Pediatric Hematology and Oncology, 31:415–424, 2014 C Informa Healthcare USA, Inc. Copyright ISSN: 0888-0018 print / 1521-0669 online DOI: 10.3109/08880018.2013.848387
ORIGINAL ARTICLE Supportive Care
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Do Proinflammatory Cytokine Levels Predict Serious Complication Risk of Infection in Pediatric Cancer Patients? Deniz Guven Karakurt,1 Ugur Demirsoy,2 Funda Corapcioglu,1 Selim Oncel,3 Meriban Karadogan,1 and Emin Sami Arisoy3 1
Department of Pediatric Oncology, Kocaeli University, Kocaeli, Turkey; 2 Department of Pediatric Hematology, Kocaeli University, Kocaeli, Turkey; 3 Department of Pediatric Infectious Diseases, Kocaeli University, Kocaeli, Turkey
Determination of risk of severe bacterial infection complication in children with cancer is important to diminish the cost of hospitalization and therapy. In this study, children with cancer (leukemia excluded) were evaluated for risk of severe infection complication, success of therapy and the relation between clinical and inflammatory parameters during neutropenic fever attacks. Children who fulfilled the criteria of neutropenic fever with cancer were enrolled in the study. During admission, together with clinical and laboratory parameters; interleukin-6, interleukin-8, soluble tumor necrosis factor receptor II, and soluble interleukin 2 reseptor ve procalcitonin levels were detected. Empirical therapy was started with piperacillin/tazobactam and relation between the inflammatory cytokine levels and therapy response parameters were evaluated. The study population included 31 children and 50 neutropenic attacks were studied. In 48% of the attacks, absolute neutrophile count was >100/mm3 and infectious agents were shown microbiologically in 12% of the attacks. In the study group with piperacillin/tazobactam monotherapy, the success rate without modification was 58%. In the therapy modified group mean duration of fever, antibiotherapy and hospitalization were significantly longer than the group without modification. Inflammatory cytokines’ levels during admission (interleukin-6, interleukin-8, soluble tumor necrosis factor reseptor II) were higher in patients with fever >3 days and in multiple regression analysis, it has been shown that they have a determinative role on fever control time. Other cytokines did not show any significant relationship with risk of severe bacterial infection complication and success of therapy. Keywords infections, neutropenia, oncology
INTRODUCTION Febrile neutropenia is a common and serious complication of cancer chemotherapy. There have been few studies inquiring the relationship between clinical parameters and proinflammatory cytokine levels during a febrile neutropenia attack in pediatric cancer patients [1–3]. There is no consensus on how to approach these patients regarding clinical and laboratory findings. These data may help to predict risk of serious infectious complications, patients’ risk groups and also duration of therapy [3, 4]. Received 6 August 2013; accepted 21 September 2013. Address correspondence to Dr Funda Corapcioglu, Department of Pediatric Oncology, Kocaeli University, Kocaeli, Turkey. E-mail: [email protected]
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
D. G. Karakurt et al.
Significant results of studies assessing clinical and laboratory features of febrile neutropenia attacks will probably shorten hospitalization and intravenous treatment periods in patients with low infectious complication risk. This patient group will also have lower risk for resistant nosocomial infections, drug toxicities, negative psychological effects and high drug expenditures. According to Infectious Diseases Society of America, high-risk patients are those with anticipated prolonged (>7 days duration) and profound neutropenia (ANC ≤100 cells/mm3 following cytotoxic chemotherapy) and/or significant medical co-morbid conditions, including hypotension, pneumonia, new-onset abdominal pain, or neurologic changes [5]. Low-risk patients, including those with anticipated brief (≤7 days duration) neutropenic periods or no or few co-morbidities, are candidates for oral or short interval intravenous empirical therapy while high-risk patients are strongly recommended to be admitted to hospital for empirical intravenous therapy [3, 4, 6]. By risk evaluation performed in many studies, it is intended to define serious infection and bacteremia risks and also associated mortality and morbidities. Determination of the risk of complication of severe bacterial infection in children with cancer is important to diminish the cost of hospitalization and therapy [1, 7, 8]. This study aims to predict risk of serious infections complications and treatment outcomes in pediatric cancer patients with lymphomas/solid tumors. We assessed clinical and laboratory findings concerning proinflammatory cytokines; which are interleukin-6 (IL-6), interleukin-8 (IL-8), soluble tumor necrosis factor receptor-2 (sTNFRII), soluble interleukin 2 receptor (sIL-2R), C-reactive protein (CRP), and procalcitonin (PCT) levels at admission for febrile neutropenia.
MATERIAL AND METHODS Our study was conducted prospectively, in Kocaeli University, Pediatric Oncology Department with approval of local research ethics committee. Leukemia patients were not included in the study as leukemia is primarily a bone marrow disease and these patients are probably high-risk neutropenic patients. Existence of both neutropenia and fever at the same time was the only criteria for febrile neutropenia [neutropenia: Absolute neutrophil count (ANC) < 500 /mm3 or initial ANC: 500–1000 /mm3 and patient is expected to have ANC below 500 /mm3 in 48–72 hours, fever: axillary temperature ≥ 38.5◦ C once or ≥38◦ C twice in 12 hours) [4]. Each febrile attack during treatment of pediatric cancer patients was included in the study. Two consecutive febrile attacks in a patient were accepted as separate attacks if there was 15 days or more after cessation of former antibiotic treatment. Patients were not included if neutropenia and fever were not concurrent or fever was thought to be caused by malignancy. Demographic characteristics (age, gender, and type of cancer) of the study group were recorded. Febrile neutropenia attacks were analyzed regarding status of cancer (remission, relapse, active disease), presence of bone marrow involvement, treatment protocol, and its myelosuppressive effect (mild: only one myelosuppressive drug, moderate: two myelosuppressive drugs, heavily: more than two), time period since last day of the treatment protocol, expected duration of neutropenia, whether radiotherapy was applied and, if so, its duration, field and degree of myelosuppressive effect and presence of any indwelling catheter. We examined all study patients during first admission and physical examination and vital findings were recorded. Full blood count, CRP, liver and renal function tests, and clean catch urine samples were studied for every febrile neutropenic attack in each patient. We also performed radiological tests when indicated, and identified both microbial agent and its’ drug susceptibility Pediatric Hematology and Oncology
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Proinflammatory Cytokines in Febrile Neutropenia
if there was any. We considered cultures with 103 colonies/mL or more positive as all patients were neutropenic. We collected 4 mL of blood sample from each patient during every febrile neutropenic attack. Thereafter, we rapidly obtained serum from blood by centrifugation. Serum samples were stored at −80◦ C and IL-6, IL-8, sIL-2R, STNFRII, PCT, and CRP were all studied at the same time after collecting samples. IL-6, IL-8, sIL-2R, STNFRII, and PCT were studied using immunoenzymatic technique (ELISA), and CRP by nephelometric method. The measurement limits, normal and cut-off values for ages for tests to be performed were assigned considering both the laboratory instruments and kits. The lowest values for ELISA kits were 0.1 mcg/L, 2 pg/mL, 10 pg/mL, 16 pg/mL, and 0.05 ng/mL for PCT, IL-6, IL-8, sIL-2R, and STNFRII, respectively. Evidence types of infections were grouped in parallel with clinical and laboratory findings. These groups were microbiologically proven infections, clinically proven infections and possible infection (fever of unknown origin). We started piperacillin/tazobactam 80/10 mg/kg every 6 hours (maximum dose 4.5 g) as empirical treatment for all patients. We stopped empirical intravenous antibiotic treatment in patients with ANC >500 /mm3 who did not have fever (for last 48 hours), any microbiological or clinical evidence of infection. The patient was discharged with oral cefixime that completed a total of 7-day treatment if empirical intravenous treatment in hospital was less than 5 days. Treatment duration was assigned considering microbiological cause and definition of infection for those with positive results. All treatment changes after 72nd hour of empirical treatment were defined as “modification of treatment”. These were grouped as addition of glycopeptides, antifungals, antivirals, or alternation of empirical treatment combination. An antibiotic of glycopeptides group was added to treatment after 72 hours of empirical treatment in patients with ongoing neutropenia and fever that had no evidence of any microbiological agent. Antifungal treatment was initiated irrespective of ANC if fever still persisted after 72nd hour of glycopeptid treatment, with radiological assessment. Addition of antiviral treatment was assigned considering clinical findings. Date of modification and new clinical, radiologic, and microbiological findings were recorded for patients who needed treatment modification. Duration of antimicrobial treatment was assigned for every patient considering microbiological evidences and response to treatment modifications. Every period of 24 hours after initiating antimicrobial treatment was accepted as “a treatment day”. The first treatment day in which ANC was >500 /mm3 was admitted as “last day of neutropenia” and time period between the first and last day of neutropenia was defined as “neutropenia period”. It was called as “fever control day”, the first day of 7-day-long time period without fever. “Hospitalization period” defined the time length patients were treated for febrile neutropenia in hospital and “treatment period” represented the total duration of both parenteral and oral antimicrobial treatment. We continued hematopoietic growth factor treatment during neutropenic attack if it was initiated before, but we did not start hematopoietic growth factor for neutropenic attack if it was not used earlier. Success of treatment was evaluated and grouped as (i) Success without modification; findings of infection disappeared without any change in empirical treatment combination, (ii) Success with modification; findings of infection disappeared with changing empirical treatment combination, (iii) General treatment success; fever disappeared with or without addition of any other antimicrobial agent to empirical combination with recovering from neutropenia and absence of fever for minimum 7 days during neutropenia period, and (iv) Treatment failure; death due to uncontrolled infection despite modification of antimicrobial treatment and toxicity of primary infection or treatment.
C Informa Healthcare USA, Inc. Copyright
D. G. Karakurt et al.
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Statistical Analysis Analyses were performed using SPSS-13.0. We calculated mean ± standard deviation (SD) for numeric variables. The chi-square test and Fisher’s exact test were used for comparison of categorical variable frequencies. The difference of mean of two groups was assigned via Student’s t test for independent groups. Kruskal–Wallis test or Mann–Whitney U test was used to compare mean values of nonparametric data between groups. Logistic regression or linear regression tests were used to perform multivariate analysis on independent variables that were significant in binary analysis.
RESULTS We evaluated 50 neutropenia attacks in 31 pediatric cancer patients. Demographic characteristics of patients are shown in Table 1. We assessed one neutropenia attack per patient in 14, two attacks in 15, and three attacks in 2 patients. We observed that 40% of neutropenia attacks were seen ≤ 5 years of age (n = 20), lymphoproliferative tumors made up 30% (n = 15) while solid tumors accounted for 70% (n = 35). A total of 88% of patients (n = 44) had active disease while 12% (n = 6) were in remission. We did not detect bone marrow involvement in any of lymphoproliferative tumor patients at diagnosis of cancer, but 8.5% (n = 3) of solid tumor patients had. The chemotherapy protocols that were followed according to diseases’ stages and subtypes were: NHL-BFM 90, LCH-III, IECES’92, Turkish Pediatric Oncology Group(TPOG) Wilms tumor, TPOG NBL 2003, IRS-IV, CCG 7921, CEV (carboplatin, etoposide, vincristin), and ICE (ifosfamide, carboplatin, etoposide). All of the patients, except two who were supposed to have moderate, were supposed to have heavily myelosuppression. Three patients (medulloblastoma, non-Hodgkin lymphoma and rhabdomyosarcoma) had 12–54 Gy radiotherapy with myelosuppressive effect in the last 4 weeks. A total of 36 patients had used hematopoietic growth factors for varying time periods before febrile neutropenia attacks. A total of 47 patients had port catheter during admittance for attacks.
TABLE 1 Demographic Characteristics of the Study Patients Number of patients Mean age ± SD (years), range Gender (M/F) Type of cancer (n) Lymphoproliferative tumors Non-Hodgkin lymphoma Langerhans cell histiocytosis Solid tumors Soft tissue sarcoma∗ Wilms tumor Central nervous system tumors∗∗ Bone tumors∗∗∗ Neuroblastoma Retinoblastoma
31 7.96 ± 5.51, 3 months-17 years 19/12 9 7 2 22 5 5 4 4 2 2
∗
Rhabdomyosarcoma (n = 4), desmoplastic small round cell tumor (n = 1). Medulloepithelioma (n = 1), medulloblastoma (n = 2), ependymoma (n = 1). ∗∗∗ Ewing sarcoma (n = 3), osteosarcoma (n = 1). ∗∗
Pediatric Hematology and Oncology
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Proinflammatory Cytokines in Febrile Neutropenia
During admittance for febrile neutropenia, mean white blood cell count, ANC, hemoglobin, and thrombocyte count were 1006 ± 877 (46–4820) /mm3 , 214 ± 252 (0–990, median: 102) /mm3 , 9.4 ± 1.8 (4–12.8) gr/dL, 107.525 ± 128.021 (8670–624.000) /mm3 , respectively. ANC was less than 100 /mm3 in 24 patients. We compared patients with ANC ≤ 100 /mm3 in lymphoproliferative and solid tumor groups (n = 5, 33% vs. n = 19, 54%) and found no significant difference between these groups at admittance (P = .147). Febrile neutropenia attacks developed averagely 8.14 ± 3.71 (0–19) days after last chemotherapy day for all patients and mean anticipated neutropenia time was 10.88 ± 5.54 (0–21) days. Infection was defined clinically in 15 [one soft tissue infection, four oral mucositis (clinical fungal mucositis), four gastroenteritis, four pneumonia], microbiologically in six patients. A total of 29 patients had fever of unknown origin. In the study group, two patients had microbial evidence in peripheral blood, where one patient had both in peripheral blood and port catheter samples and three other patients’ results were positive in urine specimens. The specified pathogens during bacteremia attacks were K. pneumoniae, S. Epidermidis, and P. Aeruginosa. E.Coli was responsible for two attacks and K. Pneumoniae for one, in urine culture positive group. Median values and ranges of IL-6, IL-8, sTNFRII, sIL-2R, and CRP at presentation of 50 attacks in study group are given in Table 2. Evaluation of CRP and proinflammatory cytokine levels considering ANC, tumor type, age groups, and definition of infection is noted in Table 3. Measurement of procalcitonin was performed categorically and was found negative in 29 attacks, below 0.5 ng/mL in 11, ≥0.5–2 ng/mL in 6, 2–10 ng/mL in 3, and ≥10 ng/mL in 2. Overall success of treatment was 100% as none of patients died in association with febrile neutropenia. Success without modification was 58%. We applied various types of modifications on the initial treatment, totally 26 in 21 attacks. These modifications were performed averagely 4.23 ± 0.83 (3–5) days after starting antimicrobial treatment. Control of fever and recovery of neutropenia in the whole group occurred after 2.0 ± 2.8 (1–16) and 6.1 ± 3.6 (1–19) days, respectively, and we observed these time periods were similar for lymphoproliferative and solid tumor patients (P = 0.586, 0.294), respectively. Mean duration of hospitalization time for all attacks in study group was 8.5 ± 5.4 (1–26) days while it was 10.5 ± 3.5 (5–24) days for duration of antibiotic use. Duration of hospitalization was longer for lymphoproliferative tumor group when compared with solid tumor patients [10.9 ± 4.3 (6–19) days vs. 8.9 ± 5.9 (1–26) days] (P = 0.047). Duration of antibiotic treatment was shorter in the group without treatment modification in contrast to treatment modification group [7.8 ± 1.7 (5–12) days vs. 10 ± 4.5 (9–24) days] (P = 0.001). Hospitalization duration was shorter in the group without treatment modification in contrast to treatment modification group [7.0 ± 3.5 (5–15) days vs. 10.0 ± 6.7 (4–26) days] (P < 0.001). There was no difference between lymphoproliferative tumor patients and solid tumor patients concerning duration of antibiotic
TABLE 2 Proinflammatory Cytokine Levels During Febrile Neutropenia Attacks in Nonleukemic Pediatric Cancer Patients Proinflammatory cytokines
Median
Ranges (of the samples)
IL-6 (pg/mL) IL-8 (pg/mL) sTNFRII (pg/mL) sIL-2R (ng/mL) CRP (mg/dL)
94.68 125.91 628.67 12.07 1.8
6.65–1729.97 18.16–3362.44 131.27–2894.34 3.09–42.58 0–18.3
C Informa Healthcare USA, Inc. Copyright
D. G. Karakurt et al.
TABLE 3 Comparison of Proinflammatory Cytokine Levels with ANC, Tumor Type, Age Groups and Infection Definition IL-6 IL-8 sTNFRII sIL-2R CRP (pg/mL) (pg/mL) (pg/mL) (ng/mL) (mg/dL)
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Features Absolute neutrophil count ≤ 100 (n = 24) > 100 (n = 26) P Tumor type Lymphoproliferative (n = 15) Solid (n = 35) P Age group ≤5 years (n = 20) >5 years (n = 30) P Type of infection documentation Fever of unknown origin (n = 29) Clinically proven infection (n = 15) Microbiologically proven infection (n = 6) P
327.18 63.16 0.01
524.13 175.01 0.01
1000.33 757.88 0.77
17.04 12.95 0.77
6.06 3.72 0.77
97.40 229.53 1.00
192.30 406.99 0.53
576.79 1001.74 0.53
17.46 13.83 0.06
3.84 5.27 1.00
268.37 137.57 0.38
429.65 284.54 0.38
1010.95 783.13 0.14
17.37 13.28 0.38
4.55 5.04 0.77
158.42 91.03 589.16 0.11
316.38 169.13 902.90 0.10
949.43 874.50 510.28 0.30
13.12 14.46 24.75 0.34
5.41 3.70 4.93 0.06
treatment [10.5 ± 2.8 (6–17) days vs. 10.6 ± 3.9 (5–24) days] (P = 0.917). We assessed the effects of inflammatory determinants over parameters of treatment response such as presence of modification, neutropenia recovery time, duration of fever control, duration of antibiotics and found that IL-8 levels were higher in the treatment modification group when compared with group without treatment modification (21 attacks, 464.53 pg/mL vs. 29 attacks, 254.28 pg/mL) (P = 0.048). There was no significant difference between groups without treatment modification and with treatment modification when IL-6, sTNFRII, sIL-2R, and CRP are noted (137.77 pg/mL, 979.28 pg/mL, 13.11 ng/mL, 3.75 mg/dl vs. 261.87 pg/ml, 729.22 pg/ml, 17.41 ng/ml, 6.35 mg/dL, P = 0.25; 0.25; 0.56; 0.25). Evaluation of effects of proinflammatory cytokine levels over parameters of treatment response is shown in Table 4. Fever control time, neutropenia recovery time, TABLE 4 Evaluation of Proinflammatory Cytokine Levels in Association with Neutropenia, Fever, Treatment and Hospitalization Durations Treatment response parameters Neutropenia period ≤7 days (n = 37) >7 days (n = 13) P Duration of fever ≤3 days (n = 39) >3 days (n = 11) P Treatment period (antibiotics) ≤7 days (n = 8) >7 days (n = 42) P Hospitalization period ≤7 days (n = 24) >7 days (n = 26) P
IL-6 (pg/mL)
IL-8 (pg/mL)
sTNFRII (pg/mL)
sIL-2R (ng/mL)
CRP (mg/dL)
45.92 118.17 0.19
97.53 172.72 0.38
628.41 1072.19 0.50
13.55 9.87 0.60
1.80 3.78 0.95
81.96 142.39 0.02
97.53 472.11 0.01
423.61 674.03 0.04
11.14 19.20 0.04
1.60 9.3 0.04
137.61 90.20 0.81
179.12 125.91 0.45
534.42 642.39 0.59
19.95 10.73 0.12
2.08 1.84 0.36
115.52 45.54 0.20
142.02 97.28 0.50
673.78 610.97 0.11
11.25 13.46 0.27
2.17 1.84 0.68
Pediatric Hematology and Oncology
Proinflammatory Cytokines in Febrile Neutropenia
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
hospitalization period and duration of antibiotic treatment were as 2.6 ± 1.4, 6.1 ± 3.7, 9.1 ± 4.8, and 10.0 ± 3.1 days in the group with procalcitonin level ≤1 ng/mL, and 5.8 ± 5.3, 5.9 ± 3.6, 11.3 ± 7.5, and 12.7 ± 4.6 days in the group procalcitonin >1 ng/mL, respectively (P = 0.014, 0.388, 0.382, 0.074). In multivariate analysis (logistic regression and linear regression tests) analysis of effects of IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels do not significantly contribute neutropenia duration (P = 0.104, 0.522, 0.701, 0.887, 0.211, 0.870), antibiotic treatment period (P = 0.423, 0.664, 0.120, 0.966, 0.935, 0.354), and hospitalization period (P = 0.913, 0.454, 0.275, 0.092, 0.408, 0.551), respectively. We also defined that IL-6, IL-8, and sTNFRII levels do interfere with duration of fever while sIL-2R, CRP, PCT levels do not (P = 0,012; 0,003; 0,051; 0,150; 0,494; 0,154). DISCUSSION A total of 70% of pediatric cancer patients have increased risk for invasive bacterial disease at first presentation. A total of 25% of them develop severe sepsis. Also, 3% of these high-risk patients eventually die. It is still not possible to predict risk of serious bacterial infection during referral to hospital for febrile neutropenia. There are some studies to determine risk of serious bacterial infection in adult patients but these studies are seldom conducted in pediatric patient groups [9–12]. We aimed to define risk of serious bacterial infection in pediatric cancer patients using clinical and laboratory parameters at first presentation of febrile neutropenia. There is still not an optimal antimicrobial treatment regimen for treatment of febrile neutropenia. Characteristics of the patient group are the most important factor influencing treatment plan. Heaviness and duration of neutropenia and presence of additional complications should be considered while deciding on empirical treatment in the beginning [13]. Studies investigating piperacillin/tazobactam monotherapy for children have been popular in recent years. In a randomized study, which was conducted in our center, we compared piperacillin/tazobactam with cefepime and success rates without modification were 56% and 48%, respectively [8]. Piperacillin/tazobactam monotherapy is the standard empirical treatment regimen used for study group in the current study. Success rate without modification is 58% in our study. Most preferred modification was adding glycopeptides to therapy amongst 26 attacks that were modified (38.4%). The main inquiry in our study was about possibility of clinical and laboratory findings as strong indicators of serious bacterial infection during febrile neutropenia attacks. “Serious infection complication” definition includes “death as a result of infection during sepsis and/or shock, bacteremia or fungemia, or febrile neutropenic attack.” It seems to be rational to evaluate clinical findings in the scope of serious infection complication as rate of microbiologically proven infection is low in our study group. We defined in our study that; tumor type, ANC at presentation and patients’ age groups didn’t play a key role in influencing microbiologically and clinically proven infection attacks. We also showed that these parameters did not affect fever control time, hospitalization, and treatment periods as noted in several other studies [13, 14]. There are limited studies evaluating importance of CRP and proinflammatory cytokine levels in detecting bacterial infection and clinical progress during febrile neutropenia attack. Persson et al. [15] investigated CRP, Serum Amyloid A (SAA), PCT, and IL-6 levels’ effect on clinical progress in 101 febrile neutropenic attacks of 79 patients. In the same study, it was shown in complicated (fever >3 days, clinically proven infections and bacteremias) patients that PCT and IL-6 levels were significantly high but CRP and SAA levels did not differ. Hodge et al. [16] investigated CRP, IL-2, IL4, IL-5, IL-10, and IL-8 levels concurrently in febrile neutropenic pediatric cancer C Informa Healthcare USA, Inc. Copyright
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
D. G. Karakurt et al.
patients. Bacteremia was detected in 8 of 31 attacks and they observed higher levels of IL-5 and IL-8 when compared with nonbacteremic group, also did not find any association between CRP, other cytokine levels and blood culture positivity. In another study, conducted with 85 febrile neutropenic pediatric patients, CRP, IL-8 and monocyte chemotactic protein were found to be significantly high in the first 24 hours of fever in bacteremic/septisemic patients [17]. Engel et al. [18] detected increase in PCT levels in the first 32 hours of fever in patients with clinically or microbiologically proven infection. They also found PCT levels were lower in patients with fever of unknown origin when compared with proven infection. Pihusch et al. [19] showed elevation of CRP, PCT, and IL-6 levels following infectious complications in 350 allogeneic stem cell transplantation patients. Only using PCT levels were reasonable to differentiate infection from transplant related complications. The cut-off value of serum PCT, that was used to predict infections with or without bacteremia, was 0.5–1.3 ng/mL (sensitivity 44–88%, specificity 61–88%) in several studies. It was reported that cut off value of serum PCT levels in the first 24 hours of febrile neutropenia could indicate low risk of serious infection if 2 ng/mL [20]. Conversely, Bont et al. [21] found no difference in serum PCT levels between bacteremic and nonbacteremic patients. But the noteworthy point of this study is bacteremic episodes depended on coagulase negative staphylococci and these microorganisms did not cause a significant increase in PCT levels [22, 23]. Also, Fleischhack et al. [24] and Svaldi et al. [25] showed higher increase of PCT in Gram (–) infections compared to Gram (+). Higher levels of IL-6 and IL-8 in ANC ≤100 /mm3 patients in our study could be a cautionary signal for serious bacterial infection in this group. We found no difference between ANC >100 /mm3 and ANC ≤100 /mm3 groups in the rate of clinically and microbiologically proven infections, so we decided to take these two parameters together as an unfavorable prognostic factor during assessment of risk factors. CRP levels, interestingly, were found lower in clinically and microbiologically proven infections when compared to attacks with UFO in our study. We analyzed treatment response parameters aiming to assess IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels’ effect during first presentation. It is not possible to compare our study results with the other studies’ results mentioned above as we did not plan to collect proinflammatory cytokine levels sequentially. We only defined significantly high IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels studied from patients with fever >3 days through the parameters above. IL-8 levels were higher in the treatment modification group when compared with group without treatment modification. We want to emphasize that duration of fever >3 days an indication for treatment modification. Additionally, we observed in multivariate analyses that only IL-6, IL-8, and sTNFRII levels affected duration of fever. It is noted in some studies that serum sIL-2R level increases in viral infections, tuberculosis and posttransplant infections [26–28]. Fleischhack et al. [24] showed major increase in serum sIL-2R level during Gram negative bacteremia of pediatric cancer patients. Studies involving sTNFIIR showed increase in serum levels especially during septicemia and pneumonia [29–31]. Moreover, sTNFIIR levels stay high for the whole fever episode and sometimes during convalescence, and this makes it insufficient for evaluating episodes’ progress and treatment response though it is useful at the time of diagnosis. In our study, considering need for modification in patients who are treated empirically and have uncontrolled fever for >72 hours, it is possible to predict outcome of febrile attacks of patients with high-proinflammatory cytokines at admission. This may be indirectly applied to the fact that duration of antibiotics and hospital duration in the treatment modification group was longer than the no treatment modification group. Our data suggest that Pediatric Hematology and Oncology
Proinflammatory Cytokines in Febrile Neutropenia
duration of fever does not provide insight into how serious the infection might be or the risk of sepsis. In conclusion, we did not find IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels as significant predictors of serious infectious complication in pediatric cancer patients during admission for febrile neutropenia, but our study data showed that IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels during admission were high in patients with fever >3 days. These findings might be helpful to prevent unnecessary use of proinflammatory cytokines on routine clinical practice.
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Declaration of Interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
REFERENCES [1] Santoloya ME, Alvarez AM, Avil´es CL, et al. Early hospital discharge followed by outpatient management versus continued hospitalization of children with cancer, fever, and neutropenia at low risk for invasive bacterial infection. J Clin Oncol. 2004;22:3784–3789. [2] Anoop P, Anjay MA. Febrile neutropenia: transition towards a risk-directed approach. Arch Dis Child. 2007;92:467–468. [3] Santolaya ME, Alvarez AM, Avil´es CL, et al. Prospective evalution of a model of prediction of invasive bacterial infection risk among children with cancer, fever and neutropenia. Clin Infect Dis. 2002;35:678–683. [4] H¨artel C, Deuster M, Lehrbecher T, Schultz C. Current approaches for risk stratification of infectious complications in pediatric oncology. Pediatr Blood Cancer. 2007;49:767–773. [5] Freifeld AG, Bow EJ, Sepkowitz KA, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis. 2011;52:56–93. [6] Wicki S, Keisker A, Aebi C, et al. Risk prediction of fever in neutropenia in children with cancer: a step towards individually tailored supportive therapy. Pediatr Blood Cancer 2008;51: 778–783. [7] C ¸ orapcıo˘glu F, Sarıalio˘glu F, Olgun N, Uysal KM. Analysis of 136 febrile neutropenic episodes in children with cancer: evaluation of treatment effectiveness and cost. Pediatr Hematol Oncol. 2004;21:535–543. [8] C ¸ orapcıo˘glu F, Sarper N, Zengin E. Monotherapy with piperacillin/tazobactam versus cefepime as empirical therapy for febrile neutropenia in pediatrıc cancer patients: a randomized comparison. Pediatr Hematol Oncol. 2006;23:177–186. [9] Ahmed N, El-Mahallawy HA, Ahmed IA, et al. Early hospital discharge versus continued hospitalization febrile cancer patients with prolonged neutropenia. Pediatr Blood Cancer. 2007;49: 786–792. [10] Klastersky J, Paesmans M, Rubenstein EB, et al. The Multinational association for supportive care in cancer risk index: a multinational scoring system for identifiying low-risk febrile neutropenic patients. J Clin Oncol. 2000;18:3038–3051. [11] Hann I, Viscoli C, Paesmans M, et al. A comprasion of outcome from febril neutropenic episodes in children compared with adults: result from four EORTC studies. International Antimicrobial Therapy Cooperative Group (IATCG) of the European Organization for Research and Treatment of Cancer (EORTC). Br J Haematol. 1997;99:580–588. [12] Boragina M, Patel H, Reiter S, Dougherty G. Management of febrile neutropenia in pediatric oncology patients: a Canadian survey. Pediatr Blood Cancer. 2007;48:521–526. [13] Duzova ¨ A, Kutluk T, Kanra G, et al. Monotherapy with meropenem versus combination therapy with piperacillin plus amikacin as empiric therapy for neutropenic fever in children with lymphoma and solid tumors. Turk J Pediatr. 2001;43:105–109. [14] Uygun V, Karasu GT, Ogunc D, et al. Piperacillin/tazobactam versus cefepime for the empirical treatment of pediatric cancer patients with neutropenia and fever: a randomized and open-label study. Pediatr Blood Cancer. 2009;53:610–614. [15] Sudhoff ¨ T, Gigounidis A, Karthaus M. Evaluation of neutropenic fever: value of serum and plasma parameters in clinical practice. Chemotherapy. 2000;46:77–85. C Informa Healthcare USA, Inc. Copyright
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
D. G. Karakurt et al.
[16] Hodge G, Osborn M, Hodge S, et al. Rapid simultaneous measurement of multiple cytokines in childhood oncology patients with febrile neutropenia: increased interleukin (IL)-8 or IL-5 correlates with culture-positive infection. Br J Haematol. 2005;132:247–248. [17] El-Maghraby SM, Moneer MM, Ismail MM, et al. The diagnostic value of C-reactive protein, Interleukin-8 and monocyte chemotactic protein in risk stratification of febrile neutropenic children with hematologic malignancies. Pediatr Hematol Oncol. 2007;29:131–136. [18] Engel A, Steinbach G, Kern P, Kern WV. Diagnostic value of procalcitonin serum levels in neutropenic patients with fever: comparison with interleukin-8. Scand J Infect Dis. 1999;31:185–189. [19] Pihusch M, Pihusch R, Fraunberger P, et al. Evaluation of C-reactive protein, interleukin-6, and procalcitonin levels in allogeneic hematopoietic stem cell recipients. Eur J Haematol. 2006;76:93–101. [20] Frei CR, Hampton SL, Burgess DS. Influence of culture site-specific MIC distributions on the pharmacokinetic and pharmacodynamic properties of piperacillin/tazobactam and piperacillin: a data analysis. Clin Ther. 2006;28:1035–1040. [21] de Bont ES, Vellenga E, Swaanenburg J, Kamps W. Procalcitonin: a diagnostic marker of bacterial infection in neutropenic cancer patients with fever? Infection. 2000;28:398–400. [22] Giamarellou H, Giamarellos-Bourboulis EJ, Repoussis P, et al. Potential use of procalcitonin as a diagnostic criterion in febrile neutropenia: experience from a multicentre study. Clin Microbiol Infect. 2004;10:628–633. [23] Persson L, Engervall P, Magnuson A, et al. Use of inflammatory markers for early detection of bacteraemia in patients with febrile neutropenia. Scand J Infect Dis. 2004; 36: 365–371. [24] Fleischhack G, Kambeck I, Cipic D, et al. Procalcitonin in paediatric cancer patient; its diagnostic relevance is superior to that of C reactive protein, interleukin 6, interleukin 8, soluble 2 receptor and soluble tumour necrosis factor receptor II. Br Haematol. 2000;111:1093–1102. [25] Svaldi M, Hirber J, Lanthaler AL, et al. Procalcitonin-reduced sensitivity and specificity in heavily leucopenic and immunosuppressed patients. Br J Haematol. 2001;115:53–57. [26] Shitrit D, Izbicki G, Bar-Gil Shitrit A, et al. Role of soluble interleukin-2 receptor levels in patients with latent tuberculosis. Lung. 2006;184:21–24. [27] Izzo F, Curley S, Maio P, et al. Correlation of soluble interleukin-2 receptor levels with severity of chronic hepatitis C virus liver injury and development of hepatocellular cancer. Surgery. 1996;120:100–105. [28] Liem LM, van Houwelingen HC, Goulmy E. Serum cytokine levels after HLA-identical bone marrow transplantation. Transplantation. 1998;66:863–871. [29] Van der Poll T, Jansen J, van Leenen D, et al. Release of soluble receptors for tumor necrosis factor in clinical sepsis and experimental endotoxemia. J Infect Dis. 1993;168:955–960. [30] Ertel W, Scholl FA, Gallati H, et al. Increased release of soluble tumor necrosis factor receptors into blood during clinical sepsis. Arch Surg. 1994;129:1330–1336. [31] De Werra I, Jaccard C, Corradin SB, et al. Cytokines, nitrite/nitrate, soluble tumor necrosis factor receptors, and procalcitonin concentrations: comparisons in patients with septic shock, cardiogenic shock, and bacterial pneumonia. Crit Care Med. 1997;25:607–613.
Pediatric Hematology and Oncology
ORIGINAL ARTICLE Supportive Care
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Do Proinflammatory Cytokine Levels Predict Serious Complication Risk of Infection in Pediatric Cancer Patients? Deniz Guven Karakurt,1 Ugur Demirsoy,2 Funda Corapcioglu,1 Selim Oncel,3 Meriban Karadogan,1 and Emin Sami Arisoy3 1
Department of Pediatric Oncology, Kocaeli University, Kocaeli, Turkey; 2 Department of Pediatric Hematology, Kocaeli University, Kocaeli, Turkey; 3 Department of Pediatric Infectious Diseases, Kocaeli University, Kocaeli, Turkey
Determination of risk of severe bacterial infection complication in children with cancer is important to diminish the cost of hospitalization and therapy. In this study, children with cancer (leukemia excluded) were evaluated for risk of severe infection complication, success of therapy and the relation between clinical and inflammatory parameters during neutropenic fever attacks. Children who fulfilled the criteria of neutropenic fever with cancer were enrolled in the study. During admission, together with clinical and laboratory parameters; interleukin-6, interleukin-8, soluble tumor necrosis factor receptor II, and soluble interleukin 2 reseptor ve procalcitonin levels were detected. Empirical therapy was started with piperacillin/tazobactam and relation between the inflammatory cytokine levels and therapy response parameters were evaluated. The study population included 31 children and 50 neutropenic attacks were studied. In 48% of the attacks, absolute neutrophile count was >100/mm3 and infectious agents were shown microbiologically in 12% of the attacks. In the study group with piperacillin/tazobactam monotherapy, the success rate without modification was 58%. In the therapy modified group mean duration of fever, antibiotherapy and hospitalization were significantly longer than the group without modification. Inflammatory cytokines’ levels during admission (interleukin-6, interleukin-8, soluble tumor necrosis factor reseptor II) were higher in patients with fever >3 days and in multiple regression analysis, it has been shown that they have a determinative role on fever control time. Other cytokines did not show any significant relationship with risk of severe bacterial infection complication and success of therapy. Keywords infections, neutropenia, oncology
INTRODUCTION Febrile neutropenia is a common and serious complication of cancer chemotherapy. There have been few studies inquiring the relationship between clinical parameters and proinflammatory cytokine levels during a febrile neutropenia attack in pediatric cancer patients [1–3]. There is no consensus on how to approach these patients regarding clinical and laboratory findings. These data may help to predict risk of serious infectious complications, patients’ risk groups and also duration of therapy [3, 4]. Received 6 August 2013; accepted 21 September 2013. Address correspondence to Dr Funda Corapcioglu, Department of Pediatric Oncology, Kocaeli University, Kocaeli, Turkey. E-mail: [email protected]
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
D. G. Karakurt et al.
Significant results of studies assessing clinical and laboratory features of febrile neutropenia attacks will probably shorten hospitalization and intravenous treatment periods in patients with low infectious complication risk. This patient group will also have lower risk for resistant nosocomial infections, drug toxicities, negative psychological effects and high drug expenditures. According to Infectious Diseases Society of America, high-risk patients are those with anticipated prolonged (>7 days duration) and profound neutropenia (ANC ≤100 cells/mm3 following cytotoxic chemotherapy) and/or significant medical co-morbid conditions, including hypotension, pneumonia, new-onset abdominal pain, or neurologic changes [5]. Low-risk patients, including those with anticipated brief (≤7 days duration) neutropenic periods or no or few co-morbidities, are candidates for oral or short interval intravenous empirical therapy while high-risk patients are strongly recommended to be admitted to hospital for empirical intravenous therapy [3, 4, 6]. By risk evaluation performed in many studies, it is intended to define serious infection and bacteremia risks and also associated mortality and morbidities. Determination of the risk of complication of severe bacterial infection in children with cancer is important to diminish the cost of hospitalization and therapy [1, 7, 8]. This study aims to predict risk of serious infections complications and treatment outcomes in pediatric cancer patients with lymphomas/solid tumors. We assessed clinical and laboratory findings concerning proinflammatory cytokines; which are interleukin-6 (IL-6), interleukin-8 (IL-8), soluble tumor necrosis factor receptor-2 (sTNFRII), soluble interleukin 2 receptor (sIL-2R), C-reactive protein (CRP), and procalcitonin (PCT) levels at admission for febrile neutropenia.
MATERIAL AND METHODS Our study was conducted prospectively, in Kocaeli University, Pediatric Oncology Department with approval of local research ethics committee. Leukemia patients were not included in the study as leukemia is primarily a bone marrow disease and these patients are probably high-risk neutropenic patients. Existence of both neutropenia and fever at the same time was the only criteria for febrile neutropenia [neutropenia: Absolute neutrophil count (ANC) < 500 /mm3 or initial ANC: 500–1000 /mm3 and patient is expected to have ANC below 500 /mm3 in 48–72 hours, fever: axillary temperature ≥ 38.5◦ C once or ≥38◦ C twice in 12 hours) [4]. Each febrile attack during treatment of pediatric cancer patients was included in the study. Two consecutive febrile attacks in a patient were accepted as separate attacks if there was 15 days or more after cessation of former antibiotic treatment. Patients were not included if neutropenia and fever were not concurrent or fever was thought to be caused by malignancy. Demographic characteristics (age, gender, and type of cancer) of the study group were recorded. Febrile neutropenia attacks were analyzed regarding status of cancer (remission, relapse, active disease), presence of bone marrow involvement, treatment protocol, and its myelosuppressive effect (mild: only one myelosuppressive drug, moderate: two myelosuppressive drugs, heavily: more than two), time period since last day of the treatment protocol, expected duration of neutropenia, whether radiotherapy was applied and, if so, its duration, field and degree of myelosuppressive effect and presence of any indwelling catheter. We examined all study patients during first admission and physical examination and vital findings were recorded. Full blood count, CRP, liver and renal function tests, and clean catch urine samples were studied for every febrile neutropenic attack in each patient. We also performed radiological tests when indicated, and identified both microbial agent and its’ drug susceptibility Pediatric Hematology and Oncology
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Proinflammatory Cytokines in Febrile Neutropenia
if there was any. We considered cultures with 103 colonies/mL or more positive as all patients were neutropenic. We collected 4 mL of blood sample from each patient during every febrile neutropenic attack. Thereafter, we rapidly obtained serum from blood by centrifugation. Serum samples were stored at −80◦ C and IL-6, IL-8, sIL-2R, STNFRII, PCT, and CRP were all studied at the same time after collecting samples. IL-6, IL-8, sIL-2R, STNFRII, and PCT were studied using immunoenzymatic technique (ELISA), and CRP by nephelometric method. The measurement limits, normal and cut-off values for ages for tests to be performed were assigned considering both the laboratory instruments and kits. The lowest values for ELISA kits were 0.1 mcg/L, 2 pg/mL, 10 pg/mL, 16 pg/mL, and 0.05 ng/mL for PCT, IL-6, IL-8, sIL-2R, and STNFRII, respectively. Evidence types of infections were grouped in parallel with clinical and laboratory findings. These groups were microbiologically proven infections, clinically proven infections and possible infection (fever of unknown origin). We started piperacillin/tazobactam 80/10 mg/kg every 6 hours (maximum dose 4.5 g) as empirical treatment for all patients. We stopped empirical intravenous antibiotic treatment in patients with ANC >500 /mm3 who did not have fever (for last 48 hours), any microbiological or clinical evidence of infection. The patient was discharged with oral cefixime that completed a total of 7-day treatment if empirical intravenous treatment in hospital was less than 5 days. Treatment duration was assigned considering microbiological cause and definition of infection for those with positive results. All treatment changes after 72nd hour of empirical treatment were defined as “modification of treatment”. These were grouped as addition of glycopeptides, antifungals, antivirals, or alternation of empirical treatment combination. An antibiotic of glycopeptides group was added to treatment after 72 hours of empirical treatment in patients with ongoing neutropenia and fever that had no evidence of any microbiological agent. Antifungal treatment was initiated irrespective of ANC if fever still persisted after 72nd hour of glycopeptid treatment, with radiological assessment. Addition of antiviral treatment was assigned considering clinical findings. Date of modification and new clinical, radiologic, and microbiological findings were recorded for patients who needed treatment modification. Duration of antimicrobial treatment was assigned for every patient considering microbiological evidences and response to treatment modifications. Every period of 24 hours after initiating antimicrobial treatment was accepted as “a treatment day”. The first treatment day in which ANC was >500 /mm3 was admitted as “last day of neutropenia” and time period between the first and last day of neutropenia was defined as “neutropenia period”. It was called as “fever control day”, the first day of 7-day-long time period without fever. “Hospitalization period” defined the time length patients were treated for febrile neutropenia in hospital and “treatment period” represented the total duration of both parenteral and oral antimicrobial treatment. We continued hematopoietic growth factor treatment during neutropenic attack if it was initiated before, but we did not start hematopoietic growth factor for neutropenic attack if it was not used earlier. Success of treatment was evaluated and grouped as (i) Success without modification; findings of infection disappeared without any change in empirical treatment combination, (ii) Success with modification; findings of infection disappeared with changing empirical treatment combination, (iii) General treatment success; fever disappeared with or without addition of any other antimicrobial agent to empirical combination with recovering from neutropenia and absence of fever for minimum 7 days during neutropenia period, and (iv) Treatment failure; death due to uncontrolled infection despite modification of antimicrobial treatment and toxicity of primary infection or treatment.
C Informa Healthcare USA, Inc. Copyright
D. G. Karakurt et al.
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Statistical Analysis Analyses were performed using SPSS-13.0. We calculated mean ± standard deviation (SD) for numeric variables. The chi-square test and Fisher’s exact test were used for comparison of categorical variable frequencies. The difference of mean of two groups was assigned via Student’s t test for independent groups. Kruskal–Wallis test or Mann–Whitney U test was used to compare mean values of nonparametric data between groups. Logistic regression or linear regression tests were used to perform multivariate analysis on independent variables that were significant in binary analysis.
RESULTS We evaluated 50 neutropenia attacks in 31 pediatric cancer patients. Demographic characteristics of patients are shown in Table 1. We assessed one neutropenia attack per patient in 14, two attacks in 15, and three attacks in 2 patients. We observed that 40% of neutropenia attacks were seen ≤ 5 years of age (n = 20), lymphoproliferative tumors made up 30% (n = 15) while solid tumors accounted for 70% (n = 35). A total of 88% of patients (n = 44) had active disease while 12% (n = 6) were in remission. We did not detect bone marrow involvement in any of lymphoproliferative tumor patients at diagnosis of cancer, but 8.5% (n = 3) of solid tumor patients had. The chemotherapy protocols that were followed according to diseases’ stages and subtypes were: NHL-BFM 90, LCH-III, IECES’92, Turkish Pediatric Oncology Group(TPOG) Wilms tumor, TPOG NBL 2003, IRS-IV, CCG 7921, CEV (carboplatin, etoposide, vincristin), and ICE (ifosfamide, carboplatin, etoposide). All of the patients, except two who were supposed to have moderate, were supposed to have heavily myelosuppression. Three patients (medulloblastoma, non-Hodgkin lymphoma and rhabdomyosarcoma) had 12–54 Gy radiotherapy with myelosuppressive effect in the last 4 weeks. A total of 36 patients had used hematopoietic growth factors for varying time periods before febrile neutropenia attacks. A total of 47 patients had port catheter during admittance for attacks.
TABLE 1 Demographic Characteristics of the Study Patients Number of patients Mean age ± SD (years), range Gender (M/F) Type of cancer (n) Lymphoproliferative tumors Non-Hodgkin lymphoma Langerhans cell histiocytosis Solid tumors Soft tissue sarcoma∗ Wilms tumor Central nervous system tumors∗∗ Bone tumors∗∗∗ Neuroblastoma Retinoblastoma
31 7.96 ± 5.51, 3 months-17 years 19/12 9 7 2 22 5 5 4 4 2 2
∗
Rhabdomyosarcoma (n = 4), desmoplastic small round cell tumor (n = 1). Medulloepithelioma (n = 1), medulloblastoma (n = 2), ependymoma (n = 1). ∗∗∗ Ewing sarcoma (n = 3), osteosarcoma (n = 1). ∗∗
Pediatric Hematology and Oncology
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Proinflammatory Cytokines in Febrile Neutropenia
During admittance for febrile neutropenia, mean white blood cell count, ANC, hemoglobin, and thrombocyte count were 1006 ± 877 (46–4820) /mm3 , 214 ± 252 (0–990, median: 102) /mm3 , 9.4 ± 1.8 (4–12.8) gr/dL, 107.525 ± 128.021 (8670–624.000) /mm3 , respectively. ANC was less than 100 /mm3 in 24 patients. We compared patients with ANC ≤ 100 /mm3 in lymphoproliferative and solid tumor groups (n = 5, 33% vs. n = 19, 54%) and found no significant difference between these groups at admittance (P = .147). Febrile neutropenia attacks developed averagely 8.14 ± 3.71 (0–19) days after last chemotherapy day for all patients and mean anticipated neutropenia time was 10.88 ± 5.54 (0–21) days. Infection was defined clinically in 15 [one soft tissue infection, four oral mucositis (clinical fungal mucositis), four gastroenteritis, four pneumonia], microbiologically in six patients. A total of 29 patients had fever of unknown origin. In the study group, two patients had microbial evidence in peripheral blood, where one patient had both in peripheral blood and port catheter samples and three other patients’ results were positive in urine specimens. The specified pathogens during bacteremia attacks were K. pneumoniae, S. Epidermidis, and P. Aeruginosa. E.Coli was responsible for two attacks and K. Pneumoniae for one, in urine culture positive group. Median values and ranges of IL-6, IL-8, sTNFRII, sIL-2R, and CRP at presentation of 50 attacks in study group are given in Table 2. Evaluation of CRP and proinflammatory cytokine levels considering ANC, tumor type, age groups, and definition of infection is noted in Table 3. Measurement of procalcitonin was performed categorically and was found negative in 29 attacks, below 0.5 ng/mL in 11, ≥0.5–2 ng/mL in 6, 2–10 ng/mL in 3, and ≥10 ng/mL in 2. Overall success of treatment was 100% as none of patients died in association with febrile neutropenia. Success without modification was 58%. We applied various types of modifications on the initial treatment, totally 26 in 21 attacks. These modifications were performed averagely 4.23 ± 0.83 (3–5) days after starting antimicrobial treatment. Control of fever and recovery of neutropenia in the whole group occurred after 2.0 ± 2.8 (1–16) and 6.1 ± 3.6 (1–19) days, respectively, and we observed these time periods were similar for lymphoproliferative and solid tumor patients (P = 0.586, 0.294), respectively. Mean duration of hospitalization time for all attacks in study group was 8.5 ± 5.4 (1–26) days while it was 10.5 ± 3.5 (5–24) days for duration of antibiotic use. Duration of hospitalization was longer for lymphoproliferative tumor group when compared with solid tumor patients [10.9 ± 4.3 (6–19) days vs. 8.9 ± 5.9 (1–26) days] (P = 0.047). Duration of antibiotic treatment was shorter in the group without treatment modification in contrast to treatment modification group [7.8 ± 1.7 (5–12) days vs. 10 ± 4.5 (9–24) days] (P = 0.001). Hospitalization duration was shorter in the group without treatment modification in contrast to treatment modification group [7.0 ± 3.5 (5–15) days vs. 10.0 ± 6.7 (4–26) days] (P < 0.001). There was no difference between lymphoproliferative tumor patients and solid tumor patients concerning duration of antibiotic
TABLE 2 Proinflammatory Cytokine Levels During Febrile Neutropenia Attacks in Nonleukemic Pediatric Cancer Patients Proinflammatory cytokines
Median
Ranges (of the samples)
IL-6 (pg/mL) IL-8 (pg/mL) sTNFRII (pg/mL) sIL-2R (ng/mL) CRP (mg/dL)
94.68 125.91 628.67 12.07 1.8
6.65–1729.97 18.16–3362.44 131.27–2894.34 3.09–42.58 0–18.3
C Informa Healthcare USA, Inc. Copyright
D. G. Karakurt et al.
TABLE 3 Comparison of Proinflammatory Cytokine Levels with ANC, Tumor Type, Age Groups and Infection Definition IL-6 IL-8 sTNFRII sIL-2R CRP (pg/mL) (pg/mL) (pg/mL) (ng/mL) (mg/dL)
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Features Absolute neutrophil count ≤ 100 (n = 24) > 100 (n = 26) P Tumor type Lymphoproliferative (n = 15) Solid (n = 35) P Age group ≤5 years (n = 20) >5 years (n = 30) P Type of infection documentation Fever of unknown origin (n = 29) Clinically proven infection (n = 15) Microbiologically proven infection (n = 6) P
327.18 63.16 0.01
524.13 175.01 0.01
1000.33 757.88 0.77
17.04 12.95 0.77
6.06 3.72 0.77
97.40 229.53 1.00
192.30 406.99 0.53
576.79 1001.74 0.53
17.46 13.83 0.06
3.84 5.27 1.00
268.37 137.57 0.38
429.65 284.54 0.38
1010.95 783.13 0.14
17.37 13.28 0.38
4.55 5.04 0.77
158.42 91.03 589.16 0.11
316.38 169.13 902.90 0.10
949.43 874.50 510.28 0.30
13.12 14.46 24.75 0.34
5.41 3.70 4.93 0.06
treatment [10.5 ± 2.8 (6–17) days vs. 10.6 ± 3.9 (5–24) days] (P = 0.917). We assessed the effects of inflammatory determinants over parameters of treatment response such as presence of modification, neutropenia recovery time, duration of fever control, duration of antibiotics and found that IL-8 levels were higher in the treatment modification group when compared with group without treatment modification (21 attacks, 464.53 pg/mL vs. 29 attacks, 254.28 pg/mL) (P = 0.048). There was no significant difference between groups without treatment modification and with treatment modification when IL-6, sTNFRII, sIL-2R, and CRP are noted (137.77 pg/mL, 979.28 pg/mL, 13.11 ng/mL, 3.75 mg/dl vs. 261.87 pg/ml, 729.22 pg/ml, 17.41 ng/ml, 6.35 mg/dL, P = 0.25; 0.25; 0.56; 0.25). Evaluation of effects of proinflammatory cytokine levels over parameters of treatment response is shown in Table 4. Fever control time, neutropenia recovery time, TABLE 4 Evaluation of Proinflammatory Cytokine Levels in Association with Neutropenia, Fever, Treatment and Hospitalization Durations Treatment response parameters Neutropenia period ≤7 days (n = 37) >7 days (n = 13) P Duration of fever ≤3 days (n = 39) >3 days (n = 11) P Treatment period (antibiotics) ≤7 days (n = 8) >7 days (n = 42) P Hospitalization period ≤7 days (n = 24) >7 days (n = 26) P
IL-6 (pg/mL)
IL-8 (pg/mL)
sTNFRII (pg/mL)
sIL-2R (ng/mL)
CRP (mg/dL)
45.92 118.17 0.19
97.53 172.72 0.38
628.41 1072.19 0.50
13.55 9.87 0.60
1.80 3.78 0.95
81.96 142.39 0.02
97.53 472.11 0.01
423.61 674.03 0.04
11.14 19.20 0.04
1.60 9.3 0.04
137.61 90.20 0.81
179.12 125.91 0.45
534.42 642.39 0.59
19.95 10.73 0.12
2.08 1.84 0.36
115.52 45.54 0.20
142.02 97.28 0.50
673.78 610.97 0.11
11.25 13.46 0.27
2.17 1.84 0.68
Pediatric Hematology and Oncology
Proinflammatory Cytokines in Febrile Neutropenia
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
hospitalization period and duration of antibiotic treatment were as 2.6 ± 1.4, 6.1 ± 3.7, 9.1 ± 4.8, and 10.0 ± 3.1 days in the group with procalcitonin level ≤1 ng/mL, and 5.8 ± 5.3, 5.9 ± 3.6, 11.3 ± 7.5, and 12.7 ± 4.6 days in the group procalcitonin >1 ng/mL, respectively (P = 0.014, 0.388, 0.382, 0.074). In multivariate analysis (logistic regression and linear regression tests) analysis of effects of IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels do not significantly contribute neutropenia duration (P = 0.104, 0.522, 0.701, 0.887, 0.211, 0.870), antibiotic treatment period (P = 0.423, 0.664, 0.120, 0.966, 0.935, 0.354), and hospitalization period (P = 0.913, 0.454, 0.275, 0.092, 0.408, 0.551), respectively. We also defined that IL-6, IL-8, and sTNFRII levels do interfere with duration of fever while sIL-2R, CRP, PCT levels do not (P = 0,012; 0,003; 0,051; 0,150; 0,494; 0,154). DISCUSSION A total of 70% of pediatric cancer patients have increased risk for invasive bacterial disease at first presentation. A total of 25% of them develop severe sepsis. Also, 3% of these high-risk patients eventually die. It is still not possible to predict risk of serious bacterial infection during referral to hospital for febrile neutropenia. There are some studies to determine risk of serious bacterial infection in adult patients but these studies are seldom conducted in pediatric patient groups [9–12]. We aimed to define risk of serious bacterial infection in pediatric cancer patients using clinical and laboratory parameters at first presentation of febrile neutropenia. There is still not an optimal antimicrobial treatment regimen for treatment of febrile neutropenia. Characteristics of the patient group are the most important factor influencing treatment plan. Heaviness and duration of neutropenia and presence of additional complications should be considered while deciding on empirical treatment in the beginning [13]. Studies investigating piperacillin/tazobactam monotherapy for children have been popular in recent years. In a randomized study, which was conducted in our center, we compared piperacillin/tazobactam with cefepime and success rates without modification were 56% and 48%, respectively [8]. Piperacillin/tazobactam monotherapy is the standard empirical treatment regimen used for study group in the current study. Success rate without modification is 58% in our study. Most preferred modification was adding glycopeptides to therapy amongst 26 attacks that were modified (38.4%). The main inquiry in our study was about possibility of clinical and laboratory findings as strong indicators of serious bacterial infection during febrile neutropenia attacks. “Serious infection complication” definition includes “death as a result of infection during sepsis and/or shock, bacteremia or fungemia, or febrile neutropenic attack.” It seems to be rational to evaluate clinical findings in the scope of serious infection complication as rate of microbiologically proven infection is low in our study group. We defined in our study that; tumor type, ANC at presentation and patients’ age groups didn’t play a key role in influencing microbiologically and clinically proven infection attacks. We also showed that these parameters did not affect fever control time, hospitalization, and treatment periods as noted in several other studies [13, 14]. There are limited studies evaluating importance of CRP and proinflammatory cytokine levels in detecting bacterial infection and clinical progress during febrile neutropenia attack. Persson et al. [15] investigated CRP, Serum Amyloid A (SAA), PCT, and IL-6 levels’ effect on clinical progress in 101 febrile neutropenic attacks of 79 patients. In the same study, it was shown in complicated (fever >3 days, clinically proven infections and bacteremias) patients that PCT and IL-6 levels were significantly high but CRP and SAA levels did not differ. Hodge et al. [16] investigated CRP, IL-2, IL4, IL-5, IL-10, and IL-8 levels concurrently in febrile neutropenic pediatric cancer C Informa Healthcare USA, Inc. Copyright
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
D. G. Karakurt et al.
patients. Bacteremia was detected in 8 of 31 attacks and they observed higher levels of IL-5 and IL-8 when compared with nonbacteremic group, also did not find any association between CRP, other cytokine levels and blood culture positivity. In another study, conducted with 85 febrile neutropenic pediatric patients, CRP, IL-8 and monocyte chemotactic protein were found to be significantly high in the first 24 hours of fever in bacteremic/septisemic patients [17]. Engel et al. [18] detected increase in PCT levels in the first 32 hours of fever in patients with clinically or microbiologically proven infection. They also found PCT levels were lower in patients with fever of unknown origin when compared with proven infection. Pihusch et al. [19] showed elevation of CRP, PCT, and IL-6 levels following infectious complications in 350 allogeneic stem cell transplantation patients. Only using PCT levels were reasonable to differentiate infection from transplant related complications. The cut-off value of serum PCT, that was used to predict infections with or without bacteremia, was 0.5–1.3 ng/mL (sensitivity 44–88%, specificity 61–88%) in several studies. It was reported that cut off value of serum PCT levels in the first 24 hours of febrile neutropenia could indicate low risk of serious infection if 2 ng/mL [20]. Conversely, Bont et al. [21] found no difference in serum PCT levels between bacteremic and nonbacteremic patients. But the noteworthy point of this study is bacteremic episodes depended on coagulase negative staphylococci and these microorganisms did not cause a significant increase in PCT levels [22, 23]. Also, Fleischhack et al. [24] and Svaldi et al. [25] showed higher increase of PCT in Gram (–) infections compared to Gram (+). Higher levels of IL-6 and IL-8 in ANC ≤100 /mm3 patients in our study could be a cautionary signal for serious bacterial infection in this group. We found no difference between ANC >100 /mm3 and ANC ≤100 /mm3 groups in the rate of clinically and microbiologically proven infections, so we decided to take these two parameters together as an unfavorable prognostic factor during assessment of risk factors. CRP levels, interestingly, were found lower in clinically and microbiologically proven infections when compared to attacks with UFO in our study. We analyzed treatment response parameters aiming to assess IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels’ effect during first presentation. It is not possible to compare our study results with the other studies’ results mentioned above as we did not plan to collect proinflammatory cytokine levels sequentially. We only defined significantly high IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels studied from patients with fever >3 days through the parameters above. IL-8 levels were higher in the treatment modification group when compared with group without treatment modification. We want to emphasize that duration of fever >3 days an indication for treatment modification. Additionally, we observed in multivariate analyses that only IL-6, IL-8, and sTNFRII levels affected duration of fever. It is noted in some studies that serum sIL-2R level increases in viral infections, tuberculosis and posttransplant infections [26–28]. Fleischhack et al. [24] showed major increase in serum sIL-2R level during Gram negative bacteremia of pediatric cancer patients. Studies involving sTNFIIR showed increase in serum levels especially during septicemia and pneumonia [29–31]. Moreover, sTNFIIR levels stay high for the whole fever episode and sometimes during convalescence, and this makes it insufficient for evaluating episodes’ progress and treatment response though it is useful at the time of diagnosis. In our study, considering need for modification in patients who are treated empirically and have uncontrolled fever for >72 hours, it is possible to predict outcome of febrile attacks of patients with high-proinflammatory cytokines at admission. This may be indirectly applied to the fact that duration of antibiotics and hospital duration in the treatment modification group was longer than the no treatment modification group. Our data suggest that Pediatric Hematology and Oncology
Proinflammatory Cytokines in Febrile Neutropenia
duration of fever does not provide insight into how serious the infection might be or the risk of sepsis. In conclusion, we did not find IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels as significant predictors of serious infectious complication in pediatric cancer patients during admission for febrile neutropenia, but our study data showed that IL-6, IL-8, sTNFRII, sIL-2R, CRP, and PCT levels during admission were high in patients with fever >3 days. These findings might be helpful to prevent unnecessary use of proinflammatory cytokines on routine clinical practice.
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
Declaration of Interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
REFERENCES [1] Santoloya ME, Alvarez AM, Avil´es CL, et al. Early hospital discharge followed by outpatient management versus continued hospitalization of children with cancer, fever, and neutropenia at low risk for invasive bacterial infection. J Clin Oncol. 2004;22:3784–3789. [2] Anoop P, Anjay MA. Febrile neutropenia: transition towards a risk-directed approach. Arch Dis Child. 2007;92:467–468. [3] Santolaya ME, Alvarez AM, Avil´es CL, et al. Prospective evalution of a model of prediction of invasive bacterial infection risk among children with cancer, fever and neutropenia. Clin Infect Dis. 2002;35:678–683. [4] H¨artel C, Deuster M, Lehrbecher T, Schultz C. Current approaches for risk stratification of infectious complications in pediatric oncology. Pediatr Blood Cancer. 2007;49:767–773. [5] Freifeld AG, Bow EJ, Sepkowitz KA, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis. 2011;52:56–93. [6] Wicki S, Keisker A, Aebi C, et al. Risk prediction of fever in neutropenia in children with cancer: a step towards individually tailored supportive therapy. Pediatr Blood Cancer 2008;51: 778–783. [7] C ¸ orapcıo˘glu F, Sarıalio˘glu F, Olgun N, Uysal KM. Analysis of 136 febrile neutropenic episodes in children with cancer: evaluation of treatment effectiveness and cost. Pediatr Hematol Oncol. 2004;21:535–543. [8] C ¸ orapcıo˘glu F, Sarper N, Zengin E. Monotherapy with piperacillin/tazobactam versus cefepime as empirical therapy for febrile neutropenia in pediatrıc cancer patients: a randomized comparison. Pediatr Hematol Oncol. 2006;23:177–186. [9] Ahmed N, El-Mahallawy HA, Ahmed IA, et al. Early hospital discharge versus continued hospitalization febrile cancer patients with prolonged neutropenia. Pediatr Blood Cancer. 2007;49: 786–792. [10] Klastersky J, Paesmans M, Rubenstein EB, et al. The Multinational association for supportive care in cancer risk index: a multinational scoring system for identifiying low-risk febrile neutropenic patients. J Clin Oncol. 2000;18:3038–3051. [11] Hann I, Viscoli C, Paesmans M, et al. A comprasion of outcome from febril neutropenic episodes in children compared with adults: result from four EORTC studies. International Antimicrobial Therapy Cooperative Group (IATCG) of the European Organization for Research and Treatment of Cancer (EORTC). Br J Haematol. 1997;99:580–588. [12] Boragina M, Patel H, Reiter S, Dougherty G. Management of febrile neutropenia in pediatric oncology patients: a Canadian survey. Pediatr Blood Cancer. 2007;48:521–526. [13] Duzova ¨ A, Kutluk T, Kanra G, et al. Monotherapy with meropenem versus combination therapy with piperacillin plus amikacin as empiric therapy for neutropenic fever in children with lymphoma and solid tumors. Turk J Pediatr. 2001;43:105–109. [14] Uygun V, Karasu GT, Ogunc D, et al. Piperacillin/tazobactam versus cefepime for the empirical treatment of pediatric cancer patients with neutropenia and fever: a randomized and open-label study. Pediatr Blood Cancer. 2009;53:610–614. [15] Sudhoff ¨ T, Gigounidis A, Karthaus M. Evaluation of neutropenic fever: value of serum and plasma parameters in clinical practice. Chemotherapy. 2000;46:77–85. C Informa Healthcare USA, Inc. Copyright
Pediatr Hematol Oncol Downloaded from informahealthcare.com by Nyu Medical Center on 05/13/15 For personal use only.
D. G. Karakurt et al.
[16] Hodge G, Osborn M, Hodge S, et al. Rapid simultaneous measurement of multiple cytokines in childhood oncology patients with febrile neutropenia: increased interleukin (IL)-8 or IL-5 correlates with culture-positive infection. Br J Haematol. 2005;132:247–248. [17] El-Maghraby SM, Moneer MM, Ismail MM, et al. The diagnostic value of C-reactive protein, Interleukin-8 and monocyte chemotactic protein in risk stratification of febrile neutropenic children with hematologic malignancies. Pediatr Hematol Oncol. 2007;29:131–136. [18] Engel A, Steinbach G, Kern P, Kern WV. Diagnostic value of procalcitonin serum levels in neutropenic patients with fever: comparison with interleukin-8. Scand J Infect Dis. 1999;31:185–189. [19] Pihusch M, Pihusch R, Fraunberger P, et al. Evaluation of C-reactive protein, interleukin-6, and procalcitonin levels in allogeneic hematopoietic stem cell recipients. Eur J Haematol. 2006;76:93–101. [20] Frei CR, Hampton SL, Burgess DS. Influence of culture site-specific MIC distributions on the pharmacokinetic and pharmacodynamic properties of piperacillin/tazobactam and piperacillin: a data analysis. Clin Ther. 2006;28:1035–1040. [21] de Bont ES, Vellenga E, Swaanenburg J, Kamps W. Procalcitonin: a diagnostic marker of bacterial infection in neutropenic cancer patients with fever? Infection. 2000;28:398–400. [22] Giamarellou H, Giamarellos-Bourboulis EJ, Repoussis P, et al. Potential use of procalcitonin as a diagnostic criterion in febrile neutropenia: experience from a multicentre study. Clin Microbiol Infect. 2004;10:628–633. [23] Persson L, Engervall P, Magnuson A, et al. Use of inflammatory markers for early detection of bacteraemia in patients with febrile neutropenia. Scand J Infect Dis. 2004; 36: 365–371. [24] Fleischhack G, Kambeck I, Cipic D, et al. Procalcitonin in paediatric cancer patient; its diagnostic relevance is superior to that of C reactive protein, interleukin 6, interleukin 8, soluble 2 receptor and soluble tumour necrosis factor receptor II. Br Haematol. 2000;111:1093–1102. [25] Svaldi M, Hirber J, Lanthaler AL, et al. Procalcitonin-reduced sensitivity and specificity in heavily leucopenic and immunosuppressed patients. Br J Haematol. 2001;115:53–57. [26] Shitrit D, Izbicki G, Bar-Gil Shitrit A, et al. Role of soluble interleukin-2 receptor levels in patients with latent tuberculosis. Lung. 2006;184:21–24. [27] Izzo F, Curley S, Maio P, et al. Correlation of soluble interleukin-2 receptor levels with severity of chronic hepatitis C virus liver injury and development of hepatocellular cancer. Surgery. 1996;120:100–105. [28] Liem LM, van Houwelingen HC, Goulmy E. Serum cytokine levels after HLA-identical bone marrow transplantation. Transplantation. 1998;66:863–871. [29] Van der Poll T, Jansen J, van Leenen D, et al. Release of soluble receptors for tumor necrosis factor in clinical sepsis and experimental endotoxemia. J Infect Dis. 1993;168:955–960. [30] Ertel W, Scholl FA, Gallati H, et al. Increased release of soluble tumor necrosis factor receptors into blood during clinical sepsis. Arch Surg. 1994;129:1330–1336. [31] De Werra I, Jaccard C, Corradin SB, et al. Cytokines, nitrite/nitrate, soluble tumor necrosis factor receptors, and procalcitonin concentrations: comparisons in patients with septic shock, cardiogenic shock, and bacterial pneumonia. Crit Care Med. 1997;25:607–613.
Pediatric Hematology and Oncology
