Association of Clinical and Para-clinical Factors With the Occurrence of Febrile Neutropenia Following Autologous Bone Marrow Transplantation in Patients With Lymphoma

Purpose: Promising results have been achieved by the administration of autologous bone marrow transplantation (BMT) in patients with lymphoma. However, infectious complications limit its positive outcomes. Therefore, better identication of related factors to the occurrence of febrile neutropenia may lead to improved management of these patients. This study aimed to evaluate the incidence and associated factors with the occurrence of febrile neutropenia in following BMT in patients with lymphoma Method: In this prospective study, 45 patients with a denite diagnosis of lymphoma who were candidates for BMT were consecutively included. Demographic characteristics, lymphoma type, and medical history, as well as the results of selected laboratory tests, were recorded. After BMT, neutrophil count and body temperature were evaluated. Results: Febrile neutropenia occurred in 28 patients (62.5 %). No statistically signicant difference was seen between those with febrile neutropenia and those without, in terms of age, gender, BMI, type of lymphoma, blood group, and the number of radiotherapy or chemotherapy courses (p>0.05 for all). Multivariable logistic regression analysis demonstrated a signicant positive relationship between the number of administered units of platelet and serum uric acid and development of febrile neutropenia. Conclusion: The results of our study showed that the incidence of febrile neutropenia is signicantly high in lymphoma patients who receive BMT. Moreover, the number of administered units of platelet and uric acid before BMT are signicantly associated with febrile neutropenia.


Introduction:
Lymphomas are classi ed into two major categories of non-Hodgkin (90%, NH) and Hodgkin (10%, HL) lymphomas [1]. The global incidence of HL has increased by 38.66% -from 72,937 in 1990 to 101,133 in 2017 -however, due to improved treatment modalities, the age-standardized death rate and the annual age-standardized disability-adjusted life year (DALY) have decreased during these years [2,3]. Bone marrow transplantation (BMT) or autologous hematopoietic stem cell transplantation (AHSCT) after high-dose chemotherapy are established as the standard treatment modality for refractory or relapsed lymphomas [4]. Patients with high-risk aggressive lymphoma who are unresponsive to the conventional regimens alone show better prognosis by receiving early consolidative transplantation. Moreover, it can signi cantly diminish the bone marrow tumor load and consequently bring about a higher rate of complete remission [4]. Despite substantial positive advances by these therapeutic modalities, transplant-related infectious complications are still causing considerable morbidity and mortality in lymphoma patients [5]. These complications occur in a varying range of 12 up to 51 percent of transplanted patients [6,7]. Higher incidences are documented in the neutropenic phase or pre-engraftment [8,9]. Moreover, the incidence of severe sepsis in AHSCT patients was reported to be ve times higher than non-AHSCT cohort, and it was associated with a mortality rate of 32.9 percent in AHSCT patients [10]. The development of fever in BMT patients is characterized as the clinical hallmark of bacteremia which is usually associated with neutropenia (febrile neutropenia) [11].
Several associated factors have been suggested for the higher incidence of infectious complications, including > 18 years age, use of unrelated graft source and myeloablative conditioning regimen, transplant-associated thrombotic microangiopathy, acute graft versus host disease, high-risk malignant disease, mucositis, and steroid administration [12,13].
However, unsuccessful results of attempts for preventing these complications necessities further investigations for other previously underappreciated associated factors. Furthermore, detecting these factors could help limit the use of antibacterial prophylaxis to high-risk patients to prevent multi-drug resistant microorganisms [14,15]. In this study, the incidence and associated factors of occurrence of febrile neutropenia were evaluated in patients with lymphoma who receive BMT after high dose chemotherapy.

Result
An overall 45 lymphoma patients were included. The baseline characteristics of the patients are described in Table 1

Study design and patients' inclusion
In this prospective study, 45 adult patients (aged between 18 and 65 years) with a de nite diagnosis of lymphoma (HL or NHL) who were candidates for BMT due to relapse following standard chemotherapy and admitted in the Ghazi Hospital in Tabriz (the only tertiary referral hospital for hemato-oncological diseases in north-west Iran) were consecutively included between April 1, 2016, andMarch 31, 2018. Patients were excluded if they had co-morbidities and those who heavily treated (received more than three chemotherapy protocols). The study was conducted in accordance with the declaration of Helsinki, and informed consent was obtained from each patient.

Medical procedure
All patients received standard medical care during hospitalization. Patients were isolated in positive pressure reverse isolation rooms and were indicated to avoid raw fruits and vegetables. No medication was administered by rectal, vaginal or intramuscular routes. Strict handwashing procedures using the disinfectant agents were employed by the hospital staff before entering patients' rooms. The procedure of BMT and standard medical care were performed following established guidelines [16]. After admission, patients received conditioning regimen including lomustine (CCNU) 200 mg/m 2 (one dose), cytarabine 400 mg/m 2 (two times daily for two days), VP16 400 mg/m 2 (two times daily) and melphalan 140 mg/m 2 (Once). Later, stem cell collection was conducted using apheresis. The following day, BMT was performed using intravenous infusion of the obtained stem cells. Moreover, all patients received prophylaxis regimen since the day of admission and also after BMT, including acyclovir 800 mg (two times daily), uconazole 150 mg (two times daily), and cipro oxacin 500 mg (two times daily during the neutropenic days). They also received trimethoprim / Sulfamethoxazole 800 mg/160 mg (two times daily for two days per week) since the day of BMT until three months later [16].

Study variables
A series of clinical and paraclinical variables were selected based on our clinical experiences and a preliminary literature review. Demographic characteristics, lymphoma type, and medical history of included patients were recorded in a preprepared questionnaire. Laboratory blood tests and liquid balance measurement (Intake -Output) were conducted 10 and 5 days before transplantation and at the transplantation day. After transplantation, the neutrophil count was measured daily, and body temperature was evaluated every four hours. A neutrophil count less than 1.5 × 10 9 /L was considered as neutropenia [17]. Body temperature (BT) was measured at 8-9 AM each day after transplantation, and the patient was considered febrile if they had BT > 38.3° C or BT > 38 if lasts more than one hour. Febrile neutropenia was de ned as a simultaneous presence of fever and neutropenia, as elucidated above.

Statistical analysis
Data are presented as mean ± standard deviation (SD) or frequency and percentage. The normal distribution of the data was evaluated by the Kolmogorov-Smirnov test. Chi-square or Fischer's exact test was conducted on categorical data and student T-test or Mann-Whitney U test for comparison between groups in parametric and non-parametric data, respectively. Simple and multiple logistic regression analyses were conducted in the context of univariate and multivariate analyses, and unadjusted and adjusted odds ratios (with 95% con dence intervals [CI]) were calculated. The analysis was performed by SPSS v 24. A p-value < 0.05 was considered as statistically signi cant.
in control group, p = 0.021). Also, higher number of units of platelet had been administered in febrile neutropenia group (median [min-max] = 18.5  in febrile neutropenia group vs 12  in control group, p = 0.043).

Discussion
There is a growing concern for the widespread emergence of multidrug-resistant microorganisms in medical centers, especially in hematology-oncology wards [18,19]. These microorganisms diminish the effectiveness of routine antibacterial prophylaxis with oral uoroquinolones in patients undergoing BMT or AHSCT [20,21,18]. In one study, discontinuing the administration of antibacterial prophylaxis in AHSCT recipients did not have a signi cant adverse effect on the early mortality of the patients after transplantation [22]. Therefore, it is suggested that bacterial prophylaxis be considered for only high-risk patients [14]. However, no risk strati cation system has been established so far for such cases. One of the reasons can be that the related and predictive factors are not fully recognized yet. Considering the close association of febrile neutropenia with infection in lymphoma patients receiving BMT [11], in this prospective study, we evaluated the incidence and associated clinical and paraclinical factors of post-transplantation development of febrile neutropenia. A considerably high incidence of febrile neutropenia occurred in these patients in our study, which was similar to the reports of previous studies [23][24][25]. No signi cant differences were detected in baseline characteristics between those patients who developed febrile neutropenia and those who did not (control); however, some of the other related factors were signi cantly different between groups including the number of administered packed cells, erythrocyte sedimentation rate (ESR) at the − 5 and − 10 days, uric acid at all three days of 0, -5, and − 10, and hemoglobin at -10 day. We also found a signi cantly lower liquid balance at 10 days before BMT in patients with febrile neutropenia which highlights the importance of adequate hydration of the patients over the days before BMT. Our further analysis by controlling for baseline characteristics revealed that only uric acid and the number of administered units of platelet were signi cant predictors of febrile neutropenia. The prognostic value of serum uric acid for both short and long term outcome of admitted medical patients has been emphasized in previous studies [26]. However, in contrast with our results, serum uric acid is demonstrated to have a positive correlation with poor outcome and mortality [27][28][29]. The lower uric acid level in those patients who would develop febrile neutropenia may be related to malnutrition which leads to lower production of uric acid. Also, the presence of a higher background in ammatory state (before BMT) may lead to impaired renal uric acid reabsorption [30]. Accordingly, in a study on 43 Taiwanese patients with SARS, 16 patients develop marked renal hypouricemia due to in ammation [31]. Overhydration could be another reason for lower uric acid levels in patients who develop febrile neutropenia. However, this hypothesis contrast with our ndings of the negative correlation of liquid balance at -10 day with the development of febrile neutropenia. Furthermore, low Na intake may result in hypouricemia in these patients who are at higher risk of febrile neutropenia [30].
In a similar attempt to identify the related factors of infection in patients receiving AHSCT, C-reactive protein and ferritin were proposed as predictive factors [14]. However, this study evaluated these factors only on the day of transplantation.
Nevertheless, we believe it can be more helpful if these factors get evaluated a few days before transplantation so that the clinicians have more time for making proper modi cations to the treatment strategies, for instance, on bacterial prophylaxis. Other studies have identi ed several other related factors including the time to platelet engraftment [32], the number of stem cells infused [33], duration of neutropenia [34,35], contamination of stem cells, presence of an indwelling central venous catheter, therapy-related mucosal damage, etc. [36,37]. However, these factors are mostly related to the post-transplantation period or during the procedure of transplantation; therefore, it may not be useful for risk strati cation and modi cation of pre-transplantation treatment.
We would like to address some limitations of our study. Due to the relatively small sample size, the results may not be generalized for the whole population, and further studies are needed to con rm our ndings. The strategy for administration of bacterial prophylaxis was patient-dependent in some cases (for example cipro oxacin was administered only during the neutropenic days), and this issue could affect our results; however, due to the established protocol of patients' treatment and ethical issues, we were not allowed to equalize the treatment in both groups.

Conclusion
The results of our study showed that the incidence of febrile neutropenia is signi cantly high in lymphoma patients who receive BMT. Moreover, some associated factors for febrile neutropenia were also investigated, and despite the limitations of our study, the results showed that the number of administered units of platelet and uric acid at ve days before BMT are signi cantly associated with febrile neutropenia.