Clinical trials using CRISPR-Cas9 in cancer patients are underway in China and anticipated to follow soon in the United Sates (1).  Is this technology being rushed into the clinic too soon?  A new set of experiments in mice (2) suggests that the gene editor may not be as precise as some investigators believed (3).

            The new study has raised some troubling concerns.  Intended to edit the T-cells of cancer patients to boost their capacity attack malignancy, few doubt the target DNA sequences will be hit successfully.  The problem is the editor may produce alterations at locations other than the intended target – perhaps many of them.  The CRISPR enzymes were designed to knock out T-cell proteins, but off-target mutations could end up inactivating other cell functions.  The problem is that no one will understand the functional implications of such changes until they become manifest and this may take time to sort out. 

            Investigators seeking to modify a single DNA base in two mice were able to achieve their goal.  However, along with their intended alteration they detected more than 1,600 additional untargeted mutations in those animals, a number shockingly higher than observed in previous CRISPR editing experiments.  Perhaps differences in methodology or conditions will ultimately explain why this extremely small scale mouse study produced so many off-target mutations.  Until it is possible to dismiss these new results as an experimental or procedural artifact or improve the precision of the CRISPR enzymes, Institutional Review Board (IRB) approvals for new clinical studies may decline.

            Should we demand CRISPR editing be perfect to permits its use in humans?  Current cancer therapies are often toxic and sometimes so mutagenic they are recognized to induce secondary malignancies in some patients (4).  If the true off-target CRISPR genome alteration rates lie somewhere between undetectable (3) and ‘alarming’ (2) we may have to calculate the risk-to-benefit ratio for given interventions and situations such as patients for whom standard therapies have failed.  Sometimes it may be necessary to resort to approaches that carry intrinsic risks.

            At the moment a great deal of basic work needs to be completed to make informed decisions about how best to harness CRISPR enzymes for medical treatments.  CRISPR evolved in bacteria, organisms with far smaller and structurally simpler genomes than human, animal or plant cells.  Intended to patrol small genomes to detect and eliminate invading viruses, the intrinsic error rates of these bacterial enzymes may be too high avoid producing large numbers of off-target mutations when they are forced to scan the enormously larger genomes of higher organisms.  It is also possible the physicochemical conditions prevalent in human cells may not allow CRISPR enzymes to function with optimal fidelity.  In addition, deploying CRISPR systems to engineer alterations that eliminate targeted sequences may demand extraordinary quality control measures.

            Additional factors may confound the easy clinical application of CRISPR genetic modification tools.  Our genomes are not totally stable.  The aging process itself – the combined impacts of environment, lifestyle, genetics and errors that creep in as DNA is replicated – may influence the emergence of diseases such as cancer (5).  This inherent mutability suggests that to some extent CRISPR manipulation of human genomes will exhibit idiosyncratic outcomes.  A protocol that succeeds in a young patient may yield greater numbers of off-target changes in older subjects.  Issuing a prognosis for success and assessing the threat of off-target changes may only be approximate for each person and situation. 

            Hopefully further work will reveal CRISPR genomic manipulation will provide a useful and safe tool to control disease.  Understanding how often the genomic editor errs and the long term consequences of those unintended changes is now critical. 



(1) A. King. 2017.  Fears that Gene-Editing Cancer Trails Are Premature.  Chemistry World, 9 January 2017.

(2). I. Haydon. 2017. CRISPR Controversy Raises Questions About Gene-Editing Technique. The Conservation, 31 May 2017.

(3) H. Ledford. 2016.  Enzyme Tweak Boosts Precision of CRISPR Genome Edits.  Nature, 6 January 2016.

(4) N. Hijiya et al. 2009. Acute Leukemia as a Secondary Malignancy in Children and Adolescents:  Current Findings and Issues.  Cancer 115(1):23-35.

(5) H. Ledford. 2017.  DNA Typos to Blame for Most Cancer Mutations.  Nature, 23 March 2017.