Survival rate in children with cancer has increased due to improved treatment options in recent years. However, some side effects related to treatment may develop. The most common side effect due to intensive chemotherapy in children with cancer is febrile neutropenia. Delay in treatment can lead to an increase in morbidity and mortality. The risk of developing infection in patients receiving chemotherapy is closely related to the duration and depth of neutropenia. The lower the neutrophil count, the higher the risk of developing infection. In addition, as the duration of neutropenia increases, the incidence rates of fungal and bacterial infections infections increase as well (8, 9).
In our study, 55 patients and 83 febrile neutropenia attacks were examined cross-sectionally. Patients with moderate and severe neutropenia were included in the study. 76% of the attacks had deep neutropenia. The median neutrophil level was 30 (0-530) / mm³. Patients had fever on the median 2nd day of neutropenia, and they were total neutropenic for an average of 11.6 days. Total duration of neutropenia was statistically significantly higher in the leukemia group than in the solid tumor group. In Alexander SW and et al. study, the presence of hypotension, tachypnea, hypoxia, emergence of new infiltration on chest X-ray, mental status changes, severe mucositis, vomiting or abdominal pain, focal infection, and other clinical indications requiring hospitalization placed the patient in the high risk group. In the presence of these findings, the patient is at high risk of developing serious medical complications (10).
The presence of every uncontrolled adverse condition (relapse, treatment refractory cases, during induction therapy), high- risk ALL and AML, consolidation or late intensification treatment, ANC < 100 / mm3 after high dose cytarabine treatment if expected neutropenia duration ≥ 7 days, toxic appearance (hypotension, shock, tachypnea, hypoxia, neurological changes), evidence of infection (pneumonia, cellulitis, abdominal pain and diarrhea, neurological changes), known MRSA colonization, previous bacteremia / sepsis history, development of mucositis after chemotherapy were included in the high risk group in our studyv ( 13). According to these criteria, 96% of the attacks in our study were high-risk.
Bone marrow suppression is different according to chemotherapy protocols. Since bone marrow suppression is higher in ALL and AML induction treatments, the risk of febrile neutropenia is higher (11).
There are some studies showing that the risk of febrile neutropenia increases during the induction period (11, 12). Yılmaz S and et al. reported that febrile neutropenia attacks were experienced most frequently during consolidation treatment in pediatric leukemia patients (13). In our study, the most common febrile neutropenia attack in leukemia patients developed during induction therapy (41%). Second most frequently FEN attack was observed during consolidation treatment (35%). The difference between studies may stem from multifactorial reasons including the difference in the chemotherapy protocols applied,, the type of solid or leukemia malignancy and the number of patients enrolled to these studies.
Mostly investigated and well known infection marker in cancer studies is CRP. However, there are some limitations in clinical practice. CRP rises within 24–48 hours and may be affected by the underlying malignancy and tissue damage (14–16). Since procalcitonin level starts to rise within 3–4 hours and reaches the highest level in 8–24 hours It is more advantageous than CRP (17, 18). Secmeer and et al. suggested that CRP and PCT levels in patients with febrile neutropenia attack were significantly higher than in nonfebrile patients (19). In our study, high CRP and PCT levels were detected in FEN patients with attack. The CRP, and PCT levels were found to be 41 (0.2–269) mg / lt, and 0.5 (0-156) ng / ml, respectively. In a study conducted, it was found that 41%, and 48% of the patients with microbiologically proven infections had CRP, and PCT positivities, respectively (20). In our study, high CRP and PCT levels were found in 9 (40%) of 22 attacks experienced by patients with microbiologically proven infections.
In many neutropenic patients, especially those taking steroids, the systemic inflammatory response is weakened, so the focus of infection may not be clearly identified. Most of the time, fever focus cannot be detected in FEN attacks in patients receiving chemotherapy (21).
Regular Kar et al. Reported that the frequency of clinically defined infections was found to be 40.5%. Among these, the most common focal mucositis has been identified (22). Özdemir ZC et al. reported that the frequency of clinically proven infections was found to be 35%, and the most common focus among them was mucositis, with a frequency of 32% (23). Clinically proven infection rate was found to be 77% in our study.
Mucositis is a common complication of cancer treatment. Mucositis risk is increased in hematological malignancies compared to solid tumors. Otmani et al. 60% of the patients with hematological malignancy developed oral mucositis, while 48% of the patients with solid tumors developed mucositis (24). In our study, the frequency of developing mucositis was not different between patients diagnosed with leukemia and solid tumors.
Although the laboratory conditions are very good in 60–70% of the cases with febrile neutropenia, the causative microorganism cannot be demonstrated (25, 26). Microbiologically proven infection rate in many studies ranges between 17–29% (27–30).
In some studies; the frequency of microbiologically proven infections varies between 16–36% (22, 31, 32). In our study, microbiologically proven infection was found during 22 (27%) of 83 attacks. The factors responsible for the infection differ according to the hospitals. While Gram-negative bacteria were the most common factors in febrile neutropenia attacks in the previous years, nowadays Gram-positive microorganisms are frequently observed (33). In studies conducted in our country, the detection rate of Gram-positive agents were reported as 69%, [Baysallar et al. (32)[, 64% [Özdemir et al. (23)], and 70%. [Akçay et al. (34)]
The cause of the increase in Gram-positive agents was the development of mucositis as a result of the use of chemotherapeutic agents such as cytosine arabinoside, deep and prolonged neutropenia attacks, long-term inserted intravenous catheters, protective treatment with fluoroquinolone and cotrimoxazole, use of antacid and histamine receptor blockers (35, 36). Demirkaya et al. found most frequently Coagulase-negative staphylococci during febrile neutropenia attacks in their study (37). In our study, bacterial growth was detected in 23% of blood cultures taken from the peripheral veins, while 47% of them were Gram- positive and 53% of them Gram- negative microorganisms. Gram- positive microorganism was grown in 20% of the cultures taken from the central catheter. No reproduction was detected in 70 % of these cases. Growth of Gram negative microorganisms was not observed in the blood culture taken from the central catheter. When we consider as a whole, in blood culture media of 55% of the patients demonstrating bacterial reproduction Gram- positive agents were grown.
Most frequently Coagulase- negative staphylococci were detected. Hann et al. reported that the growth rate of active microbial agents in the blood was lower in patients treated for solid tumors compared to leukemia patients (38). Delebarre M and et al., the growth rate of Gram-negative agents during FEN attacks was found to be higher in patients receiving leukemia treatment compared to those treated for solid tumors (39). In our study, no significant difference was found in the growth rate of Gram- negative and Gram -positive agents during FEN attacks between patients with leukemia and those with solid tumors.
Urinary tract infections are common in children. The importance of urinary tract infections (UTIs) in febrile neutropenic patients has not been clearly determined (25). Özdemir N and et al., UTI was found in 4% of patients during febrile neutropenia (40). The most frequently isolated microorganism is E.coli (41). In our study, urine culture was sent during 23% of the attacks. Bacterial growth occurred in 3 (79%) of the urine cultures sent. Candida albicans, Corynebacterium jeikeium, and Klebsiella pneumoniae were isolated separately from each one of them. The focus of fever was UTI in 4% of the attacks. The frequency of foci of fever detected was similar to the literature.
All high-risk patients with febrile neutropenia attack should receive inpatient treatment. A broad spectrum antibiotic effective against Gram-positive and Gram-negative agents, including Pseudomonas aeruginosa should be initiated empirically and in no time. This approach has decreased the infection-related mortality rates in FEN attacks in patients with cancer. Antibiotics that can be used in monotherapy can be antipseudomonal beta lactams, 4th generation cephalosporins or an antibiotic in the carbapenem group (42). It is recommended that a second antibacterial effective against Gram-negative microorganism or glycopeptide group antibiotherapy be initiated for patients with suspected resistant infection, who are clinically unstable and treated in centers with a high rate of resistant microorganisms. In one study, no difference was found between monotherapy with 4th generation cephalosporin and combined therapy containing aminoglycoside in terms of efficacy and safety (43).
Antipseudomonal penicillin and aminoglycoside; antipseudomonal cephalosporin and aminoglycoside; carbapenem and aminoglycoside are the most commonly used antibiotics in combination therapies. Combination therapy has advantages such as synergistic effect and prevention of resistance development, but the increase in side effects such as nephrotoxicity and ototoxicity due to aminoglycosides is especially seen in cisplatin, cyclosporine, and amphotericin B users (44). Kebudi R and et al., showed that the effectiveness of cefepime and ceftazidime monotherapies is comparable (45). Kamonrattana et al., detecte no significant difference between piperacillin / tazobactam monotherapy and ceftazidime / amikacin combination therapy, as for treatment response, duration of fever,neutropenia, and antibiotic use (46)
Ponraj M and et al., compared cefepime monotherapy with cefoperazone / sulbactam and amikacin combination treatment and they suggested no difference between the two groups in terms of efficacy and safety (47). In our study, for 19%, and 81% of the FEN attacks treatment with single and dual antibiotherapy were initiated, respectively. Cefepime was used most frequently as single antibiotherapy (93%). Cefepime and amikacin were used most frequently as dual antibiotherapy (42%). In Marín M and et al. study, empirically combined antibiotherapy was found to be more appropriate than single antibiotherapy in patients with hematological malignancies. In the same study, combination antibiotic treatment was found to be more effective than single antibiotic treatment in patients with solid tumors (48). In our study, no significant difference was found between the initiation of single and combined thearphy.
Since invasive fungal infection cannot be excluded in patients who do not respond to empirical antibiotic therapy and have refractory fever, empirical antifungal therapy is recommended. Caspofungin and liposomal amphotericin B are antifungals recommended for empirical therapy (42). In a study by Güneş et al. antifungal treatment was given to 20% of the patients experiencing FEN attacks and fluconazole was used most frequently as antifungal treatment (47%) (31). In a study conducted, it was revealed that the effectiveness of caspofungin and liposomal amphotericin B were similar (49). In our study, antifungal treatment was started for the management of 41% of the attacks. Mostly used antifungal agent was caspofungin(%58).
While the mortality rate due to febrile neutropenia was over 90% in the 1960s, it has now fallen below 5% as new studies and guidelines are implemented (50). In a study conducted in previous years, the mortality rate during the FEN attack was determined to be 3%, and race, age group, age of diagnosis, type of cancer and developing complications were found to be associated with the mortality rate (51). In Demirkaya et al. study, the mortality rate in FEN attack was found to be 1 percent. Bacteremia was found in one of the deceased patients, and no focus of infection was found in the other two (37). In our study, 3 patients (4%) who were examined due to FEN attacks died. Two of the patients were being treated for ALL and one for adrenocortical tumor. Growth of E.coli was detected in the blood culture of one of them. In the other two patients, presence of any microbiological infection was not confirmed.
The limitations of the study can be expressed as the small number of cases, the short follow-up period, the inability to standardize culturing methods, the simultaneous evaluation of solid tumors and leukemias, and the inability to separate leukemias according to their risk groups.