MRI has outrun other modalities in a screening trial involving high-risk women. Such research helps justify an estimated $1.4 billion a year in direct costs for the United States if new American Cancer Society guidelines
CHICAGO-MRI has outrun other modalities in a screening trial involving high-risk women. Such research helps justify an estimated $1.4 billion a year in direct costs for the United States if new American Cancer Society guidelines (see box on page 33) are followed.
At the 2007 RSNA annual meeting, Lily Kernagis, MD, a fellow in women's imaging at the University of Pennsylvania, presented results of a large prospective study comparing screen-film mammography, digital mammography, whole-breast ultrasound, and contrast-enhanced MRI in a screening population of 569 high-risk asymptomatic women. All participants had had a negative mammogram and negative physical exam.
High-risk was defined as including the following risk factors:
• A 25% lifetime risk based on genetic testing, Gail or Claus models.
• Diagnosis of lobular carcinoma in situ, atypical ductal hyperplasia, or atypical lobular hyperplasia.
• Exposure to radiation prior to puberty.
• History of cancer in the contralateral breast.
Three radiologists made the interpretations in a blinded fashion, and the results were later unblinded at conference. One physician read the screen-film and whole-breast ultrasound studies, one interpreted digital mammography, and one performed the CE-MRI readings.
Of 95 lesions recommended for biopsy, 11 (11.6%) were malignant. Half of these were invasive ductal carcinoma; the remainder included invasive lobular carcinoma, invasive cancer not otherwise specified, and DCIS. The modalities detected these 11 cancers as follows:
• Screen-film mammography, 2 of 11.
• Digital mammography, 4 of 11.
• MRI, 10 of 11.
• Whole-breast ultrasound, 2 of 11.
Seven of the 11 cancers were seen on only one modality: one with digital mammography and six with MRI alone. No cases were seen on either screen-film mammography or ultrasound alone.
"In our study, MRI detected the highest percentage of clinically occult breast cancers in a high-risk population, more than digital mammography, ultrasound, and screen-film mammography. Breast MRI may be a useful screening tool in addition to mammography in the high-risk population," Dr. Kernagis said.
Follow-up was carried out for 1 to 3 years and revealed no new breast cancer diagnoses. Based on the results and follow-up, multimodality screening had a negative predictive value of 100%, sensitivity of 100%, and specificity of 84.9%.
$1.4 billion a year
Another study presented during the same RSNA session found that 1.8 million women are at high risk according to the ACS definition, and they should be getting supplemental breast MRI. The majority of these women-1.7 million-qualify on the basis of a strong family history, said Rand Stack, MD, of White Plains Hospital in New York.
Assuming average charges of $1,038 for initial MRI with 100% screening compliance, direct costs for adding MRI as a supplemental screening tool would amount to about $1.9 billion a year. If compliance was more along the lines of mammography screening guidelines, with 75% of women participating, direct costs would reach $1.4 billion a year.
Use of breast MRI would also create indirect costs related to follow-up studies (average price, $793) and biopsies. Researchers estimated indirect costs would add another $573 million in annual expenses.
Currently, 28 million MRI exams for all organs are performed in the United States annually, so adding 1.8 million studies would equate to a 6% increase in the MRI study volume, Dr. Stack said.
Industry sources suggest there is sufficient excess capacity in the installed MR scanner base to absorb a 6% increase, although it is unclear how many machines are equipped with breast coils, he said.
Assuming the 14,000 physicians currently performing mammography all performed breast MRI, the workforce also seems to have the capacity needed to manage breast MRI. Workload would increase by two or three breast MRI interpretations per week, Dr. Stack said.
A recent survey of breast imaging specialists showed 75% were performing breast MRI, albeit at a low volume. But recent figures for the number of physicians performing mammography more generally are not available. Attendees and panelists at the RSNA session said it seems unlikely that all radiologists currently performing mammography would also offer breast MRI.