Introduction

Patients with non-small cell lung cancer (NSCLC) who have somatic activating mutations of the epidermal growth factor receptor (EGFR) gene (EGFR mutations) generally respond to EGFR-tyrosine kinase inhibitors (EGFR-TKIs, e.g., gefitinib and erlotinib) [1, 2]. Nevertheless, the majority of these patients experience eventual disease progression, despite an initial dramatic response to treatment. The central nervous system (CNS) is a common site of recurrence, which is thought to be due to penetration of the agents into the CNS [3, 4]. A previous report described refractory CNS metastases of NSCLC that were improved by high-dose gefitinib [5]. Several cases in which CNS metastases resistant to gefitinib were improved by erlotinib have been also reported [6]. In addition, in several articles, intermittent, high-dose erlotinib improved CNS metastases that were resistant to continuous, normal-dose erlotinib [7–9]. These challenges are based on the hypothesis that high cerebrospinal fluid (CSF) concentrations may be achieved by high-dose administration of EGFR-TKIs, with higher CSF concentrations achieved by erlotinib compared with gefitinib. We previously reported the CSF concentrations of erlotinib and its penetration rates in four NSCLC patients with CNS metastases [10]. In this article, we describe a Japanese woman with NSCLC and EGFR mutation having CNS metastases that are resistant to 75 mg daily erlotinib. The metastases were subsequently improved by 150 mg daily erlotinib. We investigated the plasma and CSF concentrations of erlotinib at each dose. Furthermore, including previous data, we investigated the correlation between plasma and CSF concentrations.

Patients and methods

Patients

A 66-year-old Japanese woman was diagnosed with advanced lung adenocarcinoma and EGFR mutation (L858R) 3 years prior. Erlotinib had been administered as fourth line of treatment. Before the initiation of erlotinib, the patient had no CNS metastases. Although partial response (PR) was achieved, the erlotinib dose was reduced from 150 to 75 mg daily due to grade 3 fatigue. After progression-free survival (PFS) of 12 months, the patient experienced recurrent CNS metastases. The plasma and CSF concentrations of erlotinib just before administration of the 75-mg dose were 433 and 14 nM, respectively. The patient received whole-brain radiotherapy (WBRT), but the CNS metastases did not improve (Fig. 1a). The dose was subsequently increased to 150 mg daily; the plasma and CSF concentrations just before administration of erlotinib on day 8 erlotinib were 1,117 and 44 nM, respectively. Four weeks after the initiation of 150 mg administration of erlotinib, CNS metastases improved (Fig. 1b).

Fig. 1
figure 1

Brain magnetic resonance imaging (MRI) in the 66-year-old Japanese female patient. a Brain MRI after whole-brain radiotherapy demonstrated stable brain metastases (arrowheads). At this point, 75 mg daily erlotinib was administered. b Four weeks after initiation of 150 mg daily erlotinib, the metastases were improved (arrowheads)

Full size image

We previously reported the plasma and CSF concentrations of four NSCLC patients with CNS metastases (cases 1–4) [10]. In addition to these patients, we investigated the plasma and CSF concentrations of four other NSCLC patients with CNS metastases (cases 5–8) (Table 1).

Table 1 Patient characteristics
Full size table

Methods

Blood and CSF samples were obtained just before the administration of 150 mg erlotinib on day 8 when steady-state plasma concentrations of erlotinib were assumed to be achieved. Plasma and CSF concentrations of erlotinib were determined using high-performance liquid chromatography with ultraviolet detection as previously reported [11]. Before the collection of samples and analysis, we obtained written informed consent from all patients.

Statistical analysis

Linear regression was performed on the data in Fig. 2 by fitting a straight line. The R 2 value was used to measure the strength of the relationship between the plasma and CSF concentrations of erlotinib. The 150-mg daily erlotinib data from this 66-year-old Japanese female patient were analyzed, and the 75-mg daily erlotinib data were excluded. Linear regression analysis, performed using JMP 8 software (SAS Institute, Cary, NC, USA), was two-tailed, and P values less than 0.05 were considered statistically significant.

Fig. 2
figure 2

Correlation between plasma concentrations and cerebrospinal fluid (CSF) concentrations of erlotinib. A good correlation (R 2 = 0.84) was demonstrated (P = 0.0005)

Full size image

Results

The clinical characteristics of all patients (n = 9) are summarized in Table 1. In CNS lesions, 7 of the 9 patients achieved PR (78%) and 5 of the 6 patients with EGFR mutations achieved PR (83%). Gefitinib was administered to three patients prior to erlotinib therapy; these patients achieved PR but had CNS recurrence after a long PFS of more than 6 months. Two of these three patients achieved PR after erlotinib therapy in spite of gefitinib resistance.

The plasma and CSF concentrations are summarized in Table 2. The mean ± standard deviation (SD) of CSF concentrations and penetration rates were 106 ± 59 nM and 4.5 ± 1.5%, respectively. As seen in Fig. 2, there was a good correlation (R 2 = 0.84) between plasma and CSF concentrations (P = 0.0005).

Table 2 Plasma and cerebrospinal fluid concentrations of erlotinib, and central nervous system metastases response
Full size table

Discussion

This study demonstrated that there is a good correlation (R 2 = 0.84) between plasma and CSF concentrations of erlotinib (P = 0.0005), which indicates that high CSF concentrations can be achieved by high plasma concentrations. That is, high CSF concentrations can be achieved by high-dose erlotinib administration. CNS metastases of the 66-year-old Japanese female patient responded to 150 mg daily erlotinib despite resistance to 75 mg daily erlotinib, likely because of higher CSF concentrations. There are several reports based on the hypothesis that high CSF concentrations may be achieved by high-dose erlotinib administration [5–9], and our study supports this hypothesis.

The mean ± SD of CSF penetration rates were 4.5 ± 1.5%, and this result was similar to that of recent another study [12]. However, the rates of our study were ranged from 3.2 to 7.7% and those of another study ranged from 2.5 to 13.3%, which indicates that there seem to be difference in the penetration rate between individuals. Recently, it has been reported that some transporters restrict brain penetration of erlotinib [13, 14]. Polymorphisms of these transporters can be associated with this individual difference in the penetration rate.

Since systemic chemotherapy is thought to play a limited role in CNS metastases because of the belief that the brain is a pharmacological sanctuary, patients with NSCLC who showed initial good responses to EGFR-TKIs often had recurrent CNS metastases. Considering CSF concentrations, high-dose EGFR-TKIs can be effective for such refractory metastases. Furthermore, the association between CNS responses/recurrence and CSF concentrations should be investigated in the future.