Medical Microbiology and Immunology

, Volume 207, Issue 5–6, pp 339–343 | Cite as

Lopinavir serum concentrations of critically ill infants: a pharmacokinetic investigation in South Africa

  • Michael Schultheiß
  • Sharon Kling
  • Ulrike Lenker
  • Miriam von Bibra
  • Bernd Rosenkranz
  • Hartwig Klinker
Rapid Communication


The role of therapeutic drug monitoring in pediatric antiretroviral therapy is unclear. A little pharmacokinetic datum from clinical practice exists beyond controlled approval studies including clinically stable children. The aim of this study is to quantify LPV exposure of critically ill infants in an ICU and—by identifying risk factors for inadequate exposure—to define sensible indications for TDM in pediatric HIV care; in addition, assume total drug adherence in ICU to compare LPV exposure with a setting of unknown adherence. In this prospective investigation, 15 blood samples from critically ill infants in the pediatric ICU at Tygerberg Hospital were analyzed for LPV–serum concentrations. They were then compared to those of 22 blood samples from out-patient children. Serum-level measurements were performed with an established high-performance liquid chromatography method. All LPV–serum levels of ICU patients were higher than a recommended Ctrough (= 1.000 ng/ml), 60% of levels were higher than Cmax (8.200 ng/ml). Partly, serum levels reached were extremely high (Maximum: 28.778 ng/ml). Low bodyweight and age correlated significantly with high LPV concentrations and were risk factors for serum levels higher than Cmax. Significantly fewer serum levels from infants in ICU care (mean: 11.552 ng/ml ± SD 7760 ng/ml) than from out-patient children (mean: 6.756 ng/ml ± SD 6.003 ng/ml) were subtherapeutic (0 vs. 28%, p = 0.008). Under total adherence in the ICU group, there were no subtherapeutic serum levels, while, in out-patient, children with unknown adherence 28% of serum levels were found subtherapeutic. Low bodyweight and age are risk factors for reaching potentially toxic LPV levels in this extremely fragile population. TDM can be a reasonable tool to secure sufficient and safe drug exposure in pediatric cART.


Therapeutic drug monitoring HIV Infectious diseases Pediatric intensive care Pediatric HIV care Infant HIV care Lopinavir Drug adherence 


Though the role of a general therapeutic drug monitoring (TDM) in antiretroviral therapy remains unclear [1], its usefulness as a tool for the treatment control in special patient populations is well known [2, 3, 4]. However, there are a few recommendations for the sensible use of TDM in pediatric HIV care. While the proteinase inhibitor (PI) Lopinavir (LPV) is recommended as a first-line option in combined antiretroviral therapy (cART) by different guidelines for pediatric patients [5], little is known about LPV pharmacokinetics of critically ill children in daily clinical practice. Especially, for infants and toddlers, pharmacokinetic data have mostly been generated by controlled clinical trials performed for regulatory approval, including children in clinically stable conditions. Children under the age of 2 years are at a greater risk for inadequate LPV exposure [6, 7, 8]. Earlier studies with other PIs (Nelfinavir) have shown great inter-patient variability of serum concentrations in young and small children (age < 2 years) [9].

It was the aim of this investigation to quantify LPV exposure of a very fragile population of critically ill infants in a pediatric intensive care unit in Cape Town, South Africa, and to identify risk factors for inadequate exposure. Furthermore, LPV–serum concentrations of infants in intensive care, where full adherence to drug intake can be assumed, which were compared with those of children in an out-patient setting in the same hospital, where drug application was performed mostly by the patients’ legal guardians, to explore the potential of TDM to monitor adherence.

Subjects and methods

For this prospective investigation, patients were recruited exclusively during hospitalization in the Pediatric Intensive Care Unit (PICU) of Stellenbosch University’s Tygerberg Hospital in Cape Town, South Africa. All patients received an antiretroviral treatment regimen according to South African treatment guidelines consisting of two nucleoside reverse-transcriptase inhibitors as well as the ritonavir-boosted protease-inhibitor lopinavir (LPV), a drug licensed for children under the age of 3 years. Dosage was determined by bodyweight categories [10]. Written informed consent was obtained from a legal guardian. Fifteen blood samples from critically ill infants were collected at random times during routine blood sampling. For inclusion into this investigation, the infants had to have a steady state for LPV, which is reached after a treatment span of at least five half-lives (\({t_{1/2}}\)-LPV = 5–6 h) [11].

The samples were centrifuged (4400 rpm; 10 min) and the obtained serum was deactivated in a water bath (56 °C for 60 min). Serum was stored at − 80 °C until measurement. Blood samples were analyzed for LPV–serum concentration in the TDM Laboratory of the Division of Infectious Diseases, University of Würzburg, Germany, by an established high-performance liquid chromatography (HPLC) method [12].

Additional clinical data including age, weight, height, ethnicity, sex, comorbidities, co-medication, retroviral disease clinical stage, and laboratory findings were retrieved from the patients’ clinical files or determined during investigation if unavailable. Time span between drug application and sampling was recorded according to the documentation of the PICU’s nursing staff. A full adherence to the antiretroviral treatment regimen was assumed as drug application and documentation was exclusively performed by the staff.

According to former studies, 1000 ng/ml was defined as cut-off value for inadequately low exposure. \(~{C_{{\text{max}}}}~\) (8200 ng/ml) was used as cut-off value for categorizing very high, potentially toxic serum levels [13].

The distribution of LPV–serum levels from infants in the PICU was compared to serum concentration results in out-patient children at the same hospital from another investigation of the same comprehensive study [14].

The study was approved by the Health Research Ethics Committee of the Faculty of Health Sciences, University of Stellenbosch (no. N09/05/151).


LPV–serum concentrations were analyzed in 15 blood samples from critically ill infants on an LPV-based treatment regimen, whose age ranged between 1.7 and 13.7 months (median: 4.1 months; mean: 4.7 months; SD: 3.0 months).

All infants had severe WHO stage IV disease (AIDS). Reported comorbidities were pneumonia (including pneumocystis jirovecii and cytomegalie virus pneumonia) (12), gastroenteritis (7), sepsis (3), oral candidosis (3), acute cardiac failure (2), pulmonary haemorrhage (2), meningitis (1), pulmonary tuberculosis (1), and bacterial conjunctivitis (1).

The mean body weight was 5.1 kg (median: 4.2 kg; SD: 1.5 kg; minimum: 3.6 kg; maximum: 7.8 kg). LPV was administered in a syrup formulation (Kaletra®) that included the PI-booster ritonavir with a dosage ratio of 4:1. In one case, an infant, who was treated with rifampicin for a mycobacterium tuberculosis infection, received LPV/ritonavir with a dosage ratio of 4:5. This resulted in a mean LPV dose of 328.6 mg/m2 of body surface area (median: 312.4 mg/m2; SD: 38.6 mg/m2).

The time span between drug intake and sampling ranged from 1.5 h to 26.8 h.

As shown in Fig. 1 all samples of the critically ill infants showed LPV–serum levels, which were higher than the cut-off value for inadequately low LPV exposure (1.000 ng/ml). 93.3% reached levels above \({C_{{\text{trough}}}}\)of 4.000 ng/ml recommended for patients, who are treatment-experienced. The levels ranged from 1.932 to 28.778 ng/ml. The mean LPV–serum level in this population (11.552 ng/ml; median: 9.949 ng/ml; SD: 7.760 ng/ml) was higher than \(~{C_{{\text{max}}}}~\), as described by the producer (8.200 ng/ml). Of all serum concentrations, 60% were higher than \(~{C_{{\text{max}}}}~\).

Fig. 1

Distribution of Lopinavir (LPV)–serum concentrations in infants in a pediatric intensive care unit receiving combined antiretroviral therapy (cART); line a: recommended Ctrough (1000 ng/ml) for patients naïve to cART, line b: recommended Ctrough (4000 ng/ml) for patients “experienced” with cART, and line c: Cmax (8200 ng/ml) in pediatric population as described by producing company

In this population, we found a significant negative correlation between body weight and LPV–serum concentration (p = 0.003) as shown in Fig. 2a. Of all patients with serum levels higher than \(~{C_{{\text{max}}}}~\), 93.3% had an under-average body weight (mean: 5.1 kg).

Fig. 2

Significant negative correlation between bodyweight in kilograms (left) and age in months (right) and Lopinavir (LPV)–serum concentration is shown (p = 0.003 and p = 0.008, respectively)

As shown in Fig. 2b, we found a significant negative correlation between the patients’ age at the time of sampling and their LPV–serum concentration (p = 0.007). All patients with serum levels above \({C_{{\text{max}}}}\) were younger than average in this population.

Comparing the serum concentrations of this population of critically ill infants treated in an intensive care unit (mean: 11.552 ng/ml; SD 7.760 ng/ml) with those of a population of pediatric patients treated in an out-patient setting (mean: 6.756 ng/ml; SD 6.003 ng/ml) (von Bibra et al. [14]), we found significantly fewer subtherapeutic LPV–serum levels (< 1.000 ng/ml) in the intensive care group than in the out-patient group (0%/n = 0 vs. 28% n = 6) (p = 0.008), as shown in Table 1. Of the six patients in the out-patient group with subtherapeutic serum concentrations, five had concentrations beneath LPV–HPLC–lower level of quantification. The mean LPV dose in the out-patient group was 298 mg/m2 of body surface (329 mg/m2 in the intensive care group) (p = 0.036), and the mean age and bodyweight were 22.0 months and 12.6 kg (4.7 months and 5.0 kg in the intensive care group) (p < 0.001 and p < 0.001). There were significant differences in the concomitant antiretroviral medication and CDC stage of retroviral disease in the two groups (see Table 1).

Table 1

Patients’ demographic characteristics


Intensive care patients (n = 15)

Out-patients (n = 22) [14]


Age (years)




p < 0.001*









Female (%)




Body weight (kg)




p < 0.001*









Dose (mg/m2 of body surface)




p = 0.036*









CDC category (n)




p = 0.092




p = 0.070




p = 0.472

 B3 or below



p = 0.005*

Antiretroviral drugs administered (n)




p < 0.001*




p < 0.001*




p = 0.366




p = 0.491




p = 1,000




p = 0.915

CD4+ count (n/µl)




p = 0.002*









LPV plasma levels (ng/ml)




p = 0.090









 Subtherapeutic (%)



p = 0.008*

Reasons for ICU admission (n) (multiple reasons possible)





















aRoutine visits 

*p < 0.05


There are limited pharmacokinetic data for lopinavir in infants. \({C_{{\text{max}}}}\) and \({C_{{\text{min}}}}\) have been reported by the producer to be 8.200 ± SD 2.900 ng/ml and 3.400 ± SD 2.100 ng/ml in a study with a total of 53 children, ranging in age from 6 months to 12 years [6]. In a pharmacokinetic evaluation with a total of 9 infants, Chadwick et al. have reported \({C_{{\text{max}}}}\) and \({C_{{\text{min}}}}\) to be 4.760 (range 2.840–7.280) ng/ml and 2.200 (range 990–4.900) ng/ml, respectively [15]. Serum levels above 1.000 ng/ml for treatment naïve and serum levels above 4.000 ng/ml are recommended for treatment-experienced adults.

Under the assumption of a total adherence to drug intake in the ICU population, 0% of the LPV concentrations were subtherapeutic, while 28% (n = 6) of LPV–serum levels were lower than 1.000 ng/ml in the group of out-patients at the same hospital from an investigation of von Bibra et al. (p = 0.008) [14]. In the five out-patients, who had LPV concentration below the lower limit of quantification of the HPLC measurement method applied, insufficient adherence can be suspected. In a recent study in 723 adult patients, TDM of PIs (including LPV) was used to estimate the prevalence of undetectable plasma concentrations of these substances. Being identified by TDM as insufficiently adherent was an independent risk factor for virological failure [16].

The findings of this comparison underline that TDM represents a useful tool for identifying patients in need of adherence-promoting interventions also in pediatric care.

Aside from comparing LPV concentration of the two study groups in regard to drug adherence, further conclusions from this comparison do not seem sensible because of a very complex (pharmacokinetic) setting in pediatric intensive care, which seems impossible to be comprehended sufficiently.

However, some significant differences of the two groups, found in patient characteristics, are presented in this article. In this group, patients were significantly younger and lighter, which had a significantly higher CDC stage and significantly lower CD4+ counts, as visible in Table 1. Furthermore, dosage per body surface area was significantly higher in the intensive care group using national dosage guidelines, which can be explained by relatively large age/weight categories in these guidelines [5]. The concomitant “backbone” of nucleoside reverse-transcriptase inhibitors (NRTI) in cART differed in the two groups trending towards the newer NRTI generation abacavir in the intensive care group.

While these findings do not give a scientific explanation for the differences in serum concentrations found in the two patient groups, they illustrate the fragility and distinctiveness of this group of critically ill infants.

In this investigation, none of the infants presented subtherapeutic (\({C_{{\text{min}}}}\) < 1.000 ng/ml) LPV–serum concentrations. A remarkable number of critically ill infants (9 of 15; 60%) had serum levels higher than \(~{C_{{\text{max}}}}\), as described for this age group by the producing pharmaceutical company (8.200 ng/ml) [6], reaching extremely high concentrations of up to (maximum) 28.772 ng/ml.

A low body weight and age correlated significantly (p = 0.003 and p = 0.007) with high LPV levels and were risk factors for levels higher than \(~{C_{{\text{max}}}}\). While the upper level of a therapeutic range of LPV has not been clearly defined, toxic effects by this drug such as dyslipidaemia, liver damage, and gastro-intestinal symptoms are known [17]. In other pediatric studies, though with a lower dosage regimen than in this investigation, younger and smaller children (age < 2 years and respectively < 3.5 years) where at a greater risk for inadequately low LPV exposure [8, 18].

These inconsistent study results for this age group underline the difficulties of finding the right LPV dose in a population submitted to permanent alterations through growth and development of organ function. Furthermore, the results from this investigation emphasize the potential benefit that TDM can bring as a tool for optimizing therapy in this very fragile population, which—under inadequate treatment—is at great risk for disease progression and in which drug toxicity can mean catastrophic damage [19].



Lots of gratitude to the doctors and nurses of the Pediatric Intensive Care Unit Ward A9 of Tygerberg Hospital, Cape Town, who lost their beloved consultant Dr. Louis Heynes to violent crime during the course of this investigation.

Compliance with ethical standards

Conflict of interest

The author has no conflicts of interest to declare.


  1. 1.
    Kredo T, Van der Walt J-S, Siegfried N, Cohen K (2009) Therapeutic drug monitoring of antiretrovirals for people with HIV. Cochrane Database Syst RevGoogle Scholar
  2. 2.
    Albrecht M, Mukherjee AL, Tierney C et al (2011) A randomized clinical trial evaluating therapeutic drug monitoring (TDM) for protease inhibitor-based regimens in antiretroviral-experienced HIV-infected individuals: week 48 results of the A5146 study. HIV Clin Trials 12:201–214CrossRefGoogle Scholar
  3. 3.
    Katsounas A, Frank A, Klinker H, Langmann P (2007) Efavirenz-therapy in HIV-patients with underlying liver disease: importance of continuous TDM of EFV. Eur J Med Res 12:331–336PubMedGoogle Scholar
  4. 4.
    Stek AM, Mirochnick M, Capparelli E et al (2006) Reduced lopinavir exposure during pregnancy. AIDS (London England) 20:1931–1939CrossRefGoogle Scholar
  5. 5.
    Department of Health RoSA (2015) National consolidated guidelines for the prevention of mother-to-child-transmission of HIV (PMTCT) and the management of HIV in children, adolescents and adults. Antiretroviral therapy. Pretoria 61Google Scholar
  6. 6.
    Agency EM (2012) EPAR summary for the public—Kaletra. In: EMA, ed. EMA, LondonGoogle Scholar
  7. 7.
    Chadwick EG, Capparelli EV, Yogev R et al (2008) Pharmacokinetics, safety and efficacy of lopinavir/ritonavir in infants less than 6 months of age: 24 week results. AIDS (London England) 22:249–255CrossRefGoogle Scholar
  8. 8.
    Verweel G, Burger D, Sheehan N et al (2007) Plasma concentrations of the HIV-protease inhibitor lopinavir are suboptimal in children aged 2 years and belowGoogle Scholar
  9. 9.
    Bergshoeff AS, Fraaij PL, van Rossum AM et al (2003) Pharmacokinetics of nelfinavir in children: influencing factors and dose implications. Antiviral Ther 8:215–222Google Scholar
  10. 10.
    Department of Health RoSA. National consolidated guidelines for the prevention of mother-to-child-transmission of HIV (PMTCT) and the MANAGEMENT of HIV in children, adolescents and adults. Antiretroviral Therapy. Pretoria 51–69Google Scholar
  11. 11.
    Foissac F, Urien S, Hirt D et al (2011) Pharmacokinetics and virological efficacy after switch to once-daily lopinavir-ritonavir in treatment-experienced HIV-1-infected children. Antimicrob Agents Chemother 55:4320–4325CrossRefGoogle Scholar
  12. 12.
    Langmann P, Klinker H, Schirmer D, Zilly M, Bienert A, Richter E (1999) High-performance liquid chromatographic method for the simultaneous determination of HIV-1 protease inhibitors indinavir, saquinavir and ritonavir in plasma of patients during highly active antiretroviral therapy. J Chromatogr B Biomed Sci Appl 735:41–50CrossRefGoogle Scholar
  13. 13.
  14. 14.
    von Bibra M, Rosenkranz B, Pretorius E et al (2014) Are lopinavir and efavirenz serum concentrations in HIV-infected children in the therapeutic range in clinical practice? Paediatr Int Child Health 34:138–141CrossRefGoogle Scholar
  15. 15.
    Chadwick EG, Pinto J, Yogev R et al (2009) Early initiation of lopinavir/ritonavir in infants less than 6 weeks of age: pharmacokinetics and 24-week safety and efficacy. Pediatr Infect Dis J 28:215–219CrossRefGoogle Scholar
  16. 16.
    Calcagno A, Pagani N, Ariaudo A et al (2017) Therapeutic drug monitoring of boosted PIs in HIV-positive patients: undetectable plasma concentrations and risk of virological failure. J Antimicrob Chemother 72:1741–1744CrossRefGoogle Scholar
  17. 17.
    Casado JL, Del Palacio M, Moya J et al (2011) Safety and pharmacokinetics of lopinavir in HIV/HCV coinfected patients with advanced liver disease. HIV Clinical Trials 12:235–243CrossRefGoogle Scholar
  18. 18.
    van der Lee M, Verweel G, de Groot R, Burger D (2006) Pharmacokinetics of a once-daily regimen of lopinavir/ritonavir in HIV-1-infected children. Antiviral Ther 11:439–445Google Scholar
  19. 19.
    Dunn D (2003) Short-term risk of disease progression in HIV-1-infected children receiving no antiretroviral therapy or zidovudine monotherapy: a meta-analysis. Lancet 362:1605–1611CrossRefGoogle Scholar

Copyright information

© Springer-Verlag GmbH Germany, part of Springer Nature 2018

Authors and Affiliations

  • Michael Schultheiß
    • 1
  • Sharon Kling
    • 2
  • Ulrike Lenker
    • 1
  • Miriam von Bibra
    • 1
  • Bernd Rosenkranz
    • 3
  • Hartwig Klinker
    • 1
  1. 1.Medizinische Klinik und Poliklinik II, Schwerpunkt InfektiologieUniversitätsklinikum WürzburgWürzburgGermany
  2. 2.Department of Pediatrics and Child Health, Ward A9, Tygerberg HospitalUniversity of StellenboschTygerbergSouth Africa
  3. 3.Division of Clinical Pharmacology, Department of Medicine, Tygerberg HospitalUniversity of StellenboschTygerbergSouth Africa

Personalised recommendations