Healthy tissue near the sites of lung tumors may contain genetic information that could serve as a predictor of possible cancer recurrence, according to recent research.
A study led by a team from NYU Langone Health’s Perlmutter Cancer Center, findings of which were published in the journal Nature Communications, determined that analyzing RNA collected from seemingly healthy issue that was adjacent to tumor cells accurately predicted whether or not the cancer would recur, or return, with 83% accuracy, compared to 62% accuracy when analyzing the RNA from the tumor tissue itself, according to the study.
While lung adenocarcinoma is the most prevalent non-small cell cancer and the deadliest cancer in the United States, the majority of patients who receive early-stage intervention via surgical resection, or surgical removal of their tumor, experience a cure. However, approximately 30% of patients treated with surgical resection experience disease progression, with the majority of those patients eventually dying of metastatic disease, or cancer that has spread from its original site, researchers noted.
However, the recent study data, researchers wrote, “suggest that molecular profiling of tumor-adjacent tissue can identify patients that are at high risk for progression and may help indicate appropriate neoadjuvant therapies (administered before the main treatment) for patients at risk.”
Researchers identified an inflammatory gene signature in tumor-adjacent tissue “as the strongest clinical predictor of disease progression,” according to the study, which may point the way to future treatment options via immunotherapy, which utilizes the body’s immune system to fight cancer.
“In line with our results, there is an increased appreciation of the immune microenvironment in the treatment of resectable (removable by surgery) non-small cell lung cancers, driven by the progression-free survival benefit of neoadjuvant chemo-immunotherapy,” researchers wrote. “Taken together, our findings suggest that the proposed inflammatory signature may be used as a potential indicator of future recurrence events.”
Dr. Aristotelis Tsirigos, a cancer biologist and the study’s co-senior author as well as a professor in the department of pathology at NYU Langone Health and a member of the Perlmutter Cancer Center, spoke with CURE® about the study, its findings and their implications for other cancer types.
Q: For patients and survivors, what do you think the big takeaway from this research should be?
A: What I would say is the big takeaway is that we’ve been looking at the tumor, it makes sense, for many cancer types. We’ve been looking at the tumor, the mutations, the gene expression or different parameters of the tumor for the longest time, (but) perhaps we’ve been ignoring what happens around the tumor, maybe far from the tumor in terms of predicting whether a patient will recur, in terms of cancer.
So, what the study shows is that there is a lot of information, perhaps sometimes even more information compared to the tumor, (that can be found by) looking really (at) what’s around it. And, more specifically, we show that there are different inflammatory signatures that were detected around the tumor, maybe even a few centimeters away, that hold important clues as to whether a patient’s disease is going to come back years after surgery.
And that has, I will say, two different implications. One is it’s prognostic, let’s make sure we collect this adjacent normal to the tumor, and analyze it using our approach to predict whether a patient will recur or not. If the prediction is that the patient is at high risk for recurrence, then we can treat them more aggressively, or we can monitor them more closely. So, there’s that part.
But I would say there’s also the part where you want to think ahead about discovering new therapies perhaps, or maybe another approach is to take out more tumor — maybe not just the tumor, but maybe a bigger section of the lung of these patients. … I will say the take home for now could be yes, let’s start looking at the adjacent tumor. And let’s better stratify patients by risk, identify those that are at high risk and treat them more aggressively, or monitor them more aggressively.
Q: One factor I found interesting that the study makes some mention of is, given the inflammatory response factor that was detected here, the potential for immunotherapies as a treatment avenue going forward. Can you tell me a bit about that and what direction these findings point in regarding that possibility?
A: Yes. So, it could be complicated, so let’s try to think about it together. So the fact that there is inflammation far from the tumor, so to speak, and specifically the presence of immune cell types, for example, monocytes may indicate that a few cancer cells have escaped the bulk of the tumor. So that’s one hypothesis. There are others, but that’s probably the simplest one.
So, that means the cancer cells are out there, you take out the bulk of the tumor, you think you’re cured, but there’s this presence of tumor cells that the immune system perhaps recognizes.
So, perhaps immunotherapy, in that sense, would be a good therapy right after surgery. Why? Because the tumor cells are out there, you power up your immune system to really eliminate every single tumor cell that has escaped. And it’s interesting because you might say, ‘Let’s do an MRI or CT scan to identify those cells,” (but) they’re so few perhaps that we really cannot detect them. And it’s kind of fascinating that just looking at indirect signals of inflammation would give you evidence that these cells actually have escaped and circulated in the lungs of these patients.
Q: Given these findings, and the limitations that are sort of present in every study, what do you hope in terms of future research and exploring this further?
A: First of all, this has to be replicated in bigger studies. That’s what we always say. And it’s true, and there are two reasons. First of all, you have to make sure the results are robust in bigger studies. Second, it gives you an opportunity to improve the models, because this was a retrospective study with many caveats, a relatively limited number of samples, if you think about it — we had more or less under 50 patients but still, it would be nice if we had 1,000 patients, so the more the better. So one goal would be to validate the findings.
The second goal would be to design the study with fewer caveats and make it prospective, meaning you collect samples as you go, but at the same time, use this bigger study to perhaps create a more accurate model of high-risk patients. So I should definitely emphasize that although our model is better than everything else that we’ve seen out there by looking outside the tumor, it’s still not perfect. So it’s correct, let’s say, 83% of the time. But to make this clinically accurate, actionable or (to) use it in practice, I would say it has to go above 90%, or maybe above even 95%, accurate. Otherwise, you would be misclassifying high-risk patients as low-risk and vice-versa. You want to make as few mistakes as possible, and the only way to do that would be to collect more data in a more controlled way. Now that we have a better understanding of what’s going on in the adjacent normal of these tumors, we can design much better studies to further improve our models.
Q: What sorts of implications for other cancer types do these findings have? Because this research was specifically in the realm of lung cancer, I’d imagine that some of this information could potentially eventually be useful in detecting possible recurrence for other cancer types as well.
A: You’re absolutely right, your intuition is correct. And that’s the beauty of this study. By taking the focus out of the tumor — It’s very complex tissue with multiple mutations, perhaps multiple clones, different tumors within a bigger tumor — you’re looking at signals that can be generalizable to different tissues, exactly as you said, going from lung to breast or going from lung to liver.
In fact, there’s a small part of this study that asked precisely this question: what if we look using this inflammatory signature that we discovered? What if we look at the adjacent normal of breast tumors? Would we again be able to, at least to some extent, predict whether a patient will recur or not? And the answer is yes. So there’s definitely an association we find of this inflammatory signature with worse outcomes in other cancers, not just lung.
Again, (there are) lots of caveats, because there are other studies; we didn’t make them, we use public data that is very limited when it comes to adjacent normal. We have lots of data on the tumors, but very limited data sets on the adjacent normal. But with that limited data, we were able to collect from public sources, we’re able to show that indeed in breast cancer and a few other cancers, there is this inflammatory signature that predicts bad outcomes. Now, another contribution of this study in terms of research is to really emphasize the importance of adjacent normal. So, we’ve been collecting tumors, and they’ve been very useful in understanding the landscape of oncology, but perhaps we should start a new effort collecting adjacent normals at a larger scale to really answer the question.
This transcription has been edited for clarity and conciseness.
For more news on cancer updates, research and education, don’t forget to subscribe to CURE®’s newsletters here.