For 30 years the standard of care for early stage non-small cell lung cancer has been to remove the tumor with a lobectomy—that is, taking out the entire lung lobe. This was based on a landmark trial that was published in 1995 by Robert J. Ginsberg, MD, and the lung cancer study group that showed that for lesions that are under 3 centimeters, lobectomy provides lower recurrence and superior survival when compared with sublobar resections, which include wedge resection where a pie-shaped wedge is removed from the lung.
But more recently, studies have pointed to sublobar resection as being just as good as—and maybe even better—than a lobectomy for long-term outcomes, according to Christopher Seder, MD. He is a thoracic surgeon and chief of the division of thoracic surgery at Rush University System for Health, a member of the AMA Health System Program, which provides enterprise solutions to equip leadership, physicians and care teams with resources to help drive the future of medicine.
In a recent study, Dr. Seder and the Society of Thoracic Surgeons were interested in learning whether recent trial data from the U.S. and Japan were generalizable to everyday decisions on how to best manage nonsmall cell lung cancers.
“The reason we wondered that was because in the Japanese trial, they have a population that was totally different than the American population—44% of these lung cancer patients were never smokers. They had really good lung function, and it was almost all adenocarcinomas,” he explained. “It is kind of hard to know if we can apply that because it seems like a different patient population than the U.S. routine clinical practice.”
“Then when you look at the CALGB [Cancer and Leukemia Group B] trial that was performed on what was primarily a U.S. population and fairly representative of the population that we see,” Dr. Seder said, “they had very strict inclusion criteria and intraoperative protocols that required the surgeons to dissect out all the lymph nodes, send them off for analysis and only if those lymph nodes were normal could the patients then be randomized to sublobar resection or lobectomy.”
Additionally, “patients had to be physically fit to undergo a sublobar resection or a lobectomy since they were going to be randomized,” he said. “So, you’re taking a really healthy group of patients, putting them through this strict intraoperative protocol and we know from daily practice that that is not really the way that people practice.”
“We wanted to see what happens with this group of patients—these sub-2-centimeter nonsmall cell lung cancer patients—in routine clinical practice,” Dr. Seder noted, adding that “we hypothesized that lobectomy is still going to be superior even after risk adjustment and that was going to be driven by the inferior outcomes of wedge resections more so than segmentectomies,” which is a surgical procedure that involves removing a portion of a lung lobe.
With that, for the first time ever, Dr. Seder and his team linked the Society of Thoracic Surgeons’ general thoracic surgery database to the Centers for Medicare & Medicaid Services database and the National Death Index, which connects public health and medical researchers with U.S. death records. This provided overall survival data and lung cancer-specific survival data. And then a 10-year follow-up was conducted.
“What we found was as we suspected. The overall survival of lobectomies was better than sublobar resections,” he said, noting that “lobectomies and segmentectomies were performing about the same, whereas our wedge resections were performing significantly worse for overall survival.”
Unadjusted 10-year overall survival for lobectomy reached 44.8%, while for a segmentectomy it was 44.2%. Meanwhile, for the wedge resection it was 41.4%.
In an interview with the AMA, Dr. Seder discussed his lung cancer study and what physicians should know.
AMA: Why did you decide to conduct this analysis?
Dr. Seder: We feel that this is a kind of cautionary tale—that there is a difference between what you see on trial and the true effectiveness in routine clinical practice. And that's been over and over again, that when you follow specialized protocols of specialized centers with academic surgeons who are interested in trials and then following these trial patients very closely, it oftentimes overestimates the benefits and underestimates the harm of those treatments.
So, we wanted to study the true effectiveness of this—as opposed to the efficacy that we see in trial—and show that you can't necessarily apply trial data to everybody, or that in the real world, we don't see the same outcomes.
I was surprised to see that, still, with all the information we have suggesting that nodal dissection—which is when lymph nodes are dissected out from the pulmonary hilum and mediastinum to accurately stage patients—is so important, 20% of wedge resections had zero lymph nodes sampled.
The standard of care in physician well-being recognition
Institutions in this article have been honored by the Joy in Medicine™ Health System Recognition Program.
AMA: Are there specific factors that might tilt the balance between recommending a lobectomy versus a sublobar resection in everyday clinical settings?
Dr. Seder: There are a lot of factors that go into that—and that's perhaps one of the weaknesses of the study, is that we don't collect all the data necessary that a surgeon might use to decide what operation they're going to provide. First and foremost, starting with just looking at a patient from the door, many surgeons are going to say that patient either will or will not tolerate a lobectomy or bigger operation.
If, from the outset, that patient's sick, in a wheelchair, on oxygen already and they're not going to tolerate a lobectomy, maybe they're just going to get a wedge resection no matter what. And then there's a bunch of other things that go into that decision—how solid is the tumor? How much does it light up on the PET [positron emission tomography] scan? Is it super peripheral or is it more central? Is it right on the edge of a segment or is it straddling two segments and you'd have to take out two segments in order to get the tumor out.
That's just one-tenth of the things that need to be considered with every patient that comes in ... when you're a surgeon who is deciding what operation to provide them.
AMA: Are there patient characteristics that you believe strongly predict better outcomes?
Dr. Seder: Certainly, comorbidities and patient fitness and patient-performance status. If your endpoint is overall survival, that older patient who is in a wheelchair, on oxygen, her overall survival almost certainly is going to be shorter than a 35-year-old who has lung cancer and runs marathons.
We also know that there are a bunch of different pathologic characteristics of the lung cancers that will predispose patients to have better or worse outcomes. There are also genetic drivers of lung cancer that allow patients to be treated with various targeted molecular therapies. So, yes, there are a lot of things that predict worse or better outcomes when it comes to lung cancer.
AMA: When considering a wedge resection, what steps should surgeons take to ensure they replicate the high-standard approach used in trials?
Dr. Seder: This really gets to the very core of what this paper is saying and it's that this paper does not prove that a high-quality wedge resection is worse than a high-quality lobectomy. What this shows is that there's a lot of low-quality wedge resections being performed. Now, what exactly defines a high-quality wedge resection? We don't know for sure. ...
But the things that probably matter are an adequate lymph node dissection and adequate margins. So, if you can take wide margins and take out a bunch of lymph nodes, there's a chance that your high-quality wedge resection may be a completely adequate operation.
AMA: What should physicians take from this study and how can they use this when treating patients?
Dr. Seder: The takeaway messages are the CALGB and JCOG [Japan Clinical Oncology Group] trials do not give free rein to just perform a low-quality wedge resection on anybody with a sub-2-centimeter lesion. Abiding by the standards of adequate lymph node dissections—and likely wide margins are important.
And really, the only thing we can definitively say is unless you follow trial protocol exactly, you can't expect trial-like results. So, in the real world, a wedge resection is not adequate for everybody if it's not performed in a high-quality fashion.
AMA: How might your data influence multidisciplinary tumor boards or overall decision-making processes among surgeons and oncologists?
Dr. Seder: This should be a call to action for multidisciplinary teams to consider enrolling patients in trials looking at different treatment modalities. And remember, we don't know anything until we prove it in a rigorous scientific fashion.
Everybody's got their opinions on the way they think things are, but ... we should try to be proving these things for the sake of our patients. And so, it should be a call to action for these multidisciplinary teams to enroll patients in these trials that traditionally have been getting shut down because not enough patients get enrolled.
AMA: What are your next steps?
Dr. Seder: One of the biggest limitations of this study is that the general thoracic surgery database was designed as a quality initiative focused primarily on 30-day short-term outcomes—the complications and death rates in the first 30 days.
Now that we've incorporated long-term outcome measures such as lung cancer-specific survival and overall survival, we are in the process of revising the data points collected because we know that there are a lot of things that are associated with long-term survival, such as the type of pathology, the imaging characteristics, and the biology and genomic makeup of the tumor.
There are all these things that influence long-term survival that we've never collected because it didn't really matter for short-term outcomes. So, our next steps are really to begin to beef up the data input into the general thoracic surgery database to be able to more accurately create models that perform stronger in predicting long-term outcomes for patients, now that we have those data to look at.
