This study showed no additional cancer detection through supplemental AP MRI screening for women with average risk and dense breasts. During a 2-year follow-up period, no evidence of malignancy was observed. Nevertheless, several studies have reported a significantly higher additional cancer detection rate (CDR), ranging from 11.8 to 27.4 per 1000 screenings with supplemental MRI screening (22, 23, 44, 45, 46). These studies featured different inclusion criteria, encompassing both high-risk and average-risk women and various breast density types, including C and D-type breasts and D-type breasts exclusively. However, a common finding across all these studies is the elevated additional CDR compared to mammography.
Several factors may explain the unexpected result of our study. First, the sample size in our study was relatively small. Nonetheless, in a study by Weinstein et al. (45), which included 475 women, a remarkably high CDR of 27.4 per 1000 screenings was reported after DBT screening, known to detect more cancers than mammography. Similarly, Kuhl et al. (27) found a CDR of 18.2 per 1000 in 606 women supplementary to mammography screening. Both studies closely resemble our cohort of 238 women. This suggests that with our sample size, at least two cancers should have been detected, as according to the literature, 2.8 to 6.6 additional cancers per screening should be expected for such a cohort of 238 women. However, it's important to note the very limited number of participants in our study. Considering the acceptance rate of AP MRI screening among the invited women, stood at 38.5%, with 37.8% of them completing high-quality AP MRI screening. This relatively low number of study inclusions may have introduced a bias to our results.
Second, the women in our study had undergone five rounds of mammography screening over ten years. It is well-established that after successive rounds of screening, interval cancers tend to decrease. The CDR among the women who underwent MRI was 16.5 per 1000 screenings in Bakker et al.’s study (46). The second screening round of their study showed a CDR of 5.8 per 1000 screening examinations compared with 16.5 per 1000 in the first round (47). So, Bakker et al’ results strongly support the decrease of CDR after successive screening rounds. However, in the ACRIN 6666 study (22), it was demonstrated that after three consecutive rounds of mammography screening supplemented with US screening, they could still detect 14.7 additional cancers per 1000 screenings with MRI. Conversely, we have not detected any cancer with supplementary MRI screening.
Third, our interval cancer rate was notably lower (2.03 per 1000 screening) (42). In the Dense trial, the interval cancer rate was 5, subsequently decreasing to 0.8 per 1000 screenings after MRI screening (46). However, our interval cancer rate was notably lower than the Dense trial. This difference may have diminished the significant impact of the supplementary AP-MRI screening results.
Despite high CDR, long scan acquisition, and interpretation time, high cost limits breast MRI for screening in all women-risk populations. Therefore, the 2007 American Cancer Society (ACS) has recommended additional breast MRI screening for women at high risk. This group includes women with genetics-based increased risk and their untested first-degree relatives, those with a calculated lifetime risk of 20% or greater, and those with a history of chest or mantle radiation therapy at a young age (24). A more personalized approach should be favored over offering MRI screening to every woman with risk. A recent study by Lamb LR et al. utilizing a mammography-based AI risk assessment tool demonstrated a higher CDR in women evaluated by this AI-based tool compared to traditional risk models (48). Furthermore, this study highlights that the risk assessment is contingent on the fibroglandular structure observed in mammography rather than the ACR density categories or conventional risk scores. Although many guides do not recommend MRI screening for average-risk women (49–51), EUSOBI has published a new recommendation that recommends supplemental full protocol MRI screening to women with extremely dense breasts from age 50–70 and at least every four years, preferably every 2 or 3 years in March 2022 (26). ACR guideline supports that MRI screening can be beneficial for dense breasts despite the accumulation of sufficient data (22, 23, 51). However, a study conducted by Geuzinge HA et al. (52) calibrated with the findings from the Dense trial (45), demonstrated the cost-effectiveness of MRI screening at 4-year intervals, with a cost of €15,620 per Quality-Adjusted Life Year (QALY). This assessment utilized a threshold of €22,000 per QALY gained. Consequently, a three or 2-year interval for MRI screening was deemed not cost-effective. Although our study did not specifically assess the cost-effectiveness of MRI screening, countries with limited resources need to be mindful of the limitations associated with MRI screening and carefully evaluate its cost-effectiveness.
Our study represents a feasibility assessment to explore the potential integration of AP breast MRI into a screening program. Despite numerous studies in the literature demonstrating higher CDR (44, 45), our findings did not align with these previous reports. While AP breast MRI holds promise as a screening modality, several operational considerations must be addressed to facilitate its adoption into clinical practice. For instance, it is widely accepted that reducing image acquisition and interpretation times can enhance availability and cost-effectiveness. However, the reality in real-life scenarios often diverges from these expectations, particularly within government hospital systems, where cost reduction may not be readily achievable. One of the challenges we encountered during the study was incorporating screening AP-MRI into a schedule already filled with diagnostic MRI appointments. This proved to be a daily operational challenge. Additionally, while image acquisition and interpretation times may be shorter, there remains a debate about the actual overall workflow efficiency.
A study by Borthakur et al. (53) investigated various aspects of workflow, including imaging setup, patient preparation, and MRI room turnover, revealing that the flow rate for an abbreviated MRI examination was lower than anticipated. Therefore, radiologists and clinicians should be aware of these practical challenges when considering the implementation of AP breast MRI into clinical practice. The challenges in clinical practice extend beyond longer-than-expected total procedure times. For example, incorporating screening AP MRI into an appointment schedule filled with diagnostic MRI appointments posed practical difficulties. Furthermore, it has been observed that AP MRI can increase the rate of unnecessary biopsies, leading to increased costs. This represents one of the primary disadvantages of incorporating screening MRI. However, the EUSOBI guideline recommends screening for women with extremely dense breasts from age 50–70 and at least every four years, which seems to offer potential advantages, particularly considering the findings from the Dense Trial (46). Nevertheless, in contemplating the integration of AP MRI into a screening program, one must bear in mind the possible challenges given above.
Our study has some limitations. First, the number of participants was low to make an optimal comparison with MRI and mammography. Second, the number of women who did not respond was very high, possibly due to being called up after five screening rounds.
This study represents a real-life scenario involving the implementation of supplemental breast screening for women with dense breasts. The study revealed no additional cancer detection among the 238 women with dense breasts. However, this result came at the cost of a 15.5% recall rate, a 5% biopsy recommendation, and a 3% biopsy rate. Furthermore, the study encountered several challenges, including scheduling appointments within a busy MRI schedule and managing various aspects of workflow. It is crucial to consider these limitations when considering the implementation of MRI in large-scale screening programs. Meticulous criteria should be applied when selecting women for MRI screening to ensure optimal outcomes.