Prediction of chemotherapy response by PET-CT in osteosarcoma: correlation with histologic necrosis.

BACKGROUND

The current standard for neoadjuvant chemotherapy (NACT) response evaluation in osteosarcoma is histopathologic necrosis (HN). However, it is accessible only after NACT completion and may get affected by confounding factors. Thus, noninvasive surrogate such as (18)Fluorine-Fluorodeoxyglucose-positron emission tomography-computerized tomography (PET-CT) scan would be useful to individualize therapy.

METHODS

Thirty-one treatment naive osteosarcoma patients evaluated prospectively by PET-CT scan preceding and after 3 cycles of NACT and surgery during 2006 to 2008. Various anatomic and metabolic parameters of PET-CT scan were compared with HN (good response ≥90% HN) as reference standard. Receiver operating characteristic curves were generated to assess the best threshold and predictability.

RESULTS

Median age was 17 years; 25 were male patients and 23 were nonmetastatic. Ten cases were good, whereas 21 cases were poor histologic responders. PET-CT parameters including post-NACT (2) and pre-NACT (1) standard uptake value (SUV)max ratio (SUV2:SUV1), SUV2, pre-NACT and post-NACT volumes (V1and V2), change in V after NACT, pre-NACT and post-NACT metabolic burden (MB) and change in MB after NACT correlated with HN. Two independent predictors were identified in stepwise multivariable analysis; if V1 ≤300 mL and SUV2:SUV1 ≤0.48, observed good histologic response proportions was 83%, whereas if V1 >300 mL and SUV2:SUV1 >0.48, it was 0%.

CONCLUSIONS

NACT response can be predicted reliably by PET-CT scan early in disease course (even at baseline) and PET-CT parameters correlate well with HN. MB seems to be sensitive substitute for response evaluation. Independent predictors may have wider clinical applications if further validation can be done.