Select Excerpts from the Interview

Track 3

DR LOVE: What’s your take on the issue of cisplatin versus carboplatin for metastatic disease?

DR GRALLA: At the ASCO 2005 meeting, a meta-analysis that demonstrated two-drug combinations containing a platinum agent are better than nonplatinum-containing regimens also showed the cisplatin-based combinations to be more effective (Barlesi 2005). This analysis showed approximately a 12 percent survival advantage among patients treated with a cisplatin-based regimen. A bit more toxicity occurred with the cisplatin-containing combinations, but the toxicity was classified as acceptable.

In the past year, two other meta-analyses have examined cisplatin versus carboplatin with newer agents, including a Japanese meta-analysis published by Hotta (Hotta 2005; [3.1]).

They both showed that adding the same newer agent to both platinums resulted in a 12 to 15 percent survival advantage for cisplatin. So survival appears to be a bit better with the platinum combinations versus nonplatinum, and while nobody particularly likes cisplatin, it appears to be superior to carboplatin.

Track 5

DR LOVE: What were your thoughts on the ECOG-E4599 study with bevacizumab presented by Alan Sandler (Sandler 2005)?

DR GRALLA: The big winner at ASCO 2005 was bevacizumab. The ECOG-E4599 trial was certainly an interesting study. It consisted of nearly 900 patients with nonsquamous cell lung cancers, and all the patients received paclitaxel/carboplatin. They were then randomly assigned to receive bevacizumab at 15 mg/kg every three weeks or not.

The data demonstrated approximately a 10.2-month survival with paclitaxel/ carboplatin, which is probably the best rate we have seen reported for that combination, and a 12.5-month survival with the bevacizumab, which indicates a 15 to 18 percent benefit to adding bevacizumab.

DR LOVE: What are the clinical implications of these data?

DR GRALLA: The difference in efficacy between the two arms was highly significant. Actually, both arms of the study did better than expected, so it’s not as though the bevacizumab arm did well because the chemotherapy-only arm did poorly.

I think this is likely to be a true finding, and I expect a lot of patients will receive bevacizumab as a result.

This is at least the seventh large trial, each with a minimum of 600 patients, that has examined a standard chemotherapy regimen combined with a molecularly targeted agent. All of them have been negative with the exception of the bevacizumab trial, so I think the chances of this being correct are relatively high.

DR LOVE: In the clinical setting, what agent would you combine with bevacizumab?

DR GRALLA: We have seen an advantage reported with bevacizumab and a variety of chemotherapies in several other malignancies. Its efficacy appears to be more tumor related than chemotherapy related; obviously, a second trial using different drugs from those used in E4599 would answer that question, but my prediction is that its effectiveness is not chemotherapy specific.

Track 10

DR LOVE: In your practice, how do you treat nonsmoking female patients with metastatic adenocarcinoma?

DR GRALLA: Any nonsmoking woman with adenocarcinoma has a high likelihood of doing well with a TKI. I have treated many of these patients with a TKI as monotherapy, and within three to four weeks I expect to see symptomatic relief and some hint of a response on a simple imaging study like a chest x-ray.

We don’t know whether it’s better to administer a TKI or chemotherapy first, and we need to conduct research to answer that question. As for using chemotherapy with erlotinib, the results of the two trials evaluating that weren’t so good.

DR LOVE: What new agents or combinations are being evaluated for treatment of recurrent non-small cell lung cancer?

DR GRALLA: Many second-line agents have been studied, but docetaxel remains the agent that is used as the comparator and nothing has beaten it to date, although there are other candidates. It also has shown good evidence in the first line and is one of our many reasonable choices in that setting.

An article in the Journal of Clinical Oncology indicated that the combination of bevacizumab and erlotinib can be used but we need to identify the specific patient population for this combination (Herbst 2005).

I don’t believe we can have great confidence that we have found the target population for bevacizumab, whereas with erlotinib or gefitinib, maybe we have.

Track 11

DR LOVE: Would you discuss the use of adjuvant chemotherapy in the treatment of non-small cell lung cancer?

DR GRALLA: The most recent study was reported at the 2005 ASCO meeting. The ANITA trial was a large, well-executed international trial, with more than 800 patients who were randomly assigned to postoperative vinorelbine and cisplatin or no chemotherapy (Douillard 2005).

The distribution of Stage I, II and III disease was almost equal, with slightly fewer patients having Stage II disease. This study demonstrated, as have others, a significant advantage to receiving adjuvant chemotherapy.

At the ASCO meeting in 2005, Emilio Bria presented a pooled analysis of 11 randomized trials and one meta-analysis with 6,494 patients (Bria 2005; [3.2]). No matter how the data were segregated — all studies, only those studies published in peer-review journals, earlier versus later stage — the data showed a significant advantage with adjuvant chemotherapy.

So I don’t think the question is whether adjuvant chemotherapy improves survival but rather, what is the magnitude of benefit? Dr Bria’s analysis provided evidence of an absolute benefit in the three to four percent range, which is not very much. In breast cancer, the meta-analysis of adjuvant chemotherapy showed a six percent absolute benefit.

DR LOVE: What’s the relative risk reduction in lung cancer?

DR GRALLA: Interestingly, the smaller the study, the larger the benefit and vice versa. In individual trials, relative risk benefits or hazard ratios reduced to 0.69 or 0.74. In the ANITA study, the number was 0.79 (Douillard 2005). In the largest study of all, the IALT study, it was 0.86 (Arriagada 2004). When you look at them all together, the hazard ratio is about 0.9.

Dr Bria has also observed how many patients must be treated for one patient to benefit, which is probably a good way to look at the issue. In his studies, depending on which stage and which group of patients you look at, that number is somewhere between 20 and 30 patients.

Select publications