JAC

antiviral
Journal of Antimicrobial Chemotherapy (2006) 57, 1161–1167
doi:10.1093/jac/dkl112
Advance Access publication 4 April 2006

The use of pharmacokinetically guided indinavir dose reductions in

the management of indinavir-associated renal toxicity

Mark A. Boyd1–3*, Umaporn Siangphoe1, Kiat Ruxrungtham1,4, Peter Reiss1,5,

Apicha Mahanontharit1, Joep M. A. Lange1,5, Praphan Phanuphak1,4, David A. Cooper1,2
and David M. Burger6,7
1
The HIV Netherlands Australia Thailand Research Collaboration, The Thai Red Cross AIDS Research Centre,
Bangkok, Thailand; 2National Centre in HIV Epidemiology and Clinical Research, University of New South Wales,
Sydney, Australia; 3Department of Microbiology and Infectious Diseases, Flinders Medical Centre, Flinders
University, Bedford Park 5042, South Australia, Australia; 4Faculty of Medicine, Chulalongkorn University,

Downloaded from jac.oxfordjournals.org by guest on April 4, 2011

Bangkok, Thailand; 5Centre for Poverty-related Communicable Diseases, Academic Medical Centre, University
of Amsterdam, and International Antiviral Therapy Evaluation Centre, Amsterdam, The Netherlands; 6Department of
Clinical Pharmacy, Radboud University Medical Centre, Nijmegen, The Netherlands; 7Nijmegen University Centre
for Infectious Diseases, Nijmegen, The Netherlands

Received 16 January 2006; returned 14 February 2006; revised 28 February 2006; accepted 12 March 2006

Objectives: Indinavir is associated with nephrotoxicity. Therapeutic drug monitoring of indinavir improves
clinical outcome, but there is little data regarding therapeutic drug monitoring for patients with established
indinavir-associated renal impairment. We prospectively studied the use of therapeutic drug monitoring in
patients with virological success but established nephrotoxicity on an indinavir-containing regimen.
Methods: We measured indinavir Ctrough/C2h, serum creatinine, pyuria, blood pressure (BP), weight and HIV
RNA. The major endpoint of interest was the number of patients achieving a normal creatinine level 20 weeks
following final indinavir dose adjustment. Primary analysis was by intention to treat (ITT).
Results: A total of 35 patients were enrolled; mean (SD) age 40.3 (5.8) years; mean (SD) BMI 21.5 (2.8) kg/m2.
At baseline 6/35 (17%) had a serum creatinine concentration within normal limits, but were offered enrol-
ment because of previous nephrotoxicity (nephrolithiasis and/or abnormal serum creatinine), and a screen-
ing pharmacokinetic profile associated with increased nephrotoxicity risk. By ITT analysis 11/35 (31%)
had normal creatinine at study end (P = 0.18). Of the 29 patients with abnormal creatinine at baseline, 7/29
(24.1%) had normal creatinine at study end (P = 0.016). Patients had a median (IQR) indinavir per dose
adjustment over the study of 400 (400–800) mg. We observed improvements in estimated creatinine clear-
ance, pyuria, resting BP and indinavir pharmacokinetic profile. HIV RNA control was maintained with
continued immune recovery despite lower indinavir doses.
Conclusions: Patients experiencing nephrotoxicity on an indinavir-containing regimen were safely main-
tained on indinavir by means of therapeutic drug monitoring. Parameters of renal function improved but did
not return to baseline values, at least in the short-term.

Keywords: HIV, nephrotoxicity, pyuria, pharmacokinetics (PK), therapeutic drug monitoring, serum creatinine,
creatinine clearance, blood pressure (BP), Thailand, resource-limited setting

Introduction of 800 mg three times daily taken away from food. In recent
years there has been an overwhelming trend towards combining
Indinavir is a protease inhibitor licensed for the treatment of HIV protease inhibitors with low doses of ritonavir in order to take
infection in combination with other antiretroviral agents at a dose advantage of the capacity of this agent to inhibit the cytochrome
.............................................................................................................................................................................................................................................................................................................................................................................................................................

*Correspondence address. Room 5D304.1, Department of Microbiology and Infectious Diseases, Flinders Medical Centre, Flinders University,
Bedford Park 5042, South Australia, Australia. Tel: +61-8-8204-4948; Fax: +61-8-8204-4733; E-mail: mark.boyd@fmc.sa.gov.au
.............................................................................................................................................................................................................................................................................................................................................................................................................................

1161

The Author 2006. Published by Oxford University Press on behalf of the British Society for Antimicrobial Chemotherapy. All rights reserved.
For Permissions, please e-mail: journals.permissions@oxfordjournals.org
 Boyd et al.
P-450 mediated metabolism of protease inhibitors, thereby allow-
ing for dosing regardless of food intake in twice daily, and in
some instances once daily, dosing schedules.1 Clinical experience
with indinavir has demonstrated that it has a relatively narrow
therapeutic window and is frequently associated with nephrotox-
icity, which may manifest as a syndrome of renal colic, tubulo-
interstitial nephritis or even acute renal failure. Prolonged use
of indinavir is associated with chronic elevations in serum
creatinine.2–6
There is a wide inter-individual variation in the pharmaco-
kinetics (PK) of indinavir.7,8 Relationships between the pharma-
cokinetics of indinavir and its pharmacodynamic effects have
been described. Dieleman et al.7 have demonstrated an associ-
ation between maximal indinavir concentrations and nephrotox-
icity. In a cohort of patients using indinavir 800 mg three times
daily or ritonavir-boosted indinavir 800/100 mg twice daily the
indinavir plasma concentration at 2 h post-indinavir ingestion
(C2h) was discriminating for patients with and without nephro-
toxicity (indinavir C2h > 7.5 mg/L for three times daily dosing
and indinavir C2h > 10 mg/L for twice daily dosing); similarly,
indinavir trough levels (Ctrough) correlated with antiviral response
(indinavir Ctrough > 0.1 mg/L for three times daily dosing and
indinavir Ctrough > 0.25 mg/L for twice daily doing).8 Other
studies have demonstrated that the use of therapeutic drug
monitoring of indinavir in the period immediately following
commencement of combination antiretroviral therapy leads to
improved outcomes measured in terms of toxicity, efficacy
and drug discontinuation.9,10 The outcome of patients experien-
cing nephrotoxicity who cease indinavir treatment has been
favourable, with prompt resolution of urinary abnormalities
and return of serum creatinine to normal.11–13 However, there
are no data to determine whether it is safe to continue indinavir
by optimizing the indinavir dose in patients experiencing
indinavir-associated nephrotoxicity.
Indinavir has been used in a number of studies at the HIV
Netherlands Australia Thailand Research Collaboration (HIV-
NAT), and nephrotoxicity has been observed in all.14,15 Despite
this, however, and for a number of reasons, very few patients
experiencing indinavir-associated nephrotoxicity have actually
ceased indinavir; patients demonstrating nephrotoxicity had
benefited immunologically and virologically from their antiretro-
viral therapy and had not suffered further progression of their
HIV disease; patients had nearly all previously failed a combi-
nation nucleoside regimen, and a switch to a non-nucleoside
reverse transcriptase inhibitor-containing regimen was thought
ill-advised; lifelong supply of indinavir without cost to the patient
had been guaranteed by the major study sponsor (Merck & Co.)
for all patients who participated in Merck & Co. sponsored clin-
ical trials at HIV-NAT and who continued to derive benefit; and
finally, studies had indicated that the decrease in renal function
associated with indinavir therapy was reversible, generally within
4 months of ceasing the drug.11–13 We therefore elected to invest-
igate the use of pharmacokinetically guided indinavir dose optim-
izations in patients with evidence of indinavir-associated
nephrotoxicity as part of an approved clinical protocol.
Patients and methods
Patients were eligible for enrolment if they were participating in HIV-
NAT studies in which indinavir formed part of their antiretroviral
1162
regimen. Patients who had undergone indinavir dose reduction prior
to the availability of therapeutic drug monitoring at HIV-NAT were
eligible for inclusion if study criteria were met.
Inclusion criteria were the following:
(i) ACTG grade I renal toxicity (a serum creatinine concentration
between 1.1 and 1.6 times the upper limit of normal) or higher,
determined on at least two visits with at least 3 months between
visits. The upper limit of normal serum creatinine concentration
for this study was defined as 124 mmol/L.
(ii) HIV RNA <50 copies/mL.
(iii) Patients who were receiving indinavir three times daily must have
had an indinavir level at 2 h post-ingestion >7.5 mg/L.
(iv) Patients who were using indinavir plus ritonavir twice daily must
have had an indinavir level at 2 h post-ingestion >10.0 mg/L.
(v) Patients who were using indinavir three times daily must have
had an indinavir trough level >0.10 mg/L.
(vi) Patients who were using indinavir plus ritonavir twice daily must
have had an indinavir trough level >0.25 mg/L.
(vii) Willing and able to give written informed consent.
Those patients fulfilling the inclusion criteria and giving informed
Downloaded from jac.oxfordjournals.org by guest on April 4, 2011
consent followed a protocol that included regular assessments of
indinavir PK levels (Ctrough and C2h), plasma HIV RNA, serum
creatinine and urine microscopy performed at 4 weekly intervals
while indinavir dose adjustments were undertaken, and then at
2 monthly intervals once a new stable indinavir dose with acceptable
PK parameters was established. Indinavir pharmacokinetic levels
were assessed according to the patient’s record of last indinavir
intake (indinavir Ctrough) and with (twice daily dosing regimen) or
without (three times daily dosing regimen) food for the indinavir C2h
level. Pyuria on microscopy was graded according to a scale: Grade I,
normal [0–5 cells/high powered field (HPF)]; Grade II, mild
(6–25 cells/HPF); Grade III, moderate (26–50 cells/HPF); and
Grade IV, severe (>50 cells/HPF). Because of the known association
between renal impairment and hypertension the study also mandated
routine monitoring of resting blood pressure (BP) at all study visits
according to a standard procedure. Creatinine clearance was estim-
ated from the serum creatinine according to the Cockcroft–Gault
equation [(140 – age) · weight (kg)]/[814 · serum creatinine
(mmol/L)]; the result was multiplied by 0.85 for females. Indinavir
dose reductions were made in 200 mg increments according to the
protocol-defined acceptable limits of the indinavir Ctrough and C2h,
and an indinavir dose reduction to a minimum of 200 mg per dose
(with or without ritonavir boosting) was allowed as long as PK
parameters were met. Follow-up was planned until at least
20 weeks after the final indinavir dose reduction to assess the out-
come. In the event that the clinician considered an adjustment in the
ritonavir dose (rather than the indinavir dose) to be warranted in
order to optimize the pharmacokinetics of indinavir this was allowed.
Indinavir in plasma was analysed using a previously published
HPLC method.16 Accuracy ranged from 104% to 108% and intra-day
and inter-day precision ranged from 2.12% to 7.48% and from 0.43%
to 3.46%, respectively. The calibration range was 0.04–30 mg/L.
Serum creatinine levels were measured using Vitalab Flexor equip-
ment (Vital Scientific, The Netherlands).
The primary endpoint of the study was the number of patients
achieving a serum creatinine concentration within normal limits
20 weeks following achievement of an optimal indinavir dosing
schedule. Secondary endpoints included changes in creatinine clear-
ance, pyuria, resting BP and control of HIV RNA. An absolute serum
creatinine value of 300 mmol/L was set for cessation of indinavir
therapy during the study. For comparisons regarding changes in
serum creatinine and calculated creatinine clearance we determined
 TDM in indinavir-associated renal impairment
results for three time-points: pre-indinavir exposure (at screening for
the original HIV-NAT study in which patients were first commenced
on an indinavir-including regimen), at baseline for the currently
described indinavir dose reduction study and at a time-point at
least 20 weeks after the final indinavir dose adjustment as per pro-
tocol. Data on concomitant medications were collected in order to
assess their possible contribution to nephrotoxicity. The protocol was
reviewed and approved by the Ethics Committee of The King Chu-
lalongkorn Memorial Hospital and all patients gave written informed
consent prior to participation.
The primary analysis was an intention to treat (ITT) analysis using
a last observation carried forward analysis imputation for continuous
variables and a missing = failure imputation for dichotomous-
variables. Differences over time were assessed by the measured
change in scores from baseline to last follow-up. Results were
described as either mean (–SD) or median (IQR) depending upon the
data distribution, as well as in percentages. Paired t-test and Wilcoxon
signed ranks test were used to compare the means and medians of
changes between two related time-points, respectively, while McNemar
test was used to compare the percentage of patients for each outcome.
A P value of <0.05 was used to indicate statistical significance.
Results
A total of 61 patients were screened of whom 24 were found to be
ineligible. This left 37 eligible patients of whom 2 were excluded
from the analysis, one because of withdrawal of consent and
another because of an indinavir dose reduction prior to PK mon-
itoring. Therefore 35 patients were enrolled and the analysis was
carried out based on these. Of these 35 patients not all fulfilled
the study criteria for an elevated creatinine at screening, but were
allowed to enter the study because of symptoms suggestive of
clinical nephrolithiasis and/or a previously abnormal serum cre-
atinine result, and because the screening results demonstrated a
pharmacokinetic profile associated in our experience with
indinavir-associated nephrotoxicity. All patients were enrolled
from other clinical studies conducted at HIV-NAT, and as
such they had undergone investigation for other possible causes
of renal impairment during the course of routine clinical follow-
up. Of the 35 patients enrolled, 3 were terminated during the
study. One switched to ritonavir-boosted saquinavir, one was lost
to follow-up, and the third ceased indinavir-containing therapy
after an episode of recurrent renal colic despite dose optimization
to ritonavir-boosted indinavir 400/100 mg twice daily and accept-
able indinavir pharmacokinetics (Ctrough = 0.14 mg/L, C2h = 3.46
mg/L). The trial disposition is summarized in Figure 1.
The baseline characteristics of the cohort are summarized in
Table 1. The patients in the cohort were predominantly male with
a mean age of 40.3 (5.8) years and a mean body weight of 56.7
(9.1) kg. The cohort had a median of 3.5 (1.6–3.6) years of
exposure to indinavir. A total of 27 patients (77%) of the cohort
had received concomitant medications associated with renal
toxicity prior to the study commencement, although the majority
had used these on an ‘as required’ (p.r.n.) basis only. The only
potentially nephrotoxic medication received regularly by a large
proportion of patients prior to study was co-trimoxazole prophy-
laxis 800/160 mg daily (25/35 patients, 71%). Of interest, of
patients enrolled into the study from the HIV-NAT treatment
cohort who had received indinavir the longest (>3 years)14 all
qualifying patients had continued to receive co-trimoxazole pro-
phylaxis (14 of an original enrolment of 104 patients). However,
1163
JAC
antiviral
Screening failure (n = 24)
normal creatinine (n = 15)
IDV PK acceptable (n = 8)
high viral load (n = 1)
Excluded (n = 2)
Screening
incomplete consent (n = 1)
(n = 61)
prior dose reduction (n = 1)
Enrolled (n = 35)
Discontinued (n = 3)
lost to follow-up (n = 1)
switched to ritonavir-boosted saquinavir (n = 1)
recurrent renal colic despite indinavir
dose optimization (n = 1)
Completed study (n = 32)
Downloaded from jac.oxfordjournals.org by guest on April 4, 2011
Figure 1. Study disposition. IDV, indinavir.
despite this, we did not find a relationship between receipt of
co-trimoxazole prophylaxis and the presence of renal impairment
at study baseline (data not shown).
Figure 2 expresses the numbers of patients using different
doses of indinavir and ritonavir at baseline and at final fol-
low-up. The median change of indinavir dose during the study
(per dose) was 400 (400–800) mg (P < 0.001). Of the 35 patients,
28 (80%) required a single dose adjustment, 6 (17%) underwent
two dose adjustments and 1 patient (3%) had three dose adjust-
ments. At baseline one patient was using ritonavir 200 mg twice
daily in combination with indinavir, and by study end a further
six patients had increased the ritonavir dose to 200 mg given
twice daily with indinavir. The reason given for the increased
ritonavir dose was to optimize indinavir pharmacokinetics in
situations of an unacceptably high indinavir C2h and a low,
but acceptable, indinavir Ctrough (i.e. the indinavir dose was
reduced in order to minimize nephrotoxicity, but the ritonavir
dose was concomitantly increased in order to maintain a thera-
peutic indinavir Ctrough).
Changes in serum creatinine and estimated creatinine clear-
ance are summarized in Table 2. For the primary study endpoint
(ITT analysis) 11/35 (31%) had a normal serum creatinine at
study end compared with 6/35 (16%) who had a normal
serum creatinine at baseline (P value for difference from
baseline = 0.18). Confining the analysis firstly to those patients
who completed the full study protocol (31 patients) and secondly
to only those 29 patients who had an abnormal baseline creatinine
value and who completed the study protocol (27 patients), the
number of patients with a normal creatinine value at study end
was 11/31 (35.5%; P = 0.02) and 7/27 (26%; P = 0.02), respect-
ively. Prior to indinavir exposure the cohort had a mean serum
creatinine of 82 (12.7) mmol/L, corresponding to an estimated
creatinine clearance of 84 mL/min. At baseline for the present
study the serum creatinine concentration was 149 (27.7) mmol/L,
representing an estimated creatinine clearance of 46 mL/min,
and a mean (SD) percentage reduction of creatinine clearance
from pre-indinavir exposure of 44 (14) %. After indinavir dose
 Boyd et al.

optimization the mean serum creatinine returned to 135 (29) (P = 0.001). At baseline the majority of patients had Grade 2
mmol/L, representing an estimated creatinine clearance of pyuria (48.6%) and at study end the majority (51.4%) had Grade
53 (15) mL/min. The mean change in creatinine clearance 1 pyuria (P = 0.09).
from study baseline to endpoint was 6.6 (10) mL/min Figure 3 (a and b) expresses the changes in indinavir C2h and
Ctrough at baseline and pharmacokinetic steady state 4 weeks
Table 1. Baseline characteristics following final indinavir dose reduction, respectively. The overall
change in indinavir C2h was –4.8 (–6.1 to –3.5) mg/L; (P < 0.001).
Number of The overall change in indinavir Ctrough was –0.8 (–1.1 to –0.4)
Characteristics patients (n = 35) mg/L; (P < 0.001). The percentage of patients experiencing an
indinavir C2h associated with nephrotoxicity (C2h > 10 mg/L) fell
No of patients 35 from 69% to 6% (P < 0.001); the corresponding figures for those
Gender (M:F), n (%) 26:9 (74:26) with an acceptable indinavir Ctrough (Cmin > 0.1 mg/L) before
Mean age, years (SD) 40.3 (5.8) and after dose adjustment were 97 and 86%, respectively (P =
Mean body weight, kg (SD) 56.7 (9.1) 0.05). Two patients demonstrated increases in C2h following final
Mean BMI, kgs/m2 (SD) 21.5 (2.8) indinavir dose reduction when compared with baseline; one
BMI > 25+ kgs/m2, n (%) 4 (11.4) patient switched from a three times daily regimen to a
Median CD4+ count, cells/mm3 (IQR) 111 (58–218) ritonavir-boosted twice daily regimen; the other probably repres-
Median plasma HIV-RNA, 4.3 (4.0–4.8) ents an error. One patient demonstrated a very high indinavir
log10 copies/mL (IQR) Ctrough at baseline, which responded well to dose adjustment
HIV RNA <50 copies/mL, n (%) 28 (80) from ritonavir-boosted indinavir 800/100 mg twice daily to

Downloaded from jac.oxfordjournals.org by guest on April 4, 2011

CDC categories, n (%) 400/100 mg twice daily. This case was a male with an excep-
A 5 (14.3) tionally low BMI (16 kg/m2).
B 22 (62.9) Figure 4 demonstrates changes in resting BP between baseline
C 8 (22.9) and study endpoint. The figure also includes data from 15/35
Median duration of indinavir 3.5 (1.6–3.6) (43%) patients for whom a formal BP assessment had been docu-
exposure, years (IQR) mented prior to commencement of indinavir-based therapy. In an
Antiretroviral therapy, n (%) ITT analysis of BP changes in the entire cohort from the present
indinavir/ritonavir/lamivudine/zidovudine 13 (37.1) study baseline to study end we observed a reduction in both
indinavir/ritonavir/Combivir 2 (5.7) median systolic BP [baseline 120 (110–130), endpoint 110
indinavir/ritonavir/efavirenz/lamivudine 1 (2.9) (110–120) mm Hg; P = 0.2] and diastolic BP [baseline 80
indinavir/lamivudine/zidovudine 1 (2.9) (70–90), endpoint 80 (70–80) mm Hg; P = 0.02]. Restricting
indinavir/ritonavir/lamivudine 1 (2.9) the analysis to the 15 patients for whom a pre-indinavir BP
indinavir/ritonavir/efavirenz 15 (42.9) measurement existed, we observed a substantial rise in systolic
indinavir/ritonavir/stavudine/lamivudine 1 (2.9) BP [mean (SD) = 20 (10–30) mm Hg; P = 0.013] and diastolic
indinavir/lamivudine 1 (2.9) BP [mean rise = 10 (0–20) mm Hg; P = 0.011] after exposure to
Indinavir dosage indinavir and before indinavir dose optimization took place.
800 mg 27 (77.1) At baseline 7/35 (20%) subjects had an HIV RNA >50 copies/
600 mg 6 (17.1) mL but were entered into the indinavir dose reduction protocol
400 mg 2 (5.7) because of evidence of both prior long-term virological control
Median serum creatinine, mmol/L (IQR) 159 (134–167) (the result was therefore considered likely to represent a virolo-
Creatinine > 124 mmol/L, n (%) 29 (82.6) gical ‘blip’) and indinavir pharmacokinetics consistent with
Mean indinavir C2h, mg/L (SD) 10.7 (4.1) toxicity or persistent renal symptoms at the screening assessment.
Mean indinavir Ctrough, mg/L (SD) 1.4 (1.1) At study end on an ITT analysis the same proportion (20%, but
Median systolic BP, mm Hg (IQR) 120 (110–130) not all the same patients) had HIV RNA >50 copies/mL.
Median diastolic BP, mm Hg (IQR) 80 (70–90) However, continued routine follow-up of the cohort has demon-
strated that all patients who registered an HIV RNA result
>50 copies/mL at study end subsequently registered an HIV

100 IDV dose 100 RTV dose

800 mg 100 mg
P < 0.001 600 mg
Percentage of patients

P = 0.063 200 mg
75 400 mg 75
200 mg
50 50

25 25

0 0
Baseline End point Baseline End point

Figure 2. Indinavir (IDV) and ritonavir (RTV) dose distributions (per dose) at baseline and at study end.

1164
 TDM in indinavir-associated renal impairment
RNA <50 copies/mL at the following visit, without any further
adjustment to drug dosages or regimen. Measured in terms of
immune status there was an improvement observed in CD4 count
over the course of the study despite the overall reduction in total
indinavir dose [median change in CD4 count = 32 cells/mm3
(0–86); P = 0.002].
Table 2. Mean (SD) serum creatinine, proportion of patients with
creatinine >124 mmol/L and mean calculated creatinine clearance
(SD) in all patients pre-indinavir exposure, at study baseline,
and at study end (ITT and OT analyses)
Lab
Creatinine> 124 mmol/L, n (%)
pre-indinavir exposure
baseline
endpoint, ITT
JAC
antiviral
Discussion
In this study we have demonstrated an association between the
use of pharmacokinetically guided indinavir dose reductions and
amelioration of indinavir-associated renal toxicity, at least in the
short-term. The implication of this finding is that in settings in
which there may not be an option to switch from indinavir to a
protease inhibitor associated with less nephrotoxicity, optimiza-
tion of the indinavir dosage is a reasonable alternative that is safe
and allows maintenance of antiviral efficacy.
Recent studies have suggested that indinavir may be used at
lower doses than those currently recommended without loss of
efficacy and with reduced nephrotoxicity.17,18 A pharmacokinetic
study performed in Thai patients receiving ritonavir-boosted
Results indinavir 400/100 mg twice daily demonstrated PK levels con-
sistent with reasonable efficacy and minimal nephrotoxicity, and
the clinical study of the cohort from which this patient sample
0 was drawn has demonstrated this to be the case.19,20 In the pre-
29/35 (82.6) sent study we have demonstrated benefits of indinavir dose reduc-
24/35 (68.6) tion measured in terms of improvements in serum creatinine and
thereby calculated creatinine clearance. Importantly, calculated cre-

Downloaded from jac.oxfordjournals.org by guest on April 4, 2011

endpoint, OT 20/31 (64.5)
P value between pre-indinavir exposure 0.031 atinine clearance rose by a mean of 6.6 (10.2) mL/min and attained
(n = 35) and baseline (n = 35) a mean (SD) level of >50 mL/min [52.7 (14.9)], a level consistent
P value between endpoint and baseline ITT: P = 0.180; with maintenance of reasonable clinical renal function. In associ-
OT: P = 0.016 ation with this we found that pyuria improved. A number of
Mean serum creatinine (SD), mmol/L studies have suggested that pyuria is a potential marker of renal
pre-indinavir exposure (n = 35) 81.9 (12.7) damage,11,21 and we found that the majority of patients entering
baseline (n = 35) 149.1 (27.7) this study displayed a urinary leucocytosis. Despite this improve-
endpoint (n = 35) 134.8 (29) ment the results might be considered a disappointment in that
change at baseline from pre-indinavir 67.2 (28.3), renal function improved only marginally and did not return to
exposure P < 0.001 normal. It should be remembered however that this patient cohort
change at endpoint from baseline –14.2 (26.8), had had substantial exposure to indinavir for a median (IQR) of
P = 0.003 3.5 (1.6–3.6) years and that previous reports of resolution of renal
Estimated creatinine clearance, mL/min impairment in patients receiving indinavir were in the context of
pre-indinavir exposure (n = 35) 83.5 (13.8) cessation of the drug soon after the identification of renal tox-
baseline (n = 35) 46 (11) icity. The study cohort continues in clinical follow-up and it
endpoint (n = 35) 52.7 (14.9) remains to be seen whether the short-term changes in renal status
change at baseline from pre-indinavir –37.5 (15.5), translate into more substantial long-term benefits.
exposure P < 0.001 Concern might reasonably be expressed that three patients
change at endpoint from baseline 6.6 (10.2), with HIV RNA <50 copies/mL at the beginning of the study
P = 0.001 had detectable HIV RNA following indinavir dose adjustment,
suggesting that indinavir dose adjustment may dispose patients
ITT, intention to treat; OT, on treatment. to virological failure. However, in all three cases the indinavir

20 6
IDV C2h (a) IDV Ctrough (b)
Serum indinavir concentration (mg/L)

P < 0.001
P < 0.001 5
16

4
12
3
8
2

4
1

0 0
Baseline End point Baseline End point

Figure 3. Indinavir (IDV) C2h (a) and Ctrough (b) at baseline and following dose optimization.

1165
 Boyd et al.
175 Systolic blood pressure Diastolic blood pressure
P = 0.013 P = 0.174
150
Blood pressure (mm Hg)
125
P = 0.011 P = 0.018
100
75
50
Pre IDV Baseline End point Pre IDV Baseline End point
exposure exposure
Samples 15 35 35 15 35 35
Figure 4. Systolic and diastolic blood pressures (median, IQR and range) before
exposure to indinavir (IDV), at current study baseline and at study endpoint.
pharmacokinetic parameters after dose adjustment fell well within
the limits considered to be consistent with maintenance of viro-
logical control in indinavir treatment naive patients,22 suggesting
that other temporary factors may have been responsible for the
increases in viral load. Indeed, further follow-up of this patient
cohort at HIV-NAT has demonstrated that all patients with an
HIV RNA >50 copies/mL at study end subsequently returned an
HIV RNA <50 copies/mL at the following clinic visit.
Because of the known association between renal impairment
and hypertension, we elected to monitor resting BP throughout
the study. We also managed to collect pre-indinavir BP assess-
ments on 43% of patients, which we were able to compare with
BP assessments at study baseline and following indinavir dose
adjustment. We found that patients qualifying for study in whom
measurements of BP were available from before exposure to
indinavir had experienced substantial elevations in both systolic
and diastolic BP after indinavir exposure. After indinavir dose
adjustment, systolic and diastolic BP improved in the cohort as a
whole (n = 35), although this only reached statistical significance
for the diastolic component (P = 0.02). In a restricted analysis of
the 15 patients for whom we had documented pre-indinavir BP
values, we found a clinically significant median elevation in
systolic and diastolic BP of 20 and 10 mm Hg, respectively,
from the period between indinavir exposure and baseline for
the current indinavir dose-optimization study. Changes in
systolic and diastolic BP parameters following indinavir dose-
optimization in this restricted group of 15 were similar to that
observed for the cohort as a whole. From a clinical point of view
minimization of nephrotoxicity is an important end in itself, but
the increased risk for elevated BP in the presence of renal impair-
ment makes this goal even more imperative. Given the potential
for metabolic disturbance in patients receiving indinavir, the
added risk of hypertension would likely increase the risk of
cardiovascular disease further.23,24 Recent studies suggest that
when indinavir is commenced at lower doses there is likely to
be less toxicity and therefore less risk of the additional complica-
tion of hypertension.17,20
Another feature of note in the study is that indinavir dose-
optimization necessitated not only reduction of indinavir doses,
but also concomitant increases in the ritonavir dose in some
1166
cases. This was the case in patients for whom the indinavir
C2h remained comparatively high despite an acceptable Ctrough
following indinavir dose reduction. Therefore, the indinavir dose
was decreased (by 200 mg) and ritonavir dose was increased (by
100 mg) in an attempt to lower the indinavir C2h but maintain a
therapeutic Ctrough. A number of reports have demonstrated the
large patient inter-individual PK variability of indinavir,7,8,25,26
and the one patient with dramatically high Ctrough at screening in
this study is an illustrative case in point. This patient also had an
extremely low BMI (16 kg/m2) and as a result was probably at
increased risk for indinavir toxicity.4,27 Given this physiological
and pharmacokinetic variation it follows that a variety of dif-
ferent combinations of indinavir and ritonavir will be necessary
in order to fully optimize indinavir pharmacokinetics on an
individual basis.
As a result of the limited sample size we had little statistical
power to assess other potential risk factors for nephrotoxicity in
this cohort. However, we did find that nearly three-quarters of the
cohort had been exposed to co-trimoxazole prophylaxis, and
interestingly, that every single patient from the longest running
Downloaded from jac.oxfordjournals.org by guest on April 4, 2011
HIV-NAT study of indinavir-containing therapy (>3 years) who
qualified for enrolment in the present indinavir dose optimization
study had received uninterrupted co-trimoxazole prophylaxis
since at least the time of enrolment in their original (non-dose
optimized) indinavir-containing sutdy.14 This might be inter-
preted as suggestive that co-trimoxazole prophylaxis is indeed
an important risk-factor for nephrotoxicity,4 but may also reflect
that older patients (mean age of this cohort was 40 years), and/or
those with a longer period of HIV-infection, or with poorer
immune reconstitution, are at greater risk. These confounders
cannot be disentangled in this study.
There are a number of weaknesses of our study. First, we were
unable to include a control group with nephrotoxicity who
remained on full-dose indinavir. While published data suggested
that the indinavir-associated renal impairment was reversible,
this was based on short-term experiences of renal impairment
only.11–13 We believed that many of the patients experiencing
indinavir-related nephrotoxicity were at risk for chronic, irrevers-
ible renal impairment, and given the emerging evidence at the
time regarding the utility of therapeutic drug monitoring for
optimization of indinavir dosing, it was considered unjustifiable
not to offer therapeutic drug monitoring to all patients experien-
cing renal toxicity. Second, we were unable to perform pharma-
cokinetics in those patients who had not manifested symptoms or
signs of nephrotoxicity to determine whether indinavir PK pro-
files differed significantly from those in whom toxicity was mani-
fest. In previous analyses of indinavir-associated renal toxicity,
associations have been found with factors other than pharma-
cokinetics [e.g. low body mass index (<20 kg/m2)] as well as
gender (female > male) and exposure to co-trimoxazole].4,11,27
However, these analyses were conducted without the inclusion of
pharmacokinetic data. It is possible that the determinants of renal
toxicity are unrelated to indinavir pharmacokinetics or, more
likely, that the individual pharmacokinetics of indinavir are a
necessary but not sufficient condition for the development of
nephrotoxicity [i.e. patients with low body weight/body mass
index and/or receiving co-trimoxazole prophylaxis may be at
increased risk for nephrotoxicity (sufficient conditions), but may
develop nephrotoxicity if certain key threshold pharmacokinetic
values are reached (necessary condition)]. Third, the study sam-
ple size was small, not allowing us to authoritatively investigate
 TDM in indinavir-associated renal impairment
the role of other potential risk factors for nephrotoxicity. Finally,
the follow-up period in this study is short. A longer duration of
follow-up would allow a better understanding of the long-term
implications of indinavir-associated nephrotoxicity, the improve-
ments observed in renal function following indinavir dose adjust-
ment, as well as continued virological control.
In conclusion, this study has demonstrated that dose adjust-
ments of indinavir in targeted patients experiencing indinavir-
associated nephrotoxicity can improve renal function in the
short term and allow safe continuation of indinavir-based com-
bination antiretroviral therapy in those with little option for
switch. This finding is useful because indinavir is an attractive
option as a first-line PI option in resource-limited settings, par-
ticularly in patients failing first-line therapy with an NNRTI/
2NRTI combination or in those with hypersensitivity to the
NNRTIs. Further study of ritonavir-boosted indinavir in
resource-limited settings at doses lower than those currently
recommended is warranted.
Acknowledgements
We acknowledge the patients who participated in the study as well as
the physicians, nursing staff, and laboratory and administrative
personnel who all contribute to patient care at HIV-NAT. Financial
support for the study was from HIV-NAT funds; it was not
sponsored by a pharmaceutical company or any external agency.
Transparency declarations
The authors have no conflicts of interest to declare.
References
1. Kempf DJ, Marsh KC, Kumar G et al. Pharmacokinetic enhance-
ment of inhibitors of the human immunodeficiency virus protease by
coadministration with ritonavir. Antimicrob Agents Chemother 1997; 41:
654–60.
2. Gulick RM, Meibohm A, Havlir D et al. Six-year follow-up of HIV-1-
infected adults in a clinical trial of antiretroviral therapy with indinavir,
zidovudine, and lamivudine. AIDS 2003; 17: 2345–9.
3. Kopp JB, Miller KD, Mican JA et al. Crystalluria and urinary
tract abnormalities associated with indinavir. Ann Intern Med 1997; 127:
119–25.
4. Boubaker K, Sudre P, Bally F et al. Changes in renal function
associated with indinavir. AIDS 1998; 12: F249–54.
5. Berns JS, Cohen RM, Silverman M et al. Acute renal failure due to
indinavir crystalluria and nephrolithiasis: report of two cases. Am J Kidney
Dis 1997; 30: 558–60.
6. Vigano A, Rombola G, Barbiano di Belgioioso G et al. Subtle
occurrence of indinavir-induced acute renal insufficiency. AIDS 1998;
12: 954–5.
7. Dieleman JP, Gyssens IC, van der Ende ME et al. Urological com-
plaints in relation to indinavir plasma concentrations in HIV-infected
patients. AIDS 1999; 13: 473–8.
8. Burger D, Boyd M, Duncombe C et al. Pharmacokinetics and
pharmacodynamics of indinavir with or without low-dose ritonavir in
HIV-infected Thai patients. J Antimicrob Chemother 2003; 51: 1231–8.
9. Burger D, Hugen P, Reiss P et al. Therapeutic drug monitoring
of nelfinavir and indinavir in treatment-naive HIV-1-infected individuals.
AIDS 2003; 17: 1157–65.
1167
JAC
antiviral
10. Fletcher CV, Anderson PL, Kakuda TN et al. Concentration-con-
trolled compared with conventional antiretroviral therapy for HIV infection.
AIDS 2002; 16: 551–60.
11. Sarcletti M, Petter A, Romani N et al. Pyuria in patients treated
with indinavir is associated with renal dysfunction. Clin Nephrol 2000;
54: 261–70.
12. Kopp JB, Falloon J, Filie A et al. Indinavir-associated interstitial
nephritis and urothelial inflammation: clinical and cytologic findings. Clin
Infect Dis 2002; 34: 1122–8.
13. Reilly RF, Tray K, Perazella MA. Indinavir nephropathy revisited: a
pattern of insidious renal failure with identifiable risk factors. Am J Kidney
Dis 2001; 38: E23.
14. Boyd MA, Srasuebkul P, Khongphattanayothin M et al. Boosted
versus unboosted indinavir with zidovudine and lamivudine in nucleoside
pre-treated patients: a randomized, open-label trial with 112 weeks of
follow-up (HIV-NAT 005). Antivir Ther 2006; 11: 223–32.
15. Boyd MA, Siangphoe U, Ruxrungtham K et al. Indinavir/ritonavir
800/100 mg bid and efavirenz 600 mg qd in patients failing treatment
with combination nucleoside reverse transcriptase inhibitors: 96-week
outcomes of HIV-NAT 009. HIV Med 2005; 6: 410–20.
16. Droste JA, Verweij-Van Wissen CP, Burger DM. Simultaneous
determination of the HIV drugs indinavir, amprenavir, saquinavir, ritonavir,
Downloaded from jac.oxfordjournals.org by guest on April 4, 2011
lopinavir, nelfinavir, the nelfinavir hydroxymetabolite M8, and nevirapine in
human plasma by reversed-phase high-performance liquid chromato-
graphy. Ther Drug Monit 2003; 25: 393–9.
17. Duvivier C, Myrto A, Marcelin AG et al. Efficacy and safety
of ritonavir/indinavir 100/400 mg twice daily in combination with two
nucleoside analogues in antiretroviral treatment-naive HIV-infected indi-
viduals. Antivir Ther 2003; 8: 603–9.
18. Wasmuth JC, la Porte CJ, Schneider K et al. Comparison of two
reduced-dose regimens of indinavir (600 mg vs 400 mg twice daily) and
ritonavir (100 mg twice daily) in healthy volunteers (COREDIR). Antivir
Ther 2004; 9: 213–20.
19. Boyd M, Mootsikapun P, Burger D et al. Pharmacokinetics of
reduced-dose indinavir/ritonavir 400/100 mg twice daily in HIV-1-infected
Thai patients. Antivir Ther 2005; 10: 301–7.
20. Mootsikapun P, Chetchotisakd P, Anunnatsiri S et al. Efficacy and
safety of indinavir/ritonavir 400/100 mg twice daily plus two nucleoside
analogues in treatment-naive HIV-1-infected patients with CD4+ T-cell
counts <200 cells/mm3: 96-week outcomes. Antivir Ther 2005; 10: 911–6.
21. Gagnon RF, Tecimer SN, Watters AK et al. The natural history of
leukocyturia associated with indinavir treatment in HIV+ individuals. Am J
Nephrol 2000; 20: 448–54.
22. Department of Health and Human Services. Panel on Clinical
Practices for Treatment of HIV Infection. Guidelines for the Use of
Antiretroviral Agents in HIV-1 Infected Adults and Adolescents, USA,
October 06, 2005. http://aidsinfo.nih.gov/ContentFiles/Adultand
AdolescentGL.pdf (15 January 2006, date last accessed).
23. Dragsted UB, Gerstoft J, Pedersen C et al. Randomized trial to
evaluate indinavir/ritonavir versus saquinavir/ritonavir in human immun-
odeficiency virus type 1-infected patients: the MaxCmin1 Trial. J Infect Dis
2003; 188: 635–42.
24. Friis-Moller N, Sabin CA, Weber R et al. Combination antiretroviral
therapy and the risk of myocardial infarction. N Engl J Med 2003; 349:
1993–2003.
25. van Heeswijk RP, Veldkamp AI, Hoetelmans RM et al. The steady-
state plasma pharmacokinetics of indinavir alone and in combination with a
low dose of ritonavir in twice daily dosing regimens in HIV-1-infected indi-
viduals. AIDS 1999; 13: F95–9.
26. Csajka C, Marzolini C, Fattinger K et al. Population pharmacokin-
etics of indinavir in patients infected with human immunodeficiency virus.
Antimicrob Agents Chemother 2004; 48: 3226–32.
27. Meraviglia P, Angeli E, Del Sorbo F et al. Risk factors for indinavir-
related renal colic in HIV patients: predictive value of indinavir dose/body
mass index. AIDS 2002; 16: 2089–93.