A retrospective study on correlation between SUV_(max) and influencing clinical factors when ~(18)FDG- PET applied to diagnosis of non-small cell lung cancer

YANG Jin-ji,WU Yi-long,CHEN Gang,QIAO Sui-xian,GU Li-jia,YANG Xue-ning,ZHONG Wen-zhao,HUANG Yu-juan,LIAO Ri-qiang,HUANG Yi-sheng
2008-01-01
Abstract:Background and purpose:Maximal standard uptake value (SUVmax) of primary lesion plays an important role in diagnosis of non-small cell lung cancer (NSCLC) when 18FDG-PET is applied, so we investigated the influencing factors that could impact on SUVmax of primary lesion for NSCLC patients detected by 18FDG-PET before surgery. Methods:92 NSCLC patients undergoing 18FDG-PET before complete resection were analyzed retrospectively to detect the correlation between SUVmax of primary lesion and some influencing clinical factors through partial correlation analysis,One-Way ANOVA and multiple regression and correlation. Results:Partial correlation analysis showed that SUVmax of primary lesion had significantly correlated with the longest diameter of primary lesion and pathological staging(P<0.05),.but not with age, pathological differentiation and neoadjuvant chemotherapy(P>0.05). One-Way ANOVA revealed that average SUVmaxs of primary lesion were 5.18±4.46,7.65±3.57,5.84±3.02,3.35±2.4 7,6.00±4.59,4.30±1.70 and 5.85±4.15 for adenocarcinoma(53 cases), squamous cell carcinoma(24 cases), adeno- squamous cell carcinoma(4 cases), large cell carcinoma(4 cases), sqamous-adenocarcinoma(2 cases), giant cell carcinooma(2 cases) and bronchioloalveolar cell carcinoma(2 cases) respectively, but no significant difference within groups or between groups(P>0.05). Multiple regression and correlation demonstrated that SUVmax of primary lesion had significantly correlated with the longest diameter of primary lesion(P<0.05), but not with age, subtype of NSCLC, pathological differentiation, pathological staging and neoadjuvant chemotherapy(P>0.05). Conclusion:SUVmax of primary lesion was significantly associated with the longest diameter of primary lesion through both partial correlation analysis and multiple regression and correlation, but with pathological staging only through partial correlation analysis. No significant difference of average SUVmax of primary lesion was found within or between subtypes of NSCLC.
What problem does this paper attempt to address?