A few weeks ago I found myself contemplating what topic I should present during a virtual radiology elective.  I recall recently having learned about how tumor response was measured radiologically, and felt that it was an interesting topic to read about and discuss pertinent to the field of oncology.  Having an understanding of how we quantify tumor response can give some insights on how we make an effort to identify tumor response and what that may entail.

I know that some medical students find a rotation with oncology to be quite confusing, but I believe that has more to do with how far removed the field seems in comparison to others, at least in terms of proficiency and understanding for your average medical student.  In my experience, as you become more familiar with topics in oncology you get to see the true vastness of the decision making complexity in the field.   It wasn’t just blindly following the NCCN (National Comprehensive Cancer Network) Guidelines for each unique cancer, but also knowing the adverse effects of the drugs, when to make dose adjustment and how that all plays into the goals of care for a given patient, just to name a few of the things that really started to stand out.

When treating oncology patients you may find yourself staring at the endless maze of NCCN guidelines with little understanding about how it is cobbled together.  One of the things that constitutes much of the decision making within these guidelines is understanding treatment response to a particular agent.  Treatment response is often monitored using a number of imaging modalities including most commonly CT (computerized tomography) scans but PET (positron emission tomography) scans but even MRIs (magnetic resonance imaging) are also utilized in many cases.   These radiographic studies are usually compared to prior imaging studies to look for evidence of tumor progression, stability or if the tumor is responding to the agent, and shrinks.  

In many cases you may find NCCN guidelines which indicate that an agent be discontinued in progressive disease while the same agent be maintained in the context of stable disease or in the event of treatment response.  But, is this simply that a mass does not appear to be growing, or shrunk in the context of the treatment agent?  How much shrinkage does a tumor actually need to be considered to be responding?  Or maybe the mass is slightly larger, but is that just interrater variability between radiologists or the limitations of the radiographic study?

Historically we originally started measuring solid tumor response using WHO criteria in 1981 which used a bidimensional method of measuring lesions to make comparisons.  As cancer studies moved forward, WHO criteria has been largely superseded by a more simple scoring system known as RECIST (Response Evaluation Criteria In Solid Tumors).  The RECIST scale showed high concordance with the WHO criteria using a unidimensional measurement framework.  The newer criteria also exists in lots of different variations and modifications for specific tumors.  For example RECIST-B had been used in some breast cancer studies, while immunotherapy has several of its own RECIST criteria scales.  Modifications and variations are largely similar to the standard RECIST criteria.

In the case of solid tumors, the RECIST criteria serves as a wonderful framework for defining each of these response types and ultimately guiding treatment decisions.  It also is particularly useful to have as an attempt to reach a consistent definition when studying cancer, and many investigations often use these criteria to define the responses to investigational drugs.

The RECIST criteria itself defines treatment response into 4 general categories.

Complete Response (CR) – Where the masses will disappear.

Partial Response (PR) – Where masses have a >30% reduction in SLD (sum of longest diameter of target lesions) with no new lesions and no growth of non target lesions.

Progressive Disease (PD) –  >20% increase in SLD of smallest in study or progression of non target lesions or new lesions.

Stable Disease (SD) – Anything between PR and PD.

The RECIST criteria also has rules for which lesions can be measured and considered target lesions and which cannot but in general it is those that have at least one dimension greater than 10mm, up to a maximum of 5 lesions and 2 in any given organ.  Involved lymph nodes can be considered if the short axis is up to 15mm as well.  The rules also note that nonmeasurable features are not to be included in the scoring criteria, such as pleural fluid, ascites or lymphangitis. 

In some cases, when multiple target lesions are being monitored you will see cases where some are growing, while others are shrinking leading to a confusing picture.  As it stands now, the RECIST criteria will compare the sum of longest diameters of the target lesions and compare to previous studies to get an overall picture of the cancer.  There may be criticism in this approach given that tumor heterogeneity can exist between the target lesions, and they may truly be responding differently to the therapeutic.    This could be a result of a resistance mechanism that may have cropped up in one lesion but not another, but certainly could lead to less than optimal characterization of tumor response.

As you can see, even with these definitions there is still a lack of clarity.  Even within the category considered stable disease, there may be a wide range of things that may be happening with no absolute certainty that the disease is not progressing or regressing or stable.  However, in spite of these flaws, having a standardized system to characterize tumor response radiologically is useful.  It allows us to better quantify response in studies, and helps us make decisions regarding treatment at the relevant decision points seen in NCCN guidelines. 

Looking to the future, perhaps we can find a scoring system that better accounts for the presence of tumor heterogeneity that can misrepresent tumor response.  If I were to guess, this may look largely similar to the system we have now, but with an increased emphasis on serum tumor markers including things such as free DNA.


1. Kumar R, Qi T, Cao Y, Topp B. Incorporating lesion-to-lesion heterogeneity into early oncology decision making. Front Immunol. 2023;14:1173546. doi:10.3389/fimmu.2023.1173546

2. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45(2):228-247. doi:10.1016/j.ejca.2008.10.026

3.Llewelyn R, Bell, Daniel J. Response evaluation criteria in solid tumors | Radiology Reference Article | Radiopaedia.org. Radiopaedia. doi:10.53347/rID-19147

4. Smits F, Dirksen M, Schoots, I., The Radiology Assistant : RECIST 1.1 – examples. Radiology Assistant. Accessed October 11, 2023. https://radiologyassistant.nl/more/recist-1-1/recist-1-1-examples

Categories: Blog


Leave a Reply

Avatar placeholder

Your email address will not be published. Required fields are marked *