Tobramycin pharmacokinetic and pharmacodynamic targets in people with cystic fibrosis Authors: Kiera H Harwood, Sarah Bannister, Stephen Duffull, April Martinez, Sarath Ranganathan, Shivanthan Shanthikumar, Amanda Gwee Published: 2026 ABSTRACT AbstractIntravenous tobramycin is a first‐line treatment for Pseudomonas aeruginosa infection in people with cystic fibrosis (CF). Tobramycin exhibits concentration‐dependent activity; however, excess drug exposure can lead to nephrotoxicity and ototoxicity. While dosing typically targets serum peak (Cmax) and trough (Cmin) concentrations, the area under the concentration–time curve over 24 h (AUC24) is also used to guide therapy. However, optimal exposure targets remain unclear, leading to variability in clinical practice and guidelines. This review aimed to determine the target tobramycin concentrations and exposure (Cmax, Cmin, AUC24) associated with clinical efficacy (lung function improvement) and/or toxicity (nephrotoxicity, ototoxicity) in CF. Four databases (PubMed, Medline, Embase and Cochrane) were searched and identified studies involving CF patients receiving intravenous tobramycin, recorded tobramycin drug concentration/exposure alongside either clinical efficacy or drug‐related toxicity outcomes. Studies with ≤10 patients, in vitro studies and reviews were excluded. The Newcastle–Ottawa Scale was used to assess risk of bias. A meta‐analysis was not conducted. Twenty‐six studies were included; efficacy was reported in 17, nephrotoxicity in 16 and ototoxicity in 17. Both Cmax and AUC24 were associated with efficacy, with AUC24 being the more reliable predictor. Cmin and AUC24 were associated with toxicity. Only 11 studies reported AUC24, and the overall quality of evidence was poor (21 with medium/high risk of bias, 5 low bias). Current therapeutic Cmax (20–40 mg/L) and Cmin (<1 mg/L) targets remain appropriate. Evidence suggests AUC24‐guided dosing for improved efficacy and safety; however, further research is needed to confirm the optimal AUC24 target (80–120 mg/L·h).Keywords: aminoglycosides, cystic fibrosis, pharmacodynamics, pharmacokinetics INTRODUCTION Cystic fibrosis (CF) is a life‐limiting, monogenetic disorder caused by a mutation in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, characterized by chronic lung disease and affecting approximately 1 in 3000 births. 1 People with CF are vulnerable to recurrent pulmonary infections with bacteria such as Staphylococcus aureus, Haemophilus influenzae and Pseudomonas aeruginosa. 2 The prevalence of P. aeruginosa infection increases with age, affecting around 9% of children under 7 years, 14% of those aged 12–17 years and rising to 58% in adults aged 35–44 years, with chronic infection being common. 2 Therefore, patients with CF often require multiple courses of combination antibiotic therapy to manage the P. aeruginosa bacterial load in the airways and recover baseline pulmonary function. 3 , 4 Aminoglycosides, in particular tobramycin, are recommended by the American Thoracic Society as first‐line treatment of pulmonary exacerbations in people with CF infected with P. aeruginosa in combination with another antimicrobial. 5 Tobramycin displays concentration‐ and exposure‐dependent bactericidal activity, where higher concentrations and greater overall exposure, measured by the area under the concentration–time curve over a 24‐h period (AUC24), are often described to be associated with enhanced bacterial killing and improved clinical outcomes. 6 , 7 , 8 Consequently, achieving a minimum threshold for peak concentration (Cmax) and/or AUC24 is often targeted to optimize efficacy and reduce the risk of bacterial resistance. Importantly, it is difficult to delineate the likely influence of Cmax from AUC24 for the same dosing schedule (e.g., based on once‐daily dosing) because both are linearly related to dose. Comparisons of these exposure metrics as predictors of efficacy and toxicity are therefore limited to comparisons of different dosing regimens, for example, once‐daily (q24h) versus thrice‐daily (q8h), where the Cmax is reduced on multiple daily dosing whereas the AUC24 tends to be preserved. Tobramycin use has also been associated with drug‐related toxicity, primarily associated with elevated serum concentrations. Nephrotoxicity occurs due to drug accumulation in the proximal tubular cells, which disrupts lysosomal function and activates cytotoxic pathways leading to phospholipidosis, oxidative stress and mitochondrial injury. 9 Similar mechanisms occur in the inner ear, where drug accumulation in the cochlear hair cells can trigger cell death; unlike nephrotoxicity, ototoxicity is often irreversible. 10 , 11 , 12 Therefore, as elevated serum trough concentrations (Cmin) may be an indicator of drug accumulation, Cmin is reported to be monitored as a potential predictor of toxicity. 13 , 14 Guidelines for therapeutic drug monitoring (TDM) of tobramycin recommend targeting Cmax and Cmin concentrations ranging from 20 to 40 mg/L and <0.5 to 1 mg/L, respectively. 15 , 16 , 17 , 18 , 19 , 20 More recently, there has been a shift towards AUC24 guided monitoring, 19 , 20 , 21 , 22 , 23 , 24 with an AUC24 to minimum inhibitory concentration (MIC) ratio of ≥80 suggested for efficacy. 25 , 26 However, MIC testing is not routinely performed for P. aeruginosa isolates in CF patients. When available, MIC values tend to increase over time due to colonization with multi‐drug‐resistant isolates, biofilm formation in the CF lung and disease progression. 27 Consequently, an AUC24 target range of approximately 80–120 mg/L·h is often used as a surrogate, though both this target range and the implementation of AUC‐guided dosing vary in clinical practice. 15 , 17 , 19 , 20 , 24 , 28 The lack of a clear consensus on therapeutic targets for tobramycin TDM has led to wide variability in clinical practice across centres; therefore, this systematic review aims to determine the target tobramycin concentrations or AUC24 associated with (i) clinical response as measured by either improvement in lung function and/or (ii) drug toxicity (hearing loss, renal impairment) in CF patients. METHODS This review followed the Preferred Reporting Items for Systematic Reviews and Meta‐analyses (PRISMA) guideline 29 with the protocol registered on PROSPERO (CRD42022336916). Four databases (PubMed, Medline [Ovid] [1946–present], Embase [1946–present] and The Cochrane Library) were searched on 7 March 2025. The search strategy was developed by an expert medical librarian and is outlined in Table S1. Articles that met the following criteria were included: original clinical studies involving patients of any age with CF receiving IV tobramycin for a pulmonary exacerbation; specified tobramycin drug exposure (e.g., Cmax, Cmin or AUC24) achieved; and reported either the clinical response to treatment as measured by an absolute/relative change in lung function or incidence of drug‐related toxicity (nephrotoxicity/ototoxicity). Dynamic and static in vitro studies and narrative review articles were excluded. Studies that had a sample size of ≤10 participants were also excluded. Titles and abstracts of studies retrieved were independently screened by two reviewers (K. H. and A. M. [2023]/S. B. [2025]) to identify studies that met the inclusion criteria. The full texts of all eligible studies were reviewed by the same two reviewers, and any disagreement over eligibility were resolved through consensus or through discussion with a third reviewer (A. G.). Information pertaining to study type, participants, P. aeruginosa infection status, tobramycin dosing regimen, drug exposure, clinical outcomes (absolute/relative change in forced expiratory volume in 1 s; FEV1) and drug‐related toxicity (nephrotoxicity and ototoxicity) was extracted from all eligible studies. Data extraction and assessment of risk of bias and strength of studies were conducted by a single reviewer (K. H.). The Newcastle–Ottawa Scale (NOS) for observational studies was used to assess bias, 30 and study strength was graded using the five‐point scale from the Oxford Centre for Evidence‐Based Medicine (CEBM). 31 A quantitative meta‐analysis was not done due to substantial heterogeneity in dosing regimens, exposure reporting and outcome definitions. Therefore, a structured narrative synthesis was conducted in accordance with PRISMA guidelines. 29 RESULTS A total of 10 083 manuscripts were identified from the search (5667 articles in Embase, 2720 in MEDLINE, 1178 in PubMed and 518 in Cochrane). After removal of 3830 duplicates, the titles and abstracts of 6253 articles were screened, 93 full‐text articles were reviewed and 26 papers fulfilling the inclusion criteria were identified (Figure 1). Of these, the relationship between tobramycin pharmacokinetics (PK) and clinical efficacy was explored in 17 studies, nephrotoxicity was evaluated in 16 and ototoxicity in 17. Six randomized controlled trials (RCTs) compared q24h versus q8h dosing, with differences in treatment efficacy and/or toxicity incidence in relation to tobramycin concentrations outlined in Table 1. Using the NOS tool to assess bias, five studies showed a low risk of bias with the remaining 21 studies having a medium to high risk of bias (see details in Table S2). PRISMA flowchart for systematic database search and study selection. TABLE 1Difference in clinical outcomes in RCTs comparing q24h versus q8h dosing.PaperOutcome measureDifference in clinical outcomeConcentration/exposuresRisk of bias a Grade of evidence b Smyth et al. 32 EfficacyNDq24h: Cmax 28.4 ± 5.07 mg/L, Cmin 0.1 ± 0.2 mg/LLow2Nephrotoxicityq8h greater increase in urinary biomarkers and serum creatinine compared to q24hq8h: Cmax 9.9 ± 1.4 mg/L, Cmin 1 ± 0.4 mg/LOtotoxicityNDRiethmueller et al. 33 EfficacyNDq24h: Cmax 23.9 ± 8.3 mg/L, Cmin 0.39 ± 0.3 mg/LMedium/high2NephrotoxicityNDOtotoxicityND c q8h: Cmax 8.8 ± 2.0 mg/L, Cmin 0.9 ± 0.5 mg/LVic et al. 34 EfficacyNDq24h: Cmax 42.5 ± 11.2mg/L, Cmin 0.3 ± 0.2 mg/L, AUC24 98.6 ± 32.5 mg/L·hLow2Nephrotoxicityq8h greater increase in urinary biomarkers compared to q24hOtotoxicityND c q8h: Cmax 13.2 ± 7.1 mg/L, Cmin 1.1 ± 0.8 mg/L, AUC24 121.5 ± 42 mg L·hWhitehead et al. 35 EfficacyNDq24h: Cmax 29.1 mg/LMedium/high2NephrotoxicityND c q12h/q8h: Cmax 10.5 mg/LOtotoxicityNDBurkhardt et al. 36 EfficacyNDNot reportedLow2NephrotoxicityND c OtotoxicityND c Master et al. 37 EfficacyNDq24h: AUC24 103.9 ± 19.2 mg/L·hMedium/high2NephrotoxicityNDOtotoxicityNDq8h: AUC24 94.9 ± 20.1 mg/L·hMulheran et al. 38 OtotoxicityND c q24h: Cmax 28.4 ± 5.07 mg/LMedium/high2q8h: Cmax 9.9 ± 1.4 mg/L Note: All concentration/exposure levels reported as mean or mean ± SD.Abbreviations: AUC24, area under the concentration–time curve over 24 h; Cmax, peak concentration; Cmin, trough concentration; ND, no difference; q8h, thrice‐daily dosing; q24h, once‐daily dosing; q12h, twice‐daily dosing. a Risk of bias assessed using the Newcastle–Ottawa Scale (NOS) for observational studies. b Grade of evidence calculated using Oxford CEBM. c No toxicity reported. Difference in clinical outcomes in RCTs comparing q24h versus q8h dosing. Note: All concentration/exposure levels reported as mean or mean ± SD. Abbreviations: AUC24, area under the concentration–time curve over 24 h; Cmax, peak concentration; Cmin, trough concentration; ND, no difference; q8h, thrice‐daily dosing; q24h, once‐daily dosing; q12h, twice‐daily dosing. Risk of bias assessed using the Newcastle–Ottawa Scale (NOS) for observational studies. Grade of evidence calculated using Oxford CEBM. No toxicity reported. Clinical efficacy Four clinical studies identified a correlation between serum tobramycin concentrations and improvement in lung function (Table 2). 36 , 37 , 39 , 40 An RCT of 33 adults infected with P. aeruginosa compared q24h and q8h tobramycin dosing, and found a stronger log‐linear relationship between AUC24/MIC and absolute FEV1%, than Cmax/MIC and FEV1% on day 14 (For q24h AUC24/MIC R 2 = .62 vs. Cmax/MIC R 2 = .31; for q8h AUC24/MIC R 2 = .63 vs. Cmax/MIC R 2 = 0.17). 36 A prospective PK study of 13 children receiving tobramycin in combination with ticarcillin found similar results. 40 A significant correlation was found between improvement in FEV1% and AUC24/MIC (R 2 = .77, p < .002), Cmax/MIC (R 2 = .67, p = .013) as well as time above the MIC (R 2 = .63, p = .02), however, no significant correlation between AUC24 and Cmax with FEV1% was observed. In contrast, an earlier prospective PK study involving 28 children receiving combination tobramycin and ticarcillin therapy found those who maintained a 6‐h serum tobramycin concentration (C6h) > 1 mg/L experienced significantly greater improvements in FEV1% 10 days post‐treatment despite similar Cmax concentrations. 39 Specifically, patients with a mean C6h of 0.89 ± 0.2 mg/L showed an increase in FEV1% of 42.5 ± 18.6, whereas those with a mean C6h of 0.34 ± 0.11 mg/L had a significantly smaller FEV1% increase of 8.4 ± 8.2 (p < .001). Another trial randomized 44 adults and children to q24h or q8h tobramycin in combination with ceftazidime and found no difference in the mean improvement in FEV1% between groups; however, among those who showed clinical improvement, the Cmax/MIC ratio was significantly higher compared with those who did not improve (Cmax/MIC 2 vs. 0.5, respectively, p = .01). 37 The mean ± SD AUC24 was greater with q24h dosing compared to q8h (103.9 ± 19.2 vs. 94.9 ± 20.1 mg/L·h, respectively, p < .05), though both values were within 10% of the study's AUC24 target of 100 mg/L·h. TABLE 2Studies reporting a significant relationship between tobramycin concentrations and efficacy.Paper, study designPKPD relationshipRisk of bias a Grade of evidence b Burkhardt et al., 36 prospective single‐centre RCTD14 FEV1% more strongly correlated with AUC24/MIC (R 2 = .63 for q8h, R 2 = .62 for q24h) than Cmax/MIC (R 2 = .17 for q8h, R 2 = .31 for q24h) c Low2Horrevorts et al., 39 prospective single‐centre studyCmin associated with greater FEV1% improvement (R 2 = .75 and R 2 = .6 for day 10 and 20 FEV1%, respectively) d Medium/high3Master et al., 37 prospective multi‐centre RCTCmax/MIC higher in patients with improved FEV1% (2.0 vs. 0.5, p = .01 from q8h and 6.3 vs. 3.2, p = .15 from q24h)Medium/high2Mouton et al., 40 prospective single‐centre study fAUC/MIC, fCmax/MIC, T > MIC correlated with FEV1% improvement (R 2 = .77, p = .002; R 2 = .67, p = .013; R 2 = .63, p = .02, respectively)Medium/high3Wood et al., 14 prospective single‐centre RCTHigher Cmax associated with lower readmission rates; q8h Cmax target > 10 mg/LMedium/high2Abbreviations: AUC24, area under the concentration–time curve over 24 h; Cmax, peak concentration; Cmin, trough concentration; D, day; f, free drug concentration; FEV1%, forced expiratory volume in 1 s; MIC, minimum inhibitory concentration; PKPD, pharmacokinetic/pharmacodynamic; q8h, thrice‐daily dosing; q24h, once‐daily dosing; RCT, randomized control trial; T > MIC, time over MIC. a Risk of bias assessed using the Newcastle–Ottawa Scale (NOS) for observational studies. b Grade of evidence calculated using Oxford CEBM. c Patients with Cmax/MIC < 2 and FEV1 ≥ 40 and AUC24/MIC < 20 and FEV1 ≥ 40 were excluded from correlation analysis. d Patients with Pseudomonas aeruginosa MIC > 2 mg/L were excluded. Studies reporting a significant relationship between tobramycin concentrations and efficacy. Abbreviations: AUC24, area under the concentration–time curve over 24 h; Cmax, peak concentration; Cmin, trough concentration; D, day; f, free drug concentration; FEV1%, forced expiratory volume in 1 s; MIC, minimum inhibitory concentration; PKPD, pharmacokinetic/pharmacodynamic; q8h, thrice‐daily dosing; q24h, once‐daily dosing; RCT, randomized control trial; T > MIC, time over MIC. Risk of bias assessed using the Newcastle–Ottawa Scale (NOS) for observational studies. Grade of evidence calculated using Oxford CEBM. Patients with Cmax/MIC < 2 and FEV1 ≥ 40 and AUC24/MIC < 20 and FEV1 ≥ 40 were excluded from correlation analysis. Patients with Pseudomonas aeruginosa MIC > 2 mg/L were excluded. The remaining studies reported both clinical outcomes and serum tobramycin concentrations; however, they did not specifically evaluate the relationship between them. Four of these studies compared q24h with q8h tobramycin dosing in combination with ceftazidime. 32 , 33 , 34 , 35 Two were RCTs of adults and children with CF that found no significant difference in lung function improvement despite the q24h group achieving a mean Cmax approximately three times higher than the q8h group. 32 , 35 Similarly, two paediatric RCTs by Riethmueller et al. and Vic et al. of 30 and 22 children with CF, respectively, found no difference in improvement in FEV1% following administration of the same total daily dose of tobramycin q24h versus q8h. 33 , 34 As expected, the tobramycin AUC24 was not significantly different between groups (mean ± SD 98.6 ± 32.5 mg/L·h for q24h and 121.5 ± 42 mg/L·h for q8h, p = .6). 34 Six cohort studies investigated q24h tobramycin in combination with a β‐lactam antibiotic reporting both the tobramycin concentrations achieved and clinical outcomes. 41 , 42 , 43 , 44 Two paediatric studies showed improvements in FEV1 in all children with Cmax and Cmin ranging from 7.6 to 7.9 mg/L and <1 mg/L, despite low serum AUC24 values (3.72–4.15 mg/L·h). 41 , 44 Three studies specified an AUC24 target, with two targeting an AUC24 of ≤100 mg/L·h 43 and one of 80–120 mg/L·h. 45 While another study targeted an AUC24/MIC ratio of 100. 42 In a mixed adult and paediatric cohort, clinical improvement occurred in 15/20 (75%) patients, with a mean ± SD tobramycin AUC24, Cmax and Cmin of 83.6 ± 8.89 mg/L·h, 28.7 ± 5.5 mg/L and 0 mg/L, respectively. 43 Similarly, a large retrospective study found that among those with an AUC24 ≥ 80 mg/L·h or Cmax ≥ 8 times the highest P. aeruginosa MIC, 76% and 80% had at least a return to baseline FEV1%, respectively. 45 Similar serum tobramycin exposures were reported in a prospective adult cohort by Moriarty et al., with a mean (range) AUC0–24 of 81.9 (60–115) mg/L·h and Cmax of 29.4 (23.1–35.5) mg/L, associated with an increase in mean FEV1 from 1.77 (0.59–3.3) L to 1.87 (0.79–3.2) L. 42 Higher mean AUC24 levels (141 mg/L·h) were reported in a retrospective study of 51 adults, in which 28/32 (88%) recovered their baseline FEV1%. 46 Q8h tobramycin with a β‐lactam antibiotic was assessed in two prospective studies, one in children, 41 and another in both adults and children. 47 Both reported significant overall improvements in FEV1% following treatment; however, neither study described the proportion of patients who experienced an improvement in lung function. In Pedersen et al., 41 the observed Cmax and Cmin ranged from 7.6 to 7.9 mg/L and <1 mg/L, respectively. In McLaughlin et al., the mean ± SD Cmax was 5.5 ± 1.9 mg/L and 6.5 ± 1.6 mg/L and the Cmin was 0.6 ± 0.5 mg/L and 1.0 ± 0.5 mg/L across both treatment groups, despite all patients receiving the same tobramycin dose. 47 Two studies evaluated alternative clinical outcomes to lung function. 14 , 48 A randomized prospective analysis compared a physician determined tobramycin dosing regimen (Group A) with one generated by a PK programme (Group B) and found no difference in lung function or cumulative hospital stay over 12 months between groups (mean 39 vs. 33 days, respectively), with similar Cmax and Cmin values observed in both groups. 48 Another RCT in adults found lower readmission rates with twice‐daily dosing (q12h) (33%) versus q8h dosing (65%), coupled with higher Cmax with q12h (12.5 ± 2.2 mg/L). 14 Nephrotoxicity Tobramycin‐related nephrotoxicity was reported in 94/790 (12%) patient encounters across five retrospective studies. 45 , 49 , 50 , 51 , 52 Studies reporting a significant correlation between tobramycin concentration and nephrotoxicity are outlined in Table 2. DeGrado et al. reported 2/282 (0.7%) adult admissions where tobramycin‐related nephrotoxicity occurred. 49 In this retrospective cohort, only the Cmax levels were presented, with 56% (154/277) of admissions reaching a Cmax ≥ 20 mg/L, including both cases in which nephrotoxicity occurred. A retrospective study of 34 children assessed the incidence of acute kidney injury (AKI) before and after initiation of CFTR modulator therapy. 51 A total of 15/34 (44%) patients experienced AKI (nine before vs. six after initiation of CFTR modulators). Notably, tobramycin Cmax levels were similar between groups (pre‐CFTR modulator 28.9 mg/L and post‐CFTR modulator 27.2 mg/L, respectively). Landmesser et al. reported a higher incidence of nephrotoxicity in adults and children receiving more frequent dosing compared to q24h dosing (11/22 [50%] vs. 37/129 [29%], p = .047), despite similar total daily doses. 45 However, no significant association was found between nephrotoxicity and Cmax levels when comparing subtherapeutic (<8 times highest P. aeruginosa MIC) to therapeutic (≥8 times MIC) exposures (36% vs. 27%, p = .218). Although, the mean AUC24 was statistically lower in those who developed nephrotoxicity compared to those who did not (104.3 mg/L·h vs. 119.6 mg/L·h, p = .031), the difference was within 20% and these values fell within the commonly used target range. Also, although Cmin levels were not presented, those who developed nephrotoxicity were likely to have higher Cmin levels as they received more frequent dosing. Turner et al. evaluated 103 children receiving q24h or q12h dosing and found no difference in the incidence of AKI between the two groups (20% vs. 15.9%, p = .70), despite the q24h group having higher mean ± SD Cmax (29.5 ± 11.0 mg/L) and lower AUC24 values (92.8 ± 28.7 mg/L·h) compared with the q12h group (Cmax 19.0 ± 4.9 mg/L, AUC24 128.5 ± 34.6 mg/L·h). 52 Similarly, Staubes et al. reported 18/179 (10%) cases of nephrotoxicity in adults receiving 10 mg/kg/day q24h tobramycin, with a similar incidence in those with a tobramycin Cmin < 0.05 mg/L compared with a Cmin ≥ 0.05 mg/L (9% vs. 9.5%, respectively). 50 No nephrotoxic events were reported in the remaining 11 studies, that is, 6 RCTs 14 , 32 , 33 , 34 , 36 , 37 and 5 prospective cohort studies. 39 , 43 , 48 , 53 , 54 Four studies comparing q24h versus q8h at daily doses of 9–15 mg/kg assessed changes in urinary biomarkers of renal injury, with and without serum creatinine. 32 , 33 , 34 , 37 All reported significant elevations in urinary biomarkers (N‐acetyl‐beta‐d‐glucosaminidase [NAG], α‐1‐microglobulin, β2‐microglobin and NAG/creatinine ratio) with both regimens. However, Smyth et al. found that q8h was associated with a 33% greater increase in NAG, a significant rise in serum creatinine (q8h +3.7% vs. q24h −4.5%, p < .04) and higher tobramycin Cmin (q8h 1.0 vs. q24h 0.1 mg/L). 32 This was also reported by Vic et al. who observed higher urinary β2‐microglobin levels at Day 14 with q8h versus q24h (0.87 vs. 0.18 mg/L, p < .01), along with higher Cmin (q8h 1.0 vs. q24h 0.3 mg/L) and AUC24 (q8h 121 vs. q24h 98.6 mg/L·h). 34 In contrast, Riethmueller et al. and Master et al. found no significant differences in urinary biomarkers or serum creatinine between groups. 33 , 37 Master et al. found similar AUC24 from q24h vs. q8h (103.9 vs. 94.9 mg/L·h), while Riethmueller et al. reported lower Cmin with q24h versus q8h (0.39 vs. 0.9 mg/L). Ototoxicity Ototoxicity was assessed in 17 studies with 15 studies reporting the incidence of sensorineural hearing loss through audiometric tests before, during, or following tobramycin treatment with 11 potential ototoxic events identified from 1021 patient encounters across four studies. 14 , 35 , 39 , 46 Studies reporting a significant relationship between tobramycin concentration and ototoxicity are described in Table 3. TABLE 3Studies reporting a significant relationship between tobramycin concentrations and safety.Paper, study designOutcome measurePKPD relationshipRisk of bias a Grade of evidence b Landmesser et al., 45 retrospective single‐centre studyNephrotoxicityHigher incidence of AKI in MDD vs. q24h with same total daily dose c MDD: 50% vs. q24h: 28.7%, p = .047 Medium/high3Smyth et al., 32 prospective single‐centre RCTNephrotoxicity Cmin associated with nephrotoxicity 33% greater increase in NAG from q8h vs. q24h (p < .05)3.7% vs. −4.5% change in SCr from q8h vs. q24h, respectively (p < .04) Cmin target ≤ 1 mg/L (q24h), ≤2 mg/L (q8h) Low2Vic et al., 34 prospective single‐centre RCTNephrotoxicityCmin associated with higher β2 microglobinuria on treatment D14 q8h: 0.87 mg/L vs. q24h: 0.18 mg/L, p < .001 Low2Wood et al., 14 prospective single‐centre RCTOtotoxicityCmin associated with ototoxicity q8h: 7/18 (39%) vs. q12h: 0/9 cases of ototoxicity Medium/high2Dong et al., 55 prospective single‐centre studyOtotoxicityNumber of tobramycin doses, peripheral Cmax and Cmin > 2 mg/L associated with ototoxicity (R 2 = .345, p = .017; R 2 = .510, p < .001; R 2 = .368, p = .011, respectively).Medium/high3Abbreviations: AKI, acute kidney injury; Cmin, trough concentration; Cmax, peak concentration; D, day; MDD, multiple‐daily dosing; NAG, N‐acetyl‐beta‐d‐glucosaminidase; PKPD, pharmacokinetic/pharmacodynamic; q12h, twice‐daily dosing; q8h, thrice‐daily dosing; q24h, once‐daily dosing; RCT, randomized control trial; SCr, serum creatinine. a Risk of bias assessed using the Newcastle–Ottawa Scale (NOS) for observational studies. b Grade of evidence calculated using Oxford CEBM. c Mean AUC24 was lower in those who did not experience AKI compared with those who did (p < .05). Although not measured in this study, MDD associated with higher Cmin concentrations compared with q24h dosing. Studies reporting a significant relationship between tobramycin concentrations and safety. MDD: 50% vs. q24h: 28.7%, p = .047 MDD: 50% vs. q24h: 28.7%, p = .047 Cmin associated with nephrotoxicity 33% greater increase in NAG from q8h vs. q24h (p < .05)3.7% vs. −4.5% change in SCr from q8h vs. q24h, respectively (p < .04) 33% greater increase in NAG from q8h vs. q24h (p < .05) 33% greater increase in NAG from q8h vs. q24h (p < .05) 3.7% vs. −4.5% change in SCr from q8h vs. q24h, respectively (p < .04) 3.7% vs. −4.5% change in SCr from q8h vs. q24h, respectively (p < .04) Cmin target ≤ 1 mg/L (q24h), ≤2 mg/L (q8h) q8h: 0.87 mg/L vs. q24h: 0.18 mg/L, p < .001 q8h: 0.87 mg/L vs. q24h: 0.18 mg/L, p < .001 q8h: 7/18 (39%) vs. q12h: 0/9 cases of ototoxicity q8h: 7/18 (39%) vs. q12h: 0/9 cases of ototoxicity Abbreviations: AKI, acute kidney injury; Cmin, trough concentration; Cmax, peak concentration; D, day; MDD, multiple‐daily dosing; NAG, N‐acetyl‐beta‐d‐glucosaminidase; PKPD, pharmacokinetic/pharmacodynamic; q12h, twice‐daily dosing; q8h, thrice‐daily dosing; q24h, once‐daily dosing; RCT, randomized control trial; SCr, serum creatinine. Risk of bias assessed using the Newcastle–Ottawa Scale (NOS) for observational studies. Grade of evidence calculated using Oxford CEBM. Mean AUC24 was lower in those who did not experience AKI compared with those who did (p < .05). Although not measured in this study, MDD associated with higher Cmin concentrations compared with q24h dosing. Two prospective studies identified a potential association between an increased risk of ototoxicity and higher Cmin. 14 , 39 Normal and high‐frequency audiometry were performed in an RCT by Wood et al., reporting ototoxicity in 7/18 (39%) patients receiving q8h dosing, six of whom experienced high‐frequency hearing loss, compared to none (0/9) in the q12h group (p < .05), despite similar total daily doses. 14 Q8h dosing resulted in lower mean Cmax and higher Cmin (7.9 ± 1.9 mg/L and 0.8 ± 0.3 mg/L) compared to q12h (12.5 ± 2.2 mg/L and 0.5 ± 0.2 mg/L). Similarly, Horrevorts et al. reported high‐frequency hearing loss in 1/9 children who received dosing more frequent than q8h dosing (mean Cmax 6.63 mg/L and Cmin 1.34 mg/L), whereas no cases (0/7) occurred in the q8h group (mean Cmax 7.03 mg/L and Cmin 0.51 mg/L). 39 Another RCT comparing q24h versus q8h dosing found no difference in the incidence of ototoxicity, with one case of normal‐frequency hearing loss reported in each group (mean Cmax 29.1 mg/L vs. 10.5 mg/L, respectively). 35 In a retrospective study of 51 adults (41 who received tobramycin; mean AUC24 of 141 mg/L·h), three cases of ototoxicity occurred (one sensorineural hearing loss, two vestibular events); however, it was unclear whether it was secondary to tobramycin or another aminoglycoside. 46 Tinnitus or vestibular symptoms were reported in 13/414 (3.1%) patient encounters across five prospective studies; however, tobramycin drug exposure in affected versus unaffected patients were not assessed. 32 , 37 , 43 , 53 , 56 Two RCTs comparing q24h and q8h dosing reported minimal toxicity, with one study recording a case of tinnitus in each group (mean AUC24 of 103.9 and 94.9 mg/L·h, respectively), attributed to rapid infusion rates (<15 min) rather than drug exposure and symptoms resolved after extending infusion duration. 37 The other RCT reported one case of dizziness in each group (mean ± SD Cmax and Cmin 28.4 ± 5.07 and 0.1 ± 0.2 mg/L for q24h and 9.9 ± 1.4 and 1 ± 0.4 mg/L for q8h). 32 Al‐Aloul et al. reported two cases of tinnitus and three cases of vestibular symptoms among 20 adults receiving q24h or q12h dosing (mean ± SD Cmax 7.6 mg/L, Cmin 1.4 mg/L). 53 Two (2/10, 20%) patients developed transient vestibular symptoms shortly after dose administration in Powell et al. despite no significant audiometric changes. 56 Similarly, vestibular symptoms occurred in 2/25 (8%) in Vandenbussche and Homnick, with symptom resolution occurring in both after dose reduction (Cmax reduced from 34 mg/L to 28–30 mg/L). 43 Notably, no adverse effects were seen in the remaining patients, including four with comparable peak levels; however, routine audiological monitoring was not performed. Dong et al. developed a two‐compartment PK model using data from 38 patients who received a mean (range) number of tobramycin doses of 99 (3–480). 55 High‐frequency hearing loss was significantly associated with cumulative tobramycin doses (R = .345, p = .017) and with peripheral compartment (representing peripheral tissue such as the inner ear) trough and peak concentrations exceeding 2 mg/L (R = 0.368, p = .011; R = .510, p < .001, respectively). Furthermore, patients receiving ≥6 courses with peripheral peak concentrations >3 mg/L were 5.8 times more likely to develop ototoxicity compared with those with lower exposures. No evidence of ototoxicity or vestibular symptoms were reported in the remaining six studies totalling 449 patient encounters receiving q24h or q8h dosing with Cmax levels ranging from 7.48–52.2 mg/L, Cmin from 0.3–1.1 mg/L, and AUC24 from 98.6–121.5 mg/L·h. 33 , 34 , 36 , 38 , 48 , 54 DISCUSSION This systematic review found that serum tobramycin AUC24 and Cmax were more strongly associated with treatment efficacy (4 studies, evidence Level 2–3; low to high risk of bias), whereas nephrotoxicity and ototoxicity were more closely linked to elevated Cmin and AUC24 levels (6 studies, evidence Level 2–3; low to high risk of bias) in people with CF. These findings align with current clinical guidelines, though the review highlights the need to shift from Cmax towards AUC24‐guided dosing for optimizing treatment efficacy. Although both Cmax and AUC24 have been associated with improvements in lung function, AUC24 appears to be a more reliable predictor. This is supported by four RCTs (Level 2 evidence) that reported no significant differences in FEV1 improvement between q24h and q8h dosing, despite q24h dosing achieving Cmax values approximately three times higher than q8h. With the same total daily tobramycin dose used in these studies, the AUC24 would not be expected to differ significantly as was reported in the study by Vic et al. Four of the included studies targeted an AUC24 or AUC24/MIC ratio of ≥80 mg/L·h and 100, respectively (i.e., thresholds consistent with current CF guidelines), with over 75% of patients in these studies showing improved lung function, supporting the appropriateness of these targets. 37 , 42 , 43 , 45 The current guideline‐recommended Cmax target of 20–40 mg/L for q24h dosing also appears appropriate, as observed Cmax values ranged from 23.9 to 42.5 mg/L and were associated with a 5%–15% improvement in percent‐predicted FEV1 (low grade evidence; evidence level 2/3). This is likely due to increased drug penetration into the lung tissue from higher serum concentrations, leading to greater bacterial killing. 57 Although, both Cmax and AUC24 have been identified as strong predictors of increased bacterial killing in several dynamic in vitro models of P. aeruginosa, these finding suggest that AUC24 may be a more consistent correlate of clinical efficacy. 58 , 59 , 60 , 61 The overall incidence of nephrotoxicity was relatively low, at approximately 12% across all studies, and is comparable to reported rates in non‐CF populations including critically ill or septic patients, where incidence ranges from 4% to 41%. 62 , 63 , 64 , 65 , 66 Nephrotoxicity was consistently associated with a higher Cmin, often due to more frequent dosing intervals compared to q24h dosing. Two studies reported a similar incidence of AKI despite different Cmin, though both remained in the recommended range of <0.5–1 mg/L, supporting current targets. Higher serum trough concentrations reduce the drug‐free period and promote drug accumulation in the renal tubules, both resulting in increased risk of nephrotoxicity. 67 , 68 Although serum creatinine is used as a marker of renal function, several studies have shown a rise in serum creatinine is a delayed response and occurs only after drug accumulation in renal tissue and proximal tubular damage. 65 , 69 , 70 , 71 Therefore, more recently, elevations in urinary biomarkers such as NAG and β2‐microglobulin are often used as early indicators of proximal tubular injury and aminoglycoside‐related nephrotoxicity, 11 , 72 , 73 , 74 as demonstrated by three studies reporting significant elevations in these biomarkers despite unchanged serum creatinine levels. 33 , 34 , 37 For ototoxicity, AUC24 remains an important indicator, with a high‐quality prospective pharmacokinetic study (Level 3 evidence) demonstrating a clear correlation between cumulative tobramycin doses and an increased risk of ototoxicity (R 2 = .345, p = .017). While data in people with CF are limited, these findings are consistent with studies linking aminoglycoside AUC to ototoxicity in other conditions, such as multi‐drug‐resistant tuberculosis. 75 , 76 , 77 , 78 Classification and regression tree analyses of patients with multi‐drug‐resistant tuberculosis from three clinical studies identified cumulative treatment days as the strongest predictor of amikacin‐related ototoxicity, followed by cumulative AUC and AUC24, with Cmax and Cmin levels being poor predictors. 77 Interestingly, transient vestibular symptoms and/or tinnitus were reported in four studies, likely due to short infusion durations (<15 min) or bolus injections. 32 , 37 , 43 , 56 Rapid infusions result in a significantly higher Cmax, a shorter time to reach Cmax and potentially an increased rate of drug distribution to peripheral compartments such as the vestibular system of the inner ear, all which may account for these symptoms. 79 , 80 , 81 Hence, the recommended infusion time for tobramycin is typically 20–30 min. 20 , 24 , 82 , 83 Additionally, although multiple cohort studies in our review reported vestibular symptoms and tinnitus, both indicators of ototoxicity, formal audiological testing were not conducted in some studies. This highlights the need for routine audiological monitoring, including high‐frequency testing, particularly in patients receiving repeated courses as the incidence of sensorineural hearing loss is likely underreported. Overall, this review suggests that monitoring Cmin is appropriate; however, further research examining the relationship between tobramycin Cmin and AUC24 and its relationship with ototoxicity are needed in larger cohorts. A limitation of this review is that the majority of included studies (21/26) had a moderate to high risk of bias and AUC24 was reported in only 11/26 studies and where available, values typically ranged from 80 to 120 mg/L·h, limiting assessment of the full exposure–toxicity relationship. Furthermore, the method of AUC24 calculation was not specified in the majority of studies, limiting comparability and potentially introducing further bias through variability in accuracy of AUC24 estimation. Although Cmax and Cmin data were collected across studies, variability in sampling times could not be determined or fully accounted for in the analysis. Most patients also received concomitant antibiotics (most often β‐lactams), which may have contributed to clinical improvement or nephrotoxicity but were not accounted for in analyses. In some studies, patients also received inhaled tobramycin or CFTR modulators, which may have influenced clinical response to treatment. As most studies were conducted prior to the widespread availability of CFTR modulator therapy and the number of patients who received modulators is unknown, further research is needed to assess their impact on tobramycin PK and clinical outcomes, especially given their use in ~80% of the CF population. 2 Lastly, microbiological outcomes (e.g., bacterial killing or resistance emergence) were not assessed, preventing evaluation of the relationship between drug exposure and resistance development. This systematic review evaluated the published evidence on the relationship between tobramycin exposure with clinical improvement and drug toxicity (nephrotoxicity and ototoxicity). AUC24 was identified to be a consistent predictor of treatment response more so than Cmax while both Cmin and AUC24 were important parameters for minimizing the risk of toxicity. Consistent with current clinical guidelines, q24h dosing should be used, as it is associated with lower Cmin concentrations (targeting <1 mg/L) and reduced risk of nephrotoxicity. The existing Cmax target of 20–40 mg/L also remains appropriate; however, this was only supported by studies with low grade evidence. Future studies in larger cohorts are needed to better define the optimal AUC24 target and assess whether the current narrow therapeutic range for tobramycin can be safely expanded. Nomenclature of targets and ligands Key protein targets and ligands in this article are hyperlinked to corresponding entries (in http://www.guidetopharmacology.org) and are permanently archived in the ‘Concise guide to pharmacology 2021/22’. 84 CONFLICT OF INTEREST STATEMENT None to declare. Supporting information Table S1: Tobramycin in CF search strategy. Table S2: Bias assessment using the NOS tool. Data S1. Supporting Information. ACKNOWLEDGEMENTS We would like to acknowledge Poh Chua for her assistance in designing the literature search strategy. Open access publishing facilitated by The University of Melbourne, as part of the Wiley ‐ The University of Melbourne agreement via the Council of Australasian University Librarians DATA AVAILABILITY STATEMENT The data that support the findings of this study are available from the corresponding author upon reasonable request. REFERENCES Associated Data Supplementary Materials Table S1: Tobramycin in CF search strategy. Table S2: Bias assessment using the NOS tool. Data S1. Supporting Information. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.