Erlotinib for advanced hepatocellular carcinoma

Discussion

This systematic review provides a comprehensive overview of the current evidence of the efficacy and safety of erlotinib for the treatment of advanced HCC. Although it was impossible to perform an overall meta-analysis for several reasons, in general, the results reported in these studies demonstrate that erlotinib is efficacious and well-tolerated in the treatment of advanced HCC.

When used as monotherapy, the DCR of erlotinib for advanced HCC was good, but the RR was poor,7,8 suggesting that erlotinib alone exhibited modest antitumor activity. However, the promising DCR, TTP, and OS rates in these studies encouraged clinicians to explore the use of erlotinib in conjunction with other anticancer therapies. For example, erlotinib was administered in combination with bevacizumab in 6 phase II trials;9-14 among these 6 trials, 4 phase II trials were included in another systematic review on bevacizumab performed by Fang et al.17 Erlotinib has also been combined with docetaxel in another study,15 and with sorafenib in a phase III RCT.16 Only 2 studies,9,10 with overlapping study populations, showed superior RRs for the combination of bevacizumab with erlotinib when compared with erlotinib monotherapy.7,8 However, other drug combinations did not show any significant benefits for RR.11-16 Most of the drug combinations9-11,14-16 showed similar or slightly better DCRs compared with erlotinib monotherapy,7,8 except for the combinations used in the study by Govindarajan et al12 and Yau et al,13 both of which utilized bevacizumab. All combined therapy studies, except one with bevacizumab,12 demonstrated similar or slightly improved PFS/TTP and OS rates compared with erlotinib monotherapy.7,8

None of the trials reported any unexpected toxicities, and most were grade 1 or 2 in severity. In the phase II studies, there were only 2 patient deaths; one was due to pneumonia11 and deemed unrelated to treatment, and the other was due to gastrointestinal bleeding that occurred during study, but whether it was treatment-related was not disclosed.9 Overall, the number of withdrawals due to AEs was low. In the phase III RCT, the difference in the death or withdrawal rates between the 2 study groups was not significant, suggesting that most were related to sorafenib rather than erlotinib.16 Therefore, the safety analysis in these reports suggested that erlotinib was safe and well-tolerated.

Although sorafenib is the only approved standard treatment for advanced HCC, the survival benefit is relatively modest according to 2 studies.18,19 In the SHARP trial,18 the DCR and median OS rates were 43% and 10.7 months, and in the Asia-Pacific region phase III RCT, they were 35.3% and 6.5 months.19 Overall, the effect of erlotinib for the treatment of advanced HCC in all the trials evaluated in this present study, except for that of Yau et al,12 is comparable with that of sorafenib.18,19 Yau et al12 reported that none of their enrolled 10 patients, who were resistant to sorafenib, responded to the therapeutic combination of erlotinib and bevacizumab. This conclusion suggested that patients who were resistant to sorafenib would not obtain any clinical benefit from treatment with erlotinib or bevacizumab, and the trial was terminated early due to poor patient response. For this reason, the data from that study should be interpreted with caution. However, Kaseb et al10 indicated that PFS and OS in response to erlotinib plus bevacizumab were shorter in those previously treated with sorafenib (PFS of 6.7 months and OS of 9.6 months) when compared with treatment-naive patients (PFS of 7.4 months and OS of 15.0 months). This might suggest further that in sorafenib non-responders, erlotinib plus bevacizumab may not be an appropriate treatment regimen.

Sorafenib can block vascular endothelial growth factor (VEGF) and platelet-derived growth factor (PDGF) receptor signaling,20,21 whereas erlotinib blocks EGFR signaling;6 it was hoped that the combination might have a synergistic anticancer effect. Because the combination of erlotinib and sorafenib showed promising antitumor efficacy in a phase I trial,22 and erlotinib showed moderate efficacy in phase II studies,7,8 phase II trials were skipped and a phase III trial was performed directly. Finally, this combination provided no synergistic benefit.16 In this study, we used SEARCH trial,16 erlotinib plus sorafenib as a first-line treatment in advanced HCC patients had similar OS (p=0.408) and TTP (p=0.18) outcomes compared with placebo plus sorafenib. Possible reasons for the failure of this trial to produce any meaningful effect in advanced HCC patients might be explained by the findings of a study conducted by Sieghart et al23 in a rat model of HCC. Sieghart et al23 found that treatment with erlotinib and sorafenib was not superior to the effect of sorafenib alone in vitro or in vivo, suggesting a lack of any synergistic or additive effect between the 2 drugs. However, they also suggested that erlotinib-induced VEGF expression might have contributed to the failure of the combination therapy.23 This trial demonstrates that we should perform research systematically following the study-design strategy guidelines, instead of carry out large-scale phase III trials without a clear signal of efficacy from earlier-phase studies.

HCC trials are characterized by study population heterogeneity and are, therefore, complex to interpret and dissect.18 Drugs that target all HCC subtypes will need to be explored, and there is room for such drugs, especially for the second-line or third-line treatments. There is an intensive requirement for valid biomarkers to enable selection of suitable subgroups of HCC patients who would be more likely to achieve an objective response to specific anticancer agents.24 For example, erlotinib efficacy is affected by the T790M mutation in the EGFR kinase in non-small-cell lung cancer.25 Therefore, such cases should encourage analogous studies with biomarkers to predict erlotinib efficacy in future phase III trials.

Moreover, sample augmentation or incorporation of a control arm in phase II studies would reduce selection bias and decrease the risk of random effects to obtain an impartial comparison for outcome analysis.18,26 These approaches might reduce the possibility of negative outcomes for future phase III trials.

In all of the included trails, except 2 trails,9,10 the objective response rate was low, but it may be inappropriate to apply this extremely stringent parameter. As alternative endpoints in these studies, DCR, PFS/TTP, and OS have thus been adopted. However, DCR, PFS/TTP, and OS are endpoints that reflect not only treatment effect but also tumor behavior.27 Perhaps it would be better to use primary endpoints to measure the treatment effect of erlotinib.

All of the previous observations remind us to optimize the design of trails, augment sample size, select sensitive subset populations, and use better primary endpoints in order to improve the therapeutic effect in patients with advanced HCC.

There are some limitations in the present systematic review. First, publication bias may affect this area of study and research. In this systematic review, data for one study15 were derived from an abstract because the full-text publication could not be identified, preventing a more in-depth analysis. Second, heterogeneity among trials could affect the validity of the present results. Only one study was an RCT, and most of phase II studies had no head-to-head comparisons. Furthermore, heterogeneity may exist owing to the differences in patients’ characteristics, drug dosages, and treatment regimens, which impedes comparisons between studies. Third, there could be discrepancies in the measurement of outcome measures. For example, PFS is frequently confused with TTP,28 and an unequivocal definition of PFS is seldom provided.

In conclusion, these results are encouraging, and suggest that erlotinib may be a relatively efficacious and well-tolerated treatment for advanced HCC. Owing to the small sample size and lack of controls, this conclusion is not definitive. Although the efficacy of erlotinib was not confirmed in the sole phase III trial, it should not be discarded as a possible treatment option and should be evaluated in future. Given the better overall survival observed in most phase II trials and because treatment of advanced HCC remains unstandardized and unsatisfactory, it is worth determining whether erlotinib can be of benefit to patients with advanced HCC. In future studies, further investigation of the underlying pathogenesis of HCC and exploration of sensitive subset populations are warranted to enhance the effect of erlotinib, thereby improving survival outcomes in patients with advanced HCC. Meanwhile, this systematic review has identified the need for more well-designed RCTs to evaluate the efficacy and safety of erlotinib as monotherapy and in combination with other drugs for advanced HCC.