By comparing DNA sequences in primary and recurrent breast tumors, a study led by Australian researchers revealed that about 18 percent of recurrent cases of ductal carcinoma in situ (DCIS) of the breast were de novo tumors that evolved independently of the primary lesion.
The results, presented during the 2022 American Association for Cancer Research (AACR) Annual Meeting, question the assumption that all secondary cancers found in the same breast (ipsilateral) are related to the primary tumor and whether biomarkers, by themselves, can accurately predict DCIS recurrences in these patients. The study may have significant clinical implications for how patients with recurrent DCIS lesions are managed, the researchers said.
"These tumors are new 'primaries' and these independent tumors basically suggest there is a high-risk environment or a genetic predisposition in these patients," said Tanjina Kader, PhD, a postdoctoral researcher at the Peter MacCallum Cancer Centre in the Department of Oncology at the University of Melbourne, Australia, who presented the study's findings during an AACR press briefing.
"These patients are highly susceptible to develop new tumors and, therefore, they should consider preventative therapies," she added. "It could be bilateral mastectomy or endocrine therapy or even genetic counseling."
DCIS occurs when the cells that line the milk ducts of the breast have turned cancerous, but have not spread into surrounding breast tissue. About 25 percent of cases of DCIS recur, with half of these tumors becoming invasive and potentially life-threatening.
Untill now, it was generally assumed that most recurrent tumors stem from, and are related to, the primary DCIS. If so, using a biomarker to predict such recurrences has been considered a worthwhile pursuit, with the potential to de-escalate current standard-of-care treatments-surgical excision (lumpectomy) and/or radiotherapy.
But the genetic relationship between the primary and recurrent tumors has never been proven.
Study Details
To reveal this crucial missing piece of data, the Australian-led team turned to a methodology similar to the way biologists compare DNA sequences to seek out common ancestors among living things: the study of phylogenetics. Instead of analyzing the genetic histories of plants and animals, however, Kader and colleagues performed phylogenetic analyses on primary and recurrent DCIS breast tumor samples.
An appropriate analogy for this methodology, Kader said, would be the construction of a family tree-as commonly performed in phylogeny-capable of tracking how the primary and secondary tumor pairs developed over time.
Kader referred to the process as "clonality analysis," which she said essentially compares the genetic profiles of primary tumors with secondary tumors. Tumors with shared genetic events, including chromosomal changes and/or mutations, are considered "clonal pairs," meaning they are related. Secondary tumors that do not share genetic events are considered "non-clonal" and, therefore, grew independently from the primary tumor.
"So the trunk of the tree is known as 'truncal events' or shared genetic events," she said. "And these genetic events can be chromosomal changes or mutation. And if these two tumor pairs share genetic events, we would refer this case as a 'clonal case.' And the branches of the tree resemble tumor-altering events."
In their study, Kader and colleagues extracted DNA from 65 pairs of primary DCIS and recurrences, with half of these found to be invasive breast cancer. From this cohort, the team analyzed 21 pairs through targeted sequencing or low-coverage whole-genome sequencing and 44 pairs with whole exome sequencing. Another set of 29 cases of surgically treated, non-recurrent DCIS were analyzed with a minimum 7-year follow-up.
Results showed that most recurrent DCIS tumors were related to the primary tumor in the same breast. But a significant proportion, 12 out of the 65 pairs-or about 18 percent-of the recurrent tumors, did not share genetic events with the primary tumor, suggesting they were "de novo" tumors that were unrelated to the previous cancer.
Comparing genetic profiles, the research team discovered four chromosomal changes in clonal cases that otherwise were missing in non-recurrent cases. In addition, they found frequent and specific variations of the TP53 gene in the recurrences related to the primary tumor, and not in tumors from the primary DCIS cases that don't recur or those that had non-clonal (independent) recurrence. But the genetic profiles between non-clonal cases and non-recurrent cases were found to be genetically similar.
So, while the four chromosome changes and mutational status (of TP53) might serve as a potential biomarker to predict recurrence from primary DCIS, Kader noted, the genetic similarities between non-clonal cases and non-recurrent cases cast a huge shadow over the potential accuracy of such a biomarker. She suggested this biomarker would result in the undertreatment of these patients.
"So, even though we can be excited about a predictive biomarker, this particular finding raises a question whether we should actually utilize a predictive biomarker for DCIS recurrence," Kader said.
As for the future, she said more work needs to be done to better understand the biology that separates non-recurrent cases from non-clonal cases and cloned cases.
"The interaction between stromal cells in the microenvironment and tumor cells-that's something we should definitely look at," Kader added.
Study limitations included the phylogenetic analysis based solely on chromosomal changes, which represent the main genetic changes observed in breast cancer. Also, a full phylogenetic analysis based on all the tumor-specific mutations could not be performed due to the lack of baseline genetic information from normal cells. However, the authors were able to detect the most relevant breast cancer driver mutations, such as those in TP53 or PIK3CA
Warren Froelich is a contributing writer.