Prognostic value of pretreatment 18F-FDG PET/CT in patients with metastatic malignant melanoma

来瑞鹤,蒋冲,滕月,孙一文,李爱梅
DOI: https://doi.org/10.3760/cma.j.cn121381-201909050-00110
2020-01-01
Abstract:Objective:To demonstrate whether 18F-FDG PET/CT metabolic parameters could predict prognosis in patients with metastatic maligant melanoma (MM). Methods:A retrospective analysis was conducted on a dataset composed of 47 patients who were newly diagnosed with metastatic MM and currently undergoing pretreatment 18F-FDG PET/CT in the Affiliated Hospital of Nanjing University Medical School, Nanjing Drum Tower Hospital from August 2011 to December 2018. Of 47 patients, 20 were male, 27 were female and median age 59 (23-86) years. All patients were treated with chemotherapy, immunotherapy or targeted therapy and follow-up time was 0.5 to 53.6 months. Melanoma-specific survival (MSS) and progression-free survival (PFS) were defined as the time from 18F-FDG PET/CT imaging to the patient's death and the patient's death or progression of the disease, separately. All patients underwent 18F-FDG PET/CT imaging before treatment.The maximum standardized uptake value (SUV max) was measured. Whole-body metabolic tumor volume (MTV) and whole-body total lesion glycolysis (TLG) were measured automatically and SUV>40% SUV max voxel boundary was used as threshold. The optimal thresholds of PET parameters were obtained using the receiver operating characteristic (ROC) curve, and the patients were divided into two groups separately according to the optimal thresholds of SUV max, whole-body MTV and whole-body TLG and six groups were obtained. The difference of MSS and PFS between the two groups were predicted by Kaplan-Meier method and Log-rank test. Univariate analysis was conducted to evaluate the prognostic value of PET parameters and clinical variables. The Cox proportional risk model multivariate analysis was used to determine whether the PET parameters can act as independent prognostic risk factors for MSS and PFS. Results:The cut-off values for SUV max, whole-body MTV, and whole-body TLG were 10.86, 8.12 cm 3, and 91.45, respectively, as shown in the ROC curve analysis. The PFS was significantly different in two groups divided by optimal thresholds of whole-body MTV or whole-body TLG, separately ( χ2=5.04, 5.02; both P<0.05). Similarly, the MSS was significantly different in two groups divided by optimal thresholds of SUV max or whole-body TLG, separately ( χ2=10.22, 4.38; both P<0.05). The univariate analysis results were as follows: the serum lactate dehydrogenase level≥245 U/L, lymphatic metastasis, SUV max>10.86 and whole-body TLG>91.45, which were associated with predictors of MSS. M l stage, whole-body MTV> 8.12 cm 3 and whole-body TLG>91.45, which were associated with PFS. The multivariate analysis results showed that SUV max>10.86, proving its potential as an independent prognostic risk factor for MSS. Conclusion:The 18F-FDG PET/CT metabolic parameter SUV max was the best predictive marker in metastatic MM patients, whole-body MTV and whole-body TLG helped for the prognosis of metastatic MM patients.
What problem does this paper attempt to address?