Antibiotics in Critically Ill Newborns and Children: Nephrotoxicity and Management During Renal Replacement Therapy

Epidemiology of Bacterial Infections in Newborns

Neonatal sepsis generally is classified as early onset (EOS) and late onset (LOS), according to timing of symptoms onset. This classification reflects important differences in pathogenesis: the EOS are due primarily to infections connected with delivery and vaginal contamination and the LOS are more severe, typically nosocomial infections and are associated with higher mortality resulting from resistant bacteria.

In EOS group B streptococcus (GBS) accounts for about 50% of positive cultures, followed by Escherichia coli in about 30% of the cases, particularly in very-low-birthweight newborns.

LOS are caused most often by gram-positive bacteria, in most cases coagulase-negative Staphylococcus, whereas gram-negative bacteria are far less common (around 20%) but associated with greater mortality.

Because of a high percentage of isolates resistant to more than one drug, most of neonatology practice guidelines suggest not to use a single drug alone as empiric therapy. The most rational approach is to know the local microbiologic epidemiology, as a guide to the proper prescription, with high level of surveillance to extended spectrum β-lactamase (ESBL)–producing gram-negative bacteria resistant to penicillins and cephalosporins and often also to fluoroquinolones and aminoglycosides.

Empiric Therapy in the Newborn: Risks and Adverse Effects

Although widely adopted in most neonatal ICUs, the choice to use antibiotics on an empiric basis has some relevant side effects mainly because of interference with gut colonization, appearance of resistant strains, and of Candida colonization until diffuse candidiasis. Early alteration of gut colonization can promote antibiotic resistance among normal commensal organisms just endowed in the neonatal gut. However, particularly in critically ill low- birthweight newborns, the administration of a combined antibiotic therapy is often mandatory.

Most of the neonatologists aiming at sparing wide-spectrum and new-generation antibiotics adopt a sort of stepwise increase in antibiotics starting from the oldest to the newest in case of no response and no cultural advice.

Carbapenems and third-generation cephalosporins are highly efficient in covering a broad spectrum of bacteria. On the other hand, their overuse may favor the emergence of ESBL-producing strains of gram-negative bacteria and invasive candidiasis, with a stringent correlation with the length of use of these drugs.

Agreement is broad regarding long-term antibiotic treatment in case of culture-proven sepsis in newborns, with antibiotic interruption after 72 hours from negative blood culture and clinical improvement. However, the optimal antibiotic administration duration for clinical sepsis without cultures positivity still is debated.

Some cohort studies in fact associate the prolonged duration of empiric therapy either with the occurrence of invasive candidiasis and necrotizing enterocolitis (NEC) with up to threefold increase of developing NEC for therapies lasting more than 10 days and 20% adjunctive mortality risk for each additional day of therapy, and possible development of late-onset sepsis.

Drug Absorption

Only intravenously administered drugs achieve full bioavailability. All the other routes of drug administration are hampered by the capacity of absorption, which is highly variable in the first months of life. When the oral route is used, the pharmacokinetic data are speculative because the intestinal absorption is unknown, and clearance and volume of distribution cannot be determined. The oral administration also is complicated by incomplete drug dissolution, poor solubility, instability in gastric acid, variable gastrointestinal transit time, and intestinal expression of efflux drug transporter, gut metabolism, and liver first passage. Particularly in critical newborns the oral bioavailability is therefore unpredictable.

Volume of Distribution

Most drugs diffuse from the intravascular to the extravascular space and into tissues and organs. Most drugs have a three-compartmental distribution, and drug plasma or blood levels serve as surrogates for tissue concentration at the site of the drug's action. The blood/plasma concentration is related to the total amount of drug present in the body by a relation expressed by the formula

The volume of distribution changes along with body water, decreasing from more than 90% in premature newborns to less than 60% within the first year of life.

Clearance of Drugs

Drug clearance expresses the capacity of the body to eliminate the drug and is expressed by the formula

The liver and kidney are the two main organs responsible for drug elimination. The liver metabolizes the drug and excretes it through the bile, whereas the kidneys filter and excrete drugs and metabolites. Minor sites of clearance also may be in the gut. In case of protein binding of drugs the clearance is affected by the unbound proportion. The hepatic clearance is influenced highly by age: in children some enzymes are far more expressed than adults (such as CYP3A4, CYP1A2, CYP2D6, and UGT2B7 enzymes), and uptake and efflux drug transporters on hepatocytes are instead less expressed as organic anion transporting polypeptide 1B1 (OATP1B1), OATP1B3, multidrug resistance–associated protein 2 (MRP2), and P-glycoprotein, with decreased biliary excretion of some drugs and higher-than-expected plasma concentrations. Renal clearance is due mainly to active tubular excretion through organic cation transporters (OCTs), multidrug and toxin extrusion proteins (MATEs), and organic anion transporter molecules (OATs), whose expression on the luminal and on the basolateral side of epithelial cells and activity are related to tubular maturation, which is widely unpredictable in the preterm newborn.

Renal function may be assessed routinely by serum creatinine (SCr) concentration and estimated through glomerular filtration rate (eGFR) formulas such as bedside Schwartz:

Cystatin C measurement also has been proposed for better precision and avoidance of error resulting from creatinine transplacental passage.