Simron Singh, MD: Sure. Traditionally, chemotherapy has been reserved for poorly differentiated neuroendocrine carcinomas. These are very aggressive cancers that are much related to small cell lung cancer. Typically, platinum-based treatments have been what we have used. We have used carboplatin/cisplatin with etoposide, but some people have also used other agents. Recently there has been more and more data on temozolomide. There was your retrospective series on temozolomide, and this year we had a really good study, which was our first randomized controlled prospective trial of temozolomide in pancreatic neuroendocrine tumor patients. This study showed that there was a benefit seen with the use of a temozolomide and capecitabine combination over temozolomide alone. But I think the real key to that study was that it really showed, from a prospective fashion, that temozolomide was an effective agent. So temozolomide is being used more frequently.

When we talk about grade 3 neuroendocrine patients—both poorly differentiated and well-differentiated—one of the most important trials we’re going to have is this randomized trial being done to compare cisplatin and etoposide, or platinum-based therapy of any sort, to temozolomide and capecitabine. There have been historical studies with streptozocin and 5-FU [fluorouracil] in pancreatic neuroendocrine tumors. In North America, we very rarely use streptozocin now. As we see the results of the temozolomide study become increasingly disseminated, I anticipate that streptozocin use will decrease more and more.

Jonathan R. Strosberg, MD: Yes. The study that you’re referring to is the ECOG [Eastern Cooperative Oncology Group] 2211 study, which was exclusively done in pancreatic NETs. Patients with progressive pancreatic NETs were randomized to receive temozolomide monotherapy versus capecitabine/temozolomide. This was a randomized phase II study. I think the population was 145 patients. So this was relatively large by NET standards. The primary endpoint was progression-free survival. The trial revealed a surprising improvement in progression-free survival with capecitabine/temozolomide versus temozolomide alone. With temozolomide alone, the median progression-free survival was 14 months. It was almost 23 months with capecitabine/temozolomide. There was also an improvement in overall survival, which was quite encouraging. I think it’s an important study validating the use of capecitabine/temozolomide in pancreatic NETs. That really seems to be the population where it works best. We have a bit more anecdotal information in lung NETs. I personally have not found it to be that effective in midgut NETs, at least as far as tumor shrinkage, although I sometimes use it for very aggressive midgut NETs.

Simron Singh, MD: Outside of pancreatic NETs, and perhaps even in high-grade NETs, I think the jury is still out. We need to get more information. We need more data. Perhaps we need to select the right patients to see whether the expression of MGMT [O6-methylguanine-DNA methyltransferase] methylation may help us understand who responds. I think we’re doing a better job and we need to continue studying this to try to understand which treatments may be right for certain people. There may be patients with either lung NETs or small bowel NETs who actually do quite well on capecitabine/temozolomide. I think we just need to identify them.

Jonathan R. Strosberg, MD: Yes. And when it comes to prescribing chemotherapy, it’s really important to do it right. For temozolomide, you absolutely need to give the Zofran or other 5-HT3 blocker beforehand. Often, that’s not prescribed. The main concern with chemotherapy is probably thrombocytopenia.

Simron Singh, MD: Correct.

Jonathan R. Strosberg, MD: We see significant thrombocytopenia in about 15% of patients, although the rate that they saw in the trial was a bit lower. So it is absolutely important to check the patient’s complete blood counts before each cycle. Thrombocytopenia can be quite delayed. It starts at week 4, but can be maximized week 5 or week 6. So if you see the platelet count dropping at week 4, you really want to monitor the patient closely.

Simron Singh, MD: That brings up an important question. We do know that some of these, for example, can cause thrombocytopenia. That may close the door for subsequent therapy.

Jonathan R. Strosberg, MD: Right.

Simron Singh, MD: So sequencing is always a challenge. Do you have a specific rationale for sequencing? When we are using treatments, we really want to be cautious of not shutting the door to other treatments down the road.

Jonathan R. Strosberg, MD: For patients with slow-growing somatostatin receptor-positive metastatic NETs, first-line systemic therapy is easy. It’s a somatostatin analog. But everything beyond that is complicated, particularly in pancreatic NETs. Right now, we have about 5 different options. We have everolimus and sunitinib. We have capecitabine/temozolomide chemotherapy. We have PRRT [peptide receptor radionuclide therapy]. We have liver-directed therapy. It’s anyone’s guess as to how to sequence them. I think if you go to 5 different institutions you’ll probably get 5 different answers as to what the best sequence is. But you’re right. You have to think about the long-term and potential interactions, for example, between chemotherapy and PRRT, as far as bone marrow toxicity. So it’s challenging. I don’t know if I have a great answer to that. The only thing I would say with a relative level of confidence is that if you’re dealing with relatively aggressive high tumor burden, chemotherapy is probably a good option for pancreatic NETs.

For midgut NETs, I think it’s a bit easier. I’d say PRRT is probably the appropriate second-line systemic therapy for most patients, although for patients with liver-dominant disease I still think liver embolization can be quite effective.

Simron Singh, MD: Yes. I’m not sure we know the answers to any of these questions. An important point to mention when we talk about sequencing is that we direct it on an individual patient level…sequencing may be different for different patients. I don’t think there is ever going to be 1 standard algorithm that we can follow.

Jonathan R. Strosberg, MD: True.

Simron Singh, MD: I think we really are going to individualize our therapy. One sequence may be right, as you mentioned, for a patient who has bulky disease, but it may not be right for the patient who has low-volume disseminated disease.

Transcript edited for clarity.