Researchers Use Machine Learning to Improve Breast Cancer Screening Techniques

Imagine enduring a painful, expensive and scar-inducing surgery—only to find out afterwards that it wasn’t necessary.

This is the situation for many women with high-risk breast lesions—areas of tissue that appear suspicious on a mammogram and have abnormal but not cancerous cells when tested by needle biopsy. Following surgical removal, 90% of these lesions end up being benign.

A change in the standard of care could be on the horizon thanks to researchers at Massachusetts General Hospital and MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) who have found a more precise and less invasive way to separate harmful lesions from benign ones.

MIT-AI-Cancer-Detection-01_0.jpg
From left: Manisha Bahl, director of the Massachusetts General Hospital Breast Imaging Fellowship Program; MIT Professor Regina Barzilay; and Constance Lehman, chief of the Breast Imaging Division at MGH’s Department of Radiology. Photo courtesy of MIT News

“The decision about whether or not to proceed to surgery is challenging, and the tendency is to aggressively treat these lesions [and remove them],” said Manisha Bahl, MD, Director of the Breast Imaging Fellowship Program at Mass General, in a recent interview.

Bahl, along with a team of researchers, have harnessed the power of artificial intelligence (AI) to develop a more accurate and less invasive screening method for high-risk lesions. When tested, the machine correctly diagnosed 97 percent of 335 high-risk breast lesions as malignant and reduced the number of benign surgeries by more than 30 percent compared to existing approaches. These results were recently published in Radiology.

The team developed an AI system that uses machine learning to distinguish between high-risk lesions that need to be surgically removed from those that should just be watched over time. They created this model by feeding it data on over 600 high-risk lesions, including information on the patient’s demographics and pathology reports, and then tasked it to identify patterns among the different data elements.

Through a process called deep learning, the machine uses the data to create an algorithm that can be used to predict which high-risk lesions should be surgically removed. This process differs from traditional software programming in that the researchers did not give the machine the formula for diagnosis, but rather let it analyze the data and identify patterns on its own.

“To our knowledge, this is the first study to apply machine learning to the task of distinguishing high-risk lesions that need surgery from those that don’t,” said collaborator Constance Lehman, MD, PhD, chief of the Breast Imaging Division at Mass General’s Department of Radiology, in a recent interview. “We believe this could support women to make more informed decisions about their treatment and that we could provide more targeted approaches to health care in general.”

Lehman says Mass General radiologists will begin incorporating the model into their clinical practice over the next year.

Liquid Biopsies Give Clues on When and Why Cancer Treatments Lose their Efficacy

With the advent of targeted cancer therapies and immunotherapy, and with new CAR-T therapies on the way, more cancer patients are living with their disease. However, many cancer patients find that their therapies have limitations and are faced with the potential of disease progression. Often, those who initially respond to a course of treatment eventually develop a resistance to these medications, forcing oncologists to switch therapeutic course.

Currently, one of the ways to know when a treatment stops working is by taking a biopsy of the tumor. These surgical procedures are invasive and costly, and because they can only be done sporadically, valuable treatment time can be lost. Additionally, some cancer patients may be too physically fragile for surgery.

Researchers have been looking for a safe, fast, less expensive and more accurate way to identify early signs of treatment resistance, while also searching for new insights into the genetic changes that occur within tumor cells to drive this resistance. This way, new therapy plans can be considered sooner, giving the patient a better chance for their best possible outcome.

A new diagnostic blood test known as a liquid biopsy has shown early promise in addressing these needs. Now researchers, including a team from the Mass General Cancer Center, are providing confirmatory data that may help to move liquid biopsies into clinical practice. These data were presented at the ESMO 19th World Congress on Gastrointestinal Cancer.

How do liquid biopsies work?

A liquid biopsy is a diagnostic test that detects circulating tumor DNA (ctDNA), which is genetic material released by dying tumor cells that flows through the bloodstream. These tests are less invasive than a tissue biopsy and therefore can be given with greater frequency.

Regularly monitoring ctDNA levels in a patient’s bloodstream can provide early notice when a treatment is no longer working. It could also offer a more complete picture of the genetic changes in tumor cells that are driving the resistance to treatment, which could guide new treatment courses.

Liquid biopsies and gastrointestinal cancer study

Mass General Cancer Center investigators followed nearly 40 patients with various forms of gastrointestinal cancers who had experienced initial success with targeted therapies, but then began to show signs of treatment resistance. Liquid biopsies were taken when the patients’ disease started to progress to analyze the levels and genetic profile of ctDNA in their bloodstream. Researchers identified one or more mutations or mechanisms that contributed to treatment resistance in 31 of the 40 patients. Fourteen of these patients had multiple mutations that contributed to resistance.

In patients who had both solid tissue biopsies and the liquid biopsies, the researchers found that in two-thirds of the cases, the liquid biopsies revealed the presence of more genetic mutations than tissue biopsies alone.

“Identifying what specific mutations are responsible for treatment resistance is very important in helping clinicians choosing what treatment path a patient should try next, whether it be another drug or perhaps radiation,” said study investigator Aparna Parikh, MD, from the Mass General Cancer Center.

“We have shown this approach is feasible across many different GI cancers,” she noted. “The next step is to study how best to use this new technology in daily practice. It’s important for clinicians to understand its utility as well as its limitations.”

Meet a Mass General Postdoc: Echoe Bouta

In honor of National Postdoc Appreciation Week, all this week we’ll be sharing profiles of just a few of our amazing Mass General postdocs to highlight their research and what inspires them.

Meet Echoe Bouta, PhD, a research fellow in the Department of Radiation Oncology.

Postdoc week Bouta (1)

Where did you get your PhD from?

University of Rochester

What questions are you asking in your current research? What do you hope to find out?

Lipedema is a chronic disorder that results in increased fat in the lower limbs and manifests as dramatic, painful swelling. Clinical studies demonstrate that patients present with at least partial lymphatic dysfunction, which worsens until lymphedema occurs. As the etiology is largely unknown, treatments are often ineffective, demonstrating the need to understand the relationship between fat and lymphatic capability.  The questions I hope to answer are:

      •  How does increased adiposity impair lymphatic function in an animal model of obesity?
      • Can we target key pathways in this process to improve lymphatic function?

The answer to this questions will hopefully catalyze new treatments for lipedema.

What drew you to this field?

The lymphatic research field is relatively young compared to other research fields.  I believe that we have only started to discover the importance of the lymphatic system in multiple disease states and it is exciting to be part of that.

What is a typical day like for you?

Much of my project involves imaging of the lymphatic system under different contexts, such as obesity or after delivery of a drug. Therefore, a typical day for me is a mix of in vivo experiments, data analysis and writing.

What do you like most about being a postdoc at MGH?

The people.  We are surrounded by talented scientists from a variety of fields that results in a very interdisciplinary research environment.

New Study Shows Lymph Nodes Aren’t Always to Blame for Cancer’s Progression

In a case of mistaken identity, researchers at Massachusetts General Hospital have found that lymph nodes are not always responsible for cancer’s deadly spread to other organs. These results buck many preconceived notions about lymph nodes’ role in cancer development and suggest a new pattern for the progression of certain types of cancer.

Doctors recognize that patients whose cancer spreads (metastasizes) from the original tumor to the surrounding lymph nodes have a worse prognosis than patients whose lymph nodes are cancer-free. This observation has traditionally been explained by a progression model of primary tumor to nearby lymph nodes to other organs. However, no conclusive evidence for this model has existed so far.

In a new study, researchers from the Edwin L. Steele Laboratories for Tumor Biology in the Mass General Department of Radiation Oncology investigated the “family tree” of metastases in colorectal cancer. Contrary to the prevailing belief that the spread begins in the lymph nodes, they found that the cancer could spread to both the lymph nodes and the organs simultaneously. In their report in the July 7 issue of Science, the researchers describe finding that, for the majority of colorectal cancer patients in the study, organ metastases (also called distant metastases) originated directly from the primary tumor, independent of any lymph node metastases.

Lymph node graphic 1

Lymph node graphic 2

“We now suspect that lymph node metastases simply indicate the presence of an aggressive primary tumor, rather than being directly responsible for the formation of distant metastases,” says lead and corresponding author Kamila Naxerova, PhD, Research Fellow at the Steele Labs.

The researchers analyzed more than 200 tissue samples of primary tumors, lymph node metastases and distant metastases from 17 patients with colorectal cancer. Samples from 35 percent of these patients followed the traditional progression model. In these samples, both lymph node and distant metastases came from the same cell type in the primary tumor, indicating that the cancer had spread from the primary tumor to the lymph nodes and then to other organs.

However, in 65 percent of patients, researchers found that cell types in lymph node and distant metastases were different and matched different cell types within the primary tumor, indicating independent origins for these metastasis types.

Their results suggest that although cancer progression can follow the traditional model described above, there is also a second distinct pattern of metastatic spread.

“These findings fill an important gap in our knowledge of metastatic disease evolution and have the potential to guide improvements in the clinical management of lymph node metastases,” says Naxerova.

The research team is now following up with a larger cohort of patients to confirm whether survival differences exist between patients with a traditional progression pattern vs the second progression pattern.

Rakesh K. Jain, PhD, Director of the Steele Labs, was senior author of this paper.

Women’s Health Week 2017

In honor of National Women’s Health Week this week, we put together a few highlights of the many Massachusetts General Hospital researchers who are investigating important topics pertaining to women’s health:

Eve Valera, PhD, a researcher at the Martinos Center for Biomedical Imaging, is working to learn more about the traumatic brain injuries suffered by women in abusive relationships:
http://www.massgeneral.org/research/news/ResearcherProfiles/profile-valera-TBIs.aspx

A team from the Pediatric Surgical Research Laboratories in the Department of Surgery discovered a hormone that may protect and preserve ovaries during chemotherapy:
https://mghresearchinstitute.com/2017/02/22/hormone-may-protect-and-preserve-ovaries-during-chemotherapy/

Researchers at the Cancer Center found that a specialized screening protocol may improve detection of ovarian cancer in high-risk women:
https://mghresearchinstitute.com/2017/03/06/specialized-screening-protocol-may-improve-detection-of-ovarian-cancer-in-high-risk-women/

A collaborative study between researchers at Mass General and Boston University School of Medicine found evidence implying that alcoholism may have different effects on the reward system in the brains of women than it does in men:
http://www.massgeneral.org/about/pressrelease.aspx?id=2092

Specialized Screening Protocol May Improve Detection of Ovarian Cancer in High-Risk Women

First, let’s define a few key words:

  • CA125: A protein found on the surface of many ovarian cancer cells. Most women with ovarian cancer have CA125 levels in the blood that are over 35.
  • ROCA: An abbreviation for Risk of Ovarian Cancer Algorithm, a test that assesses risk for developing ovarian cancer. This test tracks CA125 levels over time to identify significant elevations above each patient’s baseline levels.

What’s the issue researchers wanted to address? The standard advice for women at high risk of ovarian cancer, due to either family history or inherited gene mutations, is to have their ovaries and fallopian tubes surgically removed as a preventative measure once they are done having children. Some women choose to postpone this surgery and there’s a need for a screening test for these women that will detect the development of tumors while the cancer is still in a treatable stage.

How did they address this issue? Researchers from the Massachusetts General Hospital (MGH) Cancer Center and the Biostatistics Unit, and the National Cancer Institute and Anderson Cancer Center enrolled more than 3,800 women at elevated risk for ovarian cancer in two screening trials. Participants had blood tests every three months to establish their baseline levels of CA125 and to monitor for any changes using the ROCA assessment tool. Women at intermediate ROCA risk were referred for an ultrasound examination, while those at an elevated ROCA risk received both ultrasound and clinical evaluation.

What did they find? 19 malignant tumors of the ovaries or fallopian tubes were identified during the study periods. Ten cases were diagnosed during screening, and nine were diagnosed by preventive surgery. This protocol increased the proportion of tumors detected at early stages from 10 percent – which is typically seen in high-risk women who are not screened – to 50 percent.

What do these results mean? The combined results of the two screening trials suggest that a protocol involving quarterly blood testing to identify significant increases above each patient’s personal baseline in levels of CA125, followed by ultrasound examination when such elevations are detected, could improve the chance that tumors are detected at early stages when they are easier to treat.

What do researchers have to say about their findings? Researchers caution that surgery is the primary and best option for reducing the risk of ovarian cancer, and ROCA should only be considered as a promising but unproven option for patients who decide, against medical advice, to postpone their surgery. They plan to conduct more research to identify other ovarian cancer indicators and improved imaging technologies that may help to detect even more tumors at even earlier stages.

Steven Skates, PhD, of the Massachusetts General Hospital (MGH) Cancer Center and the Biostatistics Unit, is co-lead and corresponding author of the report. Learn more here.

Mass. General Study Identifies Genomic Differences Between Breast Cancers of African American and White Women

“In addition to having a higher prevalence of triple-negative breast cancers than Caucasian women – something that has been documented in previous studies – we found that African American women with breast cancer had a significantly higher prevalence of the TP53 driver mutation, basal tumor subtype and greater genomic diversity within tumors, all of which suggest more aggressive tumor biology,” says Tanya Keenan, MD, of the MGH Cancer Center, lead author of the study. “The higher risk of tumor recurrence that we observed among African American women was reduced when controlling for those factors, suggesting that these genomic differences contribute, at least partly, to the known racial disparity in the survival of African American and Caucasian breast cancer patients.”

 

READ MORE