COVID's Shadow on Lung Cancer? Study Sees Later Diagnoses


Jessica Paulus, ScD, discussed a study investigating real-world trends in lung cancer staging among community oncology practices.

Jessica Paulus, ScD

Jessica Paulus, ScD

Detecting lung cancer at an early stage results in better prognoses for patients; however, it remains unclear how health care utilization has been affected by the COVID-19 pandemic and evolving screening recommendation guidelines from the National Comprehensive Cancer Network and the US Preventive Services Task Force.

A study presented at the 2024 American Society of Clinical Oncology (ASCO) Annual Meeting investigated lung cancer stage distribution in a large network of community oncology practices in the US between 2013 and 2023. The study found an increase in the overall number of diagnosed cases and a concerning trend: a rise in the proportion of patients diagnosed with advanced stage lung cancer, both non–small cell and small cell. This could be due to changes in healthcare utilization patterns, potentially linked to the COVID-19 pandemic.

In an interview with Targeted OncologyTM, Jessica Paulus, ScD, senior director of real-world research, Ontada, a business of McKesson, discussed this study, its findings, and next steps in research.

Targeted Oncology: What are the unmet needs in this area?

Paulus: There has been lot going on in the last decade with respect to lung cancer, looking at secular or time trends in the burden of lung cancer. In the United States, the 2 events that come to mind are first, the change in screening guidelines for lung cancer set forth by the US Preventive Services Task Force that recommends regular screening for patients that are at high risk of lung cancer [getting screened] with low dose CT scans. That screening recommendation was a long time coming in terms of several clinical trials that had been conducted that indicated that screening with CT scans will reduce mortality from lung cancer. That change to screening guidelines came to the fore in 2013. Those guidelines were updated in 2021. We would hope or expect that guidelines like that would lead to a stage shift. By that I mean a reduction in the shift of a reduction in the stage of disease at presentation for patients presenting lung cancer.

The other thing is the COVID-19 pandemic, which introduced barriers to cancer screening services, both because of changes in hospital resourcing dedicated towards management of the pandemic and also because of patient behavior. Even where screening services were still online, patients may have deferred screening, as they were trying to do the cost-benefit analysis of avoiding contact with infected patients with COVID-19. Those are some of the kinds of seminal events over the last decade plus that led folks to be interested in understanding how the burden of lung cancer is changing over time.

We had another kind of objective here in that Ontada has access to all of the electronic medical records from the US Oncology Network, which is a network of community oncology clinicians in the United States that covers a huge proportion of patients seeking care in the community oncology setting. This is a huge network of health systems. There is also a health services view of this. We really need to understand, over time, how patients are presenting to this network to make sure that the right services are in place to meet their needs. If there is a shift towards more advanced age or earlier stage, we need to make sure we have the right teams and ancillary care services to meet their needs.

What is the background and methodology of this analysis presented at ASCO?

This was a large observational study of data that is emanating from that US Oncology Network. This is all real-world data that we are using for secondary purposes, for research purposes. One of the real assets to this kind of analysis is that all the data is coming from a common electronic medical record system called AI Nomad, which is designed specifically for cancer care. That has been designed to specifically elicit the information that is most important for managing the experience of patients with cancer. As research staff, we get to leverage this amazing data that is closely fit for purpose, and we can even go in and look at the [electronic medical record (EMR)] platform to see exactly what questions are being posed to clinicians to understand the source of the information that we're working with for research purposes, and that is really unique.

By leveraging kind of the AI Nomad technology, we were able to look at about 100,000 patients with lung cancer over the last decade or so. About 85% of those had non–small cell lung cancer and the other 15%, or about 15,000 patients, had small cell lung cancer. Essentially what we did is we evaluated what stage of lung cancer they had upon presentation to the US Oncology Network. I am being a little careful about not saying stage at diagnosis, because there could be some patients for whom their diagnosis was assessed outside of the US Oncology Network, and they may have come to us for a second opinion or something like that. Although we think the impacts of that is relatively nominal.

Essentially, this was a simple analysis, in that we characterize the stage at presentation to the network from 2013 until 2023. There were 100,000 patients that broke down according to the demographics that we would expect of lung cancer patients in the US Oncology Network. They were in their late 60s at diagnosis; we know this is a disease of aging. In addition, they were about 50/50 male and female.

We also noticed the stages breakdown in the way that we would expect. Unfortunately, at this point, we are still seeing a disproportionate amount of this disease diagnosed in stage III or stage IV. The reason I say unfortunate is because the treatment approaches at that point are no longer surgical. We can get curative resection when there is early-stage disease, but not when there is metastatic or distant spread. We saw that approximately, 70% non–small cell cases and more like 85% of the small cell cases were either stage III or stage IV at presentation. That is what we would have expected, but it is just not where we want to get with this disease.

In terms of the time trends that we saw, we sort of noticed 2 things. Over time, there was a greater proportion of advanced disease at presentation to the Network that was noted over the last decade. The second thing is that we noticed a more significant boost or increase in the year or years after the COVID-19 pandemic started. You can imagine a shift, a gradual albeit noticeable shift towards more advanced age, and then a more noticeable uptick around the years 2020, 2021. We also have not yet noticed in our data that uptick has returned to kind of prepandemic levels. Although our data is still going in 2024, so there is more follow up to be done.

I think this is not the direction we would want to see for this disease. I also have to say that there are a lot of caveats here and that this was not designed. The study was not designed to measure stage at true diagnosis, nor was it designed to measure incidence of lung cancer. We need to look to some of our population registries like the SEER database to really get more reliable data there because in our network, we are already seeing all patients with cancer. We are not seeing any patients who do not have cancer or who are being screened for cancer, so we just do not have the populations to be able to comment on screening, efficacy, or things like that.

To be clear, screening was not assessed in the study; it was not the objective of the study. There can be database or other types of population referral factors that are driving the trends that we are seeing here. By that, I mean, we know the US Oncology Network was growing dramatically over this time and was increasing in size, and it seemed that the US Oncology Network, the patients that were being added to the network, were more disproportionately later-stage. It is not clear why that is, whether that is a true kind of population issue or whether it is something about the types of health systems or networks that we were onboarding to our network at that time.

What do you consider to be the next steps from this research?

I think there are a few different directions, some of which are relevant for the US Oncology Network, some of which are relevant more broadly for the US population, and especially those interested in cancer prevention. Within the US Oncology Network, I think a key next step or implementation step is making sure the Network is aware of this trend. Because, as I mentioned, we need to make sure that the right supports are there for patients with stage III, stage IV disease. That is a different set of supports for patients that are early stage. It is important to be able to marshal those resources for those patients.

I think the second piece within the Network is a question that people are asking everywhere, not just for lung cancer, but are some of the stage shifts that we are seeing, perhaps secondary to the pandemic, going to come back down? Or when are they going to come back down? As I mentioned, the data that I have access to cannot directly interrogate that question; we can only see downstream effects about who is presenting to the Network. But that is a question of intense interest in the cancer epidemiology community and the cancer prevention research community, and it varies for different diseases, different types of cancer.

Every human who has breast is recommended to get breast cancer screening, and everyone with a colon is recommended to get colorectal cancer screening. That is different than lung cancer, where there is, at this point, only risk-based screening guidelines for people that are current or former smokers. That is all to say we expect different impacts of a hiatus from screening during the pandemic across different diseases. But that is an area of peak interest. What is the degree of the stage shift to the pandemic? And when is it going to return to what it was prepandemic? So those are 2 big follow-ups.

Recent Videos
Related Content