최근 면역항암제 옵디보의 전략은

 

최근 면역항암제의 임상시험에 참가하였던 전문의가 다음 글에서와 같이 치료 사례를 말한 가운데, 차세대 면역항암제로 주목받고 있는 '옵디보'가 단독요법으로 진행한 임상 3상에서 실패하였다고 전해집니다. 이는 일본의 오노약품공업과 다국적제약사 BMS제약이 공동으로 개발 중인 대표적인 면역항암제입니다.

 

BMS측은 과거 치료전력이 없는 진행성 비소세포폐암 환우를 대상으로 옵디보 단독요법을 시행한 임상 3상(CheckMate-026)에서 당초 환우의 병이 진행되지 않고 생존하는 '무진행 생존기간(progression-free survival)'을 연장하고자 하는 목표를 세웠으나 이를 달성하지 못했다는 것입니다.

 

 

 

하지만 이번 임상시험의 실패는 옵디보 면역항암제의 치료 효과의 문제라기 보다는 키트루다가 PD-L1 발현이 양성(50% 이상)으로 나타난 비소세포폐암 환우를 대상으로 한 것과는 달리, 애시당초 옵디보 임상시험에 입각하여 5% 이상의 비소세포폐암 환우 541명을 대상으로 하였다는 것입니다.

 

BMS측은 이와 같은 결과는 단지 중간 결과로서 10월 이후에 최종 결과가 나올 것이며, 현재 2가지 약제를 함께 사용하는 병용요법 임상은 여전히 진행 중이라고 밝혔습니다. 이를 두고 다음의 포브스(Forbes)지가 다룬 논평은 참조할만하다 할 것입니다.

 

Bristol-Myers Squibb's Defeat Is A Victory For Personalized Medicine

by Matthew Herper

 

Yesterday, shares of drug giant Bristol-Myers Squibb fell 16% after its breakthrough cancer drug, Opdivo, failed to slow tumor growth in previously untreated patients with non-small cell lung cancer.

The news shocked Wall Street in part because in June, Merck had announced that its very similar (possibly identical) drug, Keytruda, had extended the lives of patients with non-small cell lung cancer. There was one big difference between the two studies: the results of a diagnostic test.

Both Opdivo and Keytruda work by freeing white blood cells to kill tumors. Healthy cells protect themselves from white blood cells by emitting a "don't kill me" signal called PD-L1 (for programmed cell death ligand one). PD-L1 hooks up to a protein on the white blood cell called PD-1. Cancer cells are clever, and some of them release streams of PD-L1, leaving white blood cells to leave them alone.

Opdivo and Keytruda work by blocking PD-1. Not surprisingly, they work best against tumors that are producing lots of PD-L1. Tumors that are not doing so are probably finding other ways to evade the immune system.

Merck chose to run its lung cancer trial in patients whose tumors had PD-L1 levels of at least 50%. But Bristol-Myers Squibb chose to test patients who had PD-L1 levels of more than 5%. Although Bristol has not yet released full data from its study (that will likely come in October at a medical meeting), this difference seems the most likely explanation for its failure.

There were both ethical and business justifications for what Bristol did. The tests for PD-L1 are imperfect, and they miss patients. Setting the threshold at 50% almost certainly means that patients who would benefit from Opdivo and Keytruda don’t get it. There’s no reason a patient with 40% would not benefit, for instance.  "I think we obviously took a risk to be more bold, to try to cover a broader population of patients, because our strategy was really to never leave patients behind," Fouad Namouni, Bristol’s head of oncology drug development told me yesterday.

Until now, this strategy has benefited Bristol. Sales of Opdivo were $942 million last year, 66% more than those of Keytruda. The reason: in previous trials in melanoma and patients with lung cancer who failed other therapies, Bristol took this all-comers approach and won. That means doctors can prescribe Opdivo to any patient who might benefit without first screening them with a diagnostic test.

Merck started out behind Bristol, and has caught up by doing studies in patients whose tumors express lots of PD-L1. Until now, that has meant it lagged. But now it will have exclusive access to the 25% of non-small cell lung cancer patients whose tumors have lots of PD-L1.

Bristol could have still tried to reach a broader market and run a more conservative trial. Its study was designed to test patients first in a group that had PD-L1 of more than 5%, and then in a group with even lower PD-L1 levels. If the company had instead chosen a 20% or 30% PD-L1 threshold, it might have won on this first measure, and patients would have won, too.

It’s tempting to avoid pairing a diagnostic test with a drug because some patients will be left behind. But the alternative is that patients who don’t get any benefit will bear the side effects – and the great financial cost – of a cancer drug that won’t help them. There’s been a temptation to portray immune-system-boosting drugs as different from other gene-targeted medicines. But the goal still has to be to get patients the right medicine for them.\

Bristol-Myers is far from sunk. For one thing, insurance companies may still decide to pay for Opdivo in non-small cell lung cancer if an after-the-fact look at the data shows a benefit for patients with high PD-L1 levels. This would allow insurers to play the two drugs against each other on cost. For another, Bristol has a study ongoing that combines Opdivo with its other immune-system-booster, Yervoy. If that proves effective, it could be the first choice for many patients. But the company is paying dearly for being too cocky about the need to pair its cancer drugs with diagnostic tests. And other drugmakers will be loathe to repeat that mistake.

 

결론적으로 면역항암제인 옵디보와 키트루다는 동일한 기전으로 PD-L1 발현율이 높은 비소세포폐암 환우에게 반응율이 뛰어나타는 것은 자명하다는 것입니다. 하지만 BMS측은 이 발현율이 치료 효과를 예측하기에 유용하다는 것은 인정하지만 절대적이지 않다는 입장으로 임상시험의 근거로 5% 이상의 환우이면 누구나 처방이 가능토록 하겠다는 것입니다.

 

이러한 BMS측은 입장은 PD-L1 발현율이 비록 50% 미만인 비소세포폐암 환우들에게도 여전히 드라마틱한 치료 효과가 나타날 수 있다는 윤리적 정당화의 일환일 수도 있고, 혹은 5% 이상이면 어떤 환우도 결코 놓치기 싫다는 BMS의 비즈니스 전략일 수도 있다는 것입니다. 다만 그 전략이 윤리적이든 비즈니스적이든 PD-L1 발현율 5% 이상이지만 50% 미만인 환우들에게 급여심사를 앞두고 여전히 큰 희망이 된다는 것입니다.