QPPV.COM

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

Understand how potential drug-induced pulmonary toxicity is recognised and medically evaluated, how respiratory phenotypes are characterised, how imaging and pulmonary function testing support assessment, and how potential pulmonary safety signals are evaluated in clinical development and pharmacovigilance.

Audio Lesson 11 min
Knowledge Assessment Test your understanding of this article. Take the assessment →

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

Introduction

Drug-induced pulmonary toxicity encompasses a broad group of respiratory adverse effects caused or contributed to by medicinal products. These effects can involve the airways, alveoli, interstitium, pulmonary vasculature, pleura, respiratory muscles or central control of breathing.

The clinical spectrum is wide. A drug may cause mild cough or bronchospasm, clinically important interstitial pneumonitis, pulmonary fibrosis, pulmonary hypertension, pulmonary edema, respiratory depression or acute respiratory failure.

Among these phenotypes, drug-induced interstitial lung disease (DI-ILD) and drug-induced pneumonitis are particularly important in pharmacovigilance because they may be difficult to diagnose, have substantial clinical consequences and often require exclusion of several competing causes.

Drug-induced pulmonary toxicity is therefore not identified by a single laboratory threshold.

There is no generally applicable pulmonary equivalent of Hy's Law.

Instead, respiratory safety assessment requires integration of:

The diagnosis of drug-induced lung injury is frequently one of exclusion. A temporal relationship to treatment is important, but temporal association alone does not establish causality.

International respiratory guidance emphasises the importance of detailed medication and exposure histories when evaluating interstitial lung disease of uncertain cause. [1]

For pharmacovigilance professionals, the central task is therefore to reconstruct the respiratory phenotype and determine whether the medicinal product provides the most plausible explanation.

Learning Objectives

After reading this article, the reader should be able to:

  1. Describe the major phenotypes of drug-induced pulmonary toxicity.
  2. Explain why drug-induced interstitial lung disease and pneumonitis are particularly important safety signals.
  3. Distinguish pulmonary function abnormalities from structural pulmonary injury.
  4. Understand the roles of oxygen saturation, pulmonary function testing and DLCO.
  5. Explain the importance of chest radiography and high-resolution CT.
  6. Recognise major competing causes of pulmonary deterioration.
  7. Evaluate the temporal relationship between treatment and respiratory events.
  8. Understand the role and limitations of bronchoscopy and bronchoalveolar lavage.
  9. Assess pulmonary safety at the individual-subject and population levels.
  10. Apply a structured medical-review framework to potential drug-induced pulmonary toxicity.

Understanding Drug-Induced Pulmonary Toxicity

What Is Drug-Induced Pulmonary Toxicity?

Drug-induced pulmonary toxicity is respiratory dysfunction or structural lung injury caused or contributed to by a medicinal product.

The term encompasses several distinct biological processes.

A drug may:

The same medicinal product may produce more than one pulmonary phenotype.

For example, a treatment may cause pneumonitis in one patient and bronchospasm in another, depending on host susceptibility and mechanism.

The pharmacovigilance assessment should therefore begin with the clinical phenotype rather than the adverse-event term alone.

Why the Lung Is Vulnerable

The lung has a large surface area and is continuously exposed to the circulating drug and its metabolites.

Pulmonary toxicity may also be influenced by:

Drug-induced lung disease may be particularly difficult to recognise because respiratory symptoms are common in the general population and in patients with serious underlying disease.

Functional Versus Structural Pulmonary Toxicity

Pulmonary toxicity may produce a measurable functional abnormality before structural disease becomes obvious.

Conversely, radiographic abnormalities may occur before severe changes in routine pulmonary function.

Examples include:

The medical reviewer should therefore avoid treating a normal individual test as proof that pulmonary toxicity is absent.

Major Pulmonary Toxicity Phenotypes

Drug-Induced Interstitial Lung Disease

Drug-induced interstitial lung disease is one of the most important serious pulmonary safety phenotypes.

It represents a heterogeneous group of inflammatory and fibrotic disorders involving the pulmonary interstitium and adjacent lung structures.

Clinical manifestations may include:

Drug-induced ILD may present acutely, subacutely or chronically.

The radiographic pattern may vary substantially between drugs and patients.

A detailed review of medications and other exposures is therefore an important component of the diagnostic assessment of ILD. [1]

Drug-Induced Pneumonitis

Pneumonitis generally refers to an inflammatory pulmonary process that may be caused by a drug, infection, radiation, autoimmune disease or another exposure.

In pharmacovigilance, the term should not automatically be equated with drug causality.

A report of "pneumonitis" requires evaluation of:

This distinction is particularly important in oncology, where infection, malignancy, radiation injury and immune-mediated disease may all produce similar presentations.

Pulmonary Fibrosis

Some drug-related pulmonary injuries may progress to fibrosis.

Fibrotic disease may cause:

Fibrosis may be irreversible or only partially reversible.

The assessment therefore needs to distinguish active inflammatory injury from established structural fibrosis.

Bronchospasm and Airway Toxicity

Drugs may cause:

The clinical assessment should consider baseline respiratory disease and objective measurements where available.

Spirometry and bronchodilator response may help characterize an obstructive pattern, although interpretation depends on the clinical setting.

Pulmonary Edema

Pulmonary edema may arise through:

The finding of pulmonary edema should therefore prompt assessment of both cardiac and non-cardiac causes.

Relevant information may include:

Pulmonary Vascular Toxicity

Drug-related pulmonary vascular effects may include:

Assessment may require:

Respiratory Depression

Respiratory depression differs fundamentally from primary structural lung toxicity.

Drugs may suppress central respiratory drive or impair airway protection.

Important findings include:

The assessment should consider concomitant sedatives, opioids, alcohol and other central nervous system depressants.

Clinical Presentation and Severity

Symptoms

Common symptoms include:

Symptoms should be characterised by onset, progression and relationship to treatment.

Oxygenation

Oxygen saturation provides an important measure of respiratory severity.

The reviewer should consider:

Oxygenation should be interpreted in the context of the patient's baseline respiratory and cardiovascular status.

Severe Respiratory Events

Potential markers of severe pulmonary toxicity include:

For serious cases, the narrative should establish the chronology of respiratory deterioration and the interventions required.

Pulmonary Function Testing

Spirometry

Spirometry can identify obstructive or restrictive patterns.

Important parameters include:

However, spirometry alone does not diagnose drug-induced lung injury.

Diffusing Capacity

Diffusing capacity for carbon monoxide (DLCO) can provide information about gas transfer across the alveolar-capillary interface.

A reduction may occur with:

DLCO is therefore useful but nonspecific.

A longitudinal decline may be more informative than a single isolated abnormal value.

Exercise Testing

Exercise assessment may reveal abnormalities that are not apparent at rest.

Depending on the clinical context, assessment may include:

Results should be interpreted against baseline measurements where available.

Chest Imaging

Chest Radiography

Chest radiography may identify:

However, chest radiography can be insensitive to early or subtle interstitial abnormalities.

High-Resolution CT

High-resolution CT is particularly important in suspected interstitial lung disease.

It can help characterize:

Radiographic pattern recognition can narrow the differential diagnosis but does not by itself establish drug causality.

The ATS/ERS/JRS/ALAT framework for ILD diagnosis illustrates the importance of systematic integration of clinical history and HRCT findings. [2]

Imaging Comparison

Serial imaging may be particularly valuable.

The reviewer should compare:

A new abnormality after treatment initiation followed by improvement after withdrawal can support, but does not prove, a drug relationship.

Bronchoscopy and Bronchoalveolar Lavage

Role of Bronchoscopy

Bronchoscopy may be used when the diagnosis remains uncertain or when alternative diagnoses need to be evaluated.

Potential purposes include:

The decision should be based on clinical severity and diagnostic need.

Bronchoalveolar Lavage

BAL may provide information about:

BAL findings are generally supportive rather than pathognomonic for drug-induced pulmonary toxicity.

ATS guidance emphasises that BAL cellular analysis should be interpreted within the broader clinical and radiological context of ILD. [4]

Competing Causes

Infection

Infection is one of the most important competing diagnoses.

Relevant possibilities include:

Failure to adequately consider infection can result in incorrect attribution of pulmonary events to the medicinal product.

Malignancy

Cancer progression may cause:

This is particularly important in oncology trials.

Cardiac Disease

Heart failure can cause dyspnea, hypoxemia and pulmonary infiltrates.

The reviewer should consider:

Thromboembolism

Pulmonary embolism can present with:

It should be considered whenever clinically appropriate.

Autoimmune and Connective-Tissue Disease

Systemic autoimmune disease can produce:

A relevant autoimmune history and clinical features should therefore be reviewed.

Environmental and Occupational Exposure

The medication history should be complemented by an exposure history.

Relevant exposures may include:

A detailed exposure history is particularly important when ILD is unexplained. [3]

Temporal Relationship and Causality

Latency

Latency is an important component of assessment.

Potential patterns include:

There is no universal latency period that establishes causality.

Different mechanisms produce different temporal patterns.

Dechallenge

Improvement after treatment interruption can support a causal association.

However, interpretation depends on:

Rechallenge

Rechallenge can provide strong evidence in selected circumstances when recurrence follows re-exposure.

However, intentional rechallenge is generally inappropriate when a serious pulmonary reaction is suspected.

Any inadvertent or clinically necessary rechallenge should be evaluated carefully.

Drug-Induced Pulmonary Toxicity in Clinical Development

Individual-Subject Review

Individual medical review should be considered for subjects with:

The reviewer should reconstruct the case chronologically.

Population-Level Review

Clinical-development programmes may evaluate:

The objective is to determine whether there is an excess, clustering or coherent treatment-related pattern.

Treatment Versus Control

Control-group information is particularly important because respiratory symptoms and infections are common.

The reviewer should compare:

A finding concentrated in the treatment group may strengthen concern, but imbalance alone does not prove causality.

Exposure-Response

Exposure-response assessment may help determine whether pulmonary toxicity increases with:

A plausible exposure-response relationship can strengthen the causal hypothesis.

Pulmonary Safety in Special Populations

Patients With Pre-existing Lung Disease

Patients with:

may have increased susceptibility or reduced pulmonary reserve.

Baseline characterization is therefore important.

Oncology Populations

Cancer populations present particular challenges because pulmonary symptoms may result from:

Drug-induced ILD assessment should therefore be multidisciplinary where necessary.

Elderly and Frail Patients

Reduced respiratory reserve can increase the clinical consequences of relatively modest pulmonary injury.

Age-related changes and comorbid cardiac disease may also complicate interpretation.

Post-Marketing Pharmacovigilance

Sources of Pulmonary Safety Information

Post-marketing pulmonary safety information may arise from: [6]

Each source has different strengths and limitations.

Case-Level Medical Review

A potential pulmonary toxicity case should be reconstructed chronologically.

The reviewer should establish:

The clinical diagnosis should be established before final causality assessment whenever possible.

Signal Detection

Potential pharmacovigilance signals may involve:

A statistical or disproportionality signal should trigger clinical assessment rather than being treated as proof of drug causality.

Aggregate Assessment

Aggregate review may consider:

The objective is to determine whether the cases represent a coherent clinical syndrome.

Worked Examples

Example 1: New Pneumonitis After Treatment Initiation

A patient develops progressive dyspnea and cough several weeks after starting a new medicinal product.

CT demonstrates new bilateral ground-glass abnormalities.

Infection testing is negative.

The patient improves following treatment withdrawal and corticosteroid therapy.

The assessment should consider:

The overall evidence may support drug-induced pneumonitis, but the conclusion should reflect the strength and completeness of the available evidence.

Example 2: Pulmonary Infiltrates With Positive Infection Testing

A patient receiving immunosuppressive therapy develops fever, hypoxemia and bilateral infiltrates.

Microbiological testing identifies a clinically relevant respiratory pathogen.

Although a pulmonary adverse event occurred during treatment, the identified infection provides a strong competing explanation.

The case should not be labelled drug-induced merely because treatment preceded the event.

Example 3: Declining DLCO With Minimal Symptoms

A patient develops a progressive decline in DLCO over several months without substantial respiratory symptoms.

CT demonstrates subtle interstitial abnormalities.

The finding warrants investigation because pulmonary function can provide an early indication of evolving disease.

The reviewer should compare baseline testing, imaging and exposure over time and determine whether alternative causes are present.

Example 4: Dyspnea With Pulmonary Edema

A patient develops acute dyspnea and bilateral pulmonary opacities.

The patient also has reduced left ventricular function and elevated natriuretic peptide levels.

The clinical picture is more consistent with cardiac pulmonary edema than primary drug-induced pneumonitis.

This example illustrates why pulmonary imaging findings should not be interpreted independently of cardiovascular assessment.

Example 5: Respiratory Depression

A patient becomes somnolent with a markedly reduced respiratory rate after receiving several centrally acting medicines.

Blood gas testing demonstrates hypercapnia.

The primary safety issue is respiratory depression rather than structural lung toxicity.

The complete concomitant-medication history is therefore essential.

Limitations of Pulmonary Safety Assessment

No Single Pulmonary Biomarker

There is no universally accepted biomarker that independently establishes drug-induced pulmonary toxicity.

Potential research biomarkers may reflect:

However, these findings generally require clinical and radiological interpretation.

Imaging Is Not Pathognomonic

Ground-glass opacity, consolidation and interstitial abnormalities may occur in many diseases.

Imaging pattern can narrow the differential diagnosis but rarely establishes drug causality by itself.

Pulmonary Function Tests Are Nonspecific

A decline in FVC or DLCO may reflect:

Longitudinal assessment is often more informative than a single result.

Background Respiratory Disease Is Common

Asthma, COPD, infection and other respiratory disorders are common in the general population.

This creates substantial background event rates in clinical development and post-marketing surveillance.

Multiple Causes May Coexist

A patient may simultaneously have:

Causality may therefore be multifactorial rather than binary.

Practical Medical Review Framework

Step-by-Step Pulmonary Safety Review

A structured review can proceed as follows:

  1. Confirm the reported respiratory event.
  2. Establish the clinical phenotype.
  3. Determine symptom onset and progression.
  4. Review baseline pulmonary status.
  5. Review treatment exposure and latency.
  6. Review concomitant medicines.
  7. Assess oxygenation and respiratory severity.
  8. Review pulmonary-function testing.
  9. Review chest imaging.
  10. Evaluate infection.
  11. Evaluate cardiac disease.
  12. Evaluate thromboembolic disease where appropriate.
  13. Review autoimmune and connective-tissue disease.
  14. Review environmental and occupational exposures.
  15. Consider bronchoscopy or BAL where clinically appropriate.
  16. Assess dechallenge.
  17. Assess rechallenge where applicable.
  18. Review similar cases.
  19. Assess biological and pharmacological plausibility.
  20. Determine the most likely clinical diagnosis.
  21. Assess causality.
  22. Document the medical rationale.

Medical Documentation

The medical assessment should clearly distinguish:

For serious pulmonary cases, the narrative should allow an independent reviewer to reconstruct the clinical course.

Pharmacovigilance and Inspection Considerations

Inspection Perspective

An inspection-ready pulmonary safety process should demonstrate that the organisation can:

Pulmonary adverse events may require broader searches than a single preferred term because clinically related diagnoses can be coded differently.

FDA review materials illustrate this issue: pulmonary safety analyses may use broad ILD definitions incorporating pneumonitis, ARDS, pulmonary fibrosis, organizing pneumonia and related terms, with narratives providing information needed to evaluate etiology. [5]

Common Documentation Weaknesses

Common weaknesses include:

Common Mistakes in Pulmonary Safety Assessment

Several recurring errors deserve particular attention.

Treating pneumonitis as synonymous with drug causality.

Pneumonitis is a clinical description, not automatically an attribution.

Ignoring infection.

Infection is frequently a major competing diagnosis.

Relying on chest imaging alone.

Radiological abnormalities require clinical interpretation.

Ignoring baseline lung disease.

A small change in a patient with substantial pre-existing pulmonary disease may have a different meaning from the same finding in a previously healthy lung.

Using a single pulmonary-function result.

Longitudinal trends are generally more informative.

Ignoring cardiac causes of dyspnea.

Heart failure can closely mimic pulmonary toxicity.

Assuming improvement after withdrawal proves causality.

Dechallenge supports causality but is not definitive because many pulmonary disorders improve with time or concomitant treatment.

Over-interpreting isolated biomarkers.

No single biomarker should replace integrated clinical assessment.

What an Experienced Safety Physician Looks For

An experienced reviewer does not begin with the question:

"Is this a pulmonary adverse event?"

The more useful questions are:

This approach converts an adverse-event report into a clinically meaningful safety assessment.

Pulmonary Safety Compared With Liver, Kidney and Cardiac Safety

Pulmonary safety differs from several other organ systems.

Liver safety has a well-established population-level concept in Hy's Law and tools such as eDISH.

Kidney safety relies heavily on integration of kidney function, urine findings, injury biomarkers and clinical context.

Cardiac safety includes structured frameworks such as ICH E14/S7B for QT/QTc and proarrhythmic risk.

Pulmonary safety does not have one equivalent universal threshold.

Instead, pulmonary assessment is strongly phenotype-driven.

For suspected ILD or pneumonitis, the combination of:

is generally more informative than any single measurement.

The common principle across organ systems is therefore the same:

identify the phenotype, characterize the objective injury or dysfunction, establish the time course, evaluate competing causes, and determine whether the medicinal product is the most plausible explanation.

Key Takeaways

The central principle is:

Pulmonary safety assessment should identify the respiratory phenotype, establish its severity and time course, evaluate objective pulmonary findings, exclude important competing causes, and determine whether the medicinal product is the most plausible cause.

References

  1. Spagnolo P, Bonniaud P, Rossi G, et al. Drug-induced interstitial lung disease. European Respiratory Journal. 2022;60:2102776.

  2. Raghu G, Remy-Jardin M, Myers JL, et al. Diagnosis of idiopathic pulmonary fibrosis: an official ATS/ERS/JRS/ALAT clinical practice guideline. American Journal of Respiratory and Critical Care Medicine. 2018;198:e44-e68.

  3. Raghu G, Remy-Jardin M, Ryerson CJ, et al. Diagnosis of hypersensitivity pneumonitis in adults: an official ATS/JRS/ALAT clinical practice guideline. American Journal of Respiratory and Critical Care Medicine. 2020;202:e36-e69.

  4. Meyer KC, Raghu G, Baughman RP, et al. An official American Thoracic Society clinical practice guideline: the clinical utility of bronchoalveolar lavage cellular analysis in interstitial lung disease. American Journal of Respiratory and Critical Care Medicine. 2012;185:1004-1014.

  5. American Thoracic Society. Knowledge Gaps and Research Priorities in Immune Checkpoint Inhibitor-related Pneumonitis. Official ATS Research Statement. 2019.

  6. U.S. Food and Drug Administration. Postmarketing Adverse Event Reporting – Required Information. FDA.

Revision History

Related Articles

Drug-Induced Cardiac Toxicity: Cardiovascular Safety Assessment in Clinical Development and Pharmacovigilance

Understand how potential drug-induced cardiac toxicity is identified and medically evaluated, how ICH E14 and S7B support assessment of QT/QTc and proarrhythmic risk, and how cardiovascular safety signals are assessed across clinical development and pharmacovigilance.

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

Understand how drug-induced hematologic toxicity is recognised and medically evaluated, how cytopenias are characterised by lineage, severity and mechanism, and how hematologic safety signals are assessed across clinical development and pharmacovigilance.

Drug-Induced Kidney Injury: Renal Safety Assessment in Clinical Development and Pharmacovigilance

Understand how drug-induced kidney injury is identified and medically evaluated, why kidney function and kidney injury are not the same, how renal biomarkers support clinical-trial safety assessment, and how pharmacovigilance teams evaluate potential nephrotoxicity.