Abstract
Furazolidone is a nitrofuran antibiotic commonly used in the treatment of Helicobacter pylori infection. Although gastrointestinal symptoms are well-documented adverse effects, furazolidone-associated drug-induced interstitial lung disease is extremely rare. Herein, we report a case of a 54-year-old Chinese woman who developed fever, dry cough, and dyspnea on day 4 of furazolidone-containing quadruple therapy. Chest computed tomography revealed diffuse bilateral interlobular septal thickening and small pleural effusions. An extensive workup excluded other causes. The Naranjo adverse drug reaction probability scale score was 6, indicating a probable association with furazolidone. Discontinuation of the drug and initiation of glucocorticoid therapy resulted in rapid clinical and radiological improvement. This case highlights that furazolidone can cause drug-induced interstitial lung disease and that early recognition and intervention are crucial for a favorable outcome.
Introduction
Drug-induced interstitial lung disease (DILD) is a significant yet frequently underrecognized adverse drug reaction (ADR), accounting for approximately 3%–5% of all interstitial lung disease cases. 1 More than 500 drugs have been implicated in DILD, and this number continues to increase with the expanding use of antineoplastic agents, immunosuppressants, antibiotics, and other drug classes. 2 Among antibiotics, nitrofurantoin is a well-established causative agent of DILD. 3 Furazolidone, a structurally related nitrofuran, also carries a risk of pulmonary toxicity, although reported cases remain relatively limited. Due to its potent antibacterial activity against Helicobacter pylori and favorable resistance profile, furazolidone has become a key therapeutic option in regions with high resistance rates.4,5 Although gastrointestinal symptoms are its most commonly reported adverse effects, 6 furazolidone-associated DILD often presents with nonspecific clinical features, contributing to diagnostic delay and potential disease progression. This article presents a case of furazolidone-associated DILD occurring during H. pylori eradication therapy and provides a literature review summarizing its clinical characteristics to improve early recognition and reduce misdiagnosis and missed diagnoses.
Case report
A Chinese woman in her mid-50s presented with a 5-day history of fever, cough, and dyspnea. Prior to admission, she had been diagnosed with H. pylori infection and initiated on a standard quadruple eradication regimen in late November 2023, comprising omeprazole (40 mg/day), bismuth potassium citrate (1.2 g/day), amoxicillin (2.0 g/day), and furazolidone (0.2 g/day). On day 4 following treatment initiation, she developed a fever with a maximum temperature of 38.5°C accompanied by paroxysmal dry cough and exertional dyspnea. She initially sought care at a local community health center, where she received intravenous cefoxitin sodium for 3 days as an outpatient starting on the day of symptom onset; however, no clinical improvement was observed. Given the rapid progression of dyspnea and the need for urgent diagnostic clarification, a chest computed tomography (CT) was performed at the same facility, which revealed diffuse interlobular septal thickening in both lungs with bilateral small pleural effusions (Figure 1). Based on these radiographic findings, she was admitted and transferred to the respiratory and critical care medicine department of Taizhou Municipal Hospital (Taizhou University Affiliated Municipal Hospital), located in Taizhou, Zhejiang, China, in late November 2023 for further management. All patient details, including age and specific dates, have been deidentified to ensure patient anonymity.
Chest CT scan obtained on the day 7 of furazolidone therapy (day of hospital admission). The images show (a) diffuse interlobular septal thickening in both lungs and (b) bilateral small pleural effusions (indicated by arrows). CT: computed tomography.
The patient had no significant past medical history and denied a history of coronary heart disease, hypertension, diabetes, rheumatic connective tissue diseases, or drug allergies. There was no history of exposure to toxins, illicit drugs, birds, or poultry. Physical examination on admission revealed the following: temperature, 38.0°C; pulse rate, 85 beats/min; respiratory rate, 20 breaths/min; blood pressure, 158/79 mmHg; and oxygen saturation, 91% on room air. Coarse breath sounds with inspiratory crackles were auscultated over both lung fields. Cardiac, abdominal, and neurological examinations were unremarkable, and no lower extremity edema was observed. Admission diagnoses included the following: (a) interlobular septal thickening in both lungs, with pulmonary infection and cardiogenic pulmonary edema as differential diagnoses and (b) H. pylori infection.
Upon admission, 0.4 g empirical intravenous moxifloxacin was initiated once daily as an anti-infective therapy based on the patient’s clinical history and prior medication regimen. Initial laboratory tests on room air revealed severe hypoxemia, with an arterial partial pressure of oxygen (PaO2) of 55 mmHg and an arterial partial pressure of oxygen to fraction of inspired oxygen (PaO2/FiO2) ratio of 262 mmHg. Oxygen therapy was subsequently initiated using a Venturi mask at an FiO2 of 40%, with peripheral oxygen saturation maintained above 95%. The patient’s initial hematological, biochemical, and immunological parameters are detailed in Table 1. Notably, these results revealed marked elevations in C-reactive protein and erythrocyte sedimentation rate, with a peripheral eosinophil count of 0 × 109/L. Furthermore, testing for common viral and atypical bacterial respiratory pathogens returned negative results (Table 2). Echocardiography demonstrated preserved systolic and diastolic function, with a left ventricular ejection fraction of 72% and a ratio of early-to-late ventricular filling velocities >1. Given the patient’s severe hypoxemia, elevated D-dimer levels, lack of response to prior antibiotic therapy, and the presence of radiographic findings (interlobular septal thickening and bilateral pleural effusions) requiring exclusion of pulmonary embolism, CT pulmonary angiography (CTPA) was performed. The CTPA revealed no filling defects or other vascular abnormalities and confirmed the previously observed diffuse interlobular septal thickening in both lungs with bilateral small pleural effusions, consistent with the findings reported on the outside hospital CT scan. Bronchoscopy revealed no endobronchial lesions. Cellular analysis of bronchoalveolar lavage fluid (BALF) revealed a differential count of 51% macrophages, 46% lymphocytes, 3% neutrophils, and 1% eosinophils. Metagenomic next-generation sequencing (mNGS) of BALF was negative for bacterial, fungal, viral, parasitic, and mycobacterial pathogens. A repeat arterial blood gas analysis performed on day 2 of admission under a fraction of FiO2 of 29% showed a PaO2 of 60.2 mmHg and a PaO2/FiO2 of 208 mmHg. Additional BALF studies, including galactomannan enzyme immunoassay, GeneXpert Mycobacterium tuberculosis/rifampicin assay, and acid-fast staining, yielded negative results. Collectively, these negative results from extensive microbiological and molecular testing (including mNGS), together with the lymphocyte-predominant cellular profile of the BALF, provided strong evidence against an active infectious etiology. Consequently, empirical intravenous moxifloxacin therapy was discontinued on day 4 of hospitalization. Simultaneously, given the probable causality of furazolidone (Naranjo 7 score = 6; Table 3) and the severity of the presentation, the entire quadruple eradication regimen (furazolidone, amoxicillin, bismuth potassium citrate, and omeprazole) was discontinued, and intravenous methylprednisolone (40 mg daily) was initiated. Within 24 h, the patient’s dyspnea markedly improved, and she became afebrile. A repeat blood gas analysis at this time on room air showed a normal PaO2 of 78 mmHg (with a corresponding PaO2/FiO2 ratio of 371 mmHg). After 2 days of intravenous methylprednisolone therapy (on day 6 of hospitalization), the medication was transitioned to 15 mg oral prednisone twice daily. Follow-up chest CT on day 7 of hospitalization revealed substantial resolution of interlobular septal thickening and almost complete absorption of pleural effusions (Figure 2). Given this marked improvement, corticosteroid therapy was completely discontinued on day 8 of hospitalization. A follow-up chest CT performed approximately 6 months later (in mid-2024) revealed complete resolution of the previously observed interlobular septal thickening and bilateral pleural effusions, with no other significant abnormalities identified (Figure 3). A concomitant 13C-urea breath test was negative (δ < 2.5‰).
Initial laboratory parameters, including normal ranges.
Laboratory test results for common viral and atypical bacterial respiratory pathogens.
EA: early antigen; IgG: immunoglobulin G; IgM: immunoglobulin M; RNA: ribonucleic acid; SARS-CoV-2: severe acute respiratory syndrome coronavirus 2; VCA: viral capsid antigen.
Naranjo ADR probability scale and scoring results.
A total score ≥9 indicated a definite causal relationship between the drug and the adverse reaction, supported by objective evidence and quantitative data. A total score of 5–8 indicated a probable relationship, supported by objective evidence or quantitative results. A total score of 1–4 suggested a possible relationship, indicating that the association cannot be fully confirmed or entirely ruled out. A total score ≤0 suggested a doubtful relationship, indicating that the reaction was likely coincidental or largely unrelated.
ADR: adverse drug reaction.
Follow-up chest CT scan obtained on day 4 of glucocorticoid therapy. The images show significant improvement: (a) interlobular septal thickening in both lungs has markedly resolved and (b) bilateral pleural effusions have almost completely resolved (indicated by arrows). CT: computed tomography.
Follow-up chest CT scan obtained approximately 6 months after the initial event. The images show complete resolution: (a) interlobular septal thickening has completely resolved and (b) bilateral pleural effusions have completely resolved. CT: computed tomography.
The patient provided written informed consent for both medical treatment and publication of this case report. The study was approved by the Ethics Committee of Taizhou Municipal Hospital (Taizhou University Affiliated Municipal Hospital), School of Medicine (Approval Number: 2024-203; Date of Approval: 15 August 2024). The reporting of this study conforms to the Case Report (CARE) guidelines. 8
Discussion
The diagnosis of furazolidone-associated DILD is one of exclusion. It is primarily based on characteristic radiological pulmonary changes, a clear temporal relationship between drug use and symptom onset, and the comprehensive exclusion of alternative causes through thorough diagnostic evaluation. These include drug reaction with eosinophilia and systemic symptoms (DRESS), infections, cardiogenic pulmonary edema, interstitial lung disease related to connective tissue disorders, thromboembolic disease, and lung injury induced by other concomitant medications. This is combined with a comprehensive assessment using causality evaluation tools, such as the Naranjo ADR scale. In this case, the patient developed fever, dry cough, and dyspnea 4 days after initiating standard quadruple therapy containing furazolidone for H. pylori eradication. Chest CT revealed typical bilateral diffuse interlobular septal thickening. The patient had no other significant underlying medical conditions. BALF mNGS and conventional microbiological tests were negative, with BALF cell differentials predominantly showing lymphocytes. B-type natriuretic peptide (BNP) levels, autoimmune antibody profiles, CTPA, and echocardiography effectively ruled out other common etiologies. The Naranjo ADR score was 6 (probable). The symptoms resolved rapidly after discontinuation of the drug and administration of glucocorticoids, supporting the diagnosis of furazolidone-associated DILD. Furthermore, a literature review revealed no reported cases of DILD associated with the other drugs in the regimen (omeprazole, bismuth potassium citrate, or amoxicillin), further supporting furazolidone as the causative agent.
To systematically analyze the clinical characteristics of furazolidone-associated DILD, we conducted a literature search in PubMed and Web of Science using the query “furazolidone” OR “furoxone” AND “pulmonary” OR “lung,” with the search period extending through July 2025. Five articles,9–13 reporting a total of seven eligible cases, were included (Table 4). Together with the current case, a total of eight cases were pooled for analysis. The time from furazolidone administration to symptom onset ranged from 4 to 14 days. This included 3 cases in which symptoms developed within 4 days after drug discontinuation and an additional case that became symptomatic 3 days after drug cessation following 1 day of readministration. Clinical manifestations included fever (n = 7), dyspnea (n = 5), and cough (n = 5). Other reported symptoms comprised chest pain, headache, and fatigue. Eosinophilia demonstrated a “delayed pattern”: only two cases showed elevated baseline eosinophil levels at initial presentation. Among the 6 cases with available repeat test results, eosinophilia peaked 3–8 days after onset, with an increase of 6%–21%. Radiologic assessment revealed bilateral pulmonary infiltrates in all cases. Although three cases were evaluated by chest X-ray only, chest CT imaging in the remaining five patients (including the current case) consistently demonstrated bilateral interlobular septal thickening, suggesting this may represent a characteristic radiologic finding. Notably, the present case was accompanied by bilateral small pleural effusions, an imaging feature not previously documented in the literature. This finding initially raised strong suspicion of cardiogenic pulmonary edema, thereby delaying consideration of furazolidone-associated DILD.
Summary of furazolidone-associated DILD.
DILD: drug-induced interstitial lung disease; F: female; M: male; NR: not reported SPO2: pulse oxygen saturation.
The pathogenesis of DILD associated with furazolidone remains poorly understood. Nitrofuran antibiotics may induce DILD through nitroreductase-mediated metabolic activation. Among these agents, nitrofurantoin is the most commonly implicated in reported cases of DILD. Metabolism of its 5-nitrofuran ring generates reactive intermediates that trigger the overproduction of reactive oxygen species, thereby disrupting the pulmonary oxidant–antioxidant equilibrium and resulting in diffuse alveolar damage.14,15 Furazolidone possesses a similar 5-nitrofuran moiety, suggesting a potential mechanism of pulmonary toxicity via a type II nitroreduction process similar to that of nitrofurantoin. Furthermore, the elevated oxygen tension in the lungs may predispose this organ to increased drug-induced injury. 16
A critical step in the differential diagnosis involves distinguishing furazolidone-associated DILD from DRESS. Pulmonary involvement has been reported in approximately one-third of DRESS cases,17,18 typically manifesting as dyspnea and dry cough, with imaging features including interstitial infiltrates, pleural effusion, and acute respiratory distress syndrome.17,18 However, in the present case, neither amoxicillin nor omeprazole met the diagnostic criteria for DRESS syndrome, as evaluated by the validated European RegiSCAR scoring system, despite both agents being documented as potential triggers in the literature.17,19 Notably, the presentation was strictly confined to pulmonary injury. This is in sharp contrast to DRESS, in which pulmonary involvement is almost invariably accompanied by characteristic systemic manifestations such as cutaneous eruption, lymphadenopathy, or hepatic impairment. 19 Beyond these acute clinical distinctions, the long-term prognosis of the two conditions diverges fundamentally. DRESS syndrome is associated with significant acute mortality (5%–9%), 20 primarily due to fulminant hepatitis or eosinophilic myocarditis. Furthermore, more than 10% of survivors face long-term autoimmune sequelae, such as thyroiditis, type 1 diabetes, or systemic lupus erythematosus, necessitating extended surveillance for at least 6 months to 3 or more years.19,20 In contrast, furazolidone-related DILD, as documented in the literature, follows a benign course. It has not been associated with fatal outcomes, and patients typically achieve complete radiological and functional recovery within weeks without developing chronic pulmonary or systemic complications. Collectively, these acute clinical and prognostic findings do not support a diagnosis of DRESS.
The cornerstone of managing DILD is the immediate discontinuation of the causative agent. Current studies on the treatment of DILD remain limited. Existing guidelines are primarily based on observational reports and clinical experience and have yet to be standardized or validated through prospective clinical trials.2,15 Pharmacological management of DILD currently relies on systemic glucocorticoids, with their use guided by disease severity and clinical judgment, and specific dosing regimens typically determined at the treating physician’s discretion.1,2,15,21
Glucocorticoid therapy was documented in three reported cases of furazolidone-associated DILD. Across these cases, the decision to initiate treatment appeared to be influenced more by the severity of hypoxemia than by the eosinophil count at presentation. Cortez et al. 9 described a case of a patient with severe respiratory failure (PaO2, 47 mmHg) and 0% eosinophils who improved rapidly after steroid initiation. Kowalski et al. 11 reported a case of a patient presenting with profound hypoxemia (pulse oxygen saturation, 88% on room air) and an initial eosinophil count of only 2%. Notably, despite 3 days of supportive care, the patient’s condition failed to improve. A subsequent laboratory review revealed a marked increase in eosinophils to 21%, at which point prednisone was initiated followed by prompt clinical recovery.
In our case, the patient was diagnosed with Grade 3 (severe) DILD. 22 Despite a peripheral eosinophil count of 0 × 109/L at presentation, the severity of hypoxemia (PaO2 of 55 mmHg on room air) was the key factor determining the initiation of systemic glucocorticoids. For noncancer-related DILD, such as this case induced by furazolidone, there is currently no unified dosing standard. Therefore, dose selection was primarily informed by previous case reports of similar etiology. In reported cases of furazolidone-induced DILD, the converted prednisone-equivalent starting dose ranged from 40 to 65 mg per day.9,11,12 This range is also consistent with the commonly used dose range (10–60 mg per day) for nitrofurantoin-induced DILD, a structurally related drug. 23 Given the patient’s severe (Grade 3) presentation, methylprednisolone was initiated at a dose of 40 mg daily. This intervention resulted in rapid and significant improvement in both clinical symptoms and imaging findings. In response to this favorable outcome, treatment was transitioned to 30 mg/day oral prednisone on t day 3, with complete discontinuation of corticosteroid therapy achieved by the day 5.
Although the patient in this case did not develop corticosteroid-related complications because of the short treatment duration, clinicians should be aware that systemic glucocorticoid therapy, particularly when used long term or at high doses, may induce gastritis, osteoporosis, and opportunistic infections such as Pneumocystis jirovecii pneumonia (PJP). Current guidelines suggest considering proton pump inhibitors for gastrointestinal protection in patients receiving prednisone-equivalent doses ≥20 mg/day. They also recommend calcium and vitamin D supplementation to mitigate osteoporosis risk during extended therapy and trimethoprim-sulfamethoxazole for PJP prophylaxis in those receiving prolonged corticosteroid therapy (≥20 mg/day prednisone for ≥1 month).24,25 Notably, a retrospective study of patients on prolonged corticosteroids reported a low incidence of PJP (<1%), with breakthrough infections occurring despite prophylaxis in a subset. 26 These findings raise questions regarding the universal necessity of routine PJP prophylaxis in populations receiving steroid therapy.
Glucocorticoid therapy regimens are not standardized, typically lasting from several days to 1 month with tapering.9,11,12 Second-line agents, such as immunosuppressants or biologics (e.g. infliximab, tocilizumab, and intravenous immunoglobulin), are primarily reserved for steroid-refractory cases. Most existing evidence is derived from cancer therapy-related interstitial lung disease (CTLI) or checkpoint inhibitor pneumonitis (CIP). 25 Furthermore, antifibrotic agents pirfenidone and nintedanib have been shown to slow the progression of idiopathic pulmonary fibrosis and have also demonstrated potential in radiation-induced lung injury.27–29 Their application is also being explored in DILD, including CIP. Nevertheless, the use of these agents in furazolidone-induced DILD has not been documented in the literature, and appropriate patient selection, timing of initiation, and treatment duration remain unclear. Moreover, re-exposure to furazolidone must be strictly avoided to prevent recurrence. 13 Although no fatal cases have been reported to date and most patients achieve complete recovery with timely drug discontinuation and intervention, vigilance for rapid disease progression is still warranted. Therefore, establishing a multidisciplinary team (MDT) model, led by pulmonology and integrating expertise from radiology, pharmacy, pathology, and relevant primary specialty clinics, is crucial for achieving early recognition, precise intervention, and systematic follow-up. This approach holds promise for improving the overall prognosis of DILD.25,30
This study has limitations inherent to a single case report. The level of evidence is low, and the findings are not generalizable. Although the temporal association and exclusion of alternative causes support the diagnosis, the causal role of furazolidone remains probable in the absence of rechallenge. The corticosteroid regimen, based on limited precedents, cannot be standardized. Finally, the relatively small total number of reported cases limits broader conclusions. Future multicenter studies are needed to clarify the pathogenesis and optimal management of this condition.
In conclusion, in patients receiving furazolidone-containing H. pylori therapy who develop respiratory symptoms, DILD should be considered. Immediate drug discontinuation is critical. In the present case, glucocorticoid initiation, guided by the severity of hypoxemia (rather than eosinophil count), was associated with rapid recovery, although the optimal regimen requires further validation. Future studies including more cases are needed to better define the optimal management approach for this adverse event.
Footnotes
Acknowledgments
The authors wish to thank the radiology and pathology staff of Taizhou Municipal Hospital (Taizhou University Affiliated Municipal Hospital), School of Medicine, for their timely imaging and laboratory support.
Author contributions
MYL and CZ contributed to the conception and design of the study and drafted the manuscript; MYL and JSP acquired the data; MYL, JSP, JHL, and CZ critically revised the manuscript for important intellectual content; CZ provided final approval of the version to be submitted. All authors reviewed the manuscript.
Data availability statement
The detailed information used during the current report is available from the corresponding author on reasonable request.
Declaration of conflicting interest
The authors declare that there are no conflicts of interest.
Funding
This study was supported in part by grants from the Medical Health Science and Technology Project of the Zhejiang Provincial Health Commission (2025KY1871) and the Science and Technology Project of Taizhou (24ywa44).
