Whole-Body MRI Could Streamline Cancer Staging

By Diana Swift

Whole-body magnetic resonance imaging (WB-MRI) staging had accuracy similar to that of current standard staging pathways for lung and colon cancers, and resulted in the same treatment decisions, two British studies showed.

WB-MRI was also more efficient and reduced the number of tests, the time to complete staging, and costs, and therefore may be a viable alternative for staging in routine clinical practice, the investigators said. Compared with standard imaging, WB-MRI reduced the mean time to determining tumor size and spread by 6 days in non-small cell lung cancer (NSCLC) patients and 5 days in colorectal cancer (CRC) patients.

The studies, conducted at 16 English hospitals and published simultaneously in the Lancet Respiratory Medicine and Lancet Gastroenterology & Hepatologysuggest that WB-MRI may be a pragmatic alternative to other tests, said Stuart A. Taylor, MD, of the Centre for Medical Imaging at University College London, and colleagues.

The two trials, called Streamline C and Streamline L, found that the WB-MRI results were as accurate as those for computed tomography, positron emission tomography, site-focused MRI, and ultrasound. Per-patient costs in the U.K. National Health Service (NHS) setting fell by nearly 25% for CRC and almost 50% for NSCLC.

The sensitivity and specificity of diagnosis with WB-MRI were similar to those with standard multimodal tests for both cancers. But WB-MRI reduced the time of complete diagnostic tests from an average of 13 days to an average of 8 days in the CRC trial and from 19 to 13 days in the NSCLC trial. Costs were reduced from an average of £620 to £317 in the NSCLC cancer trial, and from an average of £285 to £216 in the CRC trial.

Moreover, in the CRC trial, agreement with the final multidisciplinary panel treatment decisions based on standard investigations and WB-MRI was similar and high (95% and 96%, respectively), as were results for the lung cancer trial (99% for standard investigations, 98% for WB-MRI).

“Our results, obtained in a real-world NHS setting, suggest that whole body MRI could be more suitable for routine clinical practice than the multiple imaging techniques recommended under current guidelines,” said Taylor in a Lancet news release. “While demands on NHS MRI scanners are currently high, MRI can image the whole body in 1 hour or less. Adopting whole-body MRI more widely could save rather than increase costs, as well as reduce the time before a patient’s treatment can begin.”

The researchers conceded, however, that MRI scanners are not as widely available as other imaging technologies are, and that eight of the recruitment hospitals had to send patients to other centers.

Lung Cancer

From February 2013 to September 2016, the Streamline L trial recruited 353 eligible patients with newly diagnosed NSCLC at 16 English hospitals. The final cohort consisted of 187 patients — 37% women, with a mean age of 67 — with 73% of the cases determined to be stage T2 or above.

Patients underwent WB-MRI, with results from this modality withheld until the completion of standard staging tests. For each patient, a multidisciplinary team made a treatment decision based on standard investigations, then on the WB-MRI staging pathway (WB-MRI plus additional tests generated), and finally on all tests.

Pathway sensitivity was 50% for WB-MRI (95% CI 37-63) and 54% for standard pathways (95% CI 41-67), a non-significant 4% difference. No adverse events related to imaging were reported. Specificity was the same in the two pathways: 93% (95% CIs 88-96 and 91-98, respectively, P=0.45). Agreement with the multidisciplinary team’s final treatment decision was 98% for WB-MRI and 99% for standard tests.

In an accompanying commentary, Mathias Meyer, MD, and Johannes Budjan, MD, both of the University Medical Center Mannheim at Heidelberg University in Germany, called Streamline L “a step in the right direction” but offered several caveats: First, occult metastatic disease at baseline staging remains an issue, given the low sensitivity of both investigative pathways for metastatic staging: WB-MRI 50% versus 54% for standard care. “A high sensitivity is crucial, especially in patients undergoing curative surgery,” Meyer and Budjan emphasized.

In addition, they said, the results of Streamline L underscore the insufficiency of all current imaging modalities for lymph node staging in NSCLC, with worse results for WB-MRI versus standard imaging, showing agreement of 65% versus 75%. Furthermore, there are certain contraindications to WB-MRI such as implanted devices, patient claustrophobia, and obesity, which would restrict universal usage. Another concern is neuronal gadolinium toxicity after MRI, Meyer and Budjan wrote.

“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,” the commentators wrote, calling for further research to improve diagnostic performance.

Study limitations, Taylor and co-authors said, included certain patient exclusions, such as pregnancy and severe systemic disease, and the high withdrawal rate in the study. In addition, the study’s waiting times for WB-MRI might not be representative of those at other hospitals and in other countries, where longer waits might delay treatment. In addition, bias may have crept in, since WB-MRI treatment decisions were made after the results of standard tests were unmasked to the multidisciplinary team, and the findings may not be relevant to non-NSCLC tumors or to costs in healthcare systems other than the NHS.

Colorectal Cancer

Taylor and colleagues also recruited 370 newly diagnosed CRC patients during March 2013 to August 2016 and tested them using a similar protocol and masking procedures to those in the lung cancer group. Of the 299 patients completing the Streamline C trial, 65% were male, and median age was 65; 96% of the patients had stage T2 or higher cancers, and 23% had baseline metastases.

Comparing the two approaches, the researchers found sensitivities of 67% (95% CI 56-78) for WB-MRI and 63% (95% CI 51-74) for standard pathways, for a difference of 4% (-5 to 13, P=0.51), and no adverse imaging-related events. Specificity was comparable: 95% for WB-MRI (95% CI 92-97) and 93% for standard pathways (95% CI 90-96, P=0.48). Agreement with the multidisciplinary team’s final treatment decision was 96% for WB-MRI and 95% for standard modalities.

In an accompanying commentary, Andreas G. Schreyer, MD, MBA, of Brandenburg University Hospital in Germany, noted that the availability of advanced MRI scanners and expert personnel will be fundamental to implementing more efficient new diagnostic paradigms and pathways. But overall, “the study gives an important evidence-based argument and shows that new diagnostic imaging approaches and techniques, which are often thought of as more expensive and complex at first sight, can eventually change clinical pathways while being more time-efficient and cost-efficient,” Schreyer wrote.

Study limitations, Taylor and co-authors said, included caveats similar to those of the Streamline L trial, stemming from patient exclusions, possible masking-related bias, and applicability of the findings to other healthcare systems and non-CRC tumors.

The Streamline trials were funded by the National Institute for Health Research (NIHR).

Taylor and several co-authors reported support from the NIHR or the Wellcome Trust; Taylor disclosed ties to Robarts Plc, and other co-authors reported ties to Guerbet and Siemens.

Meyer and Budjan reported ties to companies including Siemens Healthineers, Bracco Imaging, and Ferring Pharmaceuticals.

Schreyer reported having no competing interests.

SourceThe Lancet Respiratory Medicine