PharmaNews

AI Could Make Pharma’s Drug Testing Bottleneck Worse

Insilico used AI to select a lung fibrosis drug candidate in 18 months, and its Phase 3 trial aims to enroll 320 people to see if the early results hold up.

Answer
AI Could Make Pharma’s Drug Testing Bottleneck Worse

The brief

  1. Andreessen Horowitz (a16z) refers to the potential for drugmakers to experience 'drug candidate abundance' due to faster discovery and development processes.
  2. Formation Bio, an a16z portfolio company, has launched a $600 million fund to expand its portfolio of clinical-stage medicines.
  3. Insilico Medicine identified an enzyme called TNIK as a target for lung fibrosis and developed the drug rentosertib, which is currently in Phase 3 trials in China.

Deployment

Company
Insilico Medicine
System
AI for drug candidate selection
Stage
Live
trial enrollment
320

What happens when drugmakers can find more candidates than they have the time or money to test? Andreessen Horowitz (a16z) called the prospect “drug candidate abundance.” In its latest essay Big Pharma Can Rival Big Software, the venture firm argued that faster discovery and development could help drugmakers replenish their pipelines as blockbuster patents expire.

The brief

  1. Andreessen Horowitz (a16z) refers to the potential for drugmakers to experience 'drug candidate abundance' due to faster discovery and development processes.
  2. Formation Bio, an a16z portfolio company, has launched a $600 million fund to expand its portfolio of clinical-stage medicines.
  3. Insilico Medicine identified an enzyme called TNIK as a target for lung fibrosis and developed the drug rentosertib, which is currently in Phase 3 trials in China.

Deployment

Company
Insilico Medicine
System
AI for drug candidate selection
Stage
Live
trial enrollment
320

The idea comes from Formation Bio, an a16z portfolio company that acquires drug candidates and takes them through clinical development. “The world will have far more discovered drugs than it can actually develop,” wrote Benjamine Liu, Formation Bio’s Co-Founder and CEO. In October, the company announced a fund with $600 million in commitments to expand its portfolio of clinical-stage medicines.

What 18 Months Bought Insilico

Insilico Medicine used AI to identify an enzyme called TNIK as a potential target for lung fibrosis and design a molecule to inhibit it. The company took 18 months to go from its initial hypothesis to selecting a preclinical candidate. That candidate, rentosertib, entered human testing in 2021.

In a 71-person randomized Phase 2a trial for idiopathic pulmonary fibrosis, a disease that scars the lungs, average forced vital capacity, a measure of lung function, rose by about 98 milliliters in the highest-dose group after 12 weeks. It fell by about 20 milliliters in the placebo group.

Insilico dosed the first patient in its Phase 3 trial in China in September. The study aims to enroll 320 people and track lung function over 52 weeks. “This Phase III study is designed to validate the phase IIa study findings in a larger cohort,” said Zuojun Xu, the trial’s Lead Principal Investigator at Peking Union Medical College Hospital.

The Cost of Testing More Candidates

If AI generates more candidates, drugmakers may have to pay for more lab experiments before deciding which ones to test in patients. A16z expected that to increase demand for wet labs and contract research organizations.

Insilico Medicine Founder and CEO Alex Zhavoronkov said that even when an algorithm designs a molecule to fit its target, it cannot “guarantee a molecule won’t trigger unpredictable liver toxicity once ingested.”

“The next problem is the clinical trials still take many, many years,” said Demis Hassabis, Co-Founder and CEO of Google DeepMind, in an April interview with 20VC. He said AI could eventually help researchers model how the body processes a drug and match patients to treatments based on their genetic makeup.

The Revenue Drugmakers Have to Replace

AbbVie’s global Humira sales fell from $21.2 billion in 2022 to $4.5 billion in 2025 as biosimilars entered the U.S. market. For a16z’s bet to hold up, drugmakers would need to show across multiple programs that AI helps them bring replacement medicines to market faster or at a lower total development cost, including what they spend on candidates they ultimately drop.

Written by Amit Raja Naik

Join the discussion

0 answers

People who run these systems answer here. Your verified role is shown, never your employer unless you choose.

Sign in with LinkedIn to answer

Read next

MachineCon USA · from AIM Media House

Where enterprise AI is decided

An invitation-led day for the CIOs, CDOs and Chief AI Officers who own enterprise AI strategy.

  • New York5th edition · 583 Park Avenue
  • Dallas1st edition

Register for MachineCon USASee MachineCon USA →

  • Keynote on the main stage, MachineCon New York 2025
  • AI150 award winners on stage, MachineCon New York 2025
  • At the tables, MachineCon New York 2025
  • Networking, MachineCon New York 2025