Drug-Induced Pulmonary Toxicity: Respiratory Safety Assessment in Clinical Development and Pharmacovigilance

Drug-induced pulmonary toxicity includes adverse effects on the airways, interstitium, alveoli, pulmonary vasculature, pleura and control of breathing. Because respiratory symptoms and imaging abnormalities have many alternative causes, assessment must begin with phenotype definition, not with the assumption that temporal association proves drug causality.

Purpose and Scientific Framework

A respiratory safety assessment may integrate:

• symptoms and physical examination;
• oxygenation and respiratory severity;
• chest imaging;
• pulmonary function testing;
• exposure and latency;
• infection assessment;
• cardiac and thromboembolic causes;
• underlying lung disease;
• concomitant medicines and radiation;
• dechallenge and rechallenge where informative; and
• biologic or pharmacologic plausibility.

ICH S7A includes respiratory assessment within the nonclinical safety-pharmacology core battery. It helps identify effects on vital physiological functions before or during development, but it does not provide a complete clinical diagnostic algorithm for drug-induced lung disease.

Major Pulmonary Phenotypes

Interstitial Lung Disease and Pneumonitis

Drug-induced interstitial lung disease (DI-ILD) and pneumonitis are important because they can be serious, radiologically heterogeneous and difficult to distinguish from infection, malignancy, autoimmune disease, radiation injury or pre-existing lung disease.

A report of “pneumonitis” should therefore trigger reconstruction of symptoms, imaging, oxygen requirement, infection work-up, exposure timing and competing causes rather than automatic attribution to the product.

Pulmonary Fibrosis

Fibrotic toxicity may produce persistent dyspnoea, reduced diffusion capacity, restrictive physiology and progressive imaging abnormalities. The reviewer should distinguish active inflammatory injury from established fibrosis because reversibility and treatment implications differ.

Bronchospasm and Airway Effects

Bronchospasm, wheeze or airway hyperreactivity may reflect direct pharmacology, hypersensitivity, formulation effects or exacerbation of underlying asthma or chronic lung disease. Baseline respiratory status and objective response to treatment can materially affect interpretation.

Pulmonary Vascular Effects

Medicines can contribute to pulmonary hypertension, vascular injury or thromboembolic events through different mechanisms. These phenotypes require cardiovascular and thromboembolic evaluation rather than being grouped indiscriminately under “lung toxicity.”

Pulmonary Oedema and Respiratory Depression

Pulmonary oedema may be cardiogenic, permeability-related or secondary to fluid shifts. Respiratory depression is mechanistically different again: central or neuromuscular suppression can impair ventilation without primary structural lung injury.

Functional Versus Structural Respiratory Effects

Pulmonary function and structural injury are related but not interchangeable. A patient can have major bronchospasm with little structural imaging change, or radiographic pneumonitis before large spirometric abnormalities appear.

Normality of one test therefore does not exclude clinically important toxicity. The relevant test depends on the suspected phenotype.

Individual-Case Assessment

A practical case review should define onset, symptom pattern, severity, oxygenation, imaging, pulmonary-function findings, infection assessment, treatment exposure, concomitant medicines and underlying pulmonary disease.

Imaging

High-resolution CT can be highly informative in suspected ILD or pneumonitis, but imaging patterns are not uniquely drug specific. Organising pneumonia, diffuse alveolar damage, hypersensitivity-type patterns and other appearances can occur from several causes.

The reviewer should therefore use imaging to characterise the phenotype and disease extent rather than to assign drug causality automatically.

Pulmonary Function Testing

Spirometry, lung volumes and diffusion capacity may support assessment, especially for restrictive, obstructive or gas-transfer abnormalities. A fall in DLCO can be important in context but is non-specific and should be interpreted with haemoglobin, pulmonary vascular disease, emphysema and technical variability in mind.

Infection and Alternative Causes

Infection is one of the most important competing explanations for pneumonitis-like presentations, particularly in immunocompromised patients. Malignancy progression, radiation injury, heart failure, pulmonary embolism, autoimmune disease and pre-existing ILD may also mimic drug toxicity.

Bronchoscopy or bronchoalveolar lavage can be useful in selected cases, especially to investigate infection or inflammatory patterns, but neither is required in every case.

Clinical-Development Assessment

Population-level assessment should consider:

• incidence of defined respiratory phenotypes;
• treatment-versus-control patterns;
• time to onset;
• dose and exposure;
• baseline pulmonary risk factors;
• imaging-confirmed cases;
• oxygen requirement and serious outcomes;
• treatment interruption and recovery; and
• clustering by mechanism or combination therapy.

A few well-characterised serious ILD cases may be more informative than many non-specific cough reports.

Post-Marketing Pharmacovigilance

Follow-up should target the information most useful for differential diagnosis: imaging, oxygenation, infectious studies, pulmonary function, competing medicines, radiation, disease progression, dechallenge and outcome.

Signal evaluation should group clinically coherent phenotypes. Bronchospasm, ILD, pulmonary embolism and respiratory depression may all be respiratory adverse events but usually require different causal and mechanistic analyses.

Product-Level Interpretation

A defensible conclusion should state whether evidence supports a defined pulmonary risk, an unresolved signal, a class-related concern, an indirect respiratory effect or a more plausible alternative explanation. Uncertainty should remain visible when diagnostic information is incomplete.

Practical Medical Review Sequence

1. Define the respiratory phenotype.
2. Establish baseline lung disease and relevant exposures.
3. Reconstruct symptoms, oxygenation, imaging and pulmonary function longitudinally.
4. Exclude infection and other major competing diagnoses proportionately.
5. Assess treatment timing, dose and concomitant therapies.
6. Evaluate dechallenge/recovery and any natural rechallenge information.
7. Integrate other cases, mechanism and comparator evidence.
8. State the product-level conclusion with appropriate uncertainty.

Potential Failure Modes

The following are illustrative failure modes, not reported inspection findings.

Governance and Inspection Considerations

For an important pulmonary signal, an assessor may ask how the organisation defined the case phenotype, obtained imaging and infectious follow-up, considered competing causes, grouped similar cases and translated the conclusion into product information or risk management. No universal ILD adjudication committee or fixed imaging schedule applies to all products.

Practical Review Checklist

1. Is the respiratory phenotype clinically coherent?
2. Are baseline lung disease and smoking/environmental history considered where relevant?
3. Are imaging and pulmonary-function results interpreted within their limitations?
4. Has infection been addressed proportionately?
5. Are cardiac, thromboembolic and disease-progression alternatives considered?
6. Is treatment timing plausible?
7. Are cases grouped by mechanism rather than organ-system label alone?
8. Does the conclusion distinguish association from causality?

Relationship With Other QPPV.com Articles

This article complements [[causality-assessment-in-pharmacovigilance]], [[signal-assessment]] and the other organ-toxicity articles in the Drug Safety section.

Key Takeaways

Drug-induced pulmonary toxicity is a collection of distinct phenotypes, not a single diagnosis.

ILD and pneumonitis require particularly careful differential diagnosis because infection, malignancy, radiation and underlying disease can mimic drug toxicity.

Imaging and pulmonary function support phenotype characterisation but do not independently determine causality.

References

1. International Council for Harmonisation. ICH S7A: Safety Pharmacology Studies for Human Pharmaceuticals.
2. International Council for Harmonisation. ICH M3(R2): Nonclinical Safety Studies for the Conduct of Human Clinical Trials and Marketing Authorization for Pharmaceuticals.
3. European Medicines Agency. GVP Module IX — Signal management (Rev. 1), with current EMA implementation material.
4. Relevant contemporary respiratory-medicine guidance and peer-reviewed literature on drug-induced interstitial lung disease should be consulted for product- and phenotype-specific assessment.

Regulatory Note

ICH S7A provides a nonclinical safety-pharmacology framework for vital functions, including respiratory function. It does not mandate a universal clinical ILD diagnostic algorithm. Product-specific monitoring, adjudication and risk-minimisation measures should be proportionate to the observed risk and applicable regulatory commitments.