Radiologic assessment of response of tumors to treatment Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 1
Objective response assessment is important to describe the treatment effect of anticancer drugs. Standardization by using a common language is also important for comparison of results from different trials. In contrast to clinical results, which can be subjective, diagnostic imaging provides a greater opportunity for objectivity and standardization Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 2
It was generally accepted that a decrease in tumor size correlated with treatment effect; as a result, imaging was adopted for lesion measurement, in the World Health Organization (WHO) criteria in 1979. However, because of some limitations of the WHO criteria, the Response Evaluation Criteria in Solid Tumors (RECIST) were introduced in 2000. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 3
In RECIST, imaging was recognized as indispensable for response evaluation of solid tumors. Nevertheless, the widespread use of multidetector computed tomography and other imaging innovations have made RECIST outdated, with a concomitant need for modifications. Meanwhile, newer anticancer agents with targeted mechanisms of action have demonstrated an inherent limitation and unsuitability of anatomic tumor evaluation that assesses only lesion size. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 4
Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 5
Methods of tumor measurement according to the RECIST and WHO criteria. With the WHO criteria, the longest diameter (A) and the longest perpendicular diameter (B) are obtained and multiplied (2D measurement). With RECIST, only the longest diameter (A) is obtained (uni- or one-dimensional [1D] measurement). The location of the longest diameter is decided independently of previous study results when the tumor changes shape or rotates. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 6
Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 7
Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 8
The primary goal of RECIST was to try to unify the various modifications of the WHO criteria so that meaningful comparisons could be made among studies. This included the following: (a) the need to maintain the four categories of responses (complete response [CR], partial response [PR], stable disease [SD], and PD); (b) the need to maintain the same definition of PR so that favorable results of future therapies can be compared with WHO criteria even though the measurements will be different and (c) the need to modify the definition of PD. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 9
There are five major differences between RECIST and the WHO criteria: (a) 1D measurements are adopted, which encourages measurement of more lesions and minimizes labor (6); (b) the type of imaging to be used is stipulated; (c) the types of tumors that should or should not be chosen are defined; (d) the number of tumor lesions used for assessment is specified; and (e) the cutoff point for definition of PD is larger. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 10
RECIST Response Evaluation Criteria in Solid Tumors Standardized repeatable method for measuring response to therapy for solid tumors NOT EQUIVALENT TO A CLINICAL READ!!! RECIST is a combination of both qualitative and quantitative assessment Based on concept of target lesions and non-target lesions Target lesions are quantitatively assessed Non-target lesions are qualitatively assessed Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 11
RECIST Based on concept of target lesions and non-target lesions Target lesions are chosen based on 3 factors: Must be EASILY (and reproducibly) measurable Must be representative of the disease (clearly metastasis) Must be representative of distribution (choose measurable lesions from all involved organs) Non-target lesions are all other presumed manifestations of the disease All non-measurable lesions Measurable lesions that were not chosen as target lesions Lesions that may be (but not definitely) metastases Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 12
Target Lesions Target lesions must be measurable Definition of Measurable Lesions Size Matters Conventional CT or MRI (non-spiral): If slice collimation <10mm, minimum lesion size is 20 mm If slice collimation >10mm, minimum lesion size is 2 x collimation ex. Slice collimation = 15mm, minimum lesion size = 30mm Spiral CT If slice collimation <5mm, minimum lesion size is 10 mm If slice collimation >5mm, minimum lesion size is 2 x collimation ex. Slice collimation = 7mm, minimum lesion size = 14mm Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 13
Target Lesions Target lesions must be reproducibly measurable Definition of reproducibly measurable lesions Consistency across time points Pick lesions with well defined edges or margins Always measure longest diameter Measure lesions on same phase or same sequence (MRI) Only measure lesions that are definitely metastases (If unsure don t measure) Pick lesions that are stable in position, try to avoid mobile lesions (Avoid mesenteric masses that change in position) Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 14
Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 15
Inappropriate selection of lesions in a 56-year-old patient with breast cancer and bilateral ovarian metastases. (a) Axial contrast-enhanced CT scan from the baseline study shows enlargement and heterogeneous enhancement of both ovaries (arrowheads). (b) Corresponding CT scan obtained after therapy shows the left ovary (arrow) turned behind the uterus. Evaluation is thus more difficult due to organ movement. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 16
Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 17
Difficulty in measuring the tumor. (a) Axial contrast-enhanced CT scan of a 58-year-old patient with colon cancer shows a lobulated and ill-defined liver metastasis. The lesion seems to be the result of fusion of two metastases as well. (b) Axial contrast-enhanced CT scan of a 74-year-old patient with rectal cancer shows confluent liver metastases. Different numbers of lesions, although with a maximum of five, and different combinations of lesions can be selected for baseline measurement and follow-up according to RECIST. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 18
Target Lesions Target lesions should represent distribution of disease Definition of measurable lesions Representative of disease throughout body Pick lesions from disparate areas of the body Do not choose > 5 lesions in any one organ or anatomic location Organs are well defined Anatomic regions are up to individual interpretation (use best judgment) For lymphoma choose nodes from different nodal stations Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 19
Target Lesions Measurable lesions up to a maximum of 5 lesions per organ 10 lesions total Select based on size and reliability of measurement Sum of longest diameter (SLD) for all target lesions will be calculated at baseline and used as reference to characterize objective tumor response Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 20
Quantitative Assessment The SLD is the quantitative assessment SLD = sum of the longest diameters of target lesions This part of the evaluation is not subject to interpretation Strict rules and definitions of: Complete response = No measurable disease Partial Response = Greater than 30% decrease in score Stable Disease = Between 30% decrease and 20% increase Progression = Greater than 20% increase in score Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 21
Non Target Lesions All aspects of disease not chosen as Target Lesions All non-measurable lesions Measurable lesions that were not chosen as target lesions Lesions that may be (but not definitely) metastases Non- measurable lesions Not suitable for accurate repeated measurements Ascites Leptomeningeal disease Pleural effusions Inflammatory breast disease Cystic lesions Lymphangitis cutis/pulmonis Bone lesions Brain lesions Irradiated lesions Ground glass lung lesions Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 22
Measuring Lesions Baseline Scan Initial Review Determine if a single measurable lesion is present Once single lesion is found, no need to evaluate any further Baseline scan Full Review Determine target lesions and non-target lesions Target lesions Record site and longest diameter Measure longest diameter (LD) on slice where the lesion is largest Use magnification and appropriate window/level Non-target lesions Record site and description Will be assessed qualitatively in the future Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 23
Follow-up: Target Lesions On follow-up scans, once a lesion is identified as Target: Must continue to measure even if LD falls below size criteria Measure LD regardless of location (slice) or orientation on prior scan Choose slice where lesion is largest, even if different than baseline Measure LD regardless of poor image quality or poorly defined lesion boundaries (i.e., if target lesion is imaged, LD must be measured) If a target lesion is visible but too small to measure, list as 5mm If lesion is not imaged, enter Unknown (outside FOV) If unknown is entered, comments are required Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 24
Follow-up Scans Target lesions on follow-up: If a lesion separates to form discrete lesions, measure LD of each lesion and report separately (e.g. #3 -> 3 and 3a) If target lesion becomes confluent, measure LD of lesion and record under 1 of the lesions and enter 0 mm for other lesion(s) Non-target lesions on follow-up: All lesion region or organ that were selected will be followed and their status will be recorded as: Absent: If totally resolved (CR) Unchanged, Improved, or? increased but not clearly increased (SD) Clearly worse: Indicative of progression (PD) Not assessed: Missing, incomplete imaging (UN) Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 25
Follow-up Scans New lesions seen on follow-up: Any lesion that appears after baseline (including new lesions in irradiated areas) Any lesions that re-appear will be considered new lesions Lesions should be greater than the slice thickness (usually at least 6 mm) to be considered a new lesion Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 26
Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 27
Appearance of a new lesion in the adrenal gland in a 49-yearold man with rectal cancer. (a, b) Axial contrast-enhanced CT scans (lung window) obtained before (a) and after (b) treatment show multiple lung metastases. The target lesion (arrowhead) and many other lesions have decreased in size after treatment. (c, d) Axial contrast-enhanced CT scans of the abdomen, obtained before (c) and after (d) treatment, show a new lesion in the adrenal gland after treatment (arrow in d). This finding should be interpreted as a case of PD. Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 28
Measuring Lesions Liver lesions: Include the hypervascular peripheral component Measure in portal venous phase on CT Measured in post contrast T1 axial images (portal phase) Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 29
Tumor Response - Target Lesions Complete response (CR): Disappearance of all target lesions Partial response (PR): > 30% decrease in the SLD taking as reference the baseline SLD Stable decrease (SD): Neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for PD Progression (PD): > 20% increase in the SLD taking as reference the nadir beginning with baseline measurement (if unknown is present then that SLD cannot be used as reference) Unknown (UN): If one or more unknown is present and the SLD is not indicative of PD (explanatory comments required) Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 30
Tumor Response Non-Target Lesions Complete Response (CR): Disappearance of all non-target lesions Incomplete Response/Stable Disease (SD): If one or more is Unchanged or Improved and no PD, not assessed or not done Progression (PD): If at least one Clearly worse is present Unknown (UN): If not assessed or not imaged is present Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 31
Tumor Response New Lesions New Lesions = Progression (PD) Any new malignant lesion Any re-appearing lesion Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 32
Tumor Response - Summarized Target Lesions Non-target Lesions New Lesions Overall Response CR CR No CR CR SD No PR PR CR or SD No PR SD CR or SD No SD PD Any Yes or No PD Any PD Yes or No PD Any Any Yes (PD) PD Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 33
Surgery & Radiotherapy Note: If at all possible, lesions in areas of known radiation or surgery should not be selected as target or non-target lesions Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 34
Pleural Effusion & Ascites Note: New or enlarging pleural effusions or ascites evidenced radiographically will NOT be assumed to be malignant Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 35
Cheson Criteria Based on International Working Group Recommendations Standardized repeatable method for measuring response to therapy for NHL Response is assessed on 3 criteria: 1 Radiological Lymph nodes/ Quantitative masses 2 Clinical Physical Exam Qualitative Spleen/Liver Biochemical 3 Pathological Bone Marrow Semi-quantitative Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 36
Node Selection Must be representative of the distribution of the disease Must be clearly and reproducibly measurable in at least 2 perpendicular dimensions Definition of Target Lesions Abnormal lymph nodes and/or nodal masses and/or hepatic/splenic nodules (up to 6) >1.5 cm longest diameter and >1.0 cm transverse diameter Mediastinal and retroperitoneal areas of disease should be included whenever these sites are involved Definition of Non-target Lesions Except for splenic or hepatic nodules, involvement of other organs is considered non-measurable disease Any lymph nodes or nodal masses not selected as target lesions Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 37
Tumor Assessment Response Criteria Radiological Criteria = Lymph Nodes / Masses CR <= 1.5 cm LD if >1.5 cm at baseline, or <= 1 cm LD if between 1.1 to 1.5 cm, or >75% decrease in SPD at baseline CRu >1.5 cm LD that has regressed by >75 % in (unconfirmed) SPD at baseline PR >= 50% decrease in SPD at baseline of 6 largest dominant nodes or nodal masses No increases in other nodes PD SD >= 50% increase in SPD from nadir and/or appearance of any new lesion Less than PR but not progression Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 38
Tumor Response - Summarized Response category Physical Evaluation Lymph Nodes Lymph Node Masses Bone Marrow CR Normal Normal Normal Normal CRu Normal Normal Normal Indeterminate Normal Normal >75% decrease Normal or Indeterminate PR Normal Normal Normal Positive Normal >=50% decrease >=50% decrease Irrelevant Decrease in liver/spleen >=50% decrease >=50% decrease Irrelevant Relapse/ Progression Enlarging liver/spleen; new sites New or Increased New or Increased Reappearance Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 39
Reviewer Selection Criteria Per ICH Guidelines Chosen according to ICH Guidelines & other regulation Reviewers MUST have No financial interest in the outcome of the study No involvement in study design and conduct Agreed to no use of information learned during the course of the trial without approval by sponsor M.D. with appropriate medical expertise in clinical area Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 40
Data Form Requirements Reviewer is responsible for accuracy of data entered on the form Upon completion, reviewer must sign the form Any changes necessary (once the CRF is signed) will be considered a re-review All changes must be initialed and dated Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 41
Thank you for your attention Copyright 2008 TIMC, Matthew A. Barish M.D. All rights reserved. 42