Radiographic Versus Pathology-Integrated Response for Assessing Optimal Surgical Timing After Neoadjuvant Imatinib in Patients With Locally Advanced KIT Exon 11-Mutant GIST.

OBJECTIVE

To compare radiographic and pathology-integrated metrics for defining near-maximal treatment effect after neoadjuvant imatinib in KIT exon 11-mutant gastrointestinal stromal tumor (GIST).

SUMMARY OF BACKGROUND DATA: Operative timing after neoadjuvant imatinib is usually guided by radiographic shrinkage, although dimensional response may incompletely reflect biological treatment effect.

METHODS

We retrospectively analyzed 131 patients with locally advanced KIT exon 11-mutant GIST treated with neoadjuvant imatinib followed by curative-intent resection at 2 tertiary sarcoma centers in the United States and Italy. Radiographic response was assessed across 2-month intervals using RECIST and percent tumor shrinkage. The primary timing analysis identified the earliest interval reaching 90% of peak median response within 3 to 22 months. The same framework was applied to a pathology-integrated response score (PIRS) derived from tumor shrinkage, viable tumor percentage, and necrosis.

RESULTS

Median age was 62 years, and 60.3% of patients were male. By RECIST, 45.8% achieved partial response and 48.9% had stable disease. Radiographic response reached a near-maximal interval at >4 to 6 months, whereas PIRS reached a near-maximal interval at >10 to 12 months. PIRS discriminated across duration groups better than RECIST ( P =0.027 vs. P =0.13). Forty patients (30.5%) with stable disease had major or near-complete PIRS. Overall survival differed across PIRS categories, whereas recurrence-free survival did not differ across PIRS or RECIST categories.

CONCLUSIONS

Radiographic and pathology-integrated response identified different windows of near-maximal treatment effect, suggesting that size reduction alone may underestimate biological response. The later pathology-integrated response window was not accompanied by differences in recurrence-free survival.