Challenges Continue With Lung Cancer Screening

M. Patricia Rivera, MD

M. Patricia Rivera, MD

Screening with low-dose CT has been shown to reduce mortality from lung cancer, but increased awareness of the benefits of screening and more clarity as to which individuals should be considered for it will prompt more widespread use in clinical practice, says M. Patricia Rivera, MD.

“If lung cancer screening is implemented correctly, it is my opinion that screening is the one intervention that is likely to have the most significant impact in lung cancer mortality, aside from surgery, chemotherapy, and targeted therapy,” said Rivera.

The National Lung Screening Trial (NLST), she explains, accrued 53,453 individuals at high risk for lung cancer. Individuals were randomized to 3 annual screenings with either low-dose CT or single-view posteroanterior chest radiography. The use of low-dose CT reflected a 24.2% positive screening test versus 6.9% with radiography. There were 247 deaths per 100,000 person-years in the CT arm versus 309 deaths per 100,000 person-years in the x-ray arm. This reflected a 20% reduction in lung cancer mortality (95% CI, 6.8-26.7; P = .004).

OncLive: Please provide an overview of lung cancer screening?

In an interview during the 2018 OncLive^® State of the Science Summit^TM on Advanced Non—Small Cell Lung Cancer, Rivera, professor of medicine, University of North Carolina (UNC) Chapel Hill, UNC Lineberger Comprehensive Cancer Center, discussed the potential of lung cancer screening, challenges in its adoption, and methods to enhance its efficacy.Rivera: Lung cancer screening is fairly new. It came about after the publication of the large, randomized NLST study. The trial compared CT scan with chest x-ray in smokers over the age of 55 with a 30-pack year smoking history. It resulted in a positive study that demonstrated, for the first time, a reduction in mortality for patients who had been screened with CT scan. In 2014, the United States Preventative Task Force gave it a B recommendation; Medicare approved screening the following year. Since 2015, it’s become more available.

However, it's still pretty underutilized. As of 2015, only 3.9% of eligible patients were screened in the United States. It has to do with a lot of different barriers. One of the most important barriers is the difficulty in ascertaining an individual's smoking history, especially since screening criteria are not restricted to age. Individuals must have smoked 30-pack years; that means a pack a day for 30 years, or 2 packs for 15 years. Sometimes it's very difficult to get that information, and physicians are not very aware of the guidelines. There are a lot of guidelines for screening that have different upper age ranges.

What can initiate the widespread adoption of screening?

More importantly, the Medicare approval was contingent on the fact that patients had to have shared decision making. Physicians had to document that they reviewed the benefits and the risks with the patient. Unlike breast cancer screening, where a 40-year-old woman is told to begin annual mammograms, patients have to weigh in on their decision to be screened [for lung cancer]. Physicians have to have thorough discussions so that patients’ wishes and inputs are taken into consideration. Moreover, patients are not sure whether their insurance will pay for it; Medicaid, for example, does not pay for it. It's very challenging. There needs to be better support, specifically from institutions, communities, and the government. There is a lot that needs to be done to implement lung cancer screening in a way that is safe and respectful to the patient and cognizant of the nuances so that it's done well. One of the major required components of screening is that patients who are currently smoking be enrolled in a tobacco treatment program. However, a lot of institutions don't offer one, and a lot of insurance companies won't pay for tobacco treatment, even though it’s been shown to be incredibly important in terms of improved outcomes.

There are many patient barriers, such as fear, not wanting to know, not having access to healthcare, and not wanting to quit cigarette smoking, among others. There are also physician barriers. Primary care physicians have not bought into screening; they believe the risks outweigh the benefits. People don't have time to spend 20 minutes doing shared decision making. It's very complex to talk about lung cancer screening.

Can you recommend any screening programs that are currently available?

What are some other novel modalities used for early detection of high-risk lung cancer?

System barriers also prevent access to electronic medical records that reflect accurate smoking histories. Physicians are not easily able to identify patients who meet criteria and would benefit from screening. Oftentimes, medical records are so inaccurate in terms of smoking history. Institutions don’t provide the support that's needed so that the personnel and nurse coordinators have a way to track nodules and findings. It is a lot more complex than any other cancer screening test that we know to date. What are some other concerns in its adoption? I chaired an important workshop last year at the 2017 American Thoracic Society Annual Meeting. We identified gaps in the knowledge of how comorbidities impact screening outcomes. As with all clinical trials, patients who were enrolled in NLST were very healthy individuals. There is this healthy volunteer bias, and we're unsure if the results will translate to the general population. There are also some concerns about the potential harms of screening. This workshop was designed to think about important research questions that might lead to better research studies that identify how to incorporate chronic comorbid conditions into the decision making of lung cancer screening. Specifically, we discussed how to help physicians relay information to patients and how to increase patient understanding. It's complex, but it’s an area of huge void. For most people who are at risk for lung cancer, the high-risk individuals and heavy smokers often have COPD, heart disease, hypertension, diabetes, and all of the disease processes that are linked with tobacco and then some. It’s challenging, and we don't know what those outcomes will be. I've developed a multidisciplinary screening program at UNC Lineberger Comprehensive Cancer Center with input from thoracic radiologists and thoracic surgeons. We have a wonderful nurse coordinator, and we also have a strong research arm. We developed a research registry at our institution several years ago. My colleague, Dr Louise M. Henderson is an epidemiologist who is doing some great work. Our goal is to look at the proper implementation of lung cancer screening across UNC Health Care and North Carolina in general. We are very proud of our screening program. An area of current interest is nasal scrapings. We're going to network with a company to look at genomic profiling to see if we can identify high-risk individuals.There are investigational studies using molecular biomarkers for individuals who have nodules. If a CT scan locates a nodule, there are emerging molecular biomarkers that can help identify high-risk individuals. In NLST, 25% of individuals who were screened with a CT scan were found to have a nodule, but 96% percent of those nodules were benign. The overwhelming majority of nodules detected on screening CTs are not cancer, so a lot of these patients may undergo unnecessary procedures. Biomarkers may help.

There have been studies that have looked at prediction models. NLST enrolled smokers over the age of 55. The only 2 variables that were considered at risk were age and smoking history, but lung cancer risk is multifactorial. Prediction models that incorporate other variables, such as COPD, family and personal history of cancer, race, and gender—because women are slightly more at risk than men—have been shown to be better at identifying higher-risk individuals. It has been suggested that physicians should use prediction models to identify those at high risk, but there are problems with prediction models. They reflect a population assessment, not really an individual risk. Applying risk prediction models or any prediction model to someone can be tricky, though they have been well validated.

A scenario in which a prediction model may be useful is in a patient who has been referred for screening and meets criteria. The prediction model is performed to see what the patient’s risk of developing cancer is in the next 6 years. The test comes back with a 6% likelihood of developing cancer. That is someone who will definitely benefit from screening.

In individuals who are referred for screening and have bad heart disease or kidney disease, screening may cause more harm than good. If they come back with a 0.5% risk of developing lung cancer, it’s clearly very low. However, they have other comorbidities that put them at risk for competing causes of death. In that case, we would have to discuss the risks and benefits of screening. We're not there yet.

What is the standard treatment approach in stage I and II patients?

Does a course of treatment differ in an immunocompromised patient?

Are there any ways to predict radiation-induced pneumonitis in patients?

These are models that were reported after NLST. However, the hope is to be able to identify high-risk individuals who are healthy enough to undergo appropriate treatment if a cancer is found, and avoid screening in very low-risk patients or high-risk patients who are too sick to undergo appropriate treatment.For stage I, it is surgery. For stage II, patients with metastatic disease to hilar lymph nodes are treated with adjuvant chemotherapy after surgical resection. If a patient has no lymph node involvement, but has a tumor that’s greater than 4 cm, they should talk to a medical oncologist about adjuvant chemotherapy. The data in that subset of patients are not as strong in terms of a benefit as it is in patients with metastases to the hilar lymph nodes.It doesn't, but patients with HIV have a higher risk of lung cancer. We usually don't see immunocompromised patients with leukemia, lymphoma, or bone marrow transplants. However, being on chronic immunosuppression for transplant and undergoing solid organ transplant carries a risk of malignancy, and not just lung cancer. We have seen a few lung cancers, but that is not a significant patient population for us. Older age, underlying renal disease, interstitial lung disease, and lung fibrosis is associated with an increased risk of not only radiation induced toxicity but also of certain chemotherapy induced pneumoninities. Underlying structurally impaired lungs and primarily fibrosis is a risk factor.

Reduced lung-cancer mortality with low-dose computed tomographic screening. The National Lung Cancer Screening Trial Research Team. N Engl J Med. 2011;365:395-409. doi: 10.1056/NEJMoa1102873.