Welcome back. This is the second part of a discussion of the potential value of circulating tumor cells (CTCs) in cancer management. Please see the prior post for a general introduction to the concept of CTCs and how they have been studied in various cancers. We'll now turn to research on CTCs in the setting of lung cancer.

CTCs in SCLC
Like breast cancer cells (described in the first part of this discussion), SCLC cells are DAPI and cytokeratin positive and CD45 negative (the pattern of particular proteins on the cells), making the cellsearch technology in its current form potentially applicable to them. A British study sought to evaluate potential biomarkers in SCLC, including CTCs. 50 patients were evaluated and CTCs could be detected in all but 7—each of these 7 patients had limited stage disease. Stated another way, 86% of patients overall, and 100% of patients with extensive stage disease, had detectable CTCs. Median CTC number was much higher for patients with extensive-stage disease at 237 (range 1-44,896) compared to limited-stage disease at 2 (range 0-91). CTCs were also assayed at 22 days after therapy; they fell to a median of 1 (range 0-2960). In fact, at day 22, CTCs were detectable in only 60% of patients, compared to 86% at baseline. The median survival for patients with >300 CTCs (highest quartile) was 134 days, compared to 443 days for patients with <2 CTCs (lowest quartile). The number of CTCs at day 22 was also prognostic, with a hazard ratio of 1.43.

There is also an interesting case report about the use of CTCs for diagnosis of SCLC. A 71 year old man presented with a lung nodule, with fine-needle aspiration (FNA) biopsy revealing a low-grade neuroendocrine tumor. Staging was performed which demonstrated widespread metastatic disease, including liver and bone. Noting that this is unusual for low-grade neuroendocrine tumors, his physicians dug deeper. They ran his blood for CTCs with cellsearch and 27 CTCs were detected. Since neuroendocrine tumors (other than a rare type called Merkel carcinoma) don’t express EpCAM, they shouldn’t have CTCs detectable by cellsearch. Liver biopsy was performed demonstrating that his cancer was really SCLC; he subsequently had a partial response to carbo/etoposide.

I am currently working on a protocol to clinically test an idea for better SCLC therapy generated in the laboratory of my senior colleague in pulmonology, Dr. Steven Albelda. Prior studies of similar agents have been limited by the lack of cancer tissue for biomarker analysis. We hope that collaboration with Veridex will enable us to better understand the drug, its mechanism of action, and changes that occur during therapy of SCLC (with the ultimate aim of such understanding leading to better SCLC therapy). We also hope to further evaluate the hypothesis that enumeration of CTCs can predict progression earlier than imaging.

CTCs enumeration in NSCLC

CTC technology has also been applied to NSCLC. A NEJM paper reported on the use of a similar technology (no longer under commercial application, as best I know) to detect the T790M mutation in patients with EGFR mutation. They showed better progression-free survival in those without the mutation. Reduction in number of CTCs was associated with radiographic tumor response.

Unfortunately, many NSCLC cells do not bear the specific cytokeratin markers that the test looks for. One study was able to detect CTCs in the blood of 29.4% of resectable lung cancer patients. Similar results were found in another study that evaluated blood from both the pulmonary vein and a peripheral vein of patient undergoing lung cancer surgery. CTCs were detected in the pulmonary vein of 96.7% of patients, but only 16.7% in the peripheral vein. Unpublished results suggest that roughly 1/3 of patients with metastatic disease have detectable CTCs by cellsearch. While it is possible that there simply aren’t adequate numbers of CTCs in NSCLC, it may also be that the particular cytokeratin markers used are less sensitive for NSCLC than for SCLC and other cancers. Veridex continues to work on adjusting their method to the specific cell surface markers of NSCLC and we hope that our collaboration will help this effort to be realized. One day, we hope to use the technology for:

• An earlier test for progression: This would allow patients to avoid toxicity of ineffective chemotherapy and switch earlier to more effective therapy.
• Better evaluation for progression: Different readers can disagree on whether progression is present on imaging evaluations. CTC evaluations are thus far more reliable between observers.
• Avoid the danger of biopsy: Biopsy causes some patient discomfort and anxiety and bears risks such as pneumothorax.
• Obtain full characterization of every tumor: At the current time, the need for biopsy delays important information for some patients. Some biopsies do not obtain large amounts of material, leading the oncologist to choose which tests are most important; this will become even more relevant as we validate more markers.
• Allow researchers access to tumor cells at more time points: Evaluation of the tumor at additional points in time may allow us to better understand the changes that cells undergo as they are treated and become resistant, hopefully speeding research.

Important caveat:
I hope that you have found useful and interesting this perspective on a new technology with the potential to improve care and research. I have written before on the need to carefully vet new ideas and maintain skepticism about new ideas until they are proven (in particular see here and here). This idea applies every bit as much to the ideas that I’m excited about to those that that we routinely comment upon on the message boards as being “alternative,” “experimental,” or “unproven.” CTC technology is not yet ready for prime time in either SCLC or NSCLC; there is much more work to be done to develop the technology, evaluate the operating characteristics of CTC enumeration in SCLC and NSCLC, and validate the use of biomarkers on CTCs.