Dose-response relationship between probability of pathologic tumor control and glucose metabolic rate measured with FDG PET after preoperative chemoradiotherapy in locally advanced non-small-cell lung cancer.

PURPOSE: To determine the dose-response relationship between the probability of tumor control on the basis of pathologic tumor response (pTCP) and the residual metabolic rate of glucose (MRglc) in response to preoperative chemoradiotherapy in locally advanced non-small-cell lung cancer and to define the level of residual MRglc that corresponds to pTCP 50% and pTCP > or = 95%. METHODS AND MATERIALS: Quantitative dynamic 18F-2-fluoro-2-deoxy-D-glucose (18F-FDG) positron emission tomography was performed to measure regional MRglc at the primary lesion before and 2 weeks after preoperative chemoradiotherapy in an initial group of 13 patients with locally advanced NSCLC. A simplified kinetic method was developed subsequently from the initial dynamic study and used in the subsequent 16 patients. The preoperative radiotherapy programs consisted of (1) a split course of 42 Gy in 28 fractions within a period of 28 days using a twice-daily treatment schedule for Stage IIIA(N2) NSCLC (n = 18) and (2) standard once-daily radiation schedule of 45-63 Gy in 25-35 fractions during a 5-7-week period (n = 11). The preoperative chemotherapy regimens included two cycles of cisplatin, vinblastine, and 5-fluorouracil (n = 24), cisplatin and etoposide (n = 2), and cisplatin, Taxol, and 5-fluorouracil (n = 3). Patients free of tumor progression after preoperative chemoradiotherapy underwent surgery. The degree of residual MRglc measured 2 weeks after preoperative chemoradiotherapy and 2 weeks before surgery was correlated with the pathologic tumor response. The relationship between MRglc and pTCP was modeled using logistic regression.
RESULTS: Of 32 patients entered into the study, 29 (16 men and 13 women; 30 lesions) were evaluated for the correlation between residual MRglc and pathologic tumor response. Three patients did not participate in the second study because of a steady decline in general condition. The median age was 60 years (range 42-78). One of the 29 patients had two separate lesions, and MRglc was measured in each separately. The tumor histologic types included squamous cell carcinoma (n = 9), adenocarcinoma (n = 13), large cell carcinoma (n = 6), and poorly differentiated carcinoma (n = 2). The extent of the primary and nodal disease was as follows: Stage IIB (T3N0M0), Pancoast tumor (n = 2); Stage IIIA, T2-T3N2M0 (n = 18); Stage IIIB: T1-T3N3M0 (n = 5) and T4N0M0 (n = 2); a second lesion, T1 (n = 1); and localized stump recurrence (n = 2). A pathologically complete response was obtained in 14 (47%) of the 30 lesions. The remaining 16 lesions had residual cancer. The mean baseline value of the maximal MRglc was 0.333 +/- 0.087 micromol/min/g (n = 16), and it was reduced to 0.0957 +/- 0.059 micromol/min/g 2 weeks after chemoradiotherapy (p = 0.011). The correlation between residual MRglc and pTCP was made using an increment value of 0.02 micromol/min/g between the maximal and minimal values of MRglc. A pathologically complete response was obtained in 6 of 6 patients with residual MRglc of < or = 0.050 micromol/min/g, 3 of 4 with < or = 0.070, 4 of 7 with < or = 0.090, 0 of 4 with < or = 0.110, 1 of 3 with < or = 0.130, and 0 of 6 with > or = 0.130 micromol/min/g. The fitted logistic model showed that residual MRglc corresponding to pTCP 50% and pTCP > or = 95% was 0.076 and < or = 0.040 micromol/min/g, respectively.
CONCLUSION: The correlation between the gradient of residual MRglc after chemoradiotherapy and pTCP is an inverse dose-response relationship. Residual MRglc of 0.076 and < or = 0.040 micromol/min/g, representing pTCP 50% and pTCP > or = 95%, respectively, may be useful surrogate markers for the tumor response to radiotherapy or chemoradiotherapy in lung cancer.