Carboplatin differs from cisplatin by its pharmacokinetics and toxicity pro
file. Carboplatin is mainly eliminated by the kidneys and its dose-limiting
toxicity is the bone marrow suppression. Myelosuppression of carboplatin i
s more closely correlated with the area under the curve (AUC) of ultrafiltr
able plasma concentration versus time to which a patient is exposed, than i
t is with the administered dose. Similar relationships have been shown betw
een A UC and antitumour effect, although they are in a smaller number and l
ess close. Several method of dosage individualisation a priori (before carb
oplatin administration) have been proposed. Since carboplatin is often pres
cribed to patients with altered renal functions, this dose optimisation is
particularly justified. Dose individualisation is based on both equations a
llowing to predict the patient carboplatin clearance and the choice of targ
et A UC. The different equations proposed are bared on direct measurement o
f the renal filtration glomerular rate or an patient demographic and biolog
ical characteristics such as weight and serum creatinine. The respective ad
vantages and limits of these equations are now well known. However, the val
ues of optimal AUC that depend on cytotoxic drugs combined to carboplatin a
nd the patient hematopoietic status, are not precisely determined for each
protocol of chemotherapy. When carboplatin is given by reiterated administr
ations within each course, it is possible to adjust the last doses accordin
g to a limited number of blood samples following the first infusion and a B
ayesian analysis of the observed plasma concentrations. These methodologies
are more complex, but they may be useful for the intensification protocols
Carboplatin is still the only cytotoxic drug for which dose is individuali
sed not according to the body surface area but according to pharmacokinetic
parameters.