Conference Coverage

Lung cancer metastatic sites differ by histology, tumor factors


 

FROM ELCC

– A review of data on more than 75,000 patients with lung cancer has revealed distinct patterns of metastasis according to subtype, a finding that could help in surveillance, treatment planning, and prophylaxis, an investigator contends.

Patients with small cell lung cancer (SCLC) had significantly higher rates of liver metastases than patients with non–small cell lung cancer (NSCLC), while patients with NSCLC had significantly higher rates of metastases to bone, reported Mohamed Hendawi, MD, a visiting scholar at the Ohio State University Medical Center in Columbus.

“Predictors for liver metastasis were small cell and adenocarcinoma histology, lower and upper lobe locations, and high grade tumors. Predictors for metastasis to brain were advanced age at diagnosis, adenocarcinoma and small-cell histology, lower lobe [and] main bronchus locations, and high grade tumors,” he wrote in a scientific poster presented at the European Lung Cancer Conference.

Dr. Hendawi drew records on all patients with metastatic lung cancer included in the 2010-2013 Surveillance, Epidemiology, and End Results database. He used univariate and multivariate logistic regression models to evaluate predictors of metastasis.

The data set included a total of 76,254 patients with metastatic lung cancer, of which 17% were SCLC and 83% were NSCLC tumors. In 54% of patients, the primary tumor was in the right lung; in 38%, it was in the left lung; and, in 8% of patients, the primary tumor was bilateral.

The rates of metastases to bone were high in both major lung cancer types but, as noted before, were significantly higher in patients with NSCLC: 37% compared with 34% for patients with SCLC (P less than .001).

In contrast, the incidence of liver metastases in SCLC was more than double that of NSCLC: 46% vs. 20%, respectively (P less than .001). There were slightly, but significantly, fewer cases of brain metastases at the time of diagnosis among patients with SCLC: 25% vs. 26% (P = .003).

Histologic subtypes significantly associated with both brain and liver metastases were, in descending order, adenocarcinomas, small cell, and squamous cell cancers.

Although carcinoid lung cancers accounted for only 2.1% of all tumors, they were associated with a high rate of metastasis to brain at diagnosis (44.8%).

As noted, independent risk factors for liver metastasis were small cell and adenocarcinoma histologies (P less than .001), tumors in the upper lobe (P = .028), and high-grade tumor (P less than .001).

Independent predictors for brain metastases were advanced age at diagnosis (P less than .001), adenocarcinoma and small-cell histologies (P less than .001), lower lobe or main bronchus locations (P = .004), and higher-grade tumors (P less than .001).

In a poster discussion session, Paolo Boffetta, MD, MPH, from the Icahn School of Medicine at Mount Sinai in New York City, the invited discussant, commented that, while he thought that the data were interesting, “the main issue I had with this poster is that it’s limited to patients with metastasis, so we cannot really evaluate the risk of metastasis according to the different histological types and the absolute risk of developing metastases in one or the other organ but only the relative risk of developing metastasis in one organ versus the other having one or the other histology.”

“So, we really don’t know whether the risk is increased in one group or decreased in the other one that generates these differences,” he said.

The study did not receive outside support. Dr. Hendawi declared no conflicts of interest.

Next Article: