01Why study this case — when the hype of a "miracle drug" did the harm
Iressa (generic name gefitinib) has a slightly different color from the drug disasters so far. The cause was neither a toxic solvent nor a contaminated biological material. The entry point to harm was the excessive expectation of "a miracle drug with few side effects."
Iressa is a molecularly targeted drug for lung cancer — an oral medicine that strikes only a specific molecule (EGFR) involved in cancer-cell growth. Because it does not attack normal cells the way conventional chemotherapy does, it was received as "a breakthrough drug with little nausea or hair loss." Yet soon after launch, deaths from fatal interstitial pneumonia (acute lung injury) followed one after another. This piece traces what happened with a drug carried by runaway expectation, and how post-marketing safety measures changed as a result.
02An approval ahead of the world — Japan went first, in 2002
Iressa is an EGFR tyrosine kinase inhibitor developed by the pharmaceutical company AstraZeneca. What stands out is that Japan approved it ahead of the rest of the world, in July 2002. For patients with advanced non-small-cell lung cancer in whom surgery and conventional chemotherapy work poorly, the hope for a new option was great.
That it was an oral drug, and was held to have milder side effects than conventional chemotherapy — these combined to spread Iressa rapidly as "an easy-to-use new drug." Within just a few months of approval, it was being prescribed to large numbers of patients.
03Interstitial pneumonia — in a drug that was supposed to have "few side effects"
Soon after launch, the unexpected happened. Patients taking Iressa developed acute interstitial pneumonia — a severe lung injury in which lung tissue rapidly hardens and breathing fails — and deaths followed one after another.
Molecularly targeted drugs carried an image of being "safe because they do not attack normal cells." In reality, targeted drugs too can cause severe, fatal adverse effects. The simple equation "targeted = safe" did not hold. Interstitial pneumonia progresses quickly once it appears and has few effective treatments; many patients died within a short time of diagnosis.
04The scale of harm, and the Emergency Safety Information
Just three months after approval, in October 2002, an Emergency Safety Information notice (Yellow Letter) was issued in response to the cluster of deaths from interstitial pneumonia. The package-insert warnings were strengthened, and the importance of careful use by specialists and early detection of onset was disseminated.
Even so, reports of deaths related to interstitial pneumonia continued to accumulate, reaching, it is said, the order of several hundred. The speed of approval and the speed of uptake had generated wide exposure before the risk was adequately understood.
05Litigation — and the Supreme Court did not find liability
Victims and bereaved families sued the state and the pharmaceutical company (the Iressa litigation, in Tokyo and Osaka). Here is what makes this case distinctive. Unlike the HIV-tainted blood and CJD cases, the courts ultimately did not find legal liability on the part of the state or the company.
At the district-court stage there were rulings that partly acknowledged liability, but the high courts and then the Supreme Court's 2013 ruling held that the package-insert language and the manner of approval at the time could not be called unlawful, and denied the legal liability of the state and the company. That harm actually occurred, and whether it can be recognized as legal negligence, were separate questions — this case presses hard on the distance between "responsibility" and "harm" in drug disasters.
06Afterward — learning to choose "the patients it works for"
Iressa's story has a sequel. At first it was used broadly without clarity about "whom it works for," but research advanced, and it became clear that it works especially well in patients with a specific mutation in the EGFR gene.
Today the standard is to test for EGFR gene mutation before treatment and select patients for whom benefit can be expected and the balance against risk holds. Under appropriate patient selection (companion diagnostics), Iressa became a drug of genuine value in lung-cancer treatment. One could say that the drug itself was not the problem; the core of the harm was that it spread before "for whom, and how" had been settled.
07Four lessons that remain today
Lesson 1 — "Few side effects" is not "no serious side effects"
Targeted drugs have lighter systemic toxicity than conventional chemotherapy. But "many mild side effects" and "no fatal side effects" are entirely different. A drug with a new mechanism of action can carry severe adverse events of a new kind. The intuition that "narrowing the target = safe" was a dangerous oversimplification.
Lesson 2 — Expectation clouds risk perception
The advance reputation of a "miracle drug" relatively lowered caution in the field. The greater the expectation placed on a product, the more easily its risk is underestimated. The same holds in the world of promotional materials and advertising, where language emphasizing efficacy can steal attention from safety information. Precisely when speaking of expectation, one needs the discipline to speak of risk with equal force.
Lesson 3 — Fast approval demands strong post-marketing surveillance
Approving ahead of the world has the value of reaching patients quickly, but it presupposes a system to capture, quickly after launch, the risks unknown at the time of approval. The Iressa experience led to strengthening early post-marketing safety surveillance (Early Post-marketing Phase Vigilance, EPPV). The speed of approval and the strength of post-marketing surveillance must be designed as a set.
Lesson 4 — Defining "whom it works for" reduces harm
Iressa became a drug whose value and risk balanced only once an indicator for patient selection (EGFR mutation) was established. Taking a broad indication and "just using it" loads risk alone onto people it does not help. Biomarker-based patient selection serves not only efficacy but also the safety aim of avoiding unnecessary exposure.
08Connections to other chapters
The Iressa case connects to other chapters of this site as follows.
- Drug Disasters 03 (DES) — a shared trait: it spread before proof of efficacy and safety had been settled
- Drug Disasters 05 (HIV-tainted blood) — like it in that the state's and company's responsibility was tested in court, but a contrasting case in the conclusion (whether liability was found)
- Advertising Regulation — directly tied to the discipline of information provision: balancing the emphasis on efficacy against safety information
- Ethics, Part 3 — the ethics of clinical judgment about whom to treat and how, between expectation and reality
Iressa was approved in Japan ahead of the world as "a miracle drug with few side effects." An oral medicine that frees patients from grueling chemotherapy — that hope was real. But the size of the hope relatively thinned the vigilance toward the fatal risk of interstitial pneumonia.
The courts ultimately did not find legal liability on the part of the state or the company. That harm actually occurred, and whether it can be called negligence, were separate. This case shows that "responsibility" in drug disasters does not always match the scale of harm. At the same time, as research advanced, Iressa proved to be a drug that genuinely works when the right patients are chosen.
Anyone in modern pharma who handles new drugs has learned to speak of risk, and of "whom to use it for," with the same force as they speak of expectation. Delivering fast, and watching over safety. What a new drug can cause when it lacks both of those wheels — Iressa is the record of it.