mp_2_tumeurUnique
π Health and biotech
Personalised medicine: custom healthcare on a national scale?

Target tumours, improve survival

Agnès Vernet, Science journalist
On February 2nd, 2021 |
3 mins reading time

Alexis Gautreau
Alexis Gautreau
CNRS research director in biology at the Laboratory of Structural Biology of the Cell (BIOC*) at the Ecole Polytechnique (IP Paris)
Key takeaways
  • Oncology was the first medical field to start using personalised medicine.
  • Whilst spectacular results have shown increased survival rates for cancer patients, there are certain limits to these techniques such as development of tumour resistance to treatments.
  • Now, 20 years on, this innovative approach has delivered on its promises, but it has also raised new issues that biomedical research is yet to resolve.

Per­son­alised med­i­cine, some­times referred to as pre­ci­sion med­i­cine, owes its ori­gins to exten­sive devel­op­ments in DNA sequenc­ing tech­niques. It is an approach which aims to pro­vide treat­ments tai­lored to patients that tar­get the spe­cif­ic mech­a­nisms caus­ing their dis­ease, par­tic­u­lar­ly can­cer. Mol­e­c­u­lar pro­fil­ing of tumours, for exam­ple, has revealed their remark­able diver­si­ty. Their char­ac­ter­is­tics have been used to iden­ti­fy dis­tinct patient sub­groups from those which pre­vi­ous­ly seemed sim­i­lar in terms of oth­er clin­i­cal cri­te­ria. As such, treat­ment qual­i­ty can be vast­ly improved by treat­ing patients in each sub­group with an opti­mal med­ica­tion plan.

“In a per­fect world, we would be able to design a dif­fer­ent treat­ment for each patient based on the mol­e­c­u­lar char­ac­ter­is­tics of their tumour,” notes Alex­is Gautreau, biol­o­gy pro­fes­sor at Ecole poly­tech­nique. “But per­son­alised med­i­cine is not only about treat­ment. Rather, it is the com­bi­na­tion of mol­e­c­u­lar diag­nos­tics and med­ica­tion,” with diag­no­sis pro­vid­ing the basis for the choice of medication.

Spec­tac­u­lar results

Results have been impres­sive, as shown by trastuzum­ab used in the treat­ment of breast can­cer; one of the first “per­son­alised” med­ical treat­ments. It pre­cise­ly tar­gets some of the most aggres­sive can­cer cells, while leav­ing healthy cells intact. As a result, trastuzum­ab is now a cru­cial part of treat­ment for breast can­cer patients with a par­tic­u­lar genet­ic vari­ant. Com­pared to “stan­dard” treat­ment with emtan­sine, it has increased pro­gres­sion-free sur­vival by at least three months and over­all sur­vival by over ten per­cent. It is now admin­is­tered in com­bi­na­tion with con­ven­tion­al chemother­a­py and increas­es patient sur­vival rate by 10%.

Oth­er approach­es direct­ly tar­get genet­ic muta­tions. This is how treat­ments for cuta­neous melanoma (skin can­cer) with the BRAF-V600 muta­tion work. They specif­i­cal­ly inhib­it the mutat­ed form of an enzyme with­in the tumour, with­out affect­ing the nor­mal pro­tein in the rest of the organ­ism. In oth­er words, by cor­rect­ing the mis­func­tion caused by the muta­tion it treats the can­cer at its root.

Many of the break­throughs in per­son­alised med­i­cine are the result of col­lab­o­ra­tion between aca­d­e­m­ic researchers, start-ups and indus­try. Com­pa­nies realised that treat­ments that were unsuc­cess­ful on a large group of patients could still work on a small­er, spe­cif­ic sub­group. Hence, tak­ing anoth­er look at mol­e­cules, which had pre­vi­ous­ly been dis­card­ed due to uncon­clu­sive results. This “drug repo­si­tion­ing” approach saves time and mon­ey as it bypass­es some of the pre-clin­i­cal and ear­ly clin­i­cal devel­op­ment phases.

What’s more, using tar­get­ed treat­ments, the loca­tion of can­cer in a patient is not nec­es­sar­i­ly the deter­min­ing fac­tor in treat­ment suc­cess. For instance, the same med­ica­tion can often be used to treat patients with pan­cre­at­ic and lung can­cer, alike.

“Tar­get­ed ther­a­py works very well and pro­duces few­er side effects than con­ven­tion­al treat­ments,” Gautreau says. “After receiv­ing these treat­ments, peo­ple start believ­ing in mir­a­cles. How­ev­er, relaps­es are still commonplace.”

Not the end of the story

Indeed, after months of tumour decline, doc­tors often observe relaps­es as the can­cer evades tar­get­ed ther­a­py. One of the rea­sons for this is that all tumours are dif­fer­ent. Not only do doc­tors observe genet­ic dif­fer­ences in tumours from dif­fer­ent patients; they also find vari­a­tions with­in one tumour, as one can­cer cell dif­fers from another.

Tar­get­ed ther­a­py is effec­tive against a tumour when it is effec­tive against the major­i­ty of cells. Some cells die under the effect of the treat­ment, while oth­ers, with dif­fer­ent muta­tions, resist. Out of the over­all patch­work of can­cer­ous cells, cells that don’t respond to treat­ment even­tu­al­ly form a new tumour, which replaces the orig­i­nal one.

For this rea­son, col­lab­o­ra­tion between doc­tors and biol­o­gists is cru­cial. “Biol­o­gists have lots of mod­els, in the form of cell cul­tures or mice, for exam­ple. But these do not nec­es­sar­i­ly do a good job in show­ing how the dis­ease pro­gress­es in humans. To rep­re­sent human tumour that is 3 cm in diam­e­ter, we use a 1 mm tumour in a mouse, which has much few­er cells and there­fore less diver­si­ty. We know how to cure mouse can­cer! For humans, it’s more com­pli­cat­ed,” Gautreau acknowledges.

Still in the dark

A com­bi­na­tion of tar­get­ed ther­a­pies would seem to be a log­i­cal solu­tion. But tumours still man­age to escape. “We have observed escape phe­nom­e­na that we do not under­stand,” Gautreau empha­sizes. “It would seem that can­cer­ous cells can change when attacked by med­ica­tion. They mutate in order to sur­vive.” In real-life, this phe­nom­e­non occurs even more fre­quent­ly than math­e­mat­i­cal mod­el­ling predicts.

As such, bio­med­ical researchers are turn­ing to AI in the hopes of solv­ing the mys­tery. “If the ques­tion could be cracked using com­put­ers alone, that would be great,” he admits. “But I think that we have put enough effort in to realise that we need more exper­i­men­tal data to chew over, and a bet­ter under­stand­ing of the laws of biol­o­gy at play.”