Mount Sinai (Morningside/West), 1000 Tenth Avenue, New York, NY 10019, USA
Phone: +1 212 259 6777
E-mail: bruno.costa@mountsinai.org
Figure 1. Incidence Rates by Country of Nontyphoidal Salmonella Invasive Disease in 2017