Around the world, there are 25-40% of patients suffered from brain metastases during the course of advanced NSCLC32. Metastasis to central nervous system, primarily in the cerebral hemisphere, is a severe complication of advanced NSCLC. The prognosis of such patients is generally poor, with a median survival ranging from 2 to 6 months in the past33. Treatment options for these patients before the era of targeted therapy were quite limited, including only WBRT, stereotactic radiosurgery, surgery and chemotherapy34. Although these therapeutic methods could be combined with each other, the efficacy results is not as good as enough. Importantly, traditional chemotherapeutical methods could lead to multiple side effects including nausea, emesis, anorexia and myelosuppression. What is more, neurocognitive dysfunction and declines in quality of life is unavoidable for certain patients receiving WBRT treatment, which occurred in several months to years after initial cerebral radiotherapy35,36. Thus, novel treatment strategy with relatively tolerable toxicity is urgently needed.
As is known to us, targeted therapy could obtain better extra-cranial antitumor efficacy as compared to chemotherapy for EGFR mutated NSCLC, mainly due to inhibition of the corresponding signaling pathway. So far, there is enough evidence to show that the efficacy of gefitinib is parallel to that of erlotinib for treatment of brain metastases37. However, we still do not know whether other first generation of EGFR-TKIs could produce comparable therapeutic effect. Hence, we conducted this study comparing icotinib, one of the most frequently prescribed EGFR targeted agent in China, with gefitinib for cerebral metastases of NSCLC.
As the ICOGEN study showed to us, the efficacy of icotinib is non-inferiority when compared with gefitinib for patients with NSCLC beyond second-line treatment (median PFS 4.6 versus 3.4 months, HR 0.84)30. In this study, 24 patients (12%) diagnosed with brain metastases in the icotinib arm, while this patients number in the gefitinib arm is 26 (13%). Nevertheless, this study did not further analyze the effect of these two EGFR-TKIs on brain metastases of NSCLC. In the BRAIN study, researchers found that icotinib for first-line treatment could bring out significantly longer intracranial PFS than WBRT plus chemotherapy (median iPFS 10.0 versus 4.8 months, HR 0.56), suggesting that icotinib might be a better treatment method for NSCLC patients with brain metastases31. Still, we do not know which EGFR-TKI is more appropriate for target treatment of selected NSCLC patient harboring cerebral lesions.
In our study, we found that the whole efficacy and toxicity between icotinib and gefitinib were similar to each other (Supplementary Table 1). The results of response rates, disease control rates and median PFS in the two treatment arms were approximately 60%, 85% and 7 months, respectively. These results were in accordance with former studies, which showed that the response rate of EGFR-TKIs for intracranial lesions was up to 88%, with a disease control rate ranging from 27% to 100%38,39. No obvious difference concerning median PFS existed between the two EGFR-TKIs (Figure 2). For brain metastases, the intracranial response rate of icotinib arms was 66.6%, with 14.3% case of complete response, whereas that in gefitinib arm is 59.1% and 13.6% (Table 2). The median iPFS was 8.4 and 10.6 months, respectively, which is also similar to the previous results of EGFR-TKIs for cerebral metastases. There was also no statistically significant difference (Figure 3).
Currently, several studies have already reported the efficacy of icotinib and gefitinib for preventing brain metastasis from advanced NSCLC patients with EGFR mutation. In one study, icotinib was compared with chemotherapy as first-line therapy of advanced lung adenocarcinoma40. A total of 131 patients with EGFR mutation were treated in the icotinib group. Result showed that the cumulative risk of brain metastasis was lower in the icotinib group. In another study, the effect of gefitinib on EGFR mutated NSCLC with brain metastasis was evaluated41. There were 30 eligible patients received gefitinib according to the therapeutic strategy. It is revealed that gefitinib is effective for treating this subtype of NSCLC with cerebral metastasis. Based on the results of above studies, we conclude that EGFR-TKIs alone are effective for treatment of brain metastasis of EGFR mutated NSCLC. Nevertheless, we still do not know whether each of the first generation of EGFR-TKIs function similarly for these NSCLC patients. Our study confirmed that icotinib is as efficacious as gefitinib for brain metastasis of advanced NSCLC with EGFR mutation.
However, icotinib or gefitinib as single therapy may not be enough for EGFR-mutated NSCLC patients with cerebral metastasis, as compared with osimertinib, which has shown favorable intracranial efficacy for such patients42. Combined treatment is needed in order to improve the efficacy of first generation of EGFR targeted therapy. Several studies showed that EGFR-TKIs plus radiation therapy could produce better efficacy43,44. The combination of icotinib or gefitinib with chemotherapy or anti-angiogenesis therapy may also enhance the treatment results of first generation of EGFR-TKIs, especially for those patients who can not afford the treatment of osimertinib.
As for treatment related toxicities, there is still no obvious difference between icotinib and gefitinib. There were only 4 cases of grade 3/4 adverse events occurred in this study, 2 in the icotinib arm and 2 in the gefitinib arm. These results indicated that icotinib was as efficacious as gefitinib for brain metastasis of EGFR mutated NSCLC, with generally mild toxicities. The toxicity profile in our study was also in accordance with the ICOGEN and BRAIN studies. No other side effects related with the two EGFR-TKIs were observed during the treatment procedure. Both icotinib and gefitinib were well tolerable among NSCLC patients with cerebral metastases. None of the 43 patients investigated in the study experienced dose reduction of EGFR-TKIs. Furthermore, since icotinib is cheaper than any other targeted agents such as gefitinib, erlotinib and afatinib, physicians would be more prone to prescribe icotinib for initial treatment of NSCLC in the clinic. In a word, it is suggested that icotinib might be an ideal therapeutic option for this patient population who is intolerable for severe toxicity as a result of cerebral lesions.
In this study, we also noted that the intracranial efficacy of icotinib was superior than the whole efficacy for NSCLC. This phenomenon was mainly due to the presence of BBB. As we all know, an intact BBB could reduce intracranial drug uptake of chemotherapeutic agents and certain EGFR-TKIs. Furthermore, multidrug resistance related protein such as ATP binding cassette (ABC) transporters is also located at the surface of BBB. It is confirmed that icotinib is one of the substrates of ABC transporters45. Therefore, the cerebrospinal fluid (CSF) concentrations may be lower than the corresponding plasma concentrations, which in turn caused delayed secondary resistance mutation in brain metastases as compared to the extra-cranial sites. Ultimately, the median iPFS was longer than the whole PFS with icotinib treatment. The similar results regarding median PFS were also observed in the gefitinib arm. On the contrary, the intracranial ORR and DCR is inferior as compared to the whole efficacy in our study. The specific reason is remaining uncertain. Basic research is needed to clarify the mechanism of this phenomenon.