Screen colorectal cancer patients for lung metastases

The absence of liver involvement in patients with colorectal malignancies should not preclude a search for lung metastases, according to a review of 754 patients at a single institution.

Although pulmonary metastases from colorectal cancer usually result from hepatic metastases, the findings suggest that the true incidence of isolated lung metastases in colorectal cancer may be as high as 3%-12% in patients with rectal cancer and 1%-6% in patients with colon cancer, said at the annual meeting of the American Society of Colon and Rectal Surgeons.

Thus, staging and surveillance of all patients with colorectal cancer should include an evaluation for possible lung involvement.

In all, 754 patients at Tan Tock Seng Hospital, Singapore who were diagnosed with colorectal cancer between December 2003 and August 2007 were included in the review. The patients, whose mean age was 68 years, had rectal cancer (196 patients) or colon cancer (558, including 369 with left-side lesions and 189 with right-side lesions) said the hospital’s department of surgery.

Isolated lung metastases were determined by confirmed histology or cytology of the lung lesion or radiologic evidence suggestive of lung metastases in the absence of live lesions on CT scan up to 6 months after diagnosis of the primary cancer. Based on these criteria, isolated lung metastases were reported in 23 (12%) of the rectal patients and in 33 (6%) of the colon cancer patients, including 25 in patients with left-side lesions and 8 in patients with right-side lesions.

Of the 23 rectal cancer patients with isolated lung metastases, 19 underwent surgical resection, and of these, 16 were at stage T3 or higher and 15 were at nodal stage N1 or higher.

Isolated lung metastases were synchronous (defined as occurring within 6 months of the primary cancer) in 20 of the 33 colon cancer patients and in 9 of the 23 rectal cancer patients, noting that the lung involvement was diagnosed at presentation in 13 of the colon cancer patients and in 5 of the rectal cancer patients. An analysis of the patients with and without isolated lung metastases showed that “the site of the primary cancer was significant for isolated lung metastases, with the rectal cancer patients being twice as likely as the colon cancer patients to develop lung involvement.”

In the colon cancer patients, none of the following variables was a statistically significant predictor of isolated lung metastases: age, gender, race, location of colonic primary tumor, tumor stage (T1/T2 vs. T3/T4), or nodal status (N0 vs. N1/N2). “There was a trend toward higher incidence of isolated lung metastases in patients with left-sided vs. right-sided lesions, but the analysis did not reach statistical significance.”

Age, gender, race, and tumor stage (T1/T2 vs. T3/T4) were not statistically significant for isolated lung metastases in rectal cancer patients, “but nodal status almost reached statistical significance,” with lung involvement noted in nearly three times as many patients at N1/N2 than at N0.

Limitation of the study included its single-institution retrospective design and the small number of patients with isolated lung metastases.

Nonetheless, “the search for lung metastases is recommended in all patients with rectal cancer, especially those with N1 disease or greater, and in selected patients with colon cancer, especially those with T3 or greater lesion.”

In addition, he stated that even patients with T2 or N0 disease and those with normal postoperative serum carcinoembryonic antigen levels should be carefully monitored for lung involvement during the follow-up period.