Clin Pharmacokinet (2014) 53:185–196 DOI 10.1007/s40262-013-0108-z
ORIGINAL RESEARCH ARTICLE
Exposure–Toxicity Relationship of Sorafenib in Japanese Patients with Renal Cell Carcinoma and Hepatocellular Carcinoma Masahide Fukudo • Takuma Ito • Tomoyuki Mizuno • Keiko Shinsako • Etsuro Hatano • Shinji Uemoto • Tomomi Kamba • Toshinari Yamasaki Osamu Ogawa • Hiroshi Seno • Tsutomu Chiba • Kazuo Matsubara
•
Published online: 18 October 2013 Springer International Publishing Switzerland 2013
Abstract Background and Objectives Sorafenib has various adverse events that can cause treatment discontinuation or dose reduction. The aim of this study was to compare the safety profile between renal cell carcinoma (RCC) and hepatocellular carcinoma (HCC) patients receiving sorafenib under real-life practice conditions. Furthermore, we investigated the relationship between sorafenib exposure and clinical outcomes. Methods A total of 91 Japanese cancer patients (RCC, n = 21; HCC, n = 70) treated with sorafenib were enrolled. Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria for Adverse Events (NCI-CTCAE) version 4.0. Single blood samples were collected at each clinic visit and serum sorafenib concentrations were measured by liquid chromatography–tandem
M. Fukudo (&) T. Ito T. Mizuno K. Shinsako K. Matsubara Department of Clinical Pharmacology and Therapeutics, Kyoto University Hospital, Faculty of Medicine, Kyoto University, Sakyo-ku, Kyoto 606-8507, Japan e-mail:
[email protected] E. Hatano S. Uemoto Department of Surgery, Graduate School of Medicine, Kyoto University, Kyoto, Japan T. Kamba T. Yamasaki O. Ogawa Department of Urology, Graduate School of Medicine, Kyoto University, Kyoto, Japan H. Seno T. Chiba Department of Gastroenterology and Hepatology, Graduate School of Medicine, Kyoto University, Kyoto, Japan
mass spectrometry (LC–MS/MS). The incidence of adverse events was analyzed according to cancer type and sorafenib concentration. Results Hand-foot skin reaction (HFSR) was the most common adverse event among RCC (76 %) and HCC (66 %) patients. Elevations in hepatic transaminases and pancreatic amylase developed more frequently in patients with RCC than in those with HCC (p \ 0.05), while hyperbilirubinemia and thrombocytopenia were observed more often in HCC patients than in RCC patients (p \ 0.05). Pharmacokinetic data were available from 52 patients (RCC, n = 16; HCC, n = 36). HCC patients showed significantly higher dose-normalized concentrations than RCC patients (p = 0.0184). Sorafenib concentrations were significantly greater in patients with grade C2 HFSR and hypertension than in those not experiencing the adverse events (p = 0.0045 and 0.0453, respectively). Furthermore, receiver operating characteristic curves revealed optimal cutoff concentrations of sorafenib to predict grade C2 HFSR (5.78 lg/mL) and hypertension (4.78 lg/mL). In addition, a trend of prolonged overall survival was observed in HCC patients who achieved a maximal sorafenib concentration of C4.78 lg/mL during treatment compared with those who did not achieve the threshold concentration (12.0 vs. 6.5 months; log-rank p = 0.0824). Conclusions The results of this study suggest that the safety and pharmacokinetic profiles of sorafenib differ between Japanese cancer patients with RCC and HCC. Furthermore, the serum sorafenib concentration could be used as a guide to avoiding the development of severe HFSR while allowing prediction of the incidence of grade C2 hypertension in patients with RCC and HCC, and may potentially be related to the clinical efficacy of sorafenib for HCC.
186
1 Background Sorafenib is an orally active multikinase inhibitor that targets vascular endothelial growth factor (VEGF) receptors (VEGFRs) and platelet-derived growth factor receptors (PDGFRs) in tumor vasculature, as well as stem cell factor receptor (c-KIT), Fms-like tyrosine kinase 3 (FLT-3), and the RAF/MEK/ERK signal transduction pathway in tumor cells [1]. Sorafenib is currently used to treat patients with renal cell carcinoma (RCC) and hepatocellular carcinoma (HCC) [2–4]. The most common adverse events associated with sorafenib treatment include hand-foot skin reaction (HFSR), rash, and diarrhea [2–4]. Sorafenib-related hypertension, induced by the inhibition of the VEGF pathway, is also frequently observed, and is suggested as a potential surrogate biomarker for the efficacy [5, 6]. Grade C2 HFSR is a major clinical problem because it causes patients to have painful erythema, edema, and desquamation on their palms and soles, which can then lead to limitation in their activities of daily living (ADL) and decreased quality of life (QOL) [7]. Therefore, identifying risk factors associated with sorafenib-induced severe toxicities is of key importance to improve RCC and HCC patients’ ability to perform ADL, their QOL, and their overall treatment outcomes by avoiding undesired treatment discontinuation. Sorafenib is metabolized primarily in the liver by cytochrome P450 (CYP) 3A4 to the active metabolite sorafenib N-oxide (M-2) and by uridine diphosphate glucuronyl transferase (UGT) 1A9 to sorafenib glucuronide [8– 10]. Moreover, sorafenib is not a substrate of UGT1A1 but acts as an inhibitor of UGT1A1-mediated glucuronidation [11]. In plasma, sorafenib exists mostly in an unchanged form. The ratio of the active metabolite M-2 to the sum of sorafenib and three metabolites (i.e., M-2, M-4, and M-5) was reported to be 6–12 % in Japanese cancer patients [12]. Therefore, unlike sunitinib [13], the active metabolite M-2 may have a minor contribution to the overall efficacy and toxicity of sorafenib. Although significant inter-individual pharmacokinetic variability is found in patients treated with sorafenib, genetic variations of CYP3A4/3A5 and UGT1A1/1A9 do not seem to have a significant effect on the inter-individual variability in sorafenib exposure [11, 14, 15]. Moreover, P-glycoprotein (adenosine triphosphate [ATP]-binding cassette subfamily B member 1 [ABCB1]/multidrug resistance 1 [MDR1]) is not likely to play a role in the efflux transport of sorafenib in the intestine, and therefore does not contribute to its highly variable absorption [16, 17]. Given the saturable absorption of sorafenib [18], low solubility of sorafenib in the digestive tract can cause large intra- and inter-patient variability in bioavailability. Furthermore, a patient’s demographics such as sex [15], sarcopenia [19, 20], albuminemia [21],
M. Fukudo et al.
and a high-fat meal [22] may affect sorafenib pharmacokinetics. Increased HFSR severity has been significantly associated with the increase in systemic exposure to sorafenib, defined as area under the concentration–time curve (AUC) [11]. Furthermore, the AUC of sorafenib has been significantly greater in patients with grade C2 HFSR than that observed in patients with grade \2 HFSR [15, 23], as well as in patients with grade C2 hypertension compared with those with normal blood pressure [23]. In addition, there have been reports of patients with advanced HCC developing sorafenib-induced hepatic encephalopathy [24] or fatal liver failure [25]. Pre-existing hepatic dysfunction, secondary to HCC, may cause decreased metabolic clearance of sorafenib, which could contribute to increased sorafenib exposure, and potentially the development of severe hepatic toxicities. So far, GIDEON (Global Investigation of therapeutic DEcisions in hepatocellular carcinoma and Of its treatment with sorafeNib), a global, noninterventional, surveillance study, is ongoing to evaluate the safety and efficacy of sorafenib in patients with unresectable HCC [26]. However, little has been known about differences in sorafenib safety and pharmacokinetics between RCC and HCC patients, specifically under reallife practice conditions. Furthermore, the association between sorafenib exposure and clinical outcomes has not yet been fully elucidated. In the present study, we aimed to (1) study differences in the safety profile of sorafenib between RCC and HCC patients in clinical practice; (2) identify risk factors for early sorafenib intolerance; and (3) investigate the exposure of sorafenib and its relationships with sorafenibinduced toxicities and efficacy.
2 Patients and Methods 2.1 Patients and Treatment Ninety-one Japanese cancer patients (RCC, n = 21; HCC, n = 70) who had started sorafenib treatment between April 2008 and September 2011 at Kyoto University Hospital, Kyoto, Japan, were enrolled in this observational study (Fig. 1). Sorafenib was initiated at 400 mg twice daily (800 mg/day) and dose reduction or interruption was allowed for severe adverse events. In patients with poor liver function or co-morbidities and those with poor Eastern Cooperative Oncology Group performance status (ECOG PS), the initial starting dose was reduced to half the standard dose, 200 mg twice daily (400 mg/day); dose escalation was permitted while toxicity was acceptable. Sorafenib was administered without food (at least 1 h before or 2 h after a meal), and the administration was
PK and Safety Profile of Sorafenib
187
less than 10 %. The lower limit of quantification was 0.01 lg/mL. The non-compartmental analysis was performed using WinNonlin version 6.1 (Pharsight Corporation, Tokyo, Japan) to calculate the AUC from time zero to 12 h (AUC12), maximum concentration (Cmax), and minimum concentration (Cmin).
Study population • RCC (n = 21) • HCC (n = 70)
2.3 Safety and Efficacy Evaluation PK subgroup • RCC (n = 16) • HCC (n = 36)
Safety/tolerability analysis
PK/PD analysis
• RCC versus HCC • Risk factor
• Exposure-safety • Exposure-efficacy
Fig. 1 Schematic illustration of the analysis objects in this study. HCC hepatocellular carcinoma, PD pharmacodynamic, PK pharmacokinetic, RCC renal cell carcinoma
All adverse events were graded according to the National Cancer Institute Common Toxicity Criteria for Adverse Events (NCI-CTCAE) version 4.0. Early sorafenib intolerance was defined as discontinuation within the first month of treatment while on the initial starting dose, including treatment termination and dose reduction or interruption due to unacceptable adverse events. Overall survival was measured from the date of administration of sorafenib to the date of death. Safety and efficacy analyses were performed on the intent-to-treat population. The data cutoff date was 31 August 2012. 2.4 Statistical Analysis
continued until progressive disease, patient refusal, or intolerable adverse events. This study was conducted in accordance with the Declaration of Helsinki and its amendments, and was approved by the Kyoto University Graduate School and Faculty of Medicine Ethics Committee. Written informed consent was obtained from each patient. 2.2 Pharmacokinetic Evaluation Single blood samples (2 mL) were obtained for pharmacokinetic study at scheduled clinic visits (every 2 weeks or earlier if clinically indicated) until treatment discontinuation. The plasma elimination half-life of sorafenib in Japanese patients with advanced solid tumors was reported to be 24–30 h (mean 25.5 h) [12]; thus, a steady state of sorafenib pharmacokinetics was considered to have been achieved by day 8. When possible, serial blood samples (2 mL) before dose and at 2, 6, and 12 h post-dose were collected on day 8 to evaluate the full pharmacokinetic profile at a steady state. Serum was separated by centrifugation (3,000 rpm, 10 min), and stored at -30 C until the analysis. Serum concentrations of sorafenib were determined by liquid chromatography–tandem mass spectrometry (LC– MS/MS), according to the procedure described previously [27], with minor modifications. Analyses were performed in the multiple reaction monitoring mode at ion transitions from m/z 464.9 to 252.0 (sorafenib) and m/z 355.2 to 233.0 (roscovitine; internal standard). The intra- and inter-assay coefficients of variation and the relative bias obtained were
Bivariate correlations were assessed using the Pearson correlation coefficient. The statistical significance of differences in non-parametric values between two groups was analyzed with the Mann–Whitney U test. The Chi-square (v2) test or Fisher’s exact probability test was used to compare the proportion of patients with a given characteristic between two groups. A multivariate logistic regression model was applied to determine risk factors for early sorafenib intolerance as well as grade C2 adverse events during the first month in the study population and to estimate the adjusted odds ratio (OR) and its 95 % confidence interval. Only those variables with p \ 0.10 in the univariate analysis were included in the multivariate model using forward selection. The comparison of sorafenib concentrations between patients with and without grade C2 adverse events of interest was performed on the combined data of RCC and HCC patients in the subgroup for pharmacokinetic study. In these analyses, the sorafenib concentration adjacent to the adverse events was used, while the average value of the measurements available at a steady state during the followup period was applied for patients who did not experience any of the adverse events. A receiver operating characteristic (ROC) curve was constructed to calculate an AUC (AUCROC) to estimate an optimal threshold value of the sorafenib concentration for predicting grade C2 adverse events of interest. Median overall survival was estimated with the Kaplan–Meier method, and the log-rank test was used to examine the significance of differences between two groups. A two-sided p \ 0.05 was considered
188
significant. Statistical analyses were performed with STATA version 12 (StataCorp. LP, College Station, TX, USA) and GraphPad Prism version 5.0 (GraphPad Software, San Diego, CA, USA).
3 Results 3.1 Patient Characteristics As shown in Table 1, a majority of patients enrolled were male (81 %) and presented with ECOG PS of 0 or 1 (99 %). The study population included 78 (86 %) patients with Child-Pugh class A, and 12 (13 %) patients presented with Child-Pugh class B or C. The standard dose of sorafenib (800 mg/day) was initiated in 60 (66 %) patients, while a higher proportion of RCC patients (62 %) than HCC patients (26 %) were initiated at the reduced dose of sorafenib (400 mg/day). No significant differences in baseline characteristics between RCC and HCC patients were noted, with the exception of estimated glomerular filtration rate (p \ 0.01) and initial starting dose (p \ 0.005). 3.2 Sorafenib Dose and Pharmacokinetics The average dosage of sorafenib in RCC as well as HCC patients starting at 800 mg/day was decreased early due to intolerable toxicity, and maintained at 600 mg/day on average after the first month (Fig. 2). The daily dose of sorafenib in RCC patients starting at 400 mg/day was gradually increased after the start of treatment. However, the dosage was tapered off due to toxic effects after the first month. On average, when an HCC patient was initiated at a starting dose of 400 mg/day, this dose level was maintained throughout the first 6 months of treatment. Thereafter, the average dosage increased to 600 mg/day for the remainder of the first year of treatment. This trend was also observed among RCC patients who were initiated at the reduced starting dose. Pharmacokinetic data of sorafenib were available from 52 patients who provided written informed consent for this part of the study (RCC, n = 16; HCC, n = 36; Table 1). Full pharmacokinetic profiles were also obtained from five RCC patients receiving 800 mg/day (Fig. 3a). The mean (standard deviation) Cmin, Cmax, and AUC12 were 2.92 (0.93) lg/mL, 4.31 (1.73) lg/mL, and 39.6 (18.7) lg•h/ mL, respectively. As shown in Fig. 3b, the dose-normalized concentration of sorafenib significantly decreased over time in RCC patients (r = -0.25; p = 0.0055) as well as in HCC patients (r = -0.16; p = 0.0140). Both RCC and HCC patients showed large variability in the dose-normalized concentration of sorafenib (Fig. 3c). Notably, the median value of the sorafenib dose-normalized
M. Fukudo et al.
concentration in HCC patients was significantly greater than that in RCC patients (p = 0.0184). 3.3 Safety and Tolerability of Sorafenib No grade 5 adverse events associated with sorafenib treatment were observed. The most common adverse event of all grades was HFSR in both RCC (16/21 [76 %]) and HCC (46/70 [66 %]) patients (Table 2). Any grade elevations in aspartate aminotransferase (AST) and alanine aminotransferase (ALT) developed more frequently in RCC than HCC patients (AST, 11/21 [52 %] vs. 19/70 [27 %], p = 0.038; ALT, 8/21 [38 %] vs. 8/70 [11 %], p = 0.008). In addition, the incidence of grade 3/4 amylase elevation was significantly higher in RCC than HCC patients (3/21 [14 %] vs. 1/70 [1 %], p = 0.037). On the other hand, bilirubin elevation was noted more often in HCC than RCC patients (all grades, 38/70 [54 %] vs. 2/21 [10 %], p \ 0.0005; grade 3/4, 14/70 [20 %] vs. 0/21 [0 %], p = 0.034). Furthermore, any grade thrombocytopenia occurred at a significantly higher incidence in HCC than RCC patients (36/70 [51 %] vs. 5/21 [24 %], p = 0.044). A majority of the dose-limiting toxicities (including HFSR, rash, hypertension, diarrhea, bilirubin elevation, and thrombocytopenia) occurred within the first month of treatment (Fig. 4). Table 3 summarizes the results of univariate analyses of risk factors for the early development of grade C2 HFSR, rash, and hypertension. Female sex was identified as a potential risk factor for developing grade C2 HFSR (OR 2.77; 95 % CI 0.93–8.23; p = 0.067). In addition, age C75 years was found to marginally increase the risk of grade C2 rash (OR 4.89; 95 % CI 1.00–23.86; p = 0.050). Furthermore, ECOG PS C1 was associated with a potential decrease in the risk of grade C2 hypertension (OR 0.30; 95 % CI 0.08–1.12; p = 0.073). Early sorafenib intolerance was observed in 50 % or more of patients with RCC (4/8) as well as HCC (29/52) who were initiated at a starting dose of 800 mg/day. In contrast, early sorafenib intolerance was lower in patients starting at 400 mg/day (RCC, 4/13 [31 %]; HCC, 2/18 [11 %]) than in those starting at 800 mg/day. Multivariate logistic regression analysis showed that an initial starting dose of 800 mg/day (OR 10.5; 95 % CI 2.67–41.4; p = 0.001), age C75 years (OR 5.07; 95 % CI 1.42–18.1; p = 0.012), and body surface area \1.53 m2 (OR 3.88; 95 % CI 1.29–11.7; p = 0.016) were an independent risk factor for early sorafenib intolerance (Table 4). 3.4 Relationships of Sorafenib Exposure with Safety and Efficacy There was no significant relationship between sorafenib steady-state concentration and development of grade C2
PK and Safety Profile of Sorafenib
189
Table 1 Baseline characteristics of renal cell carcinoma and hepatocellular carcinoma patients treated with sorafenib Characteristic
Study population
p value
RCC patients (n = 21)
PK subgroup
HCC patients (n = 70)
RCC patients (n = 16)
HCC patients (n = 36)
Sex [n (%)] Male
18 (86)
56 (80 %)
Female
3 (14 %)
14 (20 %)
Median
60
70
Range
19–80
34–84
NS
15 (94 %)
31 (86 %)
1 (6 %)
5 (14 %)
62
69
19–80
47–79
1.60
1.59
1.44–2.05
1.35–2.00
Age (years)
Body surface area (m2) Median 1.59 Range
NS
1.58
1.40–2.05
NS
1.22–2.08
ECOG PS [n (%)] 0
15 (71 %)
48 (69 %)
12 (75 %)
23 (64 %)
1
6 (29 %)
21 (30 %)
NS
4 (25 %)
13 (36 %)
C2
0 (0 %)
1 (1 %)
0 (0 %)
0 (0 %)
0 (0 %)
0 (0 %)
16 (100 %)
36 (100 %)
Disease stage [n (%)] IIIb
0 (0 %)
6 (9 %)
IV
21 (100 %)
64 (91 %)
NS
Child-Pugh class [n (%)] A
20 (95 %)
58 (83 %)
16 (100 %)
30 (83 %)
B/C
1 (5 %)
11 (16 %)
NS
0 (0 %)
6 (17 %)
Unknown
0 (0 %)
1 (1 %)
0 (0 %)
0 (0 %)
eGFR (mL/min/m2) Median Range
60.2 44.1–118
74.0 33.3–130
\0.01
57.9 44.1–118
79.7 33.3–130
\0.005
10 (62.5 %)
13 (36 %)
6 (37.5 %)
23 (64 %)
Initial starting dose [n (%)] 400 mg/day
13 (62 %)
18 (26 %)
800 mg/day
8 (38 %)
52 (74 %)
ECOG PS Eastern Cooperative Oncology Group performance status, eGFR estimated glomerular filtration rate, HCC hepatocellular carcinoma, NS not significant, PK pharmacokinetic, RCC renal cell carcinoma
Sorafenib daily dose (mg/day)
a
b
1,200
1,200
1,000
1,000
800
800
600
600
400
400
200
200
0
0 0
60
120
180
240
300
360
0
60
120
180
240
300
360
Time since start of sorafenib treatment (day) Fig. 2 Changes in sorafenib daily dose in patients with a renal cell carcinoma and b hepatocellular carcinoma. Open and closed circles indicate patients starting at 400 and 800 mg/day, respectively. Data are presented as the mean with standard deviation
rash, diarrhea, bilirubin elevation, or thrombocytopenia (Fig. 5). On the other hand, sorafenib concentrations preceding grade C2 HFSR and hypertension were significantly
greater than those in patients who did not experience any of the adverse events (p = 0.0045 and 0.0453, respectively). Using an ROC curve (Fig. 6), the optimal cutoff
Sorafenib concentration (µg/mL)
a 100
10
1
0.1 0
2
4
6
8 10 12
b
Time (h)
r = −0.25, p = 0.0055 r = −0.16, p = 0.0140
Dose-normalized concentration (µg/mL/mg)
M. Fukudo et al.
Dose-normalized concentration (µg/mL/mg)
190
c p = 0.0184
RCC
HCC
Time (day)
Fig. 3 a Changes in sorafenib concentration during 12 h post-dose on day 8. Data from renal cell carcinoma patients receiving 400 mg twice daily (800 mg/day) are presented as the mean with standard deviation (n = 5). b Changes in dose-normalized concentration of sorafenib over time. Closed circles and open squares indicate patients with renal cell carcinoma and hepatocellular carcinoma, respectively.
c Comparison of the sorafenib dose-normalized concentration between renal cell carcinoma and hepatocellular carcinoma patients. Results are shown as all data from individual patients with the median and interquartile range. HCC hepatocellular carcinoma, RCC renal cell carcinoma
Table 2 Comparison of safety profile of sorafenib in patients with renal cell carcinoma and hepatocellular carcinoma Adverse eventa
All grades [n (%)]
p value
RCC patients (n = 21)
HCC patients (n = 70)
Hand-foot skin reaction
16 (76 %)
46 (66 %)
Rash
Grade 3/4 [n (%)]
p value
RCC patients (n = 21)
HCC patients (n = 70)
NS
4 (19 %)
11 (16 %)
NS
10 (48 %)
21 (30 %)
NS
0 (0 %)
0 (0 %)
NS
Hypertension Diarrhea
9 (43 %) 8 (38 %)
33 (47 %) 33 (47 %)
NS NS
6 (29 %) 0 (0 %)
11 (16 %) 5 (7 %)
NS NS
Fatigue
7 (33 %)
20 (29 %)
NS
1 (5 %)
1 (1 %)
NS
Anorexia
6 (29 %)
14 (20 %)
NS
0 (0 %)
2 (3 %)
NS
Fever
6 (29 %)
9 (13 %)
NS
0 (0 %)
2 (3 %)
NS
Alopecia
5 (24 %)
14 (20 %)
NS
0 (0 %)
0 (0 %)
NS
AST elevation
11 (52 %)
19 (27 %)
0.038
1 (5 %)
6 (9 %)
NS
ALT elevation
8 (38 %)
8 (11 %)
0.008
1 (5 %)
3 (4 %)
NS
Bilirubin elevation
2 (10 %)
38 (54 %)
\0.0005
0 (0 %)
14 (20 %)
0.034 0.037
Amylase elevation
9 (43 %)
16 (23 %)
NS
3 (14 %)
1 (1 %)
Thrombocytopenia
5 (24 %)
36 (51 %)
0.044
2 (10 %)
11 (16 %)
NS
ALT alanine aminotransferase, AST aspartate aminotransferase, HCC hepatocellular carcinoma, NS not significant, RCC renal cell carcinoma a Adverse events occurring in 10 % or more of patients for all grades
concentrations predicting grade C2 HFSR and hypertension were estimated to be 5.78 lg/mL with the AUCROC of 0.73 (95 % CI 0.60–0.87; p = 0.0044) and 4.78 lg/mL with the AUCROC of 0.66 (95 % CI 0.51–0.81; p = 0.0445). Finally, we examined whether sorafenib exposure above the threshold concentration associated with the development of grade C2 hypertension (4.78 lg/mL) was related to a survival benefit for patients with HCC. Because pharmacokinetic data were available from only 16 patients
with RCC, we did not evaluate the prognostic relevance of sorafenib exposure in RCC patients. The sorafenib Cmax during the follow-up period was used to classify HCC patients into two groups (low or high exposure group). As shown in Fig. 7, HCC patients who achieved a sorafenib Cmax of C4.78 lg/mL (high exposure group, n = 19) had longer overall survival than those who did not achieve the threshold concentration (low exposure group, n = 17), although there was no significant difference between the two groups (12.0 vs. 6.5 months; p = 0.0824).
PK and Safety Profile of Sorafenib 50
Number of patients with adverse events
Fig. 4 Time to onset of a handfoot skin reaction, b rash, c hypertension, d diarrhea, e bilirubin elevation, and f thrombocytopenia since start of sorafenib treatment. White, grey, and black columns indicate grade 1, grade 2, and grade 3/4 adverse events, respectively
191 50
a
40
40
30
30
20
20
10
10
b
0
0 0 50
60
120
180
240
300
360
0 50
c
40
40
30
30
20
20
10
10
60
120
180
240
300
360
60
120
180
240
300
360
60
120
180
240
300
360
d
0
0 0 50
60
120
180
240
300
50
e
40
0
360
f
40
30
30
20
20
10
10
0
0 0
60
120
180
240
300
360
0
Time since start of sorafenib treatment (day) Table 3 Risk factors for early development of grade C2 adverse events Factora
Hand-foot skin reaction
Rash
Hypertension
OR (95 % CI)
p value
OR (95 % CI)
p value
OR (95 % CI)
p value
2.77 (0.93–8.23)
0.067
0.71 (0.08–6.30)
0.757
0.39 (0.08–1.85)
0.233
1.23 (0.43–3.48)
0.699
4.89 (1.00–23.86)
0.050
1.83 (0.63–5.36)
0.268
0.66 (0.25–1.72)
0.395
1.18 (0.22–6.50)
0.846
2.36 (0.72–7.79)
0.158
1.63 (0.63–4.26)
0.317
0.89 (0.16–4.90)
0.896
0.30 (0.08–1.12)
0.073
0.81 (0.20–3.27)
0.769
0.99 (0.11–8.93)
0.990
0.98 (0.24–3.96)
0.981
1.63 (0.63–4.26)
0.317
3.33 (0.69–16.0)
0.133
1.17 (0.41–3.31)
0.772
0.56 (0.20–1.58)
0.274
0.36 (0.07–1.77)
0.211
0.95 (0.30–2.99)
0.928
1.59 (0.58–4.33)
0.365
1.32 (0.24–7.22)
0.750
0.80 (0.29–2.19)
0.657
Sex Female vs. male Age C75 vs. \75 years Body surface area \1.53 vs. C1.53 m2 ECOG PS C1 vs. 0 Child-Pugh class B/C vs. A eGFR \60 vs. C60 mL/min/m2 Type of cancer HCC vs. RCC Initial starting dose 800 vs. 400 mg/day
ECOG PS Eastern Cooperative Oncology Group performance status, eGFR estimated glomerular filtration rate, HCC hepatocellular carcinoma, OR odds ratio, RCC renal cell carcinoma a
The last category of each factor indicates the reference category
4 Discussion The present study revealed that sorafenib shows different safety profiles in patients with RCC and HCC in real-life
clinical practice. Elevated liver enzymes (AST/ALT) and amylase were observed more frequently in RCC patients than in HCC patients (Table 2). On the other hand, HCC patients were more susceptible to thrombocytopenia and
192
M. Fukudo et al.
Table 4 Risk factors for early sorafenib intolerance Factora
Univariate
Multivariate
OR (95 % CI)
p value
OR (95 % CI)
p value
1.23 (0.43–3.55)
0.698
3.08 (1.14–8.35)
0.027
5.07 (1.42–18.1)
0.012
2.57 (1.06–6.26)
0.037
3.88 (1.29–11.7)
0.016
0.81 (0.33–2.00)
0.646
0.23 (0.05–1.14)
0.071
0.34 (0.06–1.92)
0.221
0.84 (0.35–2.04)
0.699
1.68 (0.61–4.68)
0.317
5.09 (1.83–14.2)
0.002
10.5 (2.67–41.4)
0.001
Sex Female vs. male Age C75 vs. \75 years Body surface area \1.53 vs. C1.53 m2 ECOG PS C1 vs. 0 Child-Pugh class B/C vs. A eGFR \60 vs. C60 mL/min/m2 Type of cancer HCC vs. RCC Initial starting dose 800 vs. 400 mg/day
ECOG PS Eastern Cooperative Oncology Group performance status, eGFR estimated glomerular filtration rate, HCC hepatocellular carcinoma, OR odds ratio, RCC renal cell carcinoma The last category of each factor indicates the reference category
Fig. 5 Relationship between sorafenib concentration and development of grade C2 a hand-foot skin reaction, b rash, c hypertension, d diarrhea, e bilirubin elevation, and f thrombocytopenia. The boxes indicate the median, 25th and 75th percentiles of the data; the whiskers represent the minimum and maximum values
a
b
c
p = 0.0045
Sorafenib concentration (µg/mL)
a
0-1
p = 0.58
≥2
d
0-1
≥2
e
≥2
0-1
≥2
f
p = 0.99
0-1
p = 0.0453
p = 0.76
0-1
≥2
p = 1.00
0-1
≥2
Toxicity grade
hyperbilirubinemia than RCC patients. Since hepatic fibrosis is commonly encountered in advanced HCC, HCC patients may have the increased risk of thrombocytopenia
due to reduced production of thrombopoietin in the liver [28]. In addition, unconjugated bilirubin is generally elevated in liver cirrhosis. It has been reported that sorafenib
PK and Safety Profile of Sorafenib
193
Fig. 6 Receiver operating characteristic curves for predicting grade C2 a hand-foot skin reaction and b hypertension by sorafenib concentration. AUCROC area under the receiver operating characteristic curve
b
Sensitivity
Sensitivity
a
AUC ROC = 0.73 p = 0.0044
1 − Specificity
Overall survival rate (%)
100
≥4.78 µg/mL (n = 19)