Differentiating Pseudoprogression from True Tumor Growth

You are reading version 1 of this direction, frozen 2026-10-04. The latest version is version 2.

Version 1 of 236 papers51 conceptsfrozen 2026-10-04

In oncology, the "gold standard" for determining if a treatment is working has traditionally been simple: measure the tumor. If it shrinks, the drug is working; if it grows, the treatment has failed. However, modern therapies—ranging from immunotherapy to anti-angiogenic drugs—have broken this logic, creating a visual paradox where scans often lie about the underlying pathology.

The Visual Paradox of Treatment Response

The central challenge in modern neuro-oncology and systemic cancer care is the rise of pseudoprogression. As Spyridon Bakas and colleagues note in the AI-RANO working group recommendations, standard imaging often shows an increase in lesion size that mimics disease progression but actually represents a successful inflammatory response.

This occurs because effective treatments, particularly immunotherapies like pembrolizumab, trigger an influx of immune cells into the tumor site. As S.Y. Cindy Yang and Yongjiu Dai explain, this "flare" or immune infiltration makes the tumor appear larger or more active on a scan, even as the cancer cells themselves are dying. Relying on traditional size-based criteria (like RECIST 1.1) in these moments can lead to the "premature and mistaken discontinuation of effective clinical treatments," a danger highlighted by Jianhua Chen in the context of small-cell lung cancer.

The Biological Drivers of Ambiguity

The ambiguity isn't limited to immunotherapy. Several biological mechanisms can obscure the truth on a radiological scan:

  1. Radiation Necrosis: Following radiotherapy, dead tissue can swell and enhance on an MRI, looking identical to a returning tumor. Yang Gao and Nazanin Maleki emphasize that this "pseudo-progression" is a major hurdle in treating glioblastoma, where distinguishing between recurrence and necrosis is critical for deciding whether to perform more surgery.
  2. Pseudoresponse: This is the opposite problem. Anti-angiogenic drugs like bevacizumab can "normalize" the blood-brain barrier, stopping the leakage of contrast agents. As Artem Chaban and Birgit Waschulzik demonstrate, this makes a tumor look like it has disappeared on an MRI when it is actually still growing—a phenomenon that causes structural imaging sensitivity to fall below chance levels.
  3. Acellular Persistence: In rectal and esophageal cancers, treatments can leave behind "pools" of jelly-like mucin or benign scars. Kyubo Kim and Peng Yuan show that these non-viable remnants can mimic active malignancy, leading to overestimations of residual disease.

Beyond Size: Metabolic and Microstructural Imaging

To solve this, researchers are moving away from measuring how big a tumor is and toward measuring what it is doing.

One approach is Metabolic Imaging. While standard MRI looks at water and contrast, Amino Acid PET (such as $^{18}$F-FET PET) tracks how much protein a tumor is consuming. Robin Gutsche and Karolin J. Paprottka have shown that metabolic volume is a much more reliable predictor of survival than structural volume because inflammation doesn't "eat" amino acids the way cancer does.

Another frontier is Microstructural MRI. Thomas A. Roberts and his team utilized a technique called VERDICT MRI to noninvasively measure the actual size of cancer cells. By focusing on the "virtual biopsy" of cell density rather than gross tumor volume, they could detect a response to chemotherapy as early as three days after treatment—long before any change in tumor size was visible.

Computational Solutions and New Standards

Because the human eye often cannot distinguish these subtle patterns, the field is turning to Artificial Intelligence. Qi Yang and Julian Ziegenfeuter describe how "radiomics"—extracting high-dimensional data from pixels—can identify "habitats" within a tumor. These models can distinguish between true growth and inflammation with high accuracy by analyzing the spatial heterogeneity of the lesion.

Finally, clinical practice is evolving through new frameworks like iRANO and iRECIST. As Xin Chen explains, these criteria introduce a "wait and see" period. If a tumor appears to grow, the patient isn't immediately taken off the drug; instead, they are rescanned weeks later. If the "growth" was actually inflammation, the tumor will eventually stabilize or shrink, confirming that the initial scan was a false alarm.

By combining advanced tracers, microstructural modeling, and AI-driven analysis, the research direction is clear: moving from a "snapshot" of tumor size to a deep, multi-modal understanding of tumor biology.

Papers in this version

History & citationVersion 2 · 2026-10-04

Current · Permanent link

  1. Version 2 (latest) · 2026-10-04 · 31 papers
  2. Version 1 · 2026-10-04 · 36 papers
Cite this direction

Latest version (a living page):

@misc{lacuna_radiographic_pseudoprogression_a4fea57b,
  title = {radiographic pseudoprogression},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/radiographic-pseudoprogression/rdr_a4fea57b5fbb4f9ea1e0328afe042325},
  note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}

This exact version (version 1, frozen text):

@misc{lacuna_differentiating_pseudoprogression_from_t_a4fea57b_v1,
  title = {Differentiating Pseudoprogression from True Tumor Growth},
  author = {{Lacuna}},
  howpublished = {Lacuna research direction},
  year = {2026},
  url = {https://lacuna.tiptreesystems.com/direction/differentiating-pseudoprogression-from-true-tumor-growth/rdr_a4fea57b5fbb4f9ea1e0328afe042325/v/1},
  note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}