Skip to main content
RSNA Journals logoLink to RSNA Journals
letter
. 2024 Apr 9;311(1):e233188. doi: 10.1148/radiol.233188

Caution on the CONTRRAST Trial

Sarah Eskreis-Winkler 1,, Maxine S Jochelson 1, Katja Pinker 1, Janice S Sung 1, Christopher Comstock 1
PMCID: PMC11070605  PMID: 38591972

Editor:

We have several concerns about the recent article by Dr Phillips and colleagues, published in the November 2023 issue of Radiology (1). This reader study compared two-dimensional (2D) mammography, contrast-enhanced mammography (CEM), abbreviated MRI, and MRI, concluding that CEM was noninferior to MRI in an asymptomatic study sample. Several methodologic issues warrant clarification and call this conclusion into question.

The reader study appears to have significant selection bias. In the listed detection method (Table S1 in Phillips et al), 35 of 44 patients with cancer were recruited due to abnormal mammographic findings. If correct, this is not representative of women imaged with MRI. MRI helps to detect approximately 16 additional cancers per 1000 women screened. By including primarily cancers depicted at screening mammography and US, the study fails to address the key clinical question: Does screening CEM truly find the small early cancers missed at conventional breast imaging and depicted at MRI? To state that “the results are a key step toward the replacement of screening breast MRI [with CEM]” is unsubstantiated.

The article also does not compare CEM versus MRI and 2D mammography, the current standard of care when MRI is used for supplemental screening. In Table S1 in Phillips et al, nine of 44 cancers presented as “calcifications.” MRI is a supplement, not a replacement, to screening mammography because we know MRI will not depict some cancers manifesting as mammographic calcifications. Having microcalcifications in 20% of the cancer cohort in this report raises concerns about the fundamental validity of the study’s conclusion.

We request that the authors adjust their conclusions based on the inherent design limitations. We also ask that Figure 3 be updated to include a receiver operating characteristic curve of MRI with 2D mammography to better understand if, indeed, screening CEM is noninferior to the current standard of care in high-risk patients.

Footnotes

Disclosures of conflicts of interest: S.E.W. No relevant relationships. M.S.J. Speaker honorarium from GE HealthCare. K.P. Grants from Digital Hybrid Breast PET/MRI for Enhanced Diagnosis of Breast Cancer (HYPMED) H2020 Research and Innovation Framework Program (PHC-11-2015 #667211-2); A Body Scan for Cancer Detection using Quantum Technology (CANCERSCAN) (H2020-FETOPEN #828978); Multiparametric 18F-Fluoroestradiol PET/MRI coupled with Radiomics Analysis and Machine Learning for Prediction and Assessment of Response to Neoadjuvant Endocrine Therapy in Patients with Hormone Receptor+/HER2− Invasive Breast Cancer Jubiläumsfonds of the Austrian National Bank (18207); Vienna Science and Technology Fund (LS19-046); MSKCC 2020 Molecularly Targeted Intra-Operative Imaging Award; Breast Cancer Research Foundation June 2019 to May 2021; NIH (UG3 CA239861); NIH (R01 CA249893); NIH (1R01CA270018-01A1); supported in part through the NIH/NCI Cancer Center (P30 CA008748); MSK has institutional financial interests relative to GRAIL; consultancy from Genentech, Merantix Healthcare, Aura Health Technologies, Guerbet; speakers bureaus fees from European Society of Breast Imaging, Bayer, Siemens Healthineers, DKD, Olea Medical, and Roche; support for travel and/or meetings from European Society of Breast Imaging; participation on a DataSafety Monitoring Board from Guerbet, Bayer; executive board membership, EUSOBI. J.S.S. No relevant relationships. C.C. Personal fees for advisory board from Bayer; consultant for Guerbet; speakers bureau fees from GE Medical.

Reference

  • 1. Phillips J , Mehta TS , Portnow LH , Fishman MDC , Zhang Z , Pisano ED . Comparison of Contrast-enhanced Mammography with MRI Utilizing an Enriched Reader Study: A Breast Cancer Study (CONTRRAST Trial) . Radiology 2023. ; 309 ( 2 ): e230530 . [DOI] [PubMed] [Google Scholar]
Radiology. 2024 Apr 9;311(1):e233319. doi: 10.1148/radiol.233319

Response

Jordana Phillips *,§,, Tejas S Mehta , Leah H Portnow , Michael D C Fishman , Zheng Zhang , Etta D Pisano #

The comments from our colleagues are graciously received and present an opportunity to discuss how best to evaluate new breast imaging screening modalities (1).

To compare imaging modalities in a screening population, enough malignant cases must be included, the number of which is dictated by the expected difference between examinations being compared. Previous research comparing CEM and MRI has shown a similar sensitivity (2,3), thereby requiring large numbers of malignant cases to show a significant difference in diagnostic accuracy. This is challenging for a prospective trial given the low rate of cancers per 1000 women screened, as shown in Jochelson et al (4).

The retrospective multireader multicase reader study such as CONTRRAST allows readers to review smaller numbers of cases at prevalence rates higher than what naturally occurs. Additionally, rather than recruiting malignancies directly from screening whereby CEM and MRI would be performed in all patients but only those with imaging abnormalities would be used, we performed CEM and MRI only in screening patients who had an abnormality at imaging (at the time of the diagnostic workup or biopsy), minimizing the number of patients to recruit.

The ideal multireader multicase study would have included more MRI-based screen-detected findings. However, our practice did not have a large volume of MRI screening examinations that identified cancers not seen on other modalities, limiting our ability to recruit cancers from this population. Our inclusion of more abnormalities depicted at mammography likely improved CEM results given that low-energy images are essentially the same as 2D mammography. Yet, our results demonstrate that although these cancers were identified at traditional mammography screening, the low-energy evaluation by 12 radiologist readers did not capture all of them with a 2D sensitivity of 75% versus 89% for CEM. Investigators have managed this challenge of recruiting MRI-depicted malignancies in different ways (5).

CONTRRAST is a comparison of diagnostic performance of CEM and MRI (as well as abbreviated MRI and 2D mammography) in an asymptomatic population and not a test of how best to use these examinations in clinical practice. Further investigations are needed before CEM replaces MRI with 2D mammography in clinical practice.

Footnotes

Disclosures of conflicts of interest: J.P. Consulting fees from Hologic, GE HealthCare, and Endomag; payment for expert testimony from Medical Expert; payment for travel to an event from GE HealthCare. T.S.M. No relevant relationships. L.H.P. No relevant relationships. M.D.C.F. Association of University Radiologists strategic alignment grant; consulting fees from Hologic, Endomag, and GE HealthCare; payment for invited grand rounds from MD Anderson and Massachusetts General Hospital. Z.Z. No relevant relationships. E.D.P. Tomosynthesis Mammographic Imaging Screening Trial grant to institution; salary support through the Eastern Cooperative Oncology Group–American College of Radiology Imaging Network; advisory board member for the Harvey L. Neiman Health Policy Institute; vice president of the Academy of Radiology and Biomedical Imaging Research.

References

  • 1. Phillips J , Mehta TS , Portnow LH , Fishman MDC , Zhang Z , Pisano ED . Comparison of Contrast-enhanced Mammography with MRI Utilizing an Enriched Reader Study: A Breast Cancer Study (CONTRRAST Trial) . Radiology 2023. ; 309 ( 2 ): e230530 . [DOI] [PubMed] [Google Scholar]
  • 2. Pötsch N , Vatteroni G , Clauser P , Helbich TH , Baltzer PAT . Contrast-enhanced Mammography versus Contrast-enhanced Breast MRI: A Systematic Review and Meta-Analysis . Radiology 2022. ; 305 ( 1 ): 94 – 103 . [DOI] [PubMed] [Google Scholar]
  • 3. Xiang W , Rao H , Zhou L . A meta-analysis of contrast-enhanced spectral mammography versus MRI in the diagnosis of breast cancer . Thorac Cancer 2020. ; 11 ( 6 ): 1423 – 1432 . [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4. Jochelson MS , Pinker K , Dershaw DD , et al . Comparison of screening CEDM and MRI for women at increased risk for breast cancer: A pilot study . Eur J Radiol 2017. ; 97 : 37 – 43 . [DOI] [PubMed] [Google Scholar]
  • 5. Lawson MB , Partridge SC , Hippe DS , et al . Comparative Performance of Contrast-enhanced Mammography, Abbreviated Breast MRI, and Standard Breast MRI for Breast Cancer Screening . Radiology 2023. ; 308 ( 2 ): e230576 . [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Radiology are provided here courtesy of Radiological Society of North America

RESOURCES