top of page

Organoids vs Animals - why animal research is irreplaceable

  • Michael Florea
  • Oct 31, 2025
  • 2 min read

Updated: 5 days ago


In 2025, the FDA and NIH announced sweeping support for New Approach Methodologies (NAMs): organoids, organs-on-chips, and computational models. This should have happened years ago. NAMs could bring better human data and improve drugs before clinical trials.


But they also juxtaposed NAMs as competition for animal research.


Is it really NAMs vs animals? Will animal studies be replaced?


No.


Here's why:


Organoids are promising. They may improve drug screening before animal studies (currently done in 2D cell culture). But even if fully developed, organoids are still single tissue clumps. They fundamentally cannot model a full integrated physiological system required for systemic efficacy and toxicity testing. Ever.


There are only three technologies that potentially can: organs-on-chips, bodyoids and whole human simulation.


Organs-on-chips integrate multiple tissues, but today they are mostly individual organoids connected by microfluidics. This is a far cry from robustly and sufficiently replicating the human body. Integrating organs, blood, lymphatics, neuroendocrine, mechanical shear, nerve innervations, immune, and sensory systems would literally mean building a 90% functioning human body from scratch.


Whole human simulation is doing the above but in a computer. Do we have enough data? Probably not. Only 20% of genes have well-known functions. Only 12% have ever been targeted by all FDA recorded clinical trials combined. At the current rate, it would take estimated ~170 years to target all druggable proteins.


Bodyoids are conceptual human bodies engineered to lack consciousness through targeted neurodeficiency, grown in surrogate moms, and kept on life support until experimentation. They're technically feasible today - but even if you ethically accept them, how likely is societal acceptance at sufficient scale? There are ~100 million research animals in the US. Replacing them would require 10-20 million bodyoids yearly: 25-50% of all child-bearing-age women (44 million).


So, nothing on the horizon can realistically replace animals in the next decade.


Instead, how can we make animal studies better? First, by improving data quality. Second, by making them more predictive of human disease.


A lot can be done here: genetically engineered models, humanized models (with implanted human cells or organoids) and automated data collection systems that improve understanding of human biology with less experimental bias. And of course, adding NAM data where relevant.


NAMs aren’t here to replace animals. Animal research isn’t the enemy. The 90.4% clinical trial failure rate is. And research that’s not predictive, not reproducible or inhumane.


Let’s focus on solving that, not false dichotomies.


Original post (for comments)


References below.

 

9/10 deaths in the US today are caused by complex diseases: https://www.cdc.gov/nchs/fastats/leading-causes-of-death.htm




20% of the human genes functions is well understood: https://www.scientificamerican.com/article/mapping-the-unknome-may-reveal-critical-genes-scientists-have-ignored/#:~:text=Among the vast contents of,Oxford cell biologist Matthew Freeman.



At the current rate, it would take 170 years to target all druggable proteins: https://pmc.ncbi.nlm.nih.gov/articles/PMC10749231/




 
 
bottom of page