RSNA 2008: Mining the Landscape, Assessing the Exhibits
In assessing the imaging technologies on display at RSNA, an observer provides tools for making critical decisions regarding your capital budget for 2009 Despite reports that capital equipment budgets are frozen at some hospitals, buyers were out in force at the 2008 meeting of the RSNA in Chicago. Nonetheless, professional attendance did not reach last year’s level, and exhibit square footage purchased and vendor attendance also were down, reportedly. Clearly, the global economic crisis is having an impact, and both vendors and health care providers are feeling pressure on the capital side. Our financial analysts believe that from a capital-purchase standpoint, vendors will feel the greatest effect in the middle of 2009. For providers with technology decisions to make, it is more important than ever to make every dollar count, and we predict that there are still some great market opportunities. Providers also need to maximize the equipment that they currently own; 64-slice (and beyond) CT is a great example. Scanners capable of doing cardiac imaging are located on every corner in most cities, but we have yet to see a market where one provider has risen to the top as the cardiac imaging provider of excellence. The technology is there, but in almost all areas, the cardiac imaging program for which it was purchased has not developed. The market is wide open, with no new technology purchase necessary. One positive consequence of these changes is that hospitals, imaging centers, and now vendors are placing a renewed focus on efficiency. Surveying the offerings on the show floor makes the priorities of every vendor clear: increasing the efficiency of all departments across modalities and decreasing scheduling blocks to increase revenue from one unit dramatically. With digital mammography, for instance, halving the scanning block can more than double revenue. Top Five Trends We identified five top trends on the show floor. First, focus on superpremium technologies was diminished. We didn’t see a big-splash technology, making this one of the more sedate RSNAs in recent history. The other side of this is that we’ve seen many vendors focusing on developing budget lines: a 1.5T MRI scanner for less than $1 million and a workhorse 16-slice CT scanner costing less than $500,000. The four major manufacturers have all introduced budget CT and MRI units. Second, the focus on women’s imaging continued. This is seen not only with digital mammography and the excitement around tomosynthesis, but in a huge array of offerings in the second-line imaging space for treatment planning and diagnosis. Subject to the economy, mammography volumes may decline, but vendors perceive providers as likely to purchase. Third, interest continues to grow in breast tomosynthesis. We predict that it is the one truly disruptive technology that will be introduced in the next few years. Fourth, further advances in ultrasound generated considerable excitement, as vendors continue to add functionality. Increased attention to radiation dose in 2008 has also refocused interest on ultrasound. Fifth, we are starting to see the additional development of niche offerings, with new market entrants such as specialty, head-only PET/CT scanners and new dedicated breast MRI offerings. There is real interest in developing market-specific platforms for specific patient populations, although, as capital dollars become scarce, these niche scanners are increasingly difficult to justify over a workhorse platform. CT During the past 12 to 18 months, 64-slice CT has become the technology of choice for hospitals interested in purchasing a workhorse CT scanner. While a 64-slice CT scanner is not required to perform the vast majority of studies, prices have fallen so precipitously that 64-slice CT can now be purchased for less than $1 million (without the cardiac package). It is a justifiable purchase today if there is a possibility that coronary-artery imaging will ever be performed during the life of that scanner. Is 64-slice CT good enough for coronary-artery imaging? Evidence demonstrates that 64-slice CT is the standard of care for coronary CT angiography (CCTA) for most patients, but one serious concern is radiation dose. Dose-reduction protocols exist, but if they are not used, 64-slice CCTA is associated with a hefty radiation dose. Dose-reduction packages are very important for sites planning to offer CCTA in a programmatic way. Nonetheless, 16-slice CT remains the standard of care (see table) for nearly every exam apart from CCTA—more than 90% of diagnostic CT exams performed today. In 2007, it looked as though vendors had completely abandoned 16-slice CT, and some talked about discontinuing their 16-slice CT scanners for the US market. In 2008, however, one major vendor brought out a bare-bones (but good) 16-slice CT with a planned list price of $400,000. It is not necessary to have five superpremium scanners in a hospital. During the past year, each vendor has solidified its post–64-slice CT offerings, and all are in production today, with list prices holding steady around $2 million. Vendors are backtracking and filling in the gaps; one introduced a reasonably successful 256-slice scanner in 2007, but also offered a 128-slice scanner (upgradable to 256 slices) at RSNA 2008. The vendor that introduced dual-source CT rolled out a single-generator 128-slice scanner available in 20-, 40-, 64-, and 128-slice configurations. The 20-slice unit can be upgraded to 128 slices with as little as eight hours of downtime. While that is the best flexibility we’ve ever seen, we haven’t seen these upgrades really pay off; the option is largely an insurance policy, allowing access to premium technology as required. It may help sell the high-end technology to the CFO, but most sites don’t end up doing the upgrade. Acute stroke imaging is likely to be the next place where superpremium CT is going to have a large clinical impact. Our instinct is that 64-slice CT is probably not going to be the prime technology. A benefit of a large single 256- or 320-detector array is that it allows capture of the entire brain or heart in one rotation, offering great perfusion images. This will allow rapid assessment of the extent of a stroke and, it is hoped, lead to expedited triage of patients to interventions. Combined with angiography, it creates the true optimal cardiac exam, visualizing the clot with CCTA and then seeing, downstream, the blood-flow data that would have come from nuclear medicine in the past. In post–64-slice technology, the options available are staggering, with each vendor offering a drastically different scanner architecture. It is still difficult to assess which design offers the best long-term clinical bet. CCTA and CT Colonoscopy The slower-than-anticipated development of CCTA has been frustrating. Clinical evidence has mounted that CCTA is a good way to evaluate symptomatic, low-risk chest-pain patients in an outpatient environment or emergency department. Evidence supports 64-slice CT in this role, and operationally (and perhaps economically), CCTA makes sense. Since chest-pain patients account for 5% to 10% of emergency-department admissions, CCTA will be an operational boost—especially for crowded emergency departments and telemetry units—by allowing proper routing of low-risk patients who are probably not having myocardial infarctions. Despite the promise, reimbursement and operational hurdles have been roadblocks for fully implementing CCTA programs. Private payors would have followed CMS, had it moved forward with a more positive coverage decision, but CMS has given them reason to sit back as well. Getting the number of physicians (from radiology, cardiology, or both) necessary to provide the service around the clock is daunting; no US emergency department appears, so far, to be able to offer 24/7 CCTA on its own. CCTA also was the catalyst for the biggest imaging issue of the year: the response to radiation dose. If the 2007 study1 estimating that 1.5% to 2% of all US cancers could be due to medical radiation is to be believed, this is a significant problem. There is a large risk for the young, and especially for females, from a single CCTA exam, without a dose-reduction strategy. Media interest guarantees that patients will ask whether they need the study and whether the provider is doing something to make it safer. Providers need to be ready to answer those questions; must have technology in place, where appropriate, to reduce dose; and must ensure that frontline staff is prepared. CT colonoscopy (CTC) has been slow to take off in recent years, but the publication of the ACRIN trial2 led many to believe that coverage was imminent. The ACRIN study provided the first large data sample (5,000 patients) to show that virtual colonoscopy stands up well as a screening procedure next to standard optical colonoscopy. Nonetheless, those hopes were dashed when CMS declined to cover screening CTC in its proposed national coverage decision released in February. A few caveats: First, this may be the most operator-dependent exam in imaging today, and physicians who are very well trained are needed to replicate the ACRIN results. Second, there is still some patient misalignment here as to the needed prep and experience of a CTC exam, and distributing that knowledge is important. Third, providers should assess what virtual colonoscopy will do to scanner capacity. No volume onslaught is expected, even if CMS coverage comes through, but as volumes start to mount, it will become troublesome that these exams take longer than standard diagnostic CT. Reimbursement is not going to make up for losing the time to perform three or four more diagnostic scans for each CTC study. For portable CT, resurgence is being seen both in hospitals’ interest and in vendors’ consideration of offering technology. Our Technology Insights group has seen an upswing in hospitals looking for portable CT, whether for the operating room or, more commonly, for the ICU. There have been few offerings; a head-only portable CT unit has seen reasonable success in the ICU environment, and installations are increasing. A unit from the 1990s commands good money on the used market, and there is a CT suite available for the operating room, but at more than $1 million, its cost is comparable to that of an intraoperative MRI unit. A potentially exciting development is a work in progress: a full-body portable unit that could be the most versatile portable CT scanner available is anticipated within 18 months. 3T MRI MRI is the modality hit most acutely by the DRA, combining deep cuts in reimbursement and volumes important to imaging centers and physician offices. Due to high costs and longer replacement cycles, MRI is also expected to be hardest hit by any upcoming capital crunch. Therefore, vendors are focusing on efficiency, workflow, and increasing user friendliness. Continued upgrades in coil technology across vendors are allowing multiple scans with fewer coil switch-outs. One vendor now offers a scanner with switch-out tables, so that the patient can be prepped outside the scanning suite and wheeled in while the other patient is being wheeled out, potentially increasing throughput. Again, there is pressure on vendors to develop lower-cost workhorse units. Several vendors showed a bare-bones 1.5T unit priced at less than $1 million (a milestone). Another offered a reasonable upgrade package with which paying slightly more for a 1.5T unit that is basically a tuned-down 3T unit buys future 3T capability with just two days of downtime, at a modest cost. The pinch on MRI dollars has not softened demand for 3T scanners. Vendors are seeing a higher percentage of purchases involving 3T scanners. It took 10 years, but it is finally prime time for 3T. Nonetheless, this is not a required purchase, as 1.5T remains the standard of care for all but the most advanced neuroimaging. There is a clear benefit for all neurological imaging, but more and more other areas (such as pediatric, abdominal, and prostate imaging) also show a clear clinical advantage for 3T over 1.5T. Breast imaging, which currently accounts for 3% of all MRI procedures (Figure 1, page 36), may be the next area where it is important to have 3T. This may not happens in the next year or two, but early studies3 are showing that 3T MRI has better specificity (the Achilles heel of breast MRI). Even in cardiac imaging, some data are being developed showing some benefits at 3T. More important, the floor has risen: the full portfolio of exams can be done on 3T MRI, and it can be a workhorse scanner in a stand-alone environment, although issues in spine imaging are still being seen. Vendors are expanding their options in 3T, increasing bore size, making a more patient-friendly platform; prices have also reached a plateau at slightly less than $2 million. Looking hard at high-field open scanners that give 1.5T-quality images with a more open platform is also recommended. Not only do these scanners provide a more comfortable patient experience, they allow better access for procedures and for pediatric patients. A number of hospitals have been able to reduce the number of pediatric patients requiring sedation. Open units also have higher table weights, allowing improved access for obese patients. Patients like the experience of an open scanner, and the high-field open scanners give physicians the images they want as well; they are one of the few patient-marketable purchases in MRI. Media attention to nephrogenic systemic fibrosis and MRI contrast over the past year has been significant. Although the number of reported cases seems to have diminished, there is a clear need to reduce the amount of contrast used, especially for angiography, where three or four times the recommended dose was commonly used. Vendors are working to develop noncontrast angiography. Some niche MRI scanners also deserve recognition. For hospitals, it makes sense to own a scanner with the biggest bang for the buck. Some breast centers can make a case for dedicated breast MRI if they have enough volume, but data for other niche scanners like upright MRI are really equivocal. PET/CT PET/CT is the standard of care for oncology imaging, particularly for staging of a large majority of cancers; tumor imaging accounted for 93% of all PET/CT in 2007 (Figure 2). The National Oncologic PET Registry shared data early in 2008 that showed a change in clinical management after PET imaging for 38% of cases. An Australian registry has replicated those results. Other work showed that PET/CT does a much better job than thoracic CT of detecting lung lesions. The data are there: you need access to PET/CT if you want to have an oncology program. Despite the development of cardiac and neurological PET, oncology is going to continue to account for 90% or more of all studies performed using PET/CT over the next decade. PET/CT, however, is an area where reimbursement has been precarious. CMS continued to chip away at coverage for 2009. Reimbursement for oncology studies went down by about $20 per scan. Cardiac PET coverage has seesawed like no other exam has during the past four years, and it went down about another 20% for 2009. Because it is necessary to cover the $30,000-plus monthly cost of an on-site rubidium generator, myocardial PET remains an expensive business. Development in PET/CT technology has been somewhat stagnant, with one interesting exception: a new entry that purports to decrease PET scanning time significantly. People have long discussed using the excess downtime on PET/CT units to perform diagnostic CT scans, but operationally, it’s not simple. This new platform reportedly completes a whole-body oncology survey in five minutes. If that is true, then it will actually be possible to integrate PET and CT workflows. The scanner is available in 40-, 64-, and 128-slice configurations, so a site with a very low CT-slice capability could bring this in as PET/CT, but also as high-end CT with cardiac capabilities. Hybrid SPECT/CT systems have lagged behind PET/CT in development. Nonetheless, new SPECT and SPECT/CT systems were on display, with greater focus on noncardiac applications. SPECT/CT may be solidifying its role in the evaluation of some bone and infectious lesions, and other oncology-based applications are developing as well, though this terrain may be largely occupied by PET/CT over the next decade. SPECT/CT remains a nice-to-have, not must-have, platform. What it needs is a high-volume exam where it makes a substantial clinical impact over standard nuclear imaging, as PET/CT did for oncology imaging. SPECT/CT is still searching for the study that makes it a necessity. RSNA 2008 did not bring MRI/PET any closer to commercial readiness. Although clinicians are able to fuse MRI and PET performed separately, commercially available clinical PET/MRI (still only discussed as a head-only unit) remains three to five years away. Architectural issues associated with incorporating a PET ring into an MRI magnet historically have confounded development. One vendor introduced a prototype scanner that is currently used exclusively for investigational purposes. Women’s Imaging Now considered the gold standard for breast imaging, digital mammography may be eclipsed by breast tomosynthesis in just a few years. Currently, digital mammography is at the forefront of replacement discussions as institutions decide whether to pursue replacement of analog systems with full-field digital mammography (FFDM) scanners or to get lower-cost access to digital exams with CR mammography technology (Figure 3). At RSNA 2008, vendors showcased their flagship FFDM systems, including new platforms that will serve as the basis for future breast tomosynthesis upgrades. Indeed, as institutions consider additional digital mammography units, more are considering the market timing of future technological developments, particularly breast tomosynthesis, which provides multiple slice-like views of the breast, allowing better visualization of many breast lesions. The first commercially available product is expected to receive FDA approval in early to mid-2009. An early clinical study4 indicates that tomosynthesis demonstrates superior sensitivity, compared with conventional 2D mammography, and it is also associated with a decreased recall rate. While vendors are promising that top-of-the-line FFDM units will be upgradable for tomosynthesis, this upgrade will probably still require a full-unit switch-out, or forklift upgrade. CR mammography remains an attractive alternative, allowing hospitals access to digital technology for around half of the cost of a DR system. In April 2008, the FDA approved computer-assisted detection for CR systems, resulting in a sales boost for the one approved system. This vendor hopes to capitalize on the success of its CR platform by developing a DR FFDM solution as well, which could result in further price erosion. Second-line breast imaging exams like breast MRI and ultrasound have a well-established presence, and with recent competition from positron-emission mammography (PEM) and breast-specific gamma imaging (BSGI), the field promises to become even more diverse. PEM uses contrast enhancement with radioactive FDG to assist in the detection of lesions of less than 1 cm in diameter, and is especially effective in the detection of ductal carcinoma in situ and atypical ductal hyperplasia. BSGI is capable of detecting early-stage tumors and offers differentiation of cancerous versus benign tumors. Two vendors of PEM and BSGI displayed mobile versions of their technologies. The major PEM vendor recently received 510(k) clearance for PET-guided biopsy on the PEM platform, adding further versatility to the modality and eliminating a key disadvantage of the unit. Automated breast ultrasound (ABUS) is not yet gaining momentum. Unlike conventional handheld ultrasound, ABUS standardizes image acquisition and limits variability among operators. Such automated protocols consistently acquire a specific set of 3D volumetric representation, minimizing user error and maintaining quality standards across time and patients. While early results indicate performance equivalent to that of handheld ultrasound, further research will be required to demonstrate (potentially) superior performance. The leading vendor of ABUS has had market exclusivity to date, but one major multimodality vendor introduced a system at RSNA 2008, though its FDA clearance is still pending. Ultrasound Ultrasound introductions added to rejuvenation of this modality, although it is necessary to spend $600,000 to obtain all of the features that enable the modality to compete more effectively with CT and MRI. Vendors continued to update 3D and 4D models at RSNA 2008; 3D ultrasound and volumetric imaging offer complete assessment of morphology by stacking 2D cross-sections of the volume of interest. Currently, 3D ultrasound is best suited for obstetric and echocardiography studies, but increasing demand to view anatomical structures clearly, with more measurement parameters, is driving a continual shift away from 2D ultrasound toward 3D and volumetric imaging for other applications. Postprocessing software enhancements are allowing more detailed tissue characterization while simultaneously reducing operator dependence. DR Digital imaging remains the primary focus for vendors of radiography and radiography/fluoroscopy systems, and portability is also becoming an area of interest. Several vendors are developing new portable plates that are lightweight and durable enough to sustain routine daily use in high-volume settings. Moreover, vendors are introducing more comprehensive and versatile DR room solutions that combine multiple radiographic modalities. DR vendors also are focusing on improving workflow and reducing the steps that technologists need to take to acquire images and position patients. Auto-alignment, auto-image stitching, and tracking software are now becoming routine, as are user-friendly interfaces that are designed to enable hospitals with older systems to improve workflow by replacing one or two rooms with one new system. Two technologies on display this year, DR tomosynthesis and lung computer-aided detection, show promise in their ability potentially to improve the detection of lung cancer and lower the cost of detection. Tomosynthesis acquires 3D images using DR technology, and recent work suggests that tomosynthesis is far superior to DR alone and significantly less costly than multislice CT for the detection of pulmonary nodules. In addition, computer-aided detection with CT has the ability to detect nodules and reduce reader variation (a problem area in CT detection of lesions), even in a low-dose setting. These developments could have a potentially large impact on the way that pulmonary nodules are detected and on radiation dose, as well as reducing the cost of lung-cancer detection. Outlook for 2009 While molecular imaging is expected to play a significant future role, its near-term clinical adoption is not expected within the decade. That said, future applications include improved lesion detection and the ability to deliver targeted therapeutics. First, for 2009, expect continued price erosion across modalities. As next-generation products come to market and as broader economic issues continue to constrain capital budgets, acquisition prices will continue to erode. Second, breast tomosynthesis will become a commercial reality. With the ability to improve upon conventional mammography, tomosynthesis could prove to be a disruptive technology, and centers intending to acquire digital mammography platforms should consider their path to tomosynthesis as they make this purchase Third, look for the continued refinement—and growing adoption—of advanced CCTA applications. As clinical data continue to solidify the role of the exam, more institutions are expected to deploy CCTA in the emergency department, despite the low likelihood of broader Medicare coverage this year. Fourth, 3T MRI will further approach workhorse status through increased versatility, although the attractive prices for 1.5T scanners make it a hard sell this year. Fifth, moving forward, increasing efficiency and getting more patients through each scanner will be paramount. MRI will represent one of the primary targets of these efforts, bolstered by increased automation in each vendor’s MRI platform. Sixth, look for portable CT offerings—now catering to specific applications such as neuroimaging—to increase the versatility of their platforms with whole-body units. Additional Reading - Is Custom Reporting Difficult?