BACKGROUND: Positron emission tomography (PET) imaging using the radiotracer 18F-Fluorothymidine (FLT) has been proposed as an imaging biomarker of tumour proliferation. If FLT-PET can be established as such it will provide a non-invasive, quantitative measurement of tumour proliferation across the entire tumour. Results from validation studies have so far been conflicting with some studies confirming a good correlation between FLT uptake and Ki-67 score and others presenting negative results. METHODS: Firstly we performed a systematic review of published studies between 1998 and 2011 that explored the correlation between FLT uptake and Ki-67 score and examined possible variations in the methods used. Studies were eligible if they: (a) included patients with cancer, (b) investigated the correlation between Ki-67 measured by immunohistochemistry and FLT uptake measured with PET scanning, and (c) were published as a full paper in a peer-reviewed scientific journal. Secondly a meta-analysis of the correlation coefficient values reported from each study was performed. Correlation coefficient (r) values were extracted from each study and 95% confidence intervals (CIs) were calculated after applying Fisher's z transformation. For subgroup analysis, studies were classified by the index used to characterise Ki-67 expression (average or maximum expression), the nature of the sample (whole specimen or biopsy) and the cancer type. FINDINGS: Twenty-seven studies were identified as eligible for the meta-analysis. In the studies we examined there were variations in aspects of the methods and reporting. The meta-analysis showed that given an appropriate study design the FLT/Ki-67 correlation is significant and independent of cancer type. Specifically subgroup analysis showed that FLT/Ki-67 correlation was high in studies measuring the Ki-67 average expression regardless of use of surgery or biopsy samples (r=0.70, 95% CI=0.43-0.86, p<0.001). Of the studies that measured Ki-67 maximum expression, only those that used the whole surgical specimen provided a significant r value (r=0.72, 95% CI=0.54-0.84, p<0.001). Studies that used biopsy samples for Ki-67 maximum measurements did not produce a significant r value (r=0.04, 95% CI=-0.18-0.26, p=0.71). In terms of the cancer type subgroup analysis there is sufficient data to support a strong FLT/Ki-67 correlation for brain, lung and breast cancer. No publication bias was detected. INTERPRETATION: This systematic review and meta-analysis highlights the importance of the methods used in validation studies comparing FLT-PET imaging with the biomarker Ki-67. The correlation is significant and independent of cancer type provided a study design that uses Ki-67 average measurements, regardless of nature of sample, or whole surgical samples when measuring Ki-67 maximum expression. Sufficient data to support a strong correlation for brain, lung and breast cancer exist. However, larger, prospective studies with improved study design are warranted to validate these findings for the rest of the cancer types.
BACKGROUND: Positron emission tomography (PET) imaging using the radiotracer 18F-Fluorothymidine (FLT) has been proposed as an imaging biomarker of tumour proliferation. If FLT-PET can be established as such it will provide a non-invasive, quantitative measurement of tumour proliferation across the entire tumour. Results from validation studies have so far been conflicting with some studies confirming a good correlation between FLT uptake and Ki-67 score and others presenting negative results. METHODS: Firstly we performed a systematic review of published studies between 1998 and 2011 that explored the correlation between FLT uptake and Ki-67 score and examined possible variations in the methods used. Studies were eligible if they: (a) included patients with cancer, (b) investigated the correlation between Ki-67 measured by immunohistochemistry and FLT uptake measured with PET scanning, and (c) were published as a full paper in a peer-reviewed scientific journal. Secondly a meta-analysis of the correlation coefficient values reported from each study was performed. Correlation coefficient (r) values were extracted from each study and 95% confidence intervals (CIs) were calculated after applying Fisher's z transformation. For subgroup analysis, studies were classified by the index used to characterise Ki-67 expression (average or maximum expression), the nature of the sample (whole specimen or biopsy) and the cancer type. FINDINGS: Twenty-seven studies were identified as eligible for the meta-analysis. In the studies we examined there were variations in aspects of the methods and reporting. The meta-analysis showed that given an appropriate study design the FLT/Ki-67 correlation is significant and independent of cancer type. Specifically subgroup analysis showed that FLT/Ki-67 correlation was high in studies measuring the Ki-67 average expression regardless of use of surgery or biopsy samples (r=0.70, 95% CI=0.43-0.86, p<0.001). Of the studies that measured Ki-67 maximum expression, only those that used the whole surgical specimen provided a significant r value (r=0.72, 95% CI=0.54-0.84, p<0.001). Studies that used biopsy samples for Ki-67 maximum measurements did not produce a significant r value (r=0.04, 95% CI=-0.18-0.26, p=0.71). In terms of the cancer type subgroup analysis there is sufficient data to support a strong FLT/Ki-67 correlation for brain, lung and breast cancer. No publication bias was detected. INTERPRETATION: This systematic review and meta-analysis highlights the importance of the methods used in validation studies comparing FLT-PET imaging with the biomarker Ki-67. The correlation is significant and independent of cancer type provided a study design that uses Ki-67 average measurements, regardless of nature of sample, or whole surgical samples when measuring Ki-67 maximum expression. Sufficient data to support a strong correlation for brain, lung and breast cancer exist. However, larger, prospective studies with improved study design are warranted to validate these findings for the rest of the cancer types.
Authors: Lale Kostakoglu; Fenghai Duan; Michael O Idowu; Paul R Jolles; Harry D Bear; Mark Muzi; Jean Cormack; John P Muzi; Daniel A Pryma; Jennifer M Specht; Linda Hovanessian-Larsen; John Miliziano; Sharon Mallett; Anthony F Shields; David A Mankoff Journal: J Nucl Med Date: 2015-09-10 Impact factor: 10.057
Authors: Yu-Xiang Ye; Claudia Calcagno; Tina Binderup; Gabriel Courties; Edmund J Keliher; Gregory R Wojtkiewicz; Yoshiko Iwamoto; Jun Tang; Carlos Pérez-Medina; Venkatesh Mani; Seigo Ishino; Camilla Bardram Johnbeck; Ulrich Knigge; Zahi A Fayad; Peter Libby; Ralph Weissleder; Ahmed Tawakol; Shipra Dubey; Anthony P Belanger; Marcelo F Di Carli; Filip K Swirski; Andreas Kjaer; Willem J M Mulder; Matthias Nahrendorf Journal: Circ Res Date: 2015-09-22 Impact factor: 17.367
Authors: David M McClatchy; Henning Willers; Aaron N Hata; Zofia Piotrowska; Lecia V Sequist; Harald Paganetti; Clemens Grassberger Journal: Cancer Res Date: 2020-09-09 Impact factor: 12.701
Authors: Willem Grootjans; Lioe-Fee de Geus-Oei; Esther G C Troost; Eric P Visser; Wim J G Oyen; Johan Bussink Journal: Nat Rev Clin Oncol Date: 2015-04-28 Impact factor: 66.675