Annals of Oncology 8: 1049-1051, 1997.
© 1997 Kluwer Academic Publishers. Primed in the Netherlands.
Short report
Severe CPT-11 toxicity in patients with Gilbert's syndrome:
Two case reports
E. Wasserman,1 A. Myara,2 F. Lokiec,3 F. Goldwasser,1 F. Trivin,2 M. Mahjoubi,4 J. L. Misset1 &
E. Cvitkovic1
'FSMSIT, Hop Paul Brousse, Villejuif, 2Hop St Joseph, Paris, 3Centre Rene Huguenin, Saint Cloud, 4Rhone-Poulenc Rarer, Neuilly, France
alkaline methanolysis and HPLC.
Results: Both patients presented grade 4 neutropenia and/
Background: CPT-11 is hydrolyzed to its active metabolite or diarrhea (NCI-CTQ in every treatment cycle. Biliary index
SN-38, which is mainly eliminated through conjugation by (after Gupta et al) values were well above 4000.
hepatic undine diphosphate glucuronosyl transferases (UGTs)
Conclusion: We present the first clinical evidence linking
to the glucuronide (SN-38G) derivative. Preclinical studies bilirubin glucuronidation status and CPT-11 related toxicity.
showed that UGT*1.1 is the isozyme responsible for SN-38 The severe toxicity experienced by the two patients with
glucuronidation. Patients with Gilbert's syndrome have defi- Gilbert's syndrome treated with CPT-11 based chemotherapy
cient UGT*1.1 activity, therefore may have an increased risk has a genetic basis. Individuals with Gilbert's syndrome have
for related CPT-11 toxicity.
an enhanced risk for CPT-11 toxicity. Unconjugated serum
Patients and methods: Two patients with metastatic colon bilirubin could be predictive parameter of CPT-11 toxicity.
cancer and Gilbert's syndrome were treated with CPT-11 based
chemotherapy. CPT-11, SN-38 and SN-38G pharmacokinetics Key words: bilirubin, CPT-11, Gilbert's syndrome, glucuroniparameters were obtained. Serum bilirubin was analysed by dation, SN-38
Summary
drug disposition of molecules such as acetaminophen,
leading to enhanced toxicity as a result [6]. Thus, patients with Gilbert's syndrome may be at an increased
risk for CPT-11 related toxicity.
The two present case reports of severe CPT-11 toxicity in Gilbert's syndrome patients, underline the key
role of SN-38 glucuronidation for CPT-11 toxicity and
are the first clinical evidence linking bilirubin glucuronidation status and CPT-11 related toxicity. They call for
innovative approaches for both prediction and monitoring of CPT-11 induced clinical toxicity.
Both observations were done during the early steps
of an ongoing phase I trial, which included pharmacokinetics, of the Oxaliplatin and CPT-11 combination
regimen sequentially, given i.v. on the same day every
three weeks [7].
Patients and methods
Patient A, a 49-year-old italian woman with liver and lung metastases
arising from colon cancer had progressed on two previous fluoropyrimidine based regimens. Blood tests available in the prior five years
exhibited asymptomatic mild unconjugated hyperbilirubinemia (range
19-38 umol/1) coincident with episodes of gastroenteritis or 5-FU
based chemotherapy. At the inclusion in the Oxaliplatin/CPT-11 trial,
her baseline bilirubin and other liver parameters were normal, without
evidence of hemolysis (normal reticulocytes and haptoglobine, no
schistocvtes).
Downloaded from http://annonc.oxfordjournals.org/ by guest on September 9, 2014
CPT-11 is a DNA topoisomerase I inhibitor, derived from
Camptothecine (CPT), which acts in vivo as a prodrug
and is converted to SN-38 by a carboxylesterase. SN-38
is considered to be responsible for both the antitumoral
activity and the toxicity of CPT-11 [1].
The SN-38 plasma concentrations reflect both the
rate of CPT-11 metabolization and the rate of SN-38
elimination, which is mainly through conjugation by
hepatic uridine diphosphate glucuronosyltransferases
(UGTs) to the glucuronide (SN-38G) derivative.
Glucuronidation seems an important variable in the
determination of activity as well as in the toxicity.
Recently, Gupta et al. [2] demonstrated, in a murine
model, that UGT may be a site of modulation (induction-inhibition) of the SN-38 clearance with relevant
pharmacokinetic consequences. They have proposed a
SN-38 biliary index [3] with moderate predictive value
of severe diarrhea when CPT-11 is administered weekly.
Patients with biliary index values ^ 4000 had a significantly higher risk of grade 3-4 diarrhea.
It has been recently shown that the UGT* 1.1 isozyme,
which glucuronidates bilirubin, is also the isoform responsible for SN-38 glucuronidation [4].
Individuals with Gilbert's syndrome have deficient
UGT* 1.1 activity. This inherited disorder, characterized
by abnormal bilirubin glucuronidating activity, presents
mild, chronic, unconjugated hyperbilirubinemia [5].
UGT deficiencies are asssociated with changes in
1050
Transient Billrubin Rise after CPT-11 + Oxaliplatin
Patient A 10 cycles, Patient B 2 cycl
Max
Min
75%
25%
Median
Upper Limit Normal
Bilirubin
DiyO
Day 3-fl
Day 9-17
Figure 1. Kinetics of total bilirubin after CPT-11 based chemotherapy.
Table 1. Results.
Cycles Oxaliplatin/CPT-11 given
Bil total umol/1, baseline (median)
Bil total umol/1, maximal increase
(median)
Pharmacokinetic analysis (1 cycle)
AUC (ng • h/ml)
CPT-11
SN-38
SN-38G
Ratio SN-38G/SN-38
Serum bilirubin (AM-HPLQ"
Unconjugated bil (%)
Conjugated bil (%)b
Monoconjugates (%)
Diconjugates (%)
Ratio (diconjugates/mono +
diconjugates)0
Toxicity Gr 4/cycles
% Increase total bil over baseline
(1 cycle)
Biliary Indexd
Patient A
Patient B
10
31
2
21
52
43
15386
419
784
1.8
29975
674
2843
42
98 4
1.6
1 25
0.35
97.3
2.7
1 18
1.52
0.22
Diarrhea/1 and
Neutropenia/10
0.56
Diarrhea/1 and
Neutropenia 12
178
8222
89
7106
* Analysis performed in one cycle (patient A. day 5, cycle 5, patient B:
day 5, cycle 2).
b
Reference range: 4%, Gilbert's syndrome values: 0.9% ± 0 4%.
c
Reference range: 0.54% ±11, Gilbert's syndrome values: 0.33% ± 0.10%.
d
After Gupta et al.
Discussion
We report two patients with antecedents of defective
handling of unconjugated bilirubin, treated in a phase I
trial with CPT-11 based chemotherapy. In all the cycles
they evidenced severe toxicity (neutropenia and/or diarrhea).
The pharmacokinetic parameters in these two patients show a biliary index well above the 4000 thereshold
value proposed by Gupta et al. As a very important
interpatient variability exists for both SN-38 and SN38G pharmacokinetics (70%-90%) [10], this means that
individual deficiency in SN-38 glucuronidation, as in
Downloaded from http://annonc.oxfordjournals.org/ by guest on September 9, 2014
After the first cycle of Oxaliplatin and CPT-11, given at 85 mg/m 2
and 150 mg/m 2 respectively, the patient presented a rise in bilirubin
level, predominantly unconjugated. The bilirubin increase began on
day two (total bil 20 umol/1), reached a maximal value on day 6
(39 umol/1), and returned to normal values on day 13 (16 umol/1).
There were no changes in alkaline phosphatase, transaminases or
gammaglutamyl transferase serum concentrations.
In each subsequent cycle, the same kinetics of bilirubin nse and fall
were detected, but with some degree of hyperbilirubinemia at day 0
(range 23-31 umol/1) (Figure 1). On the seventh day after the first
treatment cycle the patient developed grade 4 diarrhea and neutropenia, which resolved on day 12, and uneventful grade 4 neutropenia in
all of the following 10 cycles. She was the only patient out of the seven
patients treated at this same dose level to experience severe neutropenia. The CPT-11 dose was not reduced and she continues treatment
with G-CSF support (300 ug s c , day 3-9). Analysis of serum bilirubin
by alkaline methanolysis and HPLC (AM-HPLQ [8] showed a decrease
in conjugated bilirubin concentrations, increased values of the relative
amount of monoconjugates, with a decreased ratio of diconjugates to
mono plus diconjugates (Table 1). This pattern is usually described in
patients with Gilbert's syndrome [9],
Her CPT-11, SN-38 and SN-38G pharmacokinetic parameters
showed a biliary index, as defined by Gupta et al. [3] of 8222 (Table 1).
She experienced a clinically meaningful minor response, with > 50%
decrease in her baseline elevated tumor markers, still mantained after
13 cycles (10 months of treatment).
Patient B is a 63-year-old Greek man with massive liver involvement
from colon cancer, treated with various fluoropyrimidine based regimens since September 1994, until progression in September 1996. He
had familiar and personal antecedents of Gilbert's syndrome, with
periodic asymptomatic increases in unconjugated bilirubin (range
40-78 umol/1). At the time of trial inclusion, his baseline bilirubine
values were: total and conjugated bil 27 umol/1 / 9 umol/1 respectively.
He was treated with Oxaliplatin and CPT-11, given at 110 mg/m 2
and 200 mg/m 2 , respectively. At day 3, total and conjugated bilirubin
increased respectively to 52 umol/1-19 umol/1, returning to baseline
values by day 9. During the second treatment cycle he repeated the
same pattern of serum bilirubin evolution (Figure 1). His biliary index,
obtained with area under the concentration-time curve ( A U Q of
CPT-11, SN-38 and SN-38G was 7106. He presented grade 4 neutropenia during the first cycle and grade 4 neutropenia and diarrhea in the
second cycle (NCI C T Q (Table 1).
The patient discontinued chemotherapy treatment with stable disease, as a consequence of a jaundice caused by acute choledocholithiasis complicated with acute cholangitis after the second cycle, which
was treated by endoscopic sphincterotomy. He returned to Greece,
where subsequent treatment with Oxaliplatin and 5-FU has rendered a
clinically meaningful stability lasting over five months.
1051
Acknowledgements
We are indebted to P. Herait, MD for the constructive
comments throughout preparation of the manuscrit, to
Mrs. D. Paumier, Mrs. J. Santoni and Mrs. S. Weill for
their technical efficient contribution to this work. We are
grateful to Mrs. M.-J. Ferreira for typing assistance.
Correspondence to •
E. Wasserman, MD
FSMSIT - Hopital Paul Brousse
12-14 Avenue Paul Vaillant Couturier
94804 VillejuifCedex
France
Downloaded from http://annonc.oxfordjournals.org/ by guest on September 9, 2014
Gilbert's syndrome, may be masked by the huge inter- References
indivual variation of the area under the curve values.
1. O'Reilly S, Rowinsky E. The clinical status of irinotecan (CPTAs far as we know, this phenomenon of severe and
11), a novel water soluble camptothecin analogue: 1996. Crit Rev
limiting toxicities in patients with hepatic conjugation
Oncol/Hematol 1996; 24: 47-70.
deficiency treated with CPT-11 based chemotherapy,
2. Gupta E, Wang X, Ramirez J, Ratain M. Modulation of glucurconcurrent with a transient rise in bilirubin has not
onidation of SN-38, the active metabolite of irinotecan, by valbeen previously reported.
proic acid and phenobarbital. Cancer Chemother Pharmacol
T9"97; 39: 440=4;
ThTs "report matcrTwith the in vitro work of Iyer et al.
3.
Gupta
E, Lestingi T, Mick R et al. Metabolic fate of irinotecan in
[4] who demonstrated, in human liver microsomes, a
humans: correlation of glucuronidation with diarrhea. Cancer
significant correlation between SN-38 and bilirubin gluRes 1994; 54: 3723-5.
curonidation, which suggests that bilirubin UGT may
4. Iyer L, King C, Tephly T and Ratain MJ. UGT isoform 1.1
also be glucuronidating SN-38.
(UGT*1.1) glucuronidates SN-38, the active metabolite of irinotecan. Proc Am Soc Clin Oncol 1997; 16: 707 (Abstr).
Additional studies with HK293 cells transfected with
5. Bosma P, Chowdhury J, Bakker C et al. The genetic basis of the
specific human UGT isozymes showed that UGT* 1.1
reduced expression of bilirubin UDP-glucuronosyltransferase 1 in
is the isoform responsible for SN-38 glucuronidation.
Gilbert's syndrome. N Engl J Med 1995; 333: 1171-5.
Patients with Gilbert's syndrome , who have deficient
6. Morais SF, Wells PG. Enhanced acetaminophen toxicity in rats
UGT* 1.1 activity may be at an increased risk of CPT-11
with bilirubin glucuronyl transferase deficiency. Hepatology 1989;
10: 163-7.
toxicity, which is supported by our report.
7. Cvitkovic E, Wasserman E, Goldwasser F et al. Preliminary
The isolated transient bilirubin rise observed in these
report on an Oxaliplatin (L-OHP)/CPT-11 phase I trial in gastrotwo patients led us to evaluate the kinetics of serum
intestinal malignancies: an active combination. Proc Am Soc Clin
bilirubin after CPT-11 administration in all the patients
Oncol 1997; 16:711 (Abstr).
of the ongoing Oxaliplatin/CPT-11 combination trial, as
8. Muraca M, Blanckaert N. Liquid-chromatographic assay and
identification of mono- and diester conjugates of bilirubin in
well as in a retrospective review of data from CPT-11
2
normal serum, Clin Chem 1983; 29: 1767-71
given as a single agent at the dose of 350 mg/m every
9. Feveri J, Blanckaert N. What can we learn from analysis of serum
three weeks.
bilirubin? J Hepatol 1986; 2: 113-21.
Our preliminary results [11] suggest that baseline 10. Canal P, Gay C, Dezeuze A et al. Pharmacokinetics and pharmacodynamics of irinotecan during a phase II clinical trial in
serum bilirubin values are a valid predictor of severe
colorectal cancer. J Clin Oncol 1996; 14: 2688-95.
CPT-11 linked toxicity episodes.
11. Wasserman E, Myara A, Paumier D et al. Baseline bilirubin and
Oxaliplatin has not shown any evidence of hepatic
its transient early increase predicts likelihood of severe neutropetoxicity in animals, nor in over 2500 treated patients at
nia and diarrhea in CPT-11 based chemotherapy. PTOC Am Soc
recommended doses [12, 13]. Moreover, its excretion is
Clin Oncol 1997; 16: 767 (Abstr).
mostly urinary, with less than 1% excreted in feces over a 12. O'Dwyer PJ, Johnson SW, Hamilton TC. Cisplatin and its analogues. In DeVita VT Jr, Hellman S, Rosenberg SA (eds): Cancerfive day determination period [14]. The presently dePrinciples and Practice of Oncology, Vol 1, 5th ed. Philadelphia:
scribed phenomenon is unlikely to be linked, therefore,
1997; 426.
to addition of Oxaliplatin to CPT-11, in the phase I-II 13. Lippincott
Brienza S, Vignoud J, Itzhaki M, Krikonan A. Oxaliplatin
experience.
(L-OHP): Global safety in 682 patients. Proc Am Soc Clin Oncol
1995; A513.
The ongoing analysis and validation of serum bilirubin as an eventual predictive parameter of severe toxicity 14. Massari C, Brienza S, Rotarski M et al. Oxaliplatin, Transplatin®
(Tp®) comparative pharmacokinetics in normal and impaired
may improve the safety of CPT-11 based chemotherapy
renal function patients. Proc Am Assoc Cancer Res 1994; A1445.
administration, and also be a tool to optimize individual
patient dosing.
Received 4 August 1997; accepted 8 September 1997.