For decades, children and adolescents diagnosed with rhabdomyosarcoma have often faced an additional invasive procedure as physicians determine how far their cancer has spread: bone marrow aspiration and biopsy. A new international study suggests that may no longer be necessary for many patients.
Researchers evaluating FDG PET/CT and PET/MRI in rhabdomyosarcoma found that PET was highly effective at ruling out bone marrow involvement when the primary tumor was FDG-avid. Among 301 patients with FDG-avid primary tumors, PET achieved 98% sensitivity and a 100% negative predictive value for detecting bone marrow metastases when traditional bone marrow sampling was used as the diagnostic gold standard.
The implications extend beyond better imaging. The findings suggest molecular imaging could eliminate an invasive procedure from routine cancer staging for a substantial group of patients while potentially providing physicians with a more complete picture of disease throughout the body.
Historically, bilateral bone marrow aspirates and biopsies taken from the iliac crests have been considered the gold standard for detecting marrow involvement in rhabdomyosarcoma. The new study, published in the Journal of Clinical Oncology, challenges that assumption.
Researchers examined patients recruited across Europe and North America through an international collaboration facilitated by the International Soft Tissue Sarcoma Consortium, or INSTRuCT. Among 301 patients with FDG-avid rhabdomyosarcoma whose PET imaging and bone marrow sampling were performed before chemotherapy and tumor resection, marrow metastases were detected by both PET and bone marrow sampling in 54 patients. PET identified another 24 patients with apparent marrow disease that was not detected through bone marrow sampling, while bone marrow sampling identified disease missed by PET in just one patient.
Overall, PET demonstrated 98% sensitivity and 90% specificity in this population, while its negative predictive value reached 100%. That last number may be the most important because it suggests that when PET shows no evidence of FDG-avid marrow disease, physicians may be able to confidently avoid routine bone marrow sampling.
The researchers concluded that bone marrow sampling "may be omitted" in patients without evidence of FDG-avid bone marrow disease on PET. When suspicious FDG-avid marrow lesions are present, the authors said additional investigation with MRI and/or biopsy should be considered.
There is another reason the findings matter. A bone marrow biopsy samples a relatively small area of the body, while PET can examine disease distribution throughout much of the skeleton. That distinction becomes important when marrow metastases are not uniformly distributed.
Among 79 patients with FDG-avid marrow disease in the study, 12 had only one FDG-avid marrow focus and another 15 had between two and five. In other words, roughly one-third of patients with PET-detected marrow disease had relatively patchy involvement, creating an obvious limitation for sampling bone marrow from predetermined locations such as the iliac crests.
The researchers illustrated that problem with patients who had FDG-avid lesions in bones outside the iliac crests but negative iliac bone marrow sampling. A biopsy can therefore be technically successful and still miss disease located somewhere else, while PET can show physicians where suspicious disease is actually located.
That could potentially change the role of biopsy itself. Instead of routinely performing bilateral marrow biopsies in nearly everyone, imaging could identify which patients need additional tissue sampling and potentially help physicians determine where that sampling should occur.
Perhaps the strongest conclusion in the study goes beyond whether some patients can avoid biopsies. The researchers questioned whether bone marrow aspiration and biopsy should continue to be considered the diagnostic gold standard at all.
Diagnostic tests are normally evaluated by comparing them against an established reference standard. In this study, PET initially was measured against bone marrow sampling, but that creates a problem if the supposed gold standard itself misses disease.
The researchers therefore recalculated diagnostic performance using evidence of marrow disease from either PET or bone marrow sampling as the reference. Under that analysis, PET reached 99% sensitivity and a 100% negative predictive value in patients with FDG-avid primary tumors. Bone marrow sampling performed substantially worse, with sensitivity calculated at 70% and a negative predictive value of 90%.
The investigators consequently proposed that tumor detection through bone marrow aspiration and biopsy should no longer be considered the gold standard for diagnosing marrow disease in rhabdomyosarcoma. That is a considerably bigger claim than simply saying PET works well: it suggests molecular imaging may provide a better way to understand the geographic distribution of metastatic disease than sampling a small portion of the body.
There is also a straightforward patient benefit. Rhabdomyosarcoma is the most common soft-tissue sarcoma in children and adolescents, and bone marrow sampling is invasive, can involve pain and stress, and may require anesthesia in pediatric patients.
The researchers noted that marrow sampling can also carry risks including bleeding, infection and anesthesia-related complications, while scheduling, performing and interpreting the procedure may potentially delay treatment. If PET imaging already being performed for cancer staging can reliably exclude marrow involvement, some of that burden could potentially disappear.
The study therefore represents an important example of nuclear medicine creating value not simply by adding another diagnostic test, but potentially by removing another procedure from the patient journey.
The findings do not mean bone marrow biopsies are about to disappear from rhabdomyosarcoma staging altogether. Timing matters, as does the FDG avidity of the primary tumor and the extent of PET coverage.
When PET was performed after chemotherapy had begun, sensitivity for detecting marrow disease fell to 86%. When PET occurred after removal of the primary tumor, sensitivity fell to 60%. Approximately 2.2% of primary tumors in the main study cohort were also not FDG-avid, meaning PET could not reliably be used to draw conclusions about metastatic disease in those cases.
Whole-body imaging is important as well. The researchers noted that 22 of the 313 primary tumors in the main cohort were located at or below the knees and could have been missed using more limited skull-base-to-thigh imaging. The authors therefore recommend that PET be performed before treatment, cover the entire body and confirm that the primary tumor is FDG-avid. When those conditions are not met, traditional bone marrow sampling may still be necessary.
There was also one particularly important exception: one patient with an FDG-avid primary tumor had marrow disease detected by bone marrow sampling but not PET. That patient had no other metastases and ultimately died of the disease. The study is retrospective, and the authors acknowledge additional limitations, meaning its findings will need to be considered carefully as clinicians determine whether and how staging guidelines should change.
The broader nuclear medicine story may be more important than rhabdomyosarcoma alone. Much of the economic argument for advanced medical imaging has historically focused on detecting disease earlier, staging patients more accurately or determining whether treatments are working, but there is another potential source of value: eliminating procedures that no longer need to be performed.
If a molecular imaging study can provide information that previously required an invasive procedure, its value is not simply the diagnostic information contained in the scan. It can potentially include avoided procedures, anesthesia, pathology, scheduling, complications and patient burden.
That changes the economics of imaging and illustrates why comparisons between the cost of an advanced scan and another diagnostic test can miss the larger picture. The relevant question increasingly may be what happens throughout the patient's diagnostic pathway because the scan was performed.
For rhabdomyosarcoma, that possibility is becoming tangible. An international study involving hundreds of patients has found that properly performed FDG PET can identify marrow disease with extremely high sensitivity in patients with FDG-avid tumors. More provocatively, researchers found that the traditional biopsy-based gold standard may itself be missing disease that whole-body molecular imaging can see.
Nuclear medicine has spent decades demonstrating that it can find disease. Its next value proposition may increasingly be demonstrating which invasive procedures patients no longer need.
