Phase II study of bevacizumab and erlotinib in the treatment of advanced hepatocellular carcinoma patients with sorafenib-refractory disease

In the past few decades, there has been a step forward in the treatment of advanced HCC. Sorafenib has demonstrated modest activity in improving the clinical outcome of this lethal disease. To date, it is the only approved targeted agent for this indication. On the other hand, the absolute magnitude of clinical benefit in terms of OS associated with sorafenib was only 2–3 months [7, 8]. Furthermore, all patients on sorafenib eventually progress on the treatment. The growing sorafenib-refractory patient population is therefore an important clinical problem and the development of second-line treatment is urgently needed in the HCC community. The current study explores the role of further biological pathway inhibition after sorafenib failure; it is the first reported in the literature evaluating the combination of B and E in patients who had received prior sorafenib treatment. The results of the present study showed that B + E was well tolerated in advanced HCC patients previously treated by sorafenib and/or other therapies. However, it did not demonstrate any signal of activity in using B + E combination in treating advanced HCC progressed after sorafenib failure.

Thus far in the literature, there are only two published studies in evaluating the efficacy of B + E in treating advanced HCC patients and all these studies were tested in the patients naïve to anti-VEGF therapies. An impressive RR of 25 %, median progression free survival (PFS) of 9 months and OS of 15.65 months were reported in a single institute phase II trial which included 40 advanced HCC patients [18]. This trial had excluded patients who had prior anti-VEGF or anti-EGFR agents, although 20 % of the enrolled patients had been treated with one line of systemic chemotherapy. Notably, with longer follow-up period and recruitment of more patients, the overall RR (28 %) and CB rate (90 %) were maintained. Nevertheless, the PFS (7.9 months) and OS (12.8 months) were not as impressive as the results shown in the initial report [22]. On the other hand, Philip et al. recently reported the results of a multi-institutional phase 2 study with similar design to the aforesaid study; only one out of 27 patients had PR while 11 % had SD, with median PFS of 3 months and OS of 9.5 months [19]. In distinct contrast to the study performed by Thomas et al. [18], the results of the phase 2 study performed by Philip et al. did not support the activity in using B + E combination in treating advanced HCC patients who were naïve to anti-VEGF therapy. In fact, the combination of B + E with different dosing schedules of B has also been tested in various solid malignancies with poor results. Ko et al. had performed the phase II study in testing the use of B + E in the treatment of gemcitabine-refractory advanced pancreatic cancer with disappointing results [23]. Dickler et al. also showed that the combination of B + E had very limited activity in treating unselected patients with metastatic breast cancer [24]. Similarly, the combination has minimal clinical activity in patients with advanced upper gastrointestinal cancers [25] and recurrent advanced squamous cell carcinoma of the head and neck [26]. All the aforementioned studies consistently demonstrated that although B + E was a fairly well tolerated regime, it had minimal clinical activity in various solid malignancies.

The mechanisms of resistance to sorafenib in HCC are still poorly understood. Proposed mechanisms of sorafenib resistance include enhanced alternative pro-angiogenic signaling, which may result either from an upregulation of alternative pathways or from the pre-existence of multiple redundant signals [27]. Importantly, increased capabilities for proliferation and invasion without angiogenesis may also be observed in resistance to anti-VEGF therapies [27]. Thus, continued anti-angiogenic therapy with B, combined with E inhibiting an additional essential tumor growth pathway in HCC, and may potentially circumvent sorafenib resistance. Nonetheless, our results showed that the combination has no activity after sorafenib failure. The negative results may be explained by few reasons. First, tumour dependence on pro-angiogenic factors may be altered after sorafenib treatment [28]. The vascular remodeling due to pericytic over-coverage renders the neovasculature less responsive to VEGF for growth dependence effectively circumventing a blocked signaling pathway with greater dependence on other alternate mechanisms [29]. Second, it has been shown that the inhibition of VEGF receptors may result in an increased propensity for metastatic dissemination as the hypoxic microenvironment associated with the sorafenib use selects for highly aggressive, invasive tumor cells [30]. Therefore, the increase in the biological aggressiveness of the tumor following sorafenib resistance may account for the poor response to subsequent systemic therapy. Third, other factors including fibroblast growth factors (FGFs), insulin like growth factors (IGFs), angiopoietins, and tumor-stromal interaction may contribute to sorafenib resistance [27]. These may enable tumors to evade inhibition by B + E. Last but not least, vascular remodeling may play a significant role in the resistance to B + E. Tumors that grow during sorafenib treatment developed as viable clusters surrounding strikingly remodeled vessels. These vessels display significant increases in active proliferation of vascular mural cells, expression of platelet-derived growth factor-B and ephrinB2. Thus, enhanced vascular stability accounts for the poor response to B + E combination [31].

There are important limitations in the current study needed to be addressed. Firstly, none within our patient cohort had derived major tumor response while on sorafenib, suggesting that these patients might be intrinsically resistant to any anti-angiogenic strategies. Whether B + E is active in patients with initial response but acquired resistance to sorafenib cannot be concluded from the current study. Secondly, the relative small number of patients recruited from a single institute in the present study is noteworthy. Thus, our present findings should be regarded as exploratory rather than confirmatory. Last but not least, the safety profile of the combination reported from this study is based on a very limited number of enrolled patients, thus these data should be regarded as preliminary and interpreted cautiously. There are few ongoing trials both in US and Asia for investigating the potential benefits in employing B + E in treating advanced HCC patients. In particular, a single arm phase II trial is currently conducted in MD Anderson Cancer Centre in evaluating the role of B + E as a second-line therapy in patients who have progressed after first-line sorafenib treatment [32]. The final results of this study will be eagerly awaited to provide more data about the benefits of B + E in advanced HCC patients progressed after sorafenib treatment. More importantly, the development of biomarkers to select patients likely to benefit from targeted therapy cannot be over-emphasized. In a recent study, high VEGF expression was correlated to longer OS and high VEGFR-2 expression to shorter PFS in upper GI cancers treated with B + E, while EGFR expression and KRAS mutation status were not predictive [33]. Prospectively designed clinical trials with rigorous tissue sampling are desperately needed.

In conclusion, the combination of B + E is inactive in an unselected population of sorafenib-refractory advanced HCC patients. Further evaluation of this combination in biomarker-selected patients may be warranted.