Study Suggests Superiority of Functional Imaging in Assessing Sarcoma Outcomes

A comparison of anatomic and functional imaging in the assessment of clinical outcomes in patients with Ewing sarcoma has shown that FDG-PET, assessed by PERCIST criteria, was superior in predicting clinical benefits and identifying responses.

Laurence H. Baker, DO

A comparison of anatomic and functional imaging in the assessment of clinical outcomes in patients with Ewing sarcoma has shown that FDG-PET, assessed by PERCIST criteria, was superior in predicting clinical benefits and identifying responses, according to a study published in the Journal of Clinical Oncology.

PERCIST criteria identified 34.8% of total patients analyzed in the response group, while anatomic imaging criteria identified significantly fewer patients in the response group, at an average of 21.7% of patients among all 4 types of criteria (volume, WHO local, WHO central, RECIST). A response rate of nearly 35% in patients with metastatic drug-refractory Ewing sarcoma is considered to be both impressive and potentially worthy of regulatory approval, according to the published results.

The imaging criteria comparison included PERCIST 1.0 for functional imaging (FDG-PET). Anatomic imaging included WHO criteria on the basis of independent assessment and on the basis of local site measurements, RECIST criteria on the basis of independent assessment, and volumetric criteria on the basis of measurements done by a central radiology group who were blinded to the PET results and clinical outcomes.

“The standard criteria are RECIST 1.1 and WHO. Those are anatomic measurements that look at either, in the case of WHO, the perpendicular diameters of lesion, or in the case RECIST, the single diameter of the sum of the unidimensional measurements on anatomic imaging,” Laurence H. Baker, DO, a professor of Medical Oncology at the University of Michigan Medical School, said in an interview with OncLive.

Of the 115 patients with Ewing sarcoma enrolled in the trial, 89 had anatomic imaging available for central review. With the use of semiautomated solid tumor segmentation software, the radiology group measured lesions from all available anatomic imaging scans at baseline and at 6 weeks.

On the basis of all these criteria, response to treatment was separated into categories of progressive disease, stable disease, and response.

The results showed that functional imaging assessment of progressive disease can be identified as early as day 9 versus at 6 weeks by using any of the anatomic imaging criteria.

Anatomic imaging criteria (volume, WHO local, WHO central, and RECIST) identified fewer patients in the response group (an average of 21.7%), whereas PERCIST identified the most patients in the response group (34.8%).

When looking at a subgroup of patients with interpretable functional imaging, 43.9% were responders according to PERCIST criteria, and 90.9% were non-progressors. Thus, in addition to having the advantage of being performed earlier in the course of a patient’s treatment, PERCIST also identified more patients with clinical response than did RECIST.

The newly defined volumetric criteria, which were selected by establishing optimal cutoffs for percentage of change in volume between baseline and week 6 that were most predictive of overall survival (OS), identified more patients in the response group (25%) than did WHO central (22.4%), WHO local (18.4%), or RECIST (21.1%). This type of criteria also identified more patients with clinical benefit from therapy. Though the time needed to assess tumors volumetrically is slightly greater than to do so uni- or bidimensionally, the process is now automated, and these results show that it may be an overall superior method of assessment of clinical response compared with the other widely used methods.

There were only slight differences found between the WHO criteria assessed locally and WHO criteria assessed by a central group of radiologists who were blinded to patient clinical status or outcome. Moreover, the added benefit of central interpretation has also been unclear in past randomized trials. However, the data generally suggest that local experts in sarcoma seem to accurately interpret anatomic imaging in the context of tumor response assessment.

Thirteen patients who received baseline anatomic imaging, but no week-6 imaging, had markedly reduced survival compared with patients who had both baseline and week-6 imaging (P <.001), with a median OS of only 1.1 months.

Prior to the current study, there is quite limited research on the ways in which common radiologic criteria, or the location of radiology reader (local vs central), compare in terms of ability to predict clinical benefit. Early trials had relied on subjective self-assessments of symptoms by patients, on the objective assessments made by investigators on the basis of radiographic shrinkage, and on physical findings to assess clinical benefit.

“The take-home message is that both volumetrics and PET are better predictors of response than the standard criteria of RECIST. This was not only in terms of response, but also overall survival,” said Baker.

Koshkin VS, Bolejack V, Schwartz LH, et al. Assessment of imaging modalities and response metrics in Ewing sarcoma: correlation with survival [published online August 29, 2016]. J Clin Oncol. doi:10.1200/JCO.2016.68.1858.

This study performed an analysis of the SARC011 trial, which was a single-arm, multicohort, multicenter, phase II study of patients with recurrent Ewing sarcoma who had been treated with an IGF1R antibody (R1507). Response was evaluated with both FDG-PET and anatomic imaging by CT or MRI, where anatomic imaging was assessed at baseline and at 6 weeks after treatment began.