Abstract
Median survival of patients with brain metastases from nonsmall cell lung cancer (NSCLC) is poor and more effective treatments are urgently needed. We have evaluated the efficacy of erlotinib in this setting and its association with activating mutations in the epidermal growth factor receptor (EGFR) gene.
We retrospectively identified patients with NSCLC and brain metastases treated with erlotinib. EGFR mutations in exons 19 and 21 were analysed by direct sequencing. Efficacy and tolerability were compared according to EGFR mutational status.
69 NSCLC patients with brain metastases were identified, 17 of whom harboured EGFR mutations. Objective response rate in patients with EGFR mutations was 82.4%; no responses were observed in unselected patients (p<0.001). Median (95% CI) time to progression within the brain for patients harbouring EGFR mutations was 11.7 (7.9–15.5) months, compared to 5.8 (5.2–6.4) months for control patients whose EGFR mutational status had not been assessed (p<0.05). Overall survival was 12.9 (6.2–19.7) months and 3.1 (2.5–3.9) months (p<0.001), respectively. The toxicity of erlotinib was as expected and no differences between cohorts were observed.
Erlotinib is active in brain metastases from NSCLC; this clinical benefit is related to the presence of activating mutations in exons 19 or 21 of the EGFR gene.
Lung cancer is the leading cause of cancer-related death worldwide. Brain metastases from nonsmall cell lung cancer (NSCLC) are present in 20–30% of patients 1 and are associated with a poor prognosis despite treatment with whole brain radiotherapy (WBRT), with a median survival of <6 months 2. Apart from WBRT, few treatment options are currently available for these patients.
Tyrosine kinase inhibitors (TKIs) of the epidermal growth factor receptor (EGFR) are novel treatment options for advanced NSCLC, with a reported response rate of 9% in an unselected non-chemotherapy-naïve population 3. Activating EGFR mutations within the tyrosine kinase (TK) domain are found to be highly associated with sensitivity to the EGFR TKIs gefitinib or erlotinib in advanced NSCLC 4–6. Almost 90% of all known mutations in the TK domain of the EGFR gene are located in exon 19 (in-frame deletion of the conserved sequence LREA) or in exon 21 (L858R point mutation). Recent studies have shown that these EGFR mutations are highly oncogenic in transgenic mice and maintenance of the lung tumours in these mice is dependent on continued expression of the EGFR mutants 7, 8. These data suggest that NSCLC expressing EGFR mutants is itself a different molecular entity 9, 10.
Although individual case reports of patients achieving objective responses to erlotinib or gefitinib have been published, the role of TKIs in patients with brain metastases remains unclear. To address this issue, we retrospectively evaluated the efficacy of erlotinib in a series of patients with brain metastases from NSCLC and its association with the presence of activating mutations in the EGFR gene. Safety was evaluated as a part of the analysis.
MATERIALS AND METHODS
Patients
We retrospectively evaluated patients with NSCLC and metastatic dissemination to the brain, who had been registered in the Spanish Lung Adenocarcinoma Data Base Study (SLADB) from April 2005 to May 2006. The SLADB is a large database sponsored by the Spanish Lung Cancer Group (SLCG), whose aim was to evaluate the feasibility of large-scale screening for EGFR mutations in NSCLC patients and to examine the association between the mutations and the outcomes of the treatment with erlotinib. Primary tumour biopsy specimens from 2,105 NSCLC patients were analysed 11 and only those harbouring EGFR mutations were included in the database.
In addition, in order to have a control population of patients with brain metastasis from lung cancer, we consulted the TargeT study database and picked patients with brain metastasis enrolled during the same time period whose EGFR mutational status was either unknown or wild-type. The TargeT study is a Spanish nonrandomised phase II trial evaluating the efficacy and safety of first- and second-line erlotinib in patients with histologically confirmed stage IIIB or IV NSCLC. Erlotinib was given at a daily dose of 150 mg until disease progression or severe toxicity.
Both the SLADB and the TargeT study were approved by the corresponding institutional review boards and patients provided written informed consent prior to enrolment.
Efficacy and safety
Assessment of treatment efficacy at the brain level was periodically performed by brain magnetic resonance imaging or computed tomography (CT) scan, according to the clinical practice of each site. Lung tumour response was evaluated by CT scan. Liver or bone metastases, if present, were evaluated by upper abdominal CT scan and bone scan, respectively. Efficacy is reported in terms of objective response rate according to the Response Evaluation Criteria in Solid Tumours 12, time to progression (TTP) and overall survival (OS). TTP of the intracranial lesions was measured from the date of first erlotinib intake until the date of progression within the brain or last follow-up. OS was measured from the date of first erlotinib intake until death or last survival follow-up. Safety data consists of the adverse events related to erlotinib according to the National Cancer Common Toxicity Criteria version 3 grading system 13.
EGFR mutation analysis
The analysis of EGFR mutations was performed at the central laboratory of the SLCG at the Catalan Institute of Oncology (Hospital Germans Trias i Pujol, Badalona, Spain). EGFR mutations in exons 19 and 21 were analysed as described previously 11. For more details on genetic analysis see the online supplementary data S1.
Statistical analysis
Patient characteristics are listed by their frequencies for qualitative variables and by median values and ranges for quantitative variables. Differences among response rates were analysed by the Chi-squared test or Fisher's exact test, as appropriate. Actuarial progression and survival curves were generated using the Kaplan–Meier method. The log-rank test was used to detect differences between subgroups. Two-sided p-values <0.05 were considered statistically significant. Statistical analyses were performed using SPSS for Windows version 13.0 (SPSS, Inc., Chicago, IL, USA).
RESULTS
Patient and tumour characteristics
This retrospective analysis includes 69 patients with NSCLC metastatic to the brain, whose main baseline and clinical characteristics are summarised in table 1. Most of the patients were current or former smokers (68.0%), with adenocarcinoma (68.0%) and an Eastern Cooperative Oncology Group performance status of one (61.5%). 37 (53.6%) patients were male.
Of the 69 patients with brain metastases, 17 (24.6%) harboured mutations in the EGFR gene. An in-frame deletion in exon 19 (E746-A750) was found in 12 (70.6%) patients, while a point mutation in exon 21 (L858R) was detected in the remaining five (29.4%) patients. The majority of patients with EGFR mutations were female (64.7%), never-smokers (64.7%) and had adenocarcinomas (82.4%).
In contrast, the 52 control cases (75.4% of the whole series) from the TargeT study were unselected patients, whose EGFR mutational status had not been assessed (50 patients) or had confirmed wild-type EGFR gene (two cases); these control patients were mainly males (59.6%) and former or current smokers (78.8%); adenocarcinoma was also the predominant histology in this group (63.5%).
Of the entire series, 55 patients were treated with standard WBRT prior to erlotinib treatment: nine (16.4%) patients with EGFR mutation and 46 (84.6%) in the control group. Approximately half (47.1%) of the patients with EGFR mutations did not receive WBRT, and oral erlotinib was the sole treatment. In contrast, all control patients with available data of treatment had received erlotinib plus radiotherapy. Median (range) time from the end of WBRT treatment until the beginning of erlotinib intake was 42 (9–270) days. None of the patients received stereotactic radiation or underwent resection of the brain lesions.
Nine (52.9%) of the 17 patients harbouring EGFR mutations and 23 (44.2%) control cases received chemotherapy after erlotinib treatment failure.
Treatment efficacy
Response was not evaluable in 16 patients due to early death; 53 patients were evaluable for response. 14 (26.4%) patients attained an objective response in the brain lesions. All of them harboured mutations in the EGFR gene. Three patients with EGFR mutations had stabilisation of the intracranial lesions. Therefore, the objective response rate in the subgroup of evaluable patients with EGFR mutations was 82.4%, with complete resolution of the brain metastases in eight cases (47.1%) and partial response in six (35.3%).
No objective response within the brain was reported among patients in the control cohort, even though they had all received WBRT. Difference in response rate between patients with EGFR mutations and unselected control patients was statistically significant (p<0.001; table 2). Remarkably, however, 77.8% of patients with the unknown EGFR mutational status showed stabilisation of the brain disease after treatment with WBRT plus erlotinib.
In the subgroup of patients with EGFR mutations, eight (47%) patients did not receive WBRT and erlotinib was the only treatment; of those, six (75%) achieved an objective response (complete response and partial response) (table 3). A representative case of brain response to erlotinib (case number 5) is shown in figure 1. All patients but one receiving erlotinib plus WBRT showed response of the intracranial disease (table 3).
Brain magnetic resonance imaging. Axial T1-weighted image after contrast (Gd) administration of patient 5, a male with adenocarcinoma of the lung and a brain metastasis, harbouring exon 19 deletion. a) A brain metastasis in the right hemisphere of the cerebellum before initiating erlotinib (white arrow). A sebaceous cyst is seen into the right subcutaneous tissue. b) Complete response of the cerebellum metastasis after 4 months of treatment with erlotinib.
In addition to the efficacy within the brain, the response of the primary tumour and extracranial metastases (if present) was also evaluated in the subgroup of patients with activating EGFR mutations (table 3). All patients with EGFR mutations showed tumour response or disease stabilisation. All patients but one achieving an objective response of the intracranial lesions also attained a response in the extracranial locations. Two of the three patients with stable disease in the brain attained a partial response in the primary tumour as well as in the extracranial metastases. One patient had stable disease at both the thoracic and brain levels.
Median (95% CI) time to progression in the brain for the entire series was 2.9 (2.3–3.5) months. Patients harbouring EGFR mutations had a median (95% CI) TTP within the brain of 11.7 (7.9–15.5) months, compared to 5.8 (5.2–6.4) months in the control cohort (p<0.05) (fig. 2a). Of the 13 progressing patients harbouring EGFR mutations, six experienced disease progression in the primary lung lesions, four within the brain and three in the liver (table S1 in the online supplementary material).
a) Median time to progression (TTP) in the brain. Patients harbouring EGFR mutations (blue line) had a median (95% CI) TTP within the brain of 11.7 (7.9–15.5) months, compared to 5.8 (5.2–6.4) months in the control cohort (green line) (p<0.05). b) Median overall survival (OS). Patients harbouring EGFR mutations (blue line) had a median (95% CI) OS of 12.9 (6.2–19.7) months while the control group (green line) showed a median OS of 3.1 (2.5–3.9) months (p<0.001). In the control groups the EGFR mutational status was either unknown or wild-type.
Median (95% CI) OS for the entire population was 4.3 (2.3–6.2) months. Patients harbouring EGFR mutations had a median OS of 12.9 (6.2–19.7) months while the control group showed a median (95% CI) OS of 3.1 (2.5–3.9) months (p<0.001) (fig. 2b). 1-yr survival was 69% in those patients with mutations and 9% in the unselected population (p<0.001) (table S2 in the online supplementary material).
No differences in response rate, TTP within the brain and OS were found according to performance status and treatment line (data not shown) among patients harbouring EGFR mutations, but the small population does not allow definitive conclusions.
Treatment toxicity
The most common side-effects of erlotinib were rash and diarrhoea. Skin disorders occurred in 37 (53.6%) cases. Grade ≥3 skin toxicity, including desquamative lesions, pruritus, acne, conjunctivitis and alopecia were more frequent in patients with EGFR mutations (18.7%) than in the control cases (11.5%), although this difference did not reach statistical significance (p = 0.17). The initial dose of erlotinib was reduced to 100 mg a day in five patients with grade 3 skin toxicity. This measure was sufficient to decrease the skin toxicity to grade 2. Gastrointestinal toxicity was mild. 17 (25%) of the 69 patients experienced some gastrointestinal symptom. Grade 3–4 diarrhoea was reported in 4% of patients in the control group, whereas none of the patients with EGFR mutations developed severe diarrhoea.
DISCUSSION
This retrospective study shows that the EGFR TKI erlotinib is active in patients with brain metastases from NSCLC. We have observed an overall response rate of 26.4% in a series of 69 NSCLC patients with metastatic dissemination to the brain treated either with WBRT plus erlotinib or erlotinib alone. Disease control was achieved in an impressive 84.9% of the patients. We have also identified a group of patients with brain metastases in whom erlotinib is particularly active. Those patients harbouring activating mutations in the EGFR TK domain attained an objective response rate of 82.4%, in some cases with highly dramatic complete responses (47.1%). In contrast, unselected patients, whose EGFR mutational status was either unknown or wild-type, showed no objective responses, even though disease control occurred in 77.8% of the population. A significant difference in TTP within the brain lesions and in OS was also observed according to EGFR mutational status. TTP within the brain for patients harbouring EGFR mutations (11.7 months) was twice that for unselected patients (5.8 months). Furthermore, patients harbouring EGFR mutations had four-fold longer OS (12.9 months) than those patients in the control group (3.1 months). 1-yr survival (69%) for patients with mutations was particularly remarkable, since median OS in unselected patients with lung cancer metastatic to the brain is normally <6 months after conventional therapy 2. Median TTP for patients with the mutated EGFR gene was of similar magnitude to that described in larger series with erlotinib. In a recent prospective study with erlotinib, reported median progression-free survival was 14 months 11, and in a pooled analysis examining five studies of first-line treatment with erlotinib or gefitinib in patients in whom EGFR mutational status was analysed, median progression-free survival for those patients harbouring activating mutations was 11.8 months 14. By contrast, median OS in our series was shorter than that reported by other authors. This result could be partly due to a shorter follow-up in our study, but it also may reflect the worse prognosis of those patients with brain metastasis and the modest results yielded by other therapeutic approaches, thus underlining the benefit provided by erlotinib. In the work from Rosell and co-workers 10, 11 treatment with erlotinib reached a median (95% CI) progression-free survival of 14 (11.3–16.7) months for patients without brain metastases and 10 (5.6–14.4) months for those with brain metastases (p = 0.31). Median (95% CI) survival was 28 (21.5–34.4) months for patients without brain metastases and 18 (4–31.9) months for patients with brain metastases (p = 0.008) (see appendix in the online supplementary material) 11.
Several reports support that stereotactic radiosurgery, Gamma Knife or linear accelerator, with or without WBRT, are interesting local therapeutic approaches for a limited number of small brain metastases and good prognosis. However, most cases require a systemic approach to provide a treatment for the extracranial disease 15. It has been suggested that EGFR mutations confer radiosensitivity in vitro 16, and recently Gow et al. 17 have concluded that the presence of EGFR mutations is an independent predictor of response to WBRT in brain metastases of lung adenocarcinoma. The impact of erlotinib on brain metastases might thus have been masked by the effects of radiation therapy to the brain. However, our study clearly shows that those patients with brain metastases and EGFR mutations are better responders to erlotinib, whether or not they had received previous WBRT. All patients with EGFR mutations obtained benefit within the brain (82.4% with objective response and 7.6% with stable disease as the best response), and 47.1% attained a complete remission of the cerebral lesions. Interestingly, six (42.9%) of the 14 patients with EGFR mutations achieving objective tumour response had not received brain radiation therapy, and four of these six attained a complete remission of brain lesions. This finding strongly supports the role of erlotinib in the response of the brain metastases.
Moreover, the efficacy of erlotinib in brain metastases was paralleled by its efficacy in the lung primary lesions and in other metastatic sites. All patients with EGFR mutations responding to treatment within the brain also responded in the extracranial lesions. In fact, brain lesions seem to be more sensitive to erlotinib than thoracic tumours: eight patients with complete responses within the brain, four of whom were treated only with erlotinib, attained partial responses in their primary tumours. Therefore, we can conclude that erlotinib is active both in brain metastases and in lung primary lesions and other metastatic sites more accessible than the brain.
In the present study, there was a difference in the number of treatment lines between patients with EGFR mutations and patients with unknown EGFR mutational status; unselected patients were more likely to have received previous therapies. While this could account for differences in outcomes between the two groups of patients, 41.2% of patients with EGFR mutations received erlotinib as a second or further line of treatment, and median TTP in this subgroup remained longer than 11 months. Moreover, among patients harbouring EGFR mutations, no significant differences in response rate, TTP within the brain and OS were detected according to line of treatment and performance status, but these data should be cautiously interpreted due to the small size of the subgroups.
Our findings support the hypothesis that erlotinib is able to cross the blood–brain barrier and displays efficacy against intracranial metastasis. In the past, the response of malignancies involving the brain has been anecdotal 18, which might reflect the absence of active medical treatments, rather than the refractoriness of brain lesions to all forms of therapy. We have previously reported that tamoxifen, which is usually regarded as ineffective in breast cancer involving the brain, induced a complete response in a patient with brain metastases from breast cancer 19.
The results observed in the present series of patients with brain metastases confirm other isolated reports of the efficacy of EGFR TKIs 20–24. Gefitinib has been reported to be active in a series of patients with brain metastases 21–24, most of them Asiatic, although a high incidence of recurrence at the brain level after treatment with gefitinib has also been addressed 25. In a prospective trial, Ceresoli et al. 24 showed efficacy of gefitinib on brain metastases from 41 patients with NSCLC, with a median overall survival of 5 months. None of the mentioned studies selected the patients for treatment according to the mutational status of the EGFR gene, or carried out this analysis. It has been pointed out that gefitinib may have an incomplete penetration though the blood–brain barrier 26 and its effectiveness for the treatment of brain metastasis may depend on the disruption of the barrier 27.
Finally, the tolerability of oral TKIs in patients with brain metastases has not been specifically addressed before, although this is particularly relevant in the case of oral drugs. Erlotinib was well tolerated overall in patients with brain metastases, with skin toxicity and diarrhoea as the most common adverse events. Skin toxicity has been associated with clinical benefit to erlotinib, but its relationship with EGFR mutations has not been evaluated 28. In the present study, a nonsignificant trend towards more severe skin toxicity in patients with EGFR mutations was observed.
In conclusion, erlotinib is well tolerated and active against brain metastases in NSCLC patients. The routine assessment of EGFR mutations in NSCLC patients with intracranial lesions is warranted.
Acknowledgments
The author affiliation details are as follows: R. Porta, Dept of Medical Oncology, Catalan Institute of Oncology, Hospital Universitari Dr. Josep Trueta, Girona and Institut de Recerca Biomèdica de Girona (IdIBGi), Hospital Universitario Dr. Josep Trueta, Girona; J.M. Sánchez, Dept of Medical Oncology, M.D. Anderson España Cancer Center, Madrid; L. Paz-Ares, Dept of Medical Oncology, Hospital Universitario Virgen del Rocío, Seville; B. Massutí, Dept of Medical Oncology, Hospital General Universitario de Alicante, Alicante; N. Reguart, Dept of Medical Oncology, CSC-Hospital Clinic de Barcelona, Barcelona; C. Mayo, Pangaea Biotech, USP Institut Universitari Dexeus, Barcelona; P. Lianes, Dept of Medical Oncology, Hospital de Mataró, Mataró; C. Queralt, Pangaea Biotech, USP Institut Universitari Dexeus, Barcelona and Dept of Medical Oncology, Catalan Institute of Oncology and Autonomous University of Barcelona, Hospital Germans Trias i Pujol, Barcelona; V. Guillem, Dept of Medical Oncology, Institut Valencià d'Oncologia – IVO, Valencia; P. Salinas, Dept of Medical Oncology, Hospital Sanitas La Zarzuela, Aravaca; S. Catot, Dept of Medical Oncology, Althaia-Xarxa Assistencial de Manresa, Manresa; D. Isla, Dept of Medical Oncology, Hospital Clínico Universitario Lozano Blesa, Zaragoza; A. Pradas, Dept of Medical Oncology, Catalan Institute of Oncology and Autonomous University of Barcelona, Hospital Germans Trias i Pujol, Barcelona; A. Gúrpide, Dept of Medical Oncology, Clínica Universitaria de Navarra, Pamplona; J. de Castro, Dept of Medical Oncology, Hospital Universitario La Paz, Madrid; E. Polo, Dept of Medical Oncology, Hospital Ernest Lluch, Calatayud; T. Puig, Institut de Recerca Biomèdica de Girona (IdIBGi), Hospital Universitario Dr. Josep Trueta, Girona and Dept of Biochemistry and Molecular Biology, Universidad de Girona, Girona; M. Tarón, Pangaea Biotech, USP Institut Universitari Dexeus, Barcelona and Dept of Medical Oncology, Catalan Institute of Oncology and Autonomous University of Barcelona, Hospital Germans Trias i Pujol, Barcelona; R. Colomer, Dept of Medical Oncology, M.D. Anderson España Cancer Center, Madrid; R. Rosell, Pangaea Biotech, USP Institut Universitari Dexeus, Barcelona and Dept of Medical Oncology, Catalan Institute of Oncology and Autonomous University of Barcelona, Hospital Germans Trias i Pujol, Barcelona, Spain.
The authors wish to thank the Spanish Lung Cancer Group (SLCG), sponsor of the SLADB study, and Roche Farma Spain, sponsor of the TargeT trial. We also acknowledge the investigators C. Santander, Dept of Medical Oncology, Hospital de San Millán, Logroño, Spain; P. Diz, Dept of Medical Oncology, Complejo Hospitalario de León, León, Spain; J. Oramas, Dept of Medical Oncology, Hospital Universitario de Canarias, Santa Cruz de Tenerife, Spain; R. Pérez, Dept of Medical Oncology, M.D. Anderson España Cancer Center, Madrid, Spain; A. García-Velasco, Dept of Medical Oncology, Catalan Institute of Oncology, Hospital Universitari Dr Josep Trueta, Girona, Spain; G. López, Dept of Medical Oncology, Hospital Universitario de Cruces, Baracaldo, Spain; A. Velasco, Dept of Medical Oncology, Hospital Universitario la Princesa, Madrid; S. del Barco, Dept of Medical Oncology, Catalan Institute of Oncology, Hospital Universitari Dr Josep Trueta; J. Montesinos, Dept of Medical Oncology, Fundacio Althaia-Xarxa Assistencial de Manresa, Manresa, Spain; P. Valero, Dept of Medical Oncology, Hospital Infanta Luisa, Seville, Spain; M. González, Hospital Universitario la Paz, Madrid, Spain; A. Montes, Dept of Medical Oncology, Catalan Institute of Oncology, L'Hospitalet, Spain; A. Izquierdo, Dept of Medical Oncology, Catalan Institute of Oncology, Hospital Universitari Dr Josep Trueta; J. Montalar, Dept of Medical Oncology, Hospital Universitario La Fe, Valencia, Spain; M. Amenedo, Dept of Medical Oncology, Centro Oncológico Regional de Galicia, La Coruña, Spain; C. Rolfo, Dept of Medical Oncology, Clínica Rotger, Palma de Mallorca, Spain; I. Bover, Dept of Medical Oncology, Hospital Son Llatzer, Mallorca, Spain; J.E. Alés, Dept of Medical Oncology, Complejo Hospitalario de Avila, Avila, Spain; P. Martínez, Dept of Medical Oncology, Hospital de Basurto, Bilbao, Spain; E. Felip, Dept of Medical Oncology, Hospital Universitari Vall d' Hebron, Barcelona, Spain; M. Cobo, Dept of Medical Oncology, Hospital Universitario Puerta de Hierro Majadahonda, Madrid; A.M. Jiménez, Dept of Medical Oncology, Hospital Universitario de Getafe, Madrid; M. Guillot, Dept of Medical Oncology, Hospital Son Dureta, Mallorca; A. Jaén, Dept of Medical Oncology, Complejo Hospitalario de Jaén, Jaén, Spain; P. Garrido, Dept of Medical Oncology, Hospital Universitario Ramón y Cajal, Madrid; C. Pallarès, Dept of Medical Oncology, Hospital Universitari de la Santa Creu i Sant Pau, Barcelona; and A. Cortijo, MD Anderson Cancer Center España, Madrid; as well as M.L. Amador and S. Figueroa, both Roche-Pharma, Basel, Switzerland; and J. Brunet, Dept of Medical Oncology, Catalan Institute of Oncology, Hospital Universitari Dr Josep Trueta. The authors also wish to thank all clinical research coordinators who collaborated in the analysis reported in this manuscript.
Footnotes
This article has supplementary material available from www.erj.ersjournals.com
Support Statement
Financial support was provided by the Susan G. Komen Breast Cancer Foundation (PDF 0504073) (R. Porta.) and the Spanish Society of Medical Oncology, Research Projects 2005 (R. Porta). The Lung Adenocarcinoma Data Base Study (SLADB) is sponsored by the Spanish Lung Cancer Group and the TargeT trial is sponsored by Roche Farma, S.A. (Spain).
Statement of Interest
A statement of interest for the present study can be found at www.erj.ersjournals.com/site/misc/statements.xhtml
- Received December 11, 2009.
- Accepted June 21, 2010.
- ©ERS 2011
REFERENCES
