Lessons for cancer drug treatment from tackling a non-cancerous overgrowth syndrome

Abnormal activity of the enzyme PI3K can drive cancer growth, and mutations in a PI3K subunit can sometimes lead to non-cancerous overgrowth. A cancer drug that inhibits PI3K dramatically reduces such overgrowth.
Robert K. Semple is in the Centre for Cardiovascular Science, University of Edinburgh, Edinburgh EH16 4TJ, UK.

Search for this author in:

Bart Vanhaesebroeck is in the UCL Cancer Institute, University College London, London WC1E 6BT, UK.

Search for this author in:

Researchers who investigate rare genetic conditions live in hope that the discovery of disease-causing mutations will lead swiftly to tailored treatments. Sadly, this is not often the case, because genetic defects usually cause impairments that are difficult or impossible to tackle using available medicines. Writing in Nature, Venot et al.1 now provide a rare exception to this rule. In severe non-cancerous overgrowth syndromes caused by mutations in an enzyme called PI3K, they show the beneficial effects of a PI3K-inhibitor drug that was initially developed to treat cancer. Their results bring the possibility of a transformative therapy for people with overgrowth conditions one step closer.

The development of humans, from a single fertilized egg to an adult body that contains around 37 trillion cells2 while maintaining symmetrical, paired body parts, is an astonishing feat that requires the lifelong coordination of cell division, survival and death. Growth-factor proteins can aid cellular coordination by acting on cell-surface receptors to stimulate intracellular signalling networks. These networks often include PI3K, which is essential for the regulation of growth and development by insulin and insulin-like growth-factor hormones.

Cancer arises from a flagrant breach of the rules of good cellular citizenship that are essential in multicellular organisms, and cancer cells acquire genetic abnormalities that subvert the checks and balances that constrain cell growth and migration. Mutations that activate PI3K signalling — mainly those in the gene PIK3CA, which encodes p110α, a catalytic subunit of PI3K — are among the most common mutations to drive solid cancers3. Such signalling can also be activated by mutations that inactivate the enzyme PTEN, which normally keeps PI3K activity in check. The link between overactive PI3K signalling and cancer motivated researchers to develop compounds known as PI3K inhibitors. However, the clinical impact of these drugs on cancer has been less impressive than hoped because of toxicity associated with high doses. And even when such drugs succeed in inhibiting activated PI3K, other proteins can compensate to provide alternative pathways that promote cancer4.

In 2012, certain PIK3CA mutations, which had previously been linked to cancer, were reported to cause rare, non-cancerous forms of overgrowth in people57. A hallmark of these overgrowth syndromes is abnormal, excessive tissue growth that affects the body in a patchy and asymmetrical manner. This overgrowth is caused by PIK3CA mutations that occur after the start of embryonic development and only in some cells5,6,8, which leads to cellular overgrowth in a mosaic-like pattern. The severity of the condition varies from person to person, and ranges from an isolated skin growth to a complex multisystem disorder called CLOVES syndrome5, which comprises considerable and often widespread overgrowth that contains an abundance of fat cells and abnormal blood vessels. PIK3CA mutations are a common feature of many overgrowth syndromes and the term PROS (for PIK3CA-related overgrowth spectrum)8 is used as a unifying description of such cases.

PROS disorders do not seem to be linked to an increase in the risk of forming the solid cancers in which PIK3CA mutations are most prevalent9. Although the reason for this is unclear, the PIK3CA mutations associated with such disorders usually occur in cell types of a different embryonic origin from those that develop cancer linked to PIK3CA mutations9.

Severe PROS disorders can be debilitating or even life-threatening. Overgrown tissue causes compression that can lead to vascular problems or organ dysfunction. Treatments aim to reduce excess tissue by surgery or by the physical blockade of enlarged blood vessels, and other therapy options are needed urgently. The availability of targeted inhibitors of p110α that had already undergone clinical testing as treatments for cancer gave researchers hope that these might offer a new therapy for PROS disorders. Yet questions arise about the effect of prolonged patient exposure to these drugs. Would this cause side effects? Would their cells adapt to dull the effect of such treatment, as occurs in cancer4? And would the overgrowth be amenable to reversal by drug-based therapy?

Venot et al.1 take an important step towards addressing these questions. Previous attempts to model PROS disorders in mice engineered to express Pik3ca containing disease-causing mutations produced excess growth in only some of the expected tissues10. Venot and colleagues engineered another mouse model of a PROS disorder, in which an artificial system was used to make the mice express constitutively active p110α in all tissues. These animals developed problems similar to those in people with PROS conditions, including the overgrowth of adipose, muscle and vascular tissue, and experienced a premature death caused by vascular complications. When the authors treated the mice with a PI3K inhibitor called alpelisib, an impressive, rapid and substantial decrease in the amount of overgrown tissue occurred, which prevented the premature death of the animals.

Crucially, Venot et al. then assessed the effects of alpelisib in 19 people with PROS disorders who had severe or life-threatening complications. In adults, the team administered the lowest dose that had been tested in trials on people with cancer (250 milligrams a day), and in children they used a dose of 50 milligrams a day. Dramatic anatomical and functional improvements occurred in all patients across many types of affected organ (Fig. 1), with some benefits noted within days of the treatment starting. The study was not randomized, blinded or subject to placebo control, yet these striking initial results suggest that this outcome is likely to have clinical importance. Resistance to alpelisib was not observed, and the drug was well tolerated by the recipients.

Figure 1 | People who have an overgrowth syndrome respond to treatment with a cancer drug. Venot et al.1 investigated a syndrome linked to abnormal activation of the enzyme PI3K, which can result in the non-cancerous overgrowth of a variety of tissues. The authors tested whether a low dose of a PI3K inhibitor called alpelisib, developed previously as a cancer therapy, could treat people who have mutations in the gene PIK3CA, which encodes the catalytic subunit of PI3K. In overgrowth syndromes, these PIK3CA mutations arise in a mosaic patchwork pattern5,6,8 in the region of the affected tissue (pink). Alpelisib treatment caused substantial improvements in the 19 recipients. Two examples of the decrease in overgrown tissue in patients after six months of drug treatment are shown.

A predicted side effect of PI3K inhibition is a high blood glucose level, caused by interference with the PI3K-mediated metabolic effects of insulin. However, blood-glucose elevation occurred in only three people, in whom the elevation was modest. In children, the drug did not have an effect on normal growth, which suggests that overgrown tissue can be targeted without harmfully blocking PI3K-dependent childhood growth. Further systematic clinical studies are now needed, and ethics committees will have to assess whether it could be justified to include a placebo in trials on patients who are severely affected.

The study of PI3K inhibition as a treatment for PROS disorders might also offer something in return towards the design of cancer therapies. The aim of PI3K-inhibitor therapy in PROS conditions would be to suppress disease-causing levels of PI3K signalling, while minimizing any side effects during the long-term, and probably lifelong, treatment. By contrast, the conventional approach of cancer therapy involves identifying differences between healthy and cancerous cells, and then hitting the cancer-specific characteristics as hard as possible to induce the death of cancer cells. In clinical trials for cancer, PI3K inhibitors are usually studied at the maximum tolerated dose. Yet whether this makes sense is unclear, given that the activity level of mutated p110α in cancer, which is low, is probably similar to the activity level of the same subunit in PROS disorders. Could a low dose of PI3K inhibitor be beneficial in treating a cancer linked to a PIK3CA mutation? This might abolish any PI3K activity that is above the usual level without completely blocking PI3K signalling, as would occur with a high dose of inhibitor. This could be tested, for example, as a strategy for preventing types of cancer in which PI3K activation or PTEN inactivation is an early event11,12, or for preventing PTEN hamartoma tumour syndrome in people who carry a mutation in PTEN and are therefore prone to cancer13.

Low-dose PI3K inhibition might also be used as an option, following conventional treatments such as chemotherapy or surgery, for slowing the evolution of cancer and its adaptation to selective pressures14. And long-term, low-dose PI3K inhibition could offer further benefits — for example, it increases the metabolic health of obese mice and rhesus monkeys15, and sustained blockade of the PI3K pathway can slow ageing in animal models16. Perhaps it is time to target abnormal signalling in cancer with a lighter touch, which could enable the use of combination therapies that are currently precluded for reasons of toxicity. After all, there is no need to use a hammer to kill a fly, and this principle might also apply to treating cancer.

Nature 558, 523-525 (2018)

doi: 10.1038/d41586-018-05365-w


  1. 1.

    Venot, Q. et al. Nature 558, 540–546 (2018).

  2. 2.

    Bianconi, E. et al. Ann. Hum. Biol. 40, 463–471 (2013).

  3. 3.

    Samuels, Y. et al. Science 304, 554 (2004).

  4. 4.

    Rozengurt, E., Soares, H. P. & Sinnet-Smith, J. Mol. Cancer Ther. 13, 2477–2488 (2014).

  5. 5.

    Kurek, K. C. et al. Am. J. Hum. Genet. 90, 1108–1115 (2012).

  6. 6.

    Lindhurst, M. J. et al. Nature Genet. 44, 928–933 (2012).

  7. 7.

    Rivière, J.-B. et al. Nature Genet. 44, 934–940 (2012).

  8. 8.

    Keppler-Noreuil, K. M. et al. Am. J. Med. Genet. A 167, 287–295 (2015).

  9. 9.

    De Santis, M. C. et al. Cancers 9, 30 (2016).

  10. 10.

    Castillo, S. D. et al. Sci. Transl. Med. 8, 332ra43 (2016).

  11. 11.

    Gerstung, M. et al. Preprint at bioRxiv (2017).

  12. 12.

    Jamal-Hanjani, M. et al. N. Engl. J. Med. 376, 2109–2121 (2017).

  13. 13.

    Yehia, L. & Eng, C. Endocr. Relat. Cancer (2018).

  14. 14.

    Sansregret, L., Vanhaesebroeck, B. & Swanton, C. Nature Rev. Clin. Oncol. 15, 139–150 (2018).

  15. 15.

    Ortega-Molina, A. et al. Cell Metab. 21, 558–570 (2015).

  16. 16.

    Bettedi, L. & Foukas, L. C. Biogerontology 18, 913–929 (2017).

Download references

Competing Financial Interests

Bart Vanhaesebroeck is a consultant to Karus Therapeutics (Oxford, UK), iOnctura (Geneva, Switzerland) and Venthera (Palo Alto, California).

Nature Briefing

An essential round-up of science news, opinion and analysis, delivered to your inbox every weekday.