The following section will discuss the role of TDM for PIs and NNRTIs only. Nucleoside analogues require intracellular activation, and levels of intracellular drug-triphosphate bears little relationship to plasma levels of parent compound – TDM is likely to be of little value in this group of drugs. The abbreviations Cmin, Cmax and AUC refer to the plasma trough (end of the dosing interval) and peak plasma levels and area under the time-concentration curve respectively.
Since the previous guidelines were last published, there has been
increasing uptake of TDM in the UK and across Europe. This is despite a
surprising lack of data to confirm any benefit of TDM in routine
clinical use. Nevertheless there are compelling arguments for TDM:
1 Low plasma drug levels correlate with virological
failure. This represents the strongest case for TDM. A
considerable body of data have now accrued to suggest that treatment
failure (as judged by viral load response) is associated with low
plasma levels of saquinavir (SQV), nelfinavir (NFV), indinavir (IDV),
ritonavir (RTV), amprenavir (APV) and efavirenz (EFV). These
derive from Phase II (SQV, IDV, APV) studies, trials of patients
commencing ART (SQV, NFV, IDV, RTV, NVP), dual PI regimens (SQV, NFV,
RTV), salvage ART (SQV, IDV, RTV) or else a broader population of HIV
patients on ART (SQV, NFV, IDV, RTV, EFV) [e.g. Acosta, Marzolini,]. A
direct relationship between plasma levels (AUC and/or Cmin) and
magnitude of viral load reduction following ART has been defined for
many of these drugs. In addition, patients randomised into both the
genotyping and standard of care arms of the VIRADAPT Study had
significantly different viral load responses depending on whether
plasma PI levels were optimal or not. These data are important since
they clearly indicate that achieving therapeutic plasma drug
concentrations had an added and separate effect upon likelihood of
success of the ART regimen.
Plasma PI concentrations are subject to a large degree of
inter-individual variability, and it is not uncommon to observe a
100-fold variability in trough concentrations of SQV, IDV, NFV and RTV.
It is therefore impossible to predict what plasma levels any given
individual will achieve on standard dosing. TDM could potentially be
used to identify patients at risk of treatment failure.
2 High plasma drug levels may predict toxicity. Although
definitive studies are lacking, there is no clear link between
abnormalities of liver function, risk of developing glucose intolerance
or lipodystrophy and plasma PI levels. Adverse effects such as rash or
hypersensitivity are likely to be related to idiosyncratic drug
reactions rather than to the amount of circulating drug in plasma.
Nevertheless there are limited data linking some toxicities to plasma
drug levels. A relationship has been observed between plasma RTV levels
(Cmax, Cmin, AUC) and elevated triglyceride levels. Gastrointestinal
intolerance and circumoral parasthesia may also be related to the C max
of RTV. High plasma IDV levels are associated with increased risk of
urological symptoms (renal colic, haematuria, dysuria). It is possible
that high plasma EFV levels may also be associated with increased risk
of CNS toxicity. Most importantly, TDM may be utilised to allow dosage
reduction in patients who are most at risk of drug toxicity (e.g.
previous intolerance, concurrent medication with overlapping
toxicities, other pre-existing disease).
3 Adherence. TDM may have a limited role in monitoring
adherence to ART [Burger]. The half-life of most PIs is short (2-10
hours), although co-administration of RTV will prolong this. Due to
wide inter-individual variability, near or complete absence of
detectable drug in plasma is a good indicator of poor adherence, but
sub-therapeutic levels are of limited usefulness. An adequate or high
plasma drug level only provides information about adherence over the
preceding few doses, rather than in the long term.
4 Drug interactions. PIs and NNRTIs are extensively
metabolised by cytochrome P450 CYP 3A4. They may not only affect the
metabolism of other drugs that share the same metabolic pathway, but
may themselves be affected by those drugs. Individual drug interactions
are beyond the scope of this discussion and are found elsewhere (e.g.
Error! Reference source not found.. TDM could potentially be
used to monitor PI/NNRTI levels in such circumstances.
5 Special groups. Certain groups of patients are
particularly susceptible to under- or overdosing with PIs/NNRTIs. These
include young children (whose liver function and metabolic rates vary
from adults, and alter with time) and patients with impaired liver
function.
Problems with TDM
PIs exhibit intra-individual (within patient) in addition to
inter-individual variability. This could pose problems for TDM. PIs are
also extensively bound (IDV 60%, APV 90%, other PIs >95%) to the acute
phase protein a
1-acid glycoprotein (AAG). Levels of AAG fluctuate with disease stage,
and acute opportunistic infection. It is important that TDM is not
performed during acute illness such as acute opportunistic infection.
There is a lack of consensus from centres offering TDM over sampling
strategy (troughs, troughs + post-dose, ‘random’ measurements in
relation to the expected population range for that drug) although most
would accept that the ‘gold standard’ of AUC is not practicable for
routine monitoring. There is an emerging consensus on what target
levels should be for each drug but differences still exist between
laboratories. An international quality assurance programme has also
been instituted.
Proposed indications for TDM:
1 Routine use (DIII). There are currently insufficient
data supporting the routine use of TDM in all patients receiving ART.
There is an urgent need for studies in this group of patients. TDM may
be performed where a drug is being used at doses outside of those
recommended by the manufacturer (as listed in the Data Sheet SPC).
2 Liver impairment (BII). TDM is likely to be of clinical
value in patients with severe liver impairment.
3 Monitoring Adherence (CII). TDM may be have a limited
role; see the caveats listed above.
4 Drug Interaction (BII). TDM should be considered in
patients on regimens including a single PI+NNRTI, or PI +
inducer/inhibitor of CYP3A4.
5 Paediatric patients (CIII). TDM is useful in children
aged
PIs/NNRTIs.
6 Minimising toxicity (CIII). TDM may be helpful in the
case of dose-related toxicities. More usually, a high level may allow
the option of dosage reduction in patients who are unlikely to have
drug resistant virus, in order to reduce the risk of toxicity.
7 Failure of ART (CIII). There is probably little point in
utilising TDM once high-level antiviral resistance has developed. TDM
may be considered when treatment intensification is an option, e.g.
when viral load reduction following a new ART regimen is sub-optimal,
or where viral resistance testing suggests that resistance is unlikely,
or to overcome a low-level virological rebound.
The Way Ahead
The next few years should see changes to these recommendations
as data emerge. Particular interest has centered on the concept of
‘inhibitory quotients’ i.e. the amount of drug (usually trough level)
that is above the IC50 of the HIV isolate for that particular patient.
This combines TDM with phenotypic (or ‘virtual phenotype’) assays and
anticipates that the risk-benefit ratio of higher drug concentrations
will vary between different groups of patients. In addition, there are
moves to make TDM more ‘user-friendly’ by moving sampling strategies
away from directly measured troughs or peaks, towards predicted troughs
derived from random sampling using sophisticated population
pharmacokinetic modelling.
References
Acosta EP, Kakuda TN,
Brundage RC, Anderson PL, Fletcher C.V. Pharmacodynamics of human
immunodeficiency virus type 1 protease inhibitors. Clin Infect Dis
2000;30 (suppl 2):S151-59
Marzolini C, Telenti A,
Decosterd LA, Greub G, Biollaz J, Buclin T. Efavirenz plasma levels can
predict treatment failure and central nervous system side effects in
HIV-1 infected patients. AIDS 2001;15:71-75