Annals of Oncology Advance Access originally published online on October 23, 2006
Annals of Oncology 2007 18(1):168-172; doi:10.1093/annonc/mdl321
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2006 European Society for Medical Oncology
supportive care |
Relationship between cytochrome 3A activity, inflammatory status and the risk of docetaxel-induced febrile neutropenia: a prospective study
1 Department of Medical Oncology, Groupe Hospitalier Cochin St Vincent de Paul, Université Paris 5, Assistance Publique—Hôpitaux de Paris, Paris, France
2 Department of Molecular Pathology, University of Texas, MD Anderson Cancer Center, Houston, USA
3 Department of Clinical Pharmacology, Groupe Hospitalier Cochin St Vincent de Paul, Université Paris 5, Assistance Publique—Hôpitaux de Paris, Paris
4 Department of Medical Oncology, Institut Curie, Paris
5 Department of Biostatistics, Groupe Hospitalier Cochin St Vincent de Paul, Université Paris 5, Assistance Publique—Hôpitaux de Paris, Paris, France
* Correspondence to: Dr J. Alexandre, Department of Molecular Pathology, unit 951, University of Texas, MD Anderson Cancer Center, PO Box 301429, Houston, TX 77230-1429, USA. Tel: +1-713-834-6070; Fax: +1-713-834-6084; E-mail: jalexand{at}mdanderson.org
 |
Abstract |
|---|
 Top Abstract introductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Background: We hypothesized that cancer-related inflammation might increase the risk of febrile neutropenia (FN) induced by docetaxel (DCX, Taxotere), by both affecting the exposure to DCX and the tissue sensitivity.
Patients and methods: Advanced cancer patients with normal liver function, performance status (PS) <3, were included. Cytochrome P450 3A (CYP 3A) activity was estimated before the first cycle of DCX by a single determination of midazolam plasma concentration, 4 hours after 0.015 mg/kg i.v. bolus. Following the first cycle of 75–100 mg/m2 DCX, clearance and area under the concentration versus time curve (AUC) were estimated using a limited sampling strategy.
Results: Among 56 assessable patients, 7 FNs occurred after first cycle (13%). In univariate analysis, high midazolam concentration and free DCX AUC were associated with severe neutropenia and FN. In addition to DCX exposure–related parameters, the risk of FN was also correlated with poor PS, baseline lymphopenia and lung cancer, while high ferritin level, indicator of an inflammatory state, reached borderline significance (P = 0.07). By multivariate analysis, total DCX AUC and baseline lymphopenia were associated with FN. High midazolam concentration was correlated with elevated ferritin level (r = 0.32; P = 0.02).
Conclusion: Inflammatory status and lymphocyte count should be included in the evaluation of the benefice/risk ratio before the initiation of DCX.
Key words: cytochrome 3A, docetaxel, febrile neutropenia, lymphocytopenia midazolam
|
introduction |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Docetaxel (DCX; TaxotereTM) is a taxoid derivative that displays high anticancer activity alone or in combination against several common solid tumors, including breast, lung and prostate cancer [1]. Its acute dose-limiting toxicity is neutropenia. Febrile neutropenia (FN) may occur in as many as 25% of patients with breast cancer treated with DCX single agent at the dose of 100 mg/m2 [2]. Moreover, with the rapidly growing fraction of elderly patients susceptible to benefit from this treatment, oncologists will need to take into consideration aging, functional status and comorbidity. Hence, to tailor cancer chemotherapy to the individual becomes an emerging field. We aimed to identify patient-dependent parameters for DCX toxicity.
The parameters susceptible to influence DCX acute toxicity may be divided into two categories: pharmacokinetic-related parameters may modify DCX exposure while tissue-dependent parameters may influence the sensitivity to a given DCX exposure.
DCX is bound to plasma proteins, mainly alfa-1 acid glycoprotein (AAG) [3] and is eliminated through liver cytochrome P450 (CYP) 3A4 and 3A5 [4]. Liver dysfunction, as clinically assessed by combined elevated transaminases >1.5 times the upper limit of normal (ULN) and alkaline phosphatases (ALPs) >2.5 x ULN, is responsible for decreased DCX clearances (CL) [5] and increased toxicity [2, 6]. However, even in patients with normal liver function, DCX CL may vary by a factor of 6 [7]. CYP 3A4 activity, assessed in vivo by various procedures, appears to be the main factor of interpatient variability for DCX CL [7–9].
However, variations of DCX CL do not fully explain the occurrence of toxic events, which could be partly due to intrinsic cell sensitivity, independent of DCX exposure. Our hypothesis was that immunological and metabolic modifications secondary to tumor proliferation, could lead to decreased DCX metabolism, as well as increased intrinsic sensitivity of normal cells to cytotoxic chemotherapy. These biological modifications are sometimes translated into a catabolic syndrome known as cachexia [10]. This syndrome is biologically characterized by an early increase of acute inflammatory proteins, mainly C-reactive protein (CRP), AAG and ferritin [11], and a decrease of albumin and prealbumin, which precedes clinical symptoms, mainly weight loss, muscle shrinkage, anorexia and altered performance status (PS). In this context, we previously showed that acute inflammatory reaction and malnutrition are correlated with an increased risk of chemotherapy-induced severe hematotoxicity [12].
We have undertaken this clinicopharmacological study to prospectively determine independent predictive factors of DCX-induced FN, with particular emphasis to inflammatory status and CYP 3A activity, using midazolam plasma concentration as a probe.
|
patients and methods |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
patients
The protocol was approved by the Ethics Committee of Hospitalier Cochin, Paris, France. All patients provided written informed consent before study enrollment. Eligible patients had histologically proven malignancies, justifying a treatment with DCX (TaxotereTM; Sanofi–Aventis, Paris, France), age
18 years and adequate bone marrow reserve with neutrophils 1.0 x 109/l and platelets 100 x 109/l. They had to complete previous anticancer chemotherapy in at least 4 weeks, and steroids in at least 2 weeks before study entry (except for DCX premedication). Pregnant or breast-feeding women and patients without effective contraception, with severe concomitant illness, participating in another clinical trial or with elevated serum bilirubin level were not included.
study treatment
DCX was administered through a central i.v. line, as a 60-minute infusion. The decision on dose was left to the investigator within the standard range of 75–100 mg/m2. Antiemetic and antiallergic premedication by i.v. ondansetron 8 mg and methylprednisolone 60 mg was given 30 minutes before DCX infusion. Prophylactic granulocyte growth factor and antibiotics were not allowed. After first cycle, treatment was continued at the discretion of the investigator. The second cycle of DCX was administered after complete hematologic recovery and at least 21 days after first cycle.
study end points
The occurrence of FN and severe neutropenia (grade 3–4 National Cancer Institute–Common Toxicity Criteria) after first infusion of DCX were primary and secondary end points, respectively. Blood cell counts were collected twice weekly until day 21 or until recovery, and additional counts were carried out in case of fever. Adverse events were sought by telephone contact twice a week until the next DCX course (minimum at day 22) or until day 30 after the first dosing. FN was defined as fever 38.5°C once or >38°C twice, 12 hours apart, or a documented infection, concomitant to a neutrophil count <1.0 x 109/l. Another secondary end point was to assess if nutritional and inflammatory status (NIS) was correlated with CYP 3A activity in cancer patients. NIS was calculated as follows: (CRP x AAG levels)/(albumin x prealbumin levels) [15]. Ferritin level was also used as a marker of inflammatory state.
pharmacokinetic studies
assessment of CYP 3A4 activity by midazolam plasma concentration.
CYP 3A activity was evaluated by a single-point determination of midazolam concentration, 4 hours after a test dose injection, as previously described [13]. Midazolam (HypnovelTM; Roche, Neuilly-sur-Seine, France) was administered as an i.v. bolus infusion at the dose of 0.015 mg/kg. Midazolam plasma assay was carried out by gas-phase chromatography and mass spectrometry.
DCX pharmacokinetics.
Individual total DCX CL were estimated using the Bayesian estimator of ADAPT II software [14]. A priori parameters and optimal sampling times, H1 (end of infusion) and H6 (5 hours after the beginning of infusion) allowing the estimation of individual pharmacokinetic parameters with a minimal bias, were obtained from a previously published study [15]. DCX plasma concentrations were determined with slight modifications by a validated high-performance liquid chromatography method [16]. Individual total DCX areas under the concentration versus time (AUC) were calculated by dividing DCX dose by CL.
estimation of unbound fraction of DCX
Plasma, collected before first infusion of DCX, was incubated with 5 µg/ml DCX during 30 minutes at 37°C [3]. Then, unbound DCX was isolated by ultrafiltration and its concentration measured. The unbound fraction of DCX (fu) was calculated by dividing the measured free concentration by the spiked concentration. The unbound DCX AUC was AUCtotal x fu.
statistical considerations
Univariate analyses were carried out using the Wilcoxon and Fisher's exact tests for continuous and discontinuous variables, respectively. Associations between continuous variables were evaluated using the non-parametric Spearman's rank correlation coefficient. Multivariate logistic regression with backward procedure was used to identify the independent factors of FN. All the variables with a P value <0.05 in the univariate analysis were included in the multivariate models. Since DCX AUC is not a baseline characteristic that could be used as a predictive parameter, two separate models were run with and without DCX AUC value as a covariate. All tests were two-sided, and P values <0.05 were considered to denote statistical significance.
|
results |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
From 1 November 2002 to 1 June 2004, 58 patients with advanced solid tumors were enrolled in the study. Two patients had unexplained very low DCX plasma concentrations not consistent with the dose administered. They did not experience FN and were excluded from the analysis. Baseline characteristics of the 56 assessable patients are summarized in Table 1. Most patients (79%) had breast, non-small-cell lung cancer (NSCLC), or prostate cancer, corresponding to approved indications of DCX. Forty-eight patients (86%) had received previous chemotherapy. No patient had concomitant elevation of transaminase >1.5 x ULN and ALPs >2.5 x ULN. All patients had normal bilirubin level.
|
The first cycle DCX dose, total DCX CL and AUC and estimation of unbound fraction of DCX are presented in Table 2. The midazolam concentration at 4 hours was correlated with total DCX AUC (Spearman's correlation coefficient: r = 0.38; P = 0.004). The unbound DCX fraction was negatively correlated to AAG blood concentration (Spearman's correlation coefficient: r = –0.72; P < 0.0001).
|
Two patients were not assessable for the occurrence of severe neutropenia because no hemogram was carried out between day 5 and day 15 following DCX injection. Among 54 assessable patients, 43 (77%) experienced severe neutropenia. Only high midazolam concentrations and unbound DCX AUC were significantly associated with increased risk of severe neutropenia (Table 3). DCX dose was not integrated in the univariate analysis given its high correlation to DCX AUC.
|
Seven of the 56 assessable patients (13%) experienced FN after the first cycle of DCX. Univariate analyses for the risk of FN are shown in Table 3. Four baseline parameters were significantly correlated to FN: lung tumors (P = 0.04), PS (P = 0.05), lymphocyte count (P = 0.01) and midazolam concentration (P = 0.03) while neutrophil count, ferritin and NIS levels were close to statistical significance (P = 0.07, 0.07 and 0.09, respectively). In addition to baseline parameters, FN was strongly correlated with total DCX AUC (P = 0.003). Lung tumors, lymphocyte count, midazolam concentration and total DCX AUC were included in the multivariate logistic regression. Total DCX AUC and baseline lymphocyte count remained independent predictive factors (Table 4). After exclusion of DCX AUC from the analysis, lymphopenia and lung cancer were the only two independent covariates in this analysis (Table 4). The comparison of baseline parameters between patients with lung cancer and those with other primary sites showed a higher median NIS ratio [9.9 (0.1–53) versus 0.7 (0.01–950), P = 0.14] and neutrophils count [6.1 x 109/l (2.0–14.8) versus 4.8 x 109/l (1.5–13.8), P = 0.13] in patients with lung cancer.
|
Midazolam concentration was correlated with baseline ferritin level (r = 0.4; P = 0.002). It was significantly higher in 24 patients with ferritin level above the limit of normal (
300 ng/ml) compared with 32 patients with ferritin level within the normal range (21.2 ± 29.3 versus 8.9 ± 13.1 ng/ml, P = 0.01 by Wilcoxon test). By contrast, neither patient age nor gender or NIS ratio was correlated to midazolam concentration. |
discussion |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
This study aimed to identify patient-related parameters influencing the risk of FN. This study came after a previous retrospective study indicating a role for malnutrition and inflammation in the acute toxicity of anticancer chemotherapy [12] and a large retrospective study underlying the role of the PS and liver metabolism on DCX acute toxicity [2]. To search further on the role of malnutrition and inflammation on chemotherapy acute toxicity, we studied the chemotherapy-related factors influencing toxicity by (i) using a single agent, DCX, to reduce chemotherapy regimen-related interpatient variability and (ii) measuring the pharmacokinetics of DCX in each patient.
We hypothesized that inflammation might modify not only CYP 3A activity and thus the exposure to any CYP 3A substrate, such as DCX, but also the bone marrow sensitivity to a given exposure.
We first observed that pharmacokinetic-related parameters namely unbound DCX AUC and midazolam concentration correlated with severe neutropenia. However, in daily practice, neutropenia is no longer a limitation, especially with the option of prophylactic granulocyte–stimulating growth factors, which may reduce the duration of grade 4 neutropenia [17]. Instead, the main obstacle is the apparently non-predictable conversion of neutropenia to FN, associated with a lethal risk. Hence, we analyzed the parameters related to FN.
The multivariate analysis confirmed DCX exposure (DCX AUC) and lymphopenia, as parameters associated with increased risk of FN. These findings raise the interesting hypothesis that biologically severe neutropenia would be mainly dependent on DCX exposure, while the occurrence of clinically significant infectious consequences may be dependent upon concomitant immune suppression. Baseline lymphopenia in cancer patients may be related to both treatment-related factors, such as the previous administration of chemotherapy, and to patient-related parameters [18]. Specifically, patients with cachexia syndrome usually exhibit lymphopenia [19], which seems to result from a direct effect of cytokines such as tumor necrosis factor-
(TNF-) [20].
Because DCX AUC is not a baseline parameter and therefore is not a clinically relevant predictive factor, it was removed from the multivariate analysis in a second step. While lymphopenia remained an independent predictive factor, tumor type (i.e. NSCLC versus other cancers) appeared to be an independent covariate. This result is in line with results of a large phase III study showing that in pretreated NSCLC patients, a dose of 100 mg/m2 of DCX was associated with an unacceptable rate of FN and toxic death, while this same dose is usually considered safe in patients with breast cancer [21]. Our study indicates that the susceptibility of lung cancer patients to FN might be related to inflammation, a common feature at diagnosis in this disease, in comparison with the other cancers treated with DCX such as breast or prostate cancers.
Moreover, our results indicate that inflammation not only has pharmacodynamic consequences (i.e. increases the conversion of neutropenia to FN) but also favors a reduction in CYP 3A activity and thus overexposure to its substrates, as many chemotherapeutic agents are. Indeed, we observed that high ferritin level was correlated with increased concentration of midazolam, indicating a decreased activity of CYP 3A. Increased ferritin plasma concentration is frequently observed in the course of advanced cancer and is directly induced by TNF- overproduction, which is involved in cachexia syndrome [22]. In addition, TNF- is able to decrease CYP 3A activity [23].
Hence, altogether, our results indicate that patients with evidence of inflammation and lymphopenia at baseline are more susceptible to FN.
By decreasing CYP 3A activity and inducing lymphopenia, immunologic modifications related to cachexia could both increase DCX exposure to normal cells and the risk of infection during neutropenia. Hence, our results support the recommendation that inflammatory status and lymphocyte count should be included in the evaluation of the benefit/risk ratio before the initiation of a cytotoxic chemotherapy.
|
Acknowledgements |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
This study was sponsored by Association Eureka, Antony, and benefited from a grant from Laboratoire Aventis, France. This study was presented in part at the 41st American Society of Clinical Oncology meeting (13–17 May 2005, Orlando, FL).
Received for publication June 13, 2006. Revision received July 28, 2006. Accepted for publication August 2, 2006.
|
References |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
1. Montero A, Fossella F, Hortobagyi G, Valero V. (2005) Docetaxel for treatment of solid tumours: a systematic review of clinical data. Lancet Oncol 6:4229–239.[CrossRef][Web of Science][Medline]
2. Alexandre J, Bleuzen P, Bonneterre J, et al. (2000) Factors predicting for efficacy and safety of docetaxel in a compassionate-use cohort of 825 heavily pretreated advanced breast cancer patients. J Clin Oncol 18:562–573.
3. Urien S, Barre J, Morin C, et al. (1996) Docetaxel serum protein binding with high affinity to alpha 1-acid glycoprotein. Invest New Drugs 14:147–151.[Web of Science][Medline]
4. Shou M, Martinet M, Korzekwa KR, et al. (1998) Role of human cytochrome P450 3A4 and 3A5 in the metabolism of taxotere and its derivatives: enzyme specificity, interindividual distribution and metabolic contribution in human liver. Pharmacogenetics 8:391–401.[Web of Science][Medline]
5. Bruno R, Vivier N, Vergniol JC, et al. (1996) A population pharmacokinetic model for docetaxel (taxotere): model building and validation. J Pharmacokinet Biopharm 24:153–172.[CrossRef][Web of Science][Medline]
6. Bruno R, Hille D, Riva A, et al. (1998) Population pharmacokinetics/pharmacodynamics of docetaxel in phase II studies in patients with cancer. J Clin Oncol 16:187–196.
7. Hirth J, Watkins PB, Strawderman M, et al. (2000) The effect of an individual's cytochrome CYP3A4 activity on docetaxel clearance. Clin Cancer Res 6:1255–1258.
8. Yamamoto N, Tamura T, Kamiya Y, et al. (2000) Correlation between docetaxel clearance and estimated cytochrome P450 activity by urinary metabolite of exogenous cortisol. J Clin Oncol 18:2301–2308.
9. Goh BC, Lee SC, Wang LZ, et al. (2002) Explaining interindividual variability of docetaxel pharmacokinetics and pharmacodynamics in Asians through phenotyping and genotyping strategies. J Clin Oncol 20:3683–3690.
10. Tisdale MJ. (2001) Cancer anorexia and cachexia. Nutrition 17:438–442.[CrossRef][Medline]
11. Gabay C and Kushner I. (1999) Acute-phase protein and other systemic responses to inflammation. N Engl J Med 340:448–454.
12. Alexandre J, Gross-Goupil M, Falissard B, et al. (2003) Evaluation of the nutritional and inflammatory status in cancer patients for the risk assessment of severe haematological toxicity following chemotherapy. Ann Oncol 14:36–41.
13. Lin YS, Lockwood GF, Graham MA, et al. (2001) In vivo phenotyping for CYP3A by a single point determination of midazolam plasma concentration. Pharmacogenetics 11:781–791.[CrossRef][Web of Science][Medline]
14. D'Argenio DZ and Schumitzky A. (1979) A program package for simulation and parameter estimation in pharmacokinetics. Comput Programs Biomed 9:115–134.[CrossRef][Web of Science][Medline]
15. Baille P, Bruno R, Schellens JH, et al. (1997) Optimal sampling strategies for Bayesian estimation of docetaxel clearance. Clin Cancer Res 3:1535–1538.[Abstract]
16. Garg MB and Ackland SP. (2000) Simple and sensitive high-performance liquid chromatography method for the determination of docetaxel in human plasma or urine. J Chromatogr B Biomed Sci Appl 748:383–388.[CrossRef][Medline]
17. Martin M, Lluch A, Segui MA, et al. (2006) Toxicity and health-related quality of life in breast cancer patients receiving adjuvant docetaxel, doxorubicin, cyclophosphamide (TAC) or 5-fluorouracil, doxorubicin and cyclophosphamide (FAC): impact of adding primary prophylactic granulocyte-colony stimulating factor to the TAC regimen. Ann Oncol 17:1205–1212.
18. Borg C, Ray-Coquard I, Philip I, et al. (2004) CD4 lymphopenia as a risk factor for FN and early death after cytotoxic chemotherapy in adult patients with cancer. Cancer 101:2675–2680.[CrossRef][Web of Science][Medline]
19. Blay JY, Negrier S, Combaret V, et al. (1992) Serum level of interkeukin-6 as a prognostic factor in metastatic renal cell carcinoma. Cancer Res 52:3317–3322.
20. Aggarwal S, Gollapudi S, Gupta S. (1999) Increased TNF-alpha-induced apoptosis in lymphocytes from aged humans: changes in TNF-alpha receptor expression and activation of caspases. J Immunol 162:2154–2161.
21. Shepherd FA, Dancey J, Ramlau R, et al. (2000) Prospective randomized trial of docetaxel versus best supportive care in patients with non-small-cell lung cancer previously treated with platinum-based chemotherapy. J Clin Oncol 18:2095–2103.
22. Torti FM and Torti SV. (2002) Regulation of ferritin genes and protein. Blood 99:3505–3516.
23. Slaviero KA, Clarke SJ, Rivory LP. (2003) Inflammatory response: an unrecognised source of variability in the pharmacokinetics and pharmacodynamics of cancer chemotherapy. Lancet Oncol 4:224–232.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  I. Ray-Coquard, C. Cropet, M. Van Glabbeke, C. Sebban, A. Le Cesne, I. Judson, O. Tredan, J. Verweij, P. Biron, I. Labidi, et al. Lymphopenia as a Prognostic Factor for Overall Survival in Advanced Carcinomas, Sarcomas, and Lymphomas Cancer Res., July 1, 2009; 69(13): 5383 - 5391. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Mir, J. Alexandre, A. Tran, J.-P. Durand, G. Pons, J.-M. Treluyer, and F. Goldwasser Relationship between GSTP1 Ile105Val polymorphism and docetaxel-induced peripheral neuropathy: clinical evidence of a role of oxidative stress in taxane toxicity Ann. Onc., April 1, 2009; 20(4): 736 - 740. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||