Vlume Measurement at CT Staging and Assessment f Respnse with Quantitative CT Lawrence Schwartz, MD Department f Radilgy Clumbia University Cllege f Physicians and Surgens LSCHWARTZ@COLUMBIA.EDU
Recmmendatins STAGING 1. PET-CT shuld be used fr staging in rutine clinical practice and in clinical trials (categry 1). 1. FDG scans can be used t image mst subtypes f lymphma and t target bipsy but is nt rutinely recmmended in lymphmas with lw FDG avidity e.g. CLL/SLL, extrandal MZL and sme cutaneus lymphmas (categry 1). 1. In HL and DLBCL staged by PET-CT there is n rle fr rutine BMB. BMB is indicated nly if it wuld change staging with a resultant change in therapy (categry 1). 1. PET-CT with cect is desirable fr staging patients likely t underg raditherapy ideally within a single scanning sessin, but a tw stage apprach using unenhanced PET-CT fllwed by reginal cect fr equivcal lesins may be preferred taking int accunt patient age, disease type and clinical stage (categry 2) 2. Bulk remains an imprtant prgnstic factr in lymphma. Vlumetric analysis f tumur bulk and ttal tumur burden as well as methds cmbining metablic activity and anatmical size r vlumes shuld be explred as ptential prgnsticatrs (categry 3). 3. Optimal reprducible methds fr vlumetric analysis are yet t be defined and will require prspective testing in multicentre studies r carefully selected retrspective datasets (categry 3).
Recmmendatins RESPONSE ASSESSMENT - QUANTITATIVE 1. Standardisatin f PET methds is mandatry fr the use f quantitative appraches (categry 1) 1. Data are emerging t suggest that quantitative measures culd be used t imprve n visual analysis fr respnse assessment in DLBCL but this requires further validatin in clinical trials (categry 2). 1. The SUV max is the nly quantitative measure with published data t indicate its pssible utility in respnse assessment but changes in tumur vlumes shuld als be explred (categry 3).
Questins Why measure vlumes? Are CT tumr measurements accurate? What type f CT acquisitin is required fr a vlume measurement? Hw d they cmpare in accuracy t PET SUV measurements? D we care hw they cmpare? Cmbinatin Metablic Tumr Vlume Can we perfrm CT vlume measurements At a single institutin In a multicenter trial
Why measure vlumes? Staging beynd Ann Arbr... Max diameter 125 mm Vlume 130590 mm3 Max diameter 142 mm Vlume 215230 mm3
Why measure vlumes? Respnse Assessment... Baseline Fllw up + 6 weeks Baseline Fllw up % CHANGE 40% 59% 72%
Are CT Measurements Accurate 97 lymph nde metastases were assessed manually (RECIST 1.1) and by vlumetry with semi-autmated sftware The quality f segmentatin after manual crrectin was acceptable t excellent in 95 % f lesins and manual crrectins were applied in 21-36 % f all lesins, mst predminantly in lymph ndes Mean precisin was 2.6-6.3 % (manual) with 0.2-1.5 % (effective) relative measurement deviatin (p <.001). Inter-reader median variatin cefficients ranged frm 9.4-12.8 % (manual) and 2.9-8.2 % (vlumetric) fr different lesin types (p <.001). The limits f agreement were ± 9.8 t ± 11.2 % fr vlumetric assessment Wulff Kiel
Are CT Measurements Accurate Materials and Methds: MSCT scans f 63 malignant lymphma patients befre and after 2 cycles f chemtherapy (307 target lymph ndes) Results: Respnse classificatin per lymph nde revealed semiautmated vlumetry and bi-dimensinal WHO t be significantly mre accurate than manual linear metric measurements. Respnse classificatin per patient based n RECIST revealed mre patients t be crrectly classified by semi-autmatic measurements, e. g. 96.0 %/92.9 % (WHO bidimensinal/vlume) cmpared t 85.7/84.1 % fr manual LAD and SAD, respectively (mean reductin in misclassified patients f 9.95 J. Weßling 1, M. Puesken 1, R. Kch 2,
Are CT Measurements Accurate Reprducibility f CT Scans Clinically indicated nn-cntrast 1.25 mm slice chest CT Up t 15 minute break Repeat same CT n same scanner Zha Radilgy July 2009
Reprducibility f CT Scans Cncrdance Crrelatin Cefficient UNI BI VOL CCC 0.9981 0.9965 0.9995 95% CI 0.9968, 0.9994 0.9940, 0.9989 0.9991, 0.9998 T estimate the reprducibility and repeatability f the tumr size measurement CCC -Used t quantify repeatability and reprducibility
Reprducibility f CT Scans Mdified Bland-Altman UNI BI VOL % difference Size in mm Size in mm 2 Size in mm 3 Mdified Bland-Altman Plt the percentage f relative difference between the repeated tumr measurements
Reprducibility f CT Scans 2 cm example UNI BI VOL 1.88 cm - 6.0% 2.62 cm 2-16.6% 3.69 cm 3-11.9% 2.00 cm 3.14 cm 2 4.19 cm 3 2.13 cm + 6.5% 3.76 cm 2 + 19.8% 4.47 cm 3 + 6.8% Fr the cmputer generated measurements, using a hypthetical 2 cm tumr, 95% cnfidence interval what wuld the secnd measurement be
What type f CT acquisitin is required fr a vlume measurement?
What type f CT acquisitin is required fr a vlume measurement?
Hw d they cmpare in accuracy t PET SUV measurements? Hw can we ptimally cmbine the metablic and anatmic infrmatin?
Cmparing CT size and PET SUV Change in Tumr size in Esphageal Cancer Changes in tumr metablism are a mre sensitive parameter fr assessing the effect f therapy Wieder H. J. Nuc Med: 46 Dec 2005
Metablic tumr vlume - MTV Vlume f tumr tissues with increased FDG uptake FDG target vlume frequently calculated by visual delineatin f tumr edge r side-by-side analysis with cntrast-enhanced CT scan Semi-autmated frm attenuatin-crrected PET/CT images by using a cnturing prgram, renders the vlume measurement mre feasible
Imprving Surrgates Autmated Segmentatin (a1) (a2) (a3) (a4) (a5) (a6) (a7) (b1) (b2) (b3) (b4) (b5) (b6) (b7) (c1) (c2) (c3) (c4) (c5) (c6) (c7) Yan et al, Medical Physics 33, 2006
Can we perfrm CT vlume measurements
Can we perfrm CT vlume measurements
Can we perfrm CT vlume measurements Baseline 6-week fllw-up
Can we perfrm CT vlume measurements Autmated Segmentatin
Can we perfrm CT vlume measurements Autmated Segmentatin
Can we perfrm CT vlume measurements
Imaging and Tumr Bilgy understanding respnse t therapy Used t determine treatment decisins fr an individual patient Used t evaluate efficacy f a nvel therapy in a clinical trial Used fr crrelative analysis t develp predictive tissue bimarkers
Opprtunities and Questins Create mre bilgically meaningful respnse criteria Are we using the crrect cut values fr PR and PD? Hw best t evaluate the spleen Is the lng axis r shrt axis a gd enugh surrgate fr true tumr burden D we need t revisit hw many lesins t measure Is tumr burden at baseline a predictive bimarker Can we measure an anatmic respnse earlier with any f these methdlgies Can these techniques help us define predictive tissue bimarkers
EVALUATION OF INTERIM RESPONSE IN CLASSICAL HL USING VOLUMETRIC CT MEASUREMENTS IN COMBINATION WITH FDG PET PARAMETERS AFTER 2 CYCLES T determine the prgressin-free survival (PFS) at 36 mnths frm enrllment fr patients with Hdgkin lymphma using CT vlumetric changes between baseline and after 2 cycles f AVG in cmbinatin with qualitative FDG PET/CT interpretatin. Using changes determined by vlumetric CT measurements, alne, between baseline and after 2-4 cycles and after 6 cycles f therapy and in cmbinatin with qualitative and quantitative FDG PET/CT interpretatin, t determine the, best verall respnse psitive and negative predictive value f each test metrics alne and in cmbinatin with each ther. cmpare the predictive values f cmbinatrial imaging (vct and FDG PET/CT) parameters with cnventinal risk factrs including IPI.
Why measure vlumes? Answers... Are CT tumr measurements accurate? What type f CT acquisitin is required fr a vlume measurement? Hw d they cmpare in accuracy t PET SUV measurements? D we care hw they cmpare? Can we perfrm CT vlume measurements At a single institutin In a multicenter trial