Whole-body MRI (WB-MRI) is more efficient and less costly than standard methods used to stage colorectal cancer and non–small cell lung cancer (NSCLC), and accuracy is similar, according to new findings.
For colorectal cancer, the time it took to complete diagnostic tests was reduced from an average of 13 days using standard multimodal diagnostics to an average of 8 days using WB-MRI. For NSCLC, the time was reduced from 19 days to 13 days.
Importantly, for both cancers, WB-MRI did not differ significantly from standard tests in diagnostic sensitivity and specificity.
"We found the accuracy of WB-MRI was not significantly different from standard multiscan staging pathways, but WB-MRI pathways were quicker and cheaper in an NHS [National Health Service] setting," said Stuart Taylor, MD, from University College London, United Kingdom. "We therefore conclude that WB-MRI is a viable first-line staging test, and the studies will be carefully considered by expert bodies such as NICE [National Institute for Health Care and Excellence] who issue guidance on NHS practice."
Taylor, who was lead author on both studies, told Medscape Medical News that the radiologists reporting on WB-MRI were representative of those who would "report it were it to be more widely disseminated, as we did not use a handful of very expert radiologists who would not be representative of usual clinical care."
The study of the use of WB-MRI for colorectal cancer was published online May 9 in the Lancet Gastroenterology and Hepatology. The study of WB-MRI for NSCLC was published May 9 in the Lancet Respiratory Medicine.
Accurate staging is imperative to ensure optimal patient outcomes and particularly to identify metastatic disease so as to determine the therapeutic strategy. But staging pathways are complex and rely on high-technology imaging platforms, such as CT, positron-emission tomographic CT (PET-CT), and MRI. In England, where both studies were conducted, the NICE publishes guidelines that indicate multiple sequential imaging tests for staging and for making decisions regarding initial treatment.
The same investigators conducted both studies, which were parallel prospective multicenter trials designed to directly compare the diagnostic accuracy and efficiency of WB-MRI-based staging pathways with the current standard in NSCLC (Streamline L) and colorectal cancer (Streamline C).
Colorectal Cancer
The Streamline C trial was conducted in 16 hospitals in England; 299 patients completed the study. Of this group, 68 (23%) had metastasis at baseline. The protocol was for patients to undergo WB-MRI, with the result being withheld until standard staging was completed and the first treatment decision was made.
The healthcare providers then recorded their treatment decision first on the basis of standard investigations, and then on the basis of the results of the WB-MRI staging pathway, and finally on the basis of all tests. The primary outcome was the difference in sensitivity for metastases between the standard staging pathway and the WB-MRI staging pathway with regard to a consensus reference standard at 12 months.
Secondary endpoints included the difference in per-patient specificity for metastatic disease detection between standard and WB-MRI staging pathways, differences in treatment decisions, staging efficiency (time taken, test number, and costs), and per-organ sensitivity.
Staging sensitivity for patients with metastatic disease was 67% for WB-MRI and 63% for standard pathways (4% difference). For the primary outcome, three perceptual errors were noted in the WB-MRI pathway and six in the standard pathway.
Specificity did not differ between the WB-MRI and standard pathways (P = .48). Sensitivity analysis showed no significant differences between WB-MRI and standard staging when lesions that were reported as equivocal were treated as either all positive or all negative or across individual organ sites.
For the WB-MRI pathway, there was 54% agreement for T stage, vs 60% for the standard staging, which was a nonsignificant difference of 6%. Agreement did not significantly differ for N stage between the pathways. Agreement for the final treatment decisions was 96% for WB-MRI and 95% for the standard pathway.
The time to staging completion was shorter for WB-MRI (median, 8 days) than for the standard pathway (13 days). In addition, WB-MRI staging required fewer tests (median, one) as compared to the standard pathway (two).
The cost of staging was also less for WB-MRI: £216 (US$281) vs £285 (US$371).
NSCLC Study
For the NSCLC study, the design, methodology, and endpoints were the same as for the colorectal trial; 187 patients completed the trial. Of this group, 52 patients (28%) had metastasis at baseline, and 137 patients (73%) had disease of stage T2 or above.
The staging sensitivity for patients with metastatic disease was 50% for WB-MRI and 54% for the standard pathway (difference of 4%, P = .73). For the study's primary endpoint, seven perceptual errors occurred in the WB-MRI pathway, and three occurred in the standard pathway.
Similar to colorectal cancer staging, specificity did not differ between the WB-MRI pathway (93%) and the standard pathway (95%). The sensitivity analysis showed no differences between the two pathways when lesions that were reported as equivocal were treated as either all positive or all negative, as well as across individual organ sites.
However, agreement for N stages was 65% for WB-MRI, vs 75% for the standard pathway, which was a significant difference of 10%. There were no significant differences in agreement for T stage between pathways.
Agreement with the final treatment decision was 98% for WB-MRI and 99% for the standard pathway. For therapeutic decisions based on WB-MRI and standard pathways, levels of agreement were similar to the retrospective consensus panel optimal treatment decision.
Standard staging involved 302 individual investigations; WB-MRI involved 232. The median number of tests did not differ between the WB-MRI and the standard pathway (one vs 0). The time to staging was shorter for WB-MRI (13 days vs 19 days), and the mean per-patient costs were also lower: £317 (US$412) vs £620 (US$806).
Some Caveats
Several experts weighed in on these articles.
In an editorial that accompaned the colorectal cancer article, Andreas Schreyer, MD, from Brandenburg University Medical School, Germany, noted that the study only represents the English NHS with regard to its national recommendations and cost calculations.
Also, the real-world approach of the study highlighted a key problem: for eight of the 16 recruitment sites, there was no local radiologic infrastructure to perform WB-MRI, and so WB-MRI had to be conducted at a different hospital.
"Therefore, the infrastructure based on availability and technical expertise of advanced MRI scanners will be a fundamental prerequisite to changing to new, more efficient paradigms and diagnostic pathways," Schreyer writes. "MRI has faced considerable backlash within the medical community due to relatively high costs and the problems involved in finding a timely slot for imaging because of the high demand for this method. This is why it is particularly important to think outside the box and look out for new medical pathways and paradigms and not to be driven by prejudices."
Scott Regenbogen, MD, MPH, an associate professor of surgery and chief of the Division of Colorectal Surgery at the University of Michigan, Ann Arbor, emphasized that the differences in cost between MRI and CT would be much higher in the United States. "The difference in cost would be substantial here," he said, "and may be higher in health systems in other countries as well."
He also questioned how many patients need an MRI. "In colon cancer, staging is usually done with a CT scan, and sometimes you may need to get a second test for more detail," he told Medscape Medical News. "For example, you may see something suspicious in the liver but can't tell if it's a benign lesion or metastasis, so an MRI is needed."
In the study, colon cancer and rectal cancer were not distinguished, and Regenbogen explained that the staging protocols for the two differ. "In rectal cancer, we use MRI routinely for staging and for making treatment decisions," he said. "If you take rectal cancers out of the equation, and they had just looked at colon cancers — how many colon cancers needed a second study? Probably not that many, and if they do, usually just an MRI and not the additional scans they mention."
He also emphasized that timing would vary in different systems and would largely depend on the availability of MRIs. "We have a good supply of CT scans and limited MRI, and we would be seeing really long delays if we were doing a whole-body scan on every cancer patient," Regenbogen said.
At his institution, which is a well-resourced academic center, the MRI runs 24 hours a day so as to accommodate all the patients who need one. "It takes 45 minutes to do a rectal cancer staging alone, and I don't know how long it would take to do a full body scan," he said. "These two papers are only looking at colorectal and lung cancer, but that in itself is a lot patients to scan."
More Questions
In an editorial that accompanied the NSCLC study, Mathias Meyer, MD, from Duke University Medical Center, Durham, North Carolina, and Johannes Budjan, MD, from the University Medical Center Mannheim, Heidelberg University, Germany, note that although the Streamline L study is a step in the right direction, caveats and questions need to be addressed. One is that baseline staging of occult metastatic disease is problematic, given the low sensitivity in both pathways for metastatic staging. In WB-MRI, it was 50%, compared to 54% for the standard pathway.
"A high sensitivity is crucial, especially in patients undergoing curative surgery," they write.
Another problem is that the study results underscore the inadequacy of all current imaging modalities for lymph node staging in NSCLC. Worse results were seen for WB-MRI compared to standard imaging, with agreement of 65% vs 75%.
In addition, the editorialists point out that there are contraindications to WB-MRI. These include implanted devices, claustrophobia, and, depending on the system, obesity. All of these can restrict universal usage, although "current developments in medical engineering are aiming to reduce those barriers to allow greater access to MRI."
There is also concern about the use of gadolinium-based contrast agents in the Streamline L study, in light of the discovery that gadolinium is retained in the brain.
"Although WB-MRI might be the imaging modality we have been hoping for in NSCLC staging, occult metastases and lymph node staging remain challenging by non-invasive imaging," they conclude. "Further research is needed to improve the diagnostic performance of non-invasive imaging, preferably a one-stop-shop approach, which allows time-efficient and accurate treatment decisions."
Future Considerations
University College London's Taylor noted that costs will vary between different healthcare systems, but in general, the ratio between test costs — such as PET-CT and WB-MRI — is similar in many settings, so the overall message should hold true. "We do also provide a full list of all staging scans patients underwent so readers can calculate the costs of both pathways in their own healthcare settings to see if our findings would be replicated," he said.
He pointed out that nearly all standard MRI machines can perform good-quality WB-MRI, so the issue is not the specification of the scanner but rather the availability and capacity of existing scanners. "To roll out WB-MRI more widely will need investment in more MRI scanners and/or prioritization of access to existing scanners," Taylor explained. "As our cost data show, in fact, in the long term, this investment would reduce the costs of staging patients with these cancers. The issue of MRI scanner capacity will differ between countries."
The Streamline trials were funded by the National Institute for Health Research (NIHR). Taylor has received consulting fees from Robarts Plc, and he and several coauthors have received support from the NIHR or the Wellcome Trust. Meyer reports receiving speaking fees from Siemens Healthineers and institutional grants from Siemens Healthineers and Bracco Imaging outside of the submitted work. Budjan reports receiving speaking fees from Siemens Healthineers and Ferring Pharmaceuticals outside of the submitted work. Schreyer has disclosed no relevant financial relationships.
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου