NEW YORK — When ductal carcinoma in situ (DCIS) does recur, most — but not all — invasive recurrences are clonally related to those initial lesions, a new study has found.
While DCIS are generally low-grade lesions that do not progress, a small portion of patients do experience recurrence, even after treatment. But as researchers from MD Anderson Cancer Center and elsewhere noted, it has been unclear whether these later recurrences are related to the initial disease or are separate occurrences, though the expectation was that most arise from the original lesion.
To examine relatedness between initial and recurrent lesions, the researchers amassed samples of primary DCIS and matched recurrences for sequencing and other analyses. As they reported in Nature Genetics on Thursday, the researchers found that three-quarters of the invasive recurrences were clonally related to the initial lesion, but that nearly 20 percent were not.
This finding suggests that a shift in thinking about the risk posed by DCIS may be needed. "By analyzing the largest DCIS cohort of its kind in the world, we discovered that a subset of invasive breast cancers following an initial DCIS are not related to the primary DCIS," co-lead author Tapsi Kumar, graduate student at MD Anderson, said in a statement. "This data challenges our prior understanding of DCIS progression and gives us a starting point to identify better predictive biomarkers to determine which DCIS lesions are most likely to progress to invasive cancer."
The researchers sequenced the whole exomes of two dozen tumor pairs in which the initial lesion was DCIS and the recurrent one was invasive breast cancer. The number of shared mutations between the pairs varied, ranging from none to 112. Overall, 83 percent of the tumor pairs exhibited evidence of clonal relatedness, but 17 percent did not. Those four pairs had no shared mutations that the researchers could detect.
Similarly, a copy-number analysis of 62 tumor pairs found 71 percent to be clonally related, 27 percent to be unrelated, and 2 percent ambiguous. Targeted sequencing of a subset of those pairs then indicated 51 percent were clonally related — including four that were considered unrelated by copy-number analysis — and 15 percent unrelated.
A combined analysis based on exome sequencing, panel-based sequencing, and copy-number data indicated that 75 percent of the DCIS-invasive breast cancer pairs were clonally related, 18 percent were unrelated, and 7 percent were ambiguous.
Analyses of tumor pairs in which the initial lesion was DCIS and the recurrent one was also DCIS found that 85 percent of those tumor pairs were related, 9 percent were unrelated, and 6 percent were ambiguous, suggesting that recurrent DCIS is likely to be due to residual disease that went undetected.
Using single-cell sequencing of tissues from four DCIS patients, the researchers further examined the clonal lineages and uncovered subclones and noted overlap in the driver mutations and chromosomal aberrations between DCIS and invasive samples.
Further, the most commonly mutated genes included PIK3CA and TP53, while the most common amplicons affected 17q12, 17q21.1, and 11q13, and their analysis indicates that these events occurred at the early stages of DCIS — years before the invasive recurrence.
"Our study indicates we can no longer consider DCIS solely as a precursor but rather also a risk factor for the development of invasive breast cancer later on in life," joint senior author Elinor Sawyer from King's College London said in a statement. "This important new information about DCIS biology and behavior, together with other findings, could change the way we manage and treat the condition in clinics in the future."