Lois Wingerson

Will we still be using mammography for breast cancer screening one fourth of the way through this century? That was the question posed in one session at ECR 2011, and speaker Christine Kuhl said she greatly fears that we actually may.

"Why not use what we already have for screening?" she asked (and not for the last time). "MRI."
People discount MRI because it's expensive, difficult to read, scary for patients, and "drives radiologists crazy," she went on. But it's "a young method, just past its teething troubles." Results from published multicenter trials downplay the use of MRI for breast screening, added Kuhl, but their results are worse than its potential because so many radiologists in those studies didn't know about the latest diagnostic criteria or how to use the magnets adequately.


Kuhl, who is a diagnostic and interventional radiologist at the University of Aachen in Germany, headed the prospective multicenter EVA trial recently published in the Journal of Clinical Oncology, which screened 687 women at high risk of breast cancer, comparing the effectiveness of mammography, ultrasound, MRI, and breast self-examination. MRI was far the most effective, tending to identify cancers at an earlier stage than either mammography or ultrasound.

"Mammography is not the perfect tool as we radiologists would like to see it or as it is portrayed to the public," she said. With its high false-positive rate, its limited sensitivity for invasive breast cancers, and its mere 36 percent sensitivity in women with dense breasts, Kuhl seemed mystified as to why it is still accepted at all. She scoffed at suggestions that adding ultrasound would make it more effective.

"The average exam time would be 21 minutes per woman," she said. At the usual reimbursement rates, she pointed out, "that would have to be considered the most expensive breast cancer screening method."

As to arguments that MRI cannot be recommended for screening without randomized trials, Kuhl countered that the most important evidence is already in after 40 years of mammography: Early detection of breast cancer does lead to better survival.

What we really need for breast cancer screening, said biophysicist Marin Yaffe of Toronto's Sunnybrook Health Sciences Center, is something like they use at the airport: a highly sensitive first-pass scanner, followed by something more specific for those pulled off to the side. He had plenty of suggestions for that second-pass test. As what that first scanner would be 14 years from now, he predicted only that it won't be mammography any more. There probably won't be mammography at all, he added.

Nor will it be the latest current innovation, tomosynthesis. "Five millimeters [the current resolution of tomosynthesis" is about 109 cells," he said. "We should be able to do better than that."
On the other hand, unlike Kuhl, Yaffe could not come up with a suitable alternative. Targeted imaging with nanotech probes is "promising" but much farther down the road than 2025, he said. Other markers, such as diffuse optical tomography of oxygenated and deoxyhemoglobin, may be useful as markers that correlate to treatment results, but probably not for screening. Scanning for hyperpolarized 13C-containing molecules is "probably not" a screening method, either.

"The thing that annoys me," said Kuhl later, "is that we do have [the answer] at our fingertips, which we choose to ignore. If in 2025 we still ignore the huge potential of MRI for screening for cancer, we will have missed the opportunity to save many women's lives."

One session attendee suggested a different possibility: that by 2025 epidemiologists could stratify by low risk those people who don't need screening at all. Nobody accepted the invitation to discuss whether by 2025 breast cancer screening might be not only better, but less prevalent.

Articles

Breast Cancer Screening in 2025: Mammography, MRI, or More?

Final results of the prospective French CIRTACI trial are in: 

are very safe in general, as long as radiologists are vigilant for patients with allergies, asthma, or previous reactions to contrast agents. The vast majority are more itchy than serious.

This must have come as no surprise to the audience at ECR 2011 this week. But session chairman Tim Leiner, associate professor of radiology at the University Medical Center Utrecht, said it would be useful information nonetheless, given the level of their concern about the risks of regulation in Europe and litigation in America related to contrast media reactions.

Two months ago, Omniscan (now part of GE) settled a closely watched lawsuit over nephrogenic system fibrosis (NSF) rather than letting it going to jury trial. Radiologists know well how to avoid the exceedingly rare reactions to contrast media, Leiner said, but regulators and litigators are "too late to the party."

The 31-center CIRTACI study included 330 patients who had reactions to imaging contrast media, but after exclusions 275 remained for analysis. Of these people, 40 percent had allergic reactions, 51 percent had non-allergic reactions, and the remainder had problems such as edema, nausea, and allergy to latex. "Always remember to ask about allergy to latex," said Clement, who works at the French health agency INSERM.

Clement began by showing an image of what he described as a typical adverse response to a contrast medium: a well-defined itchy red spot roughly two inches in diameter on a patient's back. A full 80 percent of the reactions in the CIRTACI trial were grade 1 or grade 2 (minor, by the study's criteria). The proportions for severity of response were nearly identical for gadolinium-containing contrast media (15 percent of cases) as for the remainder, the iodine-containing agents.

Most reactions appeared within eight to 10 minutes, and none appeared after 20 minutes. The more severe the reaction, the more likely it is allergic, Clement said. All grade 4 reactions and 90 percent of grade 3 reactions showed signs of being allergic. He added that 30 percent of low-grade reactions were also "truly allergic." (The study seems to have lain to rest any doubts that contrast agents can cause allergic reactions, as all responses were assessed by both a dermatologist and an allergist.)

It's very important to follow up with tryptase testing as well as skin tests for any patient who shows a reaction, Clement said, to avoid repeating exposure to that contrast agent in the future. In one case out of five, a patient with a reaction gave a negative result on a skin test, so tryptase testing is justified. (But not every patient who undergoes contrast imaging for the first time should have allergy testing, he said during the discussion. That would cause far too many false positive reactions. It's only necessary to allergy-test patients who have already shown a reaction.)

Fully half of people who reacted to one contrast agent showed a cross-reaction to another contrast agent on allergy testing. Thus, Clement said, it's important to fully explore all reactions with blood testing and allergologic testing.

Another speaker at the same session presented results of a 555-center study from Germany that found a scant 0.6 percent adverse reaction rate among more than 160,000 patients given the contrast medium iobitridol (Xenetix). But as patients were not accrued consecutively, one member of the audience observed, there is no way to know whether this is an underestimate.

Final results of the prospective French CIRTACI trial are in: contrast media for diagnostic imaging are very safe in general, as long as radiologists are vigilant for patients with allergies, asthma, or previous reactions to contrast agents. The vast majority are more itchy than serious.

Contrast Media Reactions Are Allergic, but Minor

PET hybrid technologies are already hot, and two innovations discussed during the final session of ECR 2011 are aimed at making them more even powerful and efficient.

A team based at Memorial Sloan-Kettering Cancer Center in New York has expanded on preclinical testing of a method that detects Cerenkov radiation (the same emission that gives nuclear reactor their eerie blue color) using PET scans, which could enable a unique dual-modality form of PET. And microprocessor-based detectors now under development in Spain could dramatically improve PET's resolution and efficiency.

Cerenkov Luminescence Imaging had its launch in a 

in 2009 by a team of imaging engineers at the University of California-Davis and at Millennium Pharmaceuticals, Inc. Although these signals (related to interference by emitted particles that travel faster than the speed of light) are relatively weak, the developers pointed out that they can be detected by ordinary optical imaging equipment. The New York team has followed up by testing Cerenkov imaging, first on phantoms labeled with an 89Zr-containing contrast medium and then on mice bearing similarly labeled tumors.

 radiolabeled with a zirconium-containing form of the drug trastuzumab injected into the flanks of mice. These appeared similar to the corresponding images from standard PET. At ECR 2011, they reported using the mouse model to visualize the difference between HER2-positive breast cancer cells flagged with radiolabeled trastuzumab on one flank and unlabeled HER2-negative cells on the opposite flank. They also demonstrated that they could use Cerenkov imaging of the radiolabel to guide resection of the tumor, after which the signal disappeared.

Cerenkov radiation imaging could be used to guide resection of tumors either endoscopically or intraoperatively, predicted team member Alessandro Ruggiero, who now works in the department of nuclear medicine at Erasmus Medical Center in Rotterdam. The researchers also envision that it could be used for low-cost, high-throughput in vivo testing of radioactive tracers.

Would surgeons really be willing to adopt a technology that required them to use a radioactive tracer in the operative environment? Martin Pomper, deputy director of the In Vivo Cellular and Molecular Imaging Center at Johns Hopkins University, expressed skepticism, given the large number of existing fluorescent tracers for PET. Ruggiero responded that they hoped to develop agents with a longer half-life than zirconium. If they can achieve that, Pomper said privately, "there's a chance, but it's a little bit remote."

Whatever the source of the signal in the PET scanner, engineers at the Institute for High-Energy Physics in Barcelona are confident that they have found a much better way to detect it than the scintillating crystals used in current scanners. Their 

 designed with a 1-mm thick CdTe detector has a high detection efficiency and a far superior signal-to-noise ratio compared to existing PET, said research professor Mokhtar Chmeissani.

Because the detector is solid state, it can be made small enough to fit right inside the gantry of a hybrid MRI-PET scanner, he said, which makes its signal completely impervious to drift due to the strong magnetic field. Built on a Medipix2 integrated circuit, the detector has a 1 mm x 1 mm x 2 mm resolution. With a signal scatter of a mere 2 percent, said Chmeissani, its signal-to-noise ratio is roughly 30 times better than that of a state-of-the-art clinical PET scanner. Next, the Barcelona team intends to attach it to a gamma camera, predicting that it will increase the sensitivity by a factor of 100 over a current pinhole gamma camera.

Commercial development of the detector is sponsored by a 20 million-Euro grant from the European Research Council. Chmeissani declined to speculate when it would be ready for clinical use, but he did anticipate that its use will be limited.

"This is specifically designed for the brain," he said. "The market for PET scanners is about one per million people. I expect this would be used for about one in every ten million people. It's a very dedicated device.”

Improving on PET: Adding a Blue Glow, Removing a Crystal

Kuhl, who is a diagnostic and interventional radiologist at the University of Aachen in Germany, headed the prospective multicenter EVA trial recently published in the Journal of Clinical Oncology, which screened 687 women at high risk of breast cancer, comparing the effectiveness of mammography, ultrasound, MRI, and breast self-examination. MRI was far the most effective, tending to identify cancers at an earlier stage than either mammography or ultrasound.