Intravenous Ceftazidime-Avibactam in Extremely Premature Neonates With Carbapenem-Resistant Enterobacteriaceae: Two Case Reports

Abstract

The increasing use of carbapenems has contributed to a notable distribution of carbapenem-resistant Enterobacteriaceae (CRE). Recently, the incidence of CRE-associated infections is increasing significantly in NICUs, which pose a grave challenge to clinical treatment. We report 2 cases of IV ceftazidimeavibactam use to treat CRE infections in extremely premature neonates. The first case was diagnosed with bacteraemia and meningitis and the second one was diagnosed with bacteraemia only. Due to the lack of neonatal-specific information for IV ceftazidime-avibactam, the usual pediatric dose (62.5 mg/kg/dose every 8 hours) was used in these patients. Clinical cure occurred in these 2 patients. Although blood cultures became sterile after starting ceftazidime-avibactam in the second case, the patient died, presumably owing to sepsis or various causes, such as prematurity and chronic lung disease. Large and randomized studies are necessary to ensure the safety and efficacy of IV ceftazidime-avibactam for the treatment of neonates with sepsis caused by multidrug resistant organisms.

Introduction

Neonatal sepsis remains one of the main causes of neonatal morbidity and mortality, especially among premature infants.^1–3 Although several modalities for early diagnosis and treatment of neonatal sepsis have been established, the incidence of late-onset sepsis (LOS) has not been reduced and subsequent antibiotic resistance has become more prevalent.^4–6 A study conducted in 2018 in a large tertiary referral hospital in Saudi Arabia found that the incidence of early-onset sepsis (EOS) in the first 3 days of life was approximately 1.7%, whereas that of LOS occurring after 3 days of life, was approximately 18%.⁷ Group B Streptococcus, followed by Escherichia coli, were the most common causes of EOS. Late-onset sepsis was mostly due to coagulase-negative Staphylococcus.^7,8 Gram-negatives, E coli, Klebsiella species, Enterobacter cloacae and Pseudomonas aeruginosa were the most commonly isolated pathogens (18.5% each), whereas Serratia marcescens, Acinetobacter species and Kluyvera ascorbata were less common.⁷ The incidence of CREs on average is 0.27% per year.⁷

As a result, increasing bacterial resistance has contributed to the creation of multidrug resistant organisms (MDROs). In a recent study, about 20% of episodes of Gram-negative sepsis were caused by Gram-negative MDROs, which contributed to the reintroduction of antibiotics without a clear safety record.⁹ Infections with carbapenem-resistant Enterobacteriaceae (CRE) have emerged as a critical threat to public health, especially those due to carbapenem-resistant Klebsiella pneumonia.¹⁰

Limited data existed about treatment of CRE infection in neonates. Indeed, a combination of available regimens (polymyxins, aminoglycosides, tigecycline, and carbapenems) was restricted due to inferior efficacy, safety, resistance, and suboptimal pharmacokinetics.^10–12

Yet, no published case reports or studies have evaluated the use of ceftazidime-avibactam in extremely premature neonates with CRE. Herein, we report 2 unique cases of premature neonates who received ceftazidimeavibactam for the treatment of CRE.

Case 1

A 920-g 27-week female infant was born to a 16-year-old female by cesarean section. Apgar scores were 5 and 7 at 1 and 5 minutes, respectively. The infant had respiratory distress and was transported to a NICU Level-III; surfactant was given by the endotracheal route and the infant was placed on a mechanical ventilator. An umbilical venous catheter (UVC) was placed. Blood culture and a CBC were drawn, and penicillin G and gentamicin therapies were started. Antibiotics were discontinued on the third day of life (DOL) after a negative blood culture result. She was extubated on DOL 3. A brain ultrasound was also done and showed a grade I intraventricular hemorrhage. On DOL 6, cloxacillin (50 mg/kg/dose every 12 hours) and amikacin (15 mg/kg/dose every 36 hours) were started for suspected sepsis because of apnea, and lethargy. In addition, the UVC was removed. Blood tests showed that WBC count was elevated at 34.6 × 10⁹/L, neutrophils 58.4 %, lymphocytes 21.9%, and platelet (PLT) count was 239 × 10⁹/L. No C-reactive protein (CRP) was drawn. Antibiotics were discontinued at DOL 8 after a negative blood culture result.

On DOL 9, the peripherally inserted central catheter (PICC) line was inserted. On DOL 11, the infant developed episodes of apnea, thus required reintubation. Meropenem (40 mg/kg/dose every 8 hours) and vancomycin (15 mg/kg/dose every 12 hours) were started empirically due to suspected late onset of neonatal sepsis and full sepsis tests were conducted. Her laboratory results were as follows: WBC and CRP count were high 71.18 × 10⁹/L, 66.7 mg/L, respectively, neutrophils 80.6%, lymphocytes 5.1%, and PLT 311 × 10⁹/L. The blood culture showed K pneumonia sensitive to colistin, gentamicin, tigecycline, and resistant to cephalosporins, carbapenems, fluoroquinolones, amikacin, and cotrimoxazole. Lumbar cerebrospinal fluid (CSF) analysis was turbid, showed 1761 WBC count/mm³, and no red blood cells (RBCs). The CSF protein was 378.4 mg/dL and glucose was 1 mg/dL. The CSF culture grew K pneumonia and the isolate was sensitive to colistin and resistant to all other antibiotics. On DOL 15, the patient was diagnosed with meningitis and bacteremia. Based on culture and sensitivity results, meropenem and vancomycin were switched to IV colistin (50,000 IU/kg/dose every 8 hours) and IV gentamicin (4 mg/kg/dose every 24 hours) and the PICC was removed. On DOL 17, day 2 of IV colistin and gentamicin therapy, the patient did not show any clinical improvement. Thus, blood culture and CBC were repeated. Blood tests showed WBC count 21.87 × 10⁹/L, PLT 15 × 10⁹/L, and blood culture still showed K pneumonia. Due to the instability of the patient's condition, thrombocytopenia, and prolonged coagulation profile: prothrombin time was 26.2 seconds, international normalized ratio was 2.5, partial thromboplastin time was 52.6 seconds, we decided not to repeat a lumbar puncture. Additionally, due to severe thrombocytopenia and lack of clinical improvement as well as poor penetration of IV gentamicin in CSF, both ceftazidime-avibactam and fosfomycin sensitivity were requested.

The organism was reported to be sensitive to ceftazidime-avibactam and fosfomycin. Fosfomycin has better CSF penetration; however, it is a non-formulary medication in our hospital. Therefore, the infectious disease team recommended to add IV ceftazidime-avibactam (62.5 mg/kg/dose every 8 hours) to the colistin therapy and discontinue gentamicin with close monitoring of renal function. After 48 hours, the repeated blood culture showed growth of methicillin-resistant Staphylococcus epidermidis (MRSE) only. Thus, on DOL 19, IV linezolid (10 mg/kg/dose every 8 hours) was added to colistin and ceftazidime-avibactam. Intravenous linezolid was continued for 54 days due to the persistence of MRSE infection. The total duration of colistin was 14 days and that of ceftazidime-avibactam was 21 days. The results of cultures, organisms, and sensitivity are shown in Table 1. No adverse renal effects were observed during the entire therapy. On DOL 48, the infant was extubated. On DOL 75, the nasal cannula was removed, and the infant was maintained on room air. The infant was discharged at the corrected age of 42 weeks plus 5 days with a weight of 1960 grams.

Table 1.

Clinical Course of Neonate Described in Case 1

DOL	Culture	Organism	Antibiotics
0	Blood	No growth	Penicillin G, gentamicin for 2 days
6	Blood	No growth	Cloxacillin, amikacin for 2 days
11	Blood CSF	Klebsiella pneumoniae (CRE) K pneumoniae (CRE)	Meropenem, vancomycin for 3 days Colistin, gentamicin for 2 days, then colistin, ceftazidime-avibactam for 12 days, then ceftazidime-avibactam for 9 days
19	Blood	Staphylococcus epidermidis (MRSE)	Linezolid for 2 days, then linezolid, rifampicin for 52 days
26	Blood	Vancomycin-resistant enterococci
28	Blood	S epidermidis (MRSE)
29	Bronchial	No growth
32	Blood	S epidermidis (MRSE)
44	Blood	No growth

CRE, carbapenem-resistant Enterobacteriaceae; CSF, cerebrospinal fluid; DOL, day of life; MRSE, methicillin-resistant Staphylococcus epidermidis

Case 2

A 925-g 28-week female infant was born, as a triplet, to a 28-year-old female by cesarean section. Apgar scores were 4 and 6 at 1 and 5 minutes, respectively. The infant had respiratory distress and was transported to a NICU Level-III; surfactant was given by the endotracheal route and the infant was placed on a mechanical ventilator. A UVC was placed. Blood culture and a CBC were drawn, and ampicillin and gentamicin therapies were started empirically and were administered for 48 hours. She was extubated on DOL 6, the UVC was removed and a PICC line was inserted. On DOL 11, cloxacillin (50 mg/kg/dose every 12 hours) and amikacin (15 mg/kg/day every 36 hours) were started for suspected sepsis because of increased oxygen requirements and episodes of desaturation. Blood tests showed WBC count was high at 29.82 × 10⁹/L, neutrophils 47.2%, and lymphocytes 13.7%, PLT count was 50 × 10⁹/L, and no CRP was drawn. The antibiotic was changed to IV linezolid instead of cloxacillin (10 mg/kg/dose every 8 hours), when blood cultures showed MRSE, and this antibiotic was continued for a total of 21 days due to the persistent MRSE infection. Due to the MRSE positive blood cultures, amikacin was continued for 3 days and then discontinued. On DOL 19, PICC line was removed. On DOL 27, day 15 of linezolid therapy, blood culture and CBC were obtained. Piperacillin-tazobactam (100 mg/kg/dose every 8 hours) was added because the patient had apnea and required reintubation. Blood tests showed WBC count 18.64 × 10⁹/L, PLT 342 × 10⁹/L, and blood culture was negative for bacterial growth. However, piperacillin-tazobactam continued for 7 days to treat the presumed LOS.

On DOL 37, sepsis tests were conducted because the patient was experiencing an increase in the frequency of bradycardia and desaturation episodes. The infant was shifted from a nasal canal to bilevel positive airway pressure. Thus, cefepime (50 mg/kg/dose every 12 hours) was started. Her laboratory data were as follows: WBC count and CRP were high 82 × 10⁹/L, >191 mg/L, respectively, neutrophils 90%, lymphocytes 7.8%, and PLT was low 79 × 10⁹/L. The Gram stain showed Gram-negative bacilli. Thus, amikacin (15 mg/kg/dose every 24 hours) was added. The blood culture grew extended-spectrum β-lactamases producing K pneumonia. Initially, K pneumonia was sensitive to meropenem and amikacin but resistant to cefepime and gentamicin. Based on culture sensitivity results, cefepime was changed to meropenem (40 mg/kg/dose every 8 hours). Furthermore, the patient required intubation and invasive ventilation due to continued apnea. Despite the maximum meropenem and amikacin dose, and appropriate post- and pre-level of amikacin: 24.6 mcg/mL and 4.03 mcg/mL, respectively, repeated blood cultures on DOL 39, 40, 42, 48, and 51 remained positive with K pneumonia.

Later, on DOL 54, K pneumonia isolated from blood culture became resistant to meropenem. The PLT count was extremely low 7 × 10⁹/L, WBC count was 5.2 × 10⁹/L, and the CRP on this day was elevated at 172 mg/L. Thus, additional sensitivities were requested, including ceftazidime-avibactam to determine resistance patterns of the organism. The organism was reported to be sensitive to ceftazidime-avibactam. Therefore, we decided to add IV ceftazidime-avibactam (62.5 mg/kg/dose every 8 hours) to the amikacin therapy and discontinued meropenem with close monitoring of renal function. On DOL 60, day 4 of ceftazidime-avibactam therapy, a blood culture was obtained and showed no growth. Her laboratory tests showed a significant improvement in WBC count 9.72 × 10⁹/L and PLT 60 × 10⁹/L. Thus, amikacin was discontinued. The results of cultures, organisms, and sensitivity are shown in Table 2. During therapy, serum creatinine increased from base line 87.1 μmol/L to 107.2 μmol/L on DOL 60. Thus, we changed the ceftazidime-avibactam frequency to every 24 hours. The patient died on DOL 61.

Table 2.

Clinical Course of Neonate Described in Case 2

DOL	Culture	Organism	Antibiotics
0	Blood	No growth	Ampicillin, gentamicin for 2 days
11	CSF A Blood	No growth Staphylococcus epidermidis (MRSE)	Cloxacillin, amikacin for 1 day then linezolid, amikacin for 2 days, then linezolid for 19 days
13	Blood	No growth	No change
15	Blood	S epidermidis (MRSE)	No change
22	Blood	No growth	No change
27	Blood	No growth	Piperacillin-tazobactam for 7 days
37	Blood	Klebsiella pneumoniae (ESBL)	Cefepime for 1 day then cefepime, amikacin for 2 days, then meropenem, amikacin for 15 days
56	Blood	K pneumoniae (CRE)	Ceftazidime-avibactam, amikacin for 3 days, then ceftazidime-avibactam for 2 days
60	Blood	No growth	None

CRE, carbapenem-resistant Enterobacteriaceae; CSF, cerebrospinal fluid; DOL, day of life; ESBL, extended-spectrum β-lactamases; MRSE, methicillin-resistant Staphylococcus epidermidis

Discussion

We reported 2 unique cases in which ceftazidimeavibactam 62.5 mg/kg/dose every 8 hours was administered to treat infections caused by MDROs in premature neonates. Ceftazidime-avibactam, a novel combination, contains a third-generation cephalosporin and non-β-lactam β-lactamase inhibitor.¹³ Although avibactam protects ceftazidime from degradation by a variety of β-lactamases, avibactam does not decrease the activity of ceftazidime against ceftazidime-susceptible organisms.¹³ Initially, ceftazidime-avibactam was used to treat complicated intra-abdominal infections, urinary tract infections (UTIs) including pyelonephritis, and hospital-acquired as well as ventilator-associated pneumonia in adults; however, it is now being approved for pediatric patients aged 3 months and older for the treatment of complicated UTI and complicated intra-abdominal infections in combination with metronidazole.^14,15 Limited data were available on the use of ceftazidime-avibactam in critically ill pediatric patients and 1 recent case report about using ceftazidime-avibactam in a premature infant with pandrug-resistant K pneumoniae UTI was published in Turkey.¹⁶ They used 50 mg/kg/dose (ceftazidime 40 mg/kg and avibactam 10 mg/kg) every 8 hours for 10 days with transient glucosuria and normal renal function test.¹⁶ Tamma et al¹⁷ published a case study of successful treatment of persistent Burkholderia cepacia in a 2-month-old term infant. Vargas et al¹⁸ outlined a 14-year-old boy's successful treatment of septic shock due to K pneumonia carbapenemase-multidrug resistant. Recently, Hobson et al¹⁹ reported their successful experience in the treatment of bacteraemia due to New Delhi metallo-β-lactamases-1 producing Morganella morganii with aztreonam and ceftazidime-avibactam combination in a 3-year-old pediatric patient with hematologic malignancy. Iosifidis et al²⁰ published a report of successful treatment of extensively drug-resistant or pan drug-resistant K pneumoniae in children <5 years of age.

Treatment of neonatal sepsis caused by CRE is a real challenge for neonatologists and infectious disease physicians due to the serious complications and lack of choices for treatment.²¹ Indeed, the latest recommendations for CRE treatment in pediatrics and neonates are driven by case reports and case series with substantial patient variation and evaluated treatment options.¹⁸ Overall, to treat CRE, available data supports a combination therapy with at least 2 agents, given significant reductions in mortality with this approach compared with monotherapy, with cited mortality rates as low as 13%.^22–25 In adults, polymyxin or tigecycline combined with aminoglycosides, fluoroquinolones, minocycline, and sulfamethoxazole-trimethoprim are preferred choices and should be guided by clinical source of infection, susceptibility patterns, and expert consultation based on drug susceptibility tests, the antibiotic's ability to cross the blood–brain barrier, and capacity to reach the minimum inhibitory concentration in the brain.²⁶ In pediatrics, tigecycline is not FDA-approved. Additionally, fluoroquinolones, minocycline and trimethoprim/sulfamethoxazole are not usually preferred in neonates due to potential side effects.^27,28

In our NICU, the appropriate selection of antibiotics for the treatment of proven sepsis (initiation, duration, and discontinuation) depends on an interprofessional team approach involving infectious disease experts, clinical pharmacists, and neonatologists. Taking into consideration various factors, such as the type and duration of previous antibiotics, source of infection, antimicrobial sensitivity results, presence of Gram-negative MDRO outbreak, persistent infection, hemodynamic stability of patients, the status of laboratory tests before initiation, pharmacokinetic properties, and availability of medications in the pharmacy. Our NICU guideline of neonatal sepsis requires that we start ampicillin and gentamicin as the first-line treatment of EOS and amikacin and cloxacillin for LOS. Sometimes the preliminary result of the blood culture is negative; however, the patient is still unstable or has developed new signs or symptoms of sepsis. Thus, in these cases, the physicians start broad-spectrum antibiotics, such as cefotaxime, meropenem and vancomycin.²⁹ Additionally, the stewardship program team rounds daily to review patient medications, as well as adjust the treatment course according to culture results, clinical conditions, and other laboratory results. The average monthly utilization of the most commonly used antibiotics in our NICU (day of therapy [DOT] per 1000 patient-days) as follows: ampicillin: 98.3 DOT per 1000 patient-days; gentamicin: 114.3 DOT per 1000 patient-days; amikacin: 48.2 DOT per 1000 patient-days; cloxacillin: 96.4 DOT per 1000 patient-days; cefotaxime: 48.2 DOT per 1000 patient-days; and meropenem, 128.6 DOT per 1000 patient-days.

In this report, we describe 2 premature neonates with CRE producing K pneumoniae. The first one was diagnosed with bacteraemia and meningitis, whereas the second one was diagnosed with bacteraemia. Far less is known about CRE epidemiology and risk factors in neonates and children. The risk factors associated with CRE infection in neonates include previous exposure to meropenem and ICU admission.³⁰ In our first case, due to poor penetration of aminoglycoside in CSF and based on drug susceptibility tests, the multidisciplinary team recommended anti-infective treatment with IV ceftazidime-avibactam combined with IV colistin. Ceftazidimeavibactam (Avycaz; Allergan, NJ) at a concentration of 12.5 mg/mL was administered over 60 minutes via either peripheral or central line. Although the blood culture became sterile after starting ceftazidime-avibactam in the second case, the patient died, probably owing to multiple factors, such as prematurity, sepsis, and/or chronic lung disease. No data were available regarding the most appropriate dose of ceftazidime-avibactam for the treatment of neonatal sepsis due to Gram-negative MDROs.

Renal function test results obtained before and after ceftazidime-avibactam initiation were studied. We noticed a significant increase in creatinine results in the second patient from 87.1 μmol/L to 107.2 μmol/L; however, the results were unchanged in the first patient. It is unknown whether the renal impairment in the affected patient is secondary to the medication or to the sepsis. A randomized controlled phase I trial tested the safety and tolerability of ceftazidime-avibactam in ill patients who are less than 18 years old. They found most of adverse effects are non-renal manifestations.³¹ Although ceftazidime-avibactam exhibits linear pharmacokinetics, it has low serum protein binding with 80% to 90% of the dose eliminated as an unchanged drug through the kidney.³¹ Thus, a dose adjustment is required in patients with renal impairment to avoid drug accumulation and toxicity.³¹ The pharmacokinetic analysis of ceftazidimeavibactam in 32 children was evaluated as part of a phase 1 safety and tolerability study.³¹ Newborns have different renal function than older infants or children. Therefore, the safety, effectiveness, and pharmacokinetics of ceftazidime-avibactam as an anti-infective treatment in neonates are not yet known.

This report remains a study of 2 cases collected retrospectively with short follow up till discharge or death time. Additionally, the studied medication was used in combination with other antibiotics to strengthen its efficacy. Another limitation, we were unable to measure the serum drug level and it is unknown whether the dose has to be adjusted according to serum drug concentration or not. Therefore, a large, well-designed, and prospective study is required to investigate the safety, efficacy, and the pharmacokinetic characteristics of ceftazidime-avibactam in premature infants. We thought during the COVID-19 crisis, the incidence of LOS would decline secondary to the application of visiting restrictions, reducing in the number of health practitioners who were on duty, wearing of face masks, and the emphasis on hand hygiene. To our knowledge, this is the first report describing ceftazidime-avibactam use in extremely premature neonates with CRE infections. Two different diagnoses were included in this report (meningitis and bacteraemia).

Conclusion

Based on our findings, we recommend that a large and randomized controlled trial should be conducted to determine the appropriate dose and safety of ceftazidime-avibactam for the treatment of infants with sepsis caused by Gram-negative MDROs.

ABBREVIATIONS

CBC

complete blood count

CRE

carbapenem-resistant Enterobacteriaceae

CRP

C-reactive protein

CSF

cerebrospinal fluid

DOL

day of life

DOT

day of therapy

EOS

early-onset sepsis

FDA

US Food and Drug Administration

ICU

intensive care unit

IV

intravenous

LOS

late-onset sepsis

MDROs

multidrug resistant organisms

MRSE

methicillin-resistant Staphylococcus epidermidis

NICU

neonatal intensive care unit

PICC

peripherally inserted central catheter

PLT

platelet

RBC

red blood cell

UTI

urinary tract infection

UVC

umbilical venous catheter

WBC

white blood cell