Drug-induced Pleuritis Associated with Hachimijiogan

Abstract

A 78-year-old man presented to our hospital with left-sided pleural effusion. A diagnosis could not be established based on blood or pleural fluid analyses. A re-evaluation of his medical history revealed that he had begun taking hachimijiogan five months prior to presentation. Pleural effusion resolved after discontinuation of hachimijiogan, resulting in a diagnosis of drug-induced pleuritis. Subsequently, an accidental rechallenge was performed to confirm the diagnosis. Drug-induced pleuritis due to Kampo medicines is uncommon, with no prior reports of hachimijiogan being associated with this condition. Obtaining a detailed medication history is essential for diagnosing pleural effusions of unknown etiology.

Introduction

The most common causes of exudative pleural effusion include malignancy, tuberculosis, and bacterial infections (1). Although uncommon, exudative pleural effusions can be associated with certain medications. Drug-induced pleuritis has been documented with more than 30 different drugs, including dasatinib, amiodarone, and sodium valproate (2). No standardized diagnostic criteria exist for drug-induced pleuritis; however, this condition is typically identified by symptom resolution following discontinuation of the suspected drug (2,3). Discontinuing the offending drug serves as both a diagnostic and therapeutic approach, underscoring the importance of obtaining a detailed medication history when assessing pleural effusions of unknown etiology.

Kampo medicine consists of a combination of crude drugs. Originating in traditional Chinese medicine, it has been refined independently in Japan. The most frequent adverse effects of Kampo medicine are liver and lung injuries, whereas pleuritis is exceptionally rare (4).

We herein report a case of drug-induced pleuritis associated with hachimijiogan, a Kampo medicine.

Case Report

A 78-year-old man presented to our hospital with exertional dyspnea that had persisted for 3 weeks. His medical history included hypertension, atrial fibrillation, and hyperuricemia. He had been taking losartan potassium, edoxaban tosylate hydrate, bisoprolol fumarate, and febuxostat regularly for over four years. Five months before his hospital visit, his family physician prescribed hachimijiogan for lower back pain, which he continued to take at the time of the visit.

Upon an initial examination, the physical findings were recorded as follows: body temperature, 36.0°C; blood pressure, 126/86 mmHg; pulse rate, 82 beats per minute; respiratory rate, 16 breaths per minute; and oxygen saturation (SpO₂), 95% on room air. Lower leg edema was not observed, although chest auscultation revealed diminished breath sounds in the left lower lung field. Chest radiography revealed moderate left pleural effusion (Figure A). The results of the blood and pleural fluid analyses are shown in Table. No elevation in the white blood cell count or C-reactive protein (CRP) levels was observed, and the liver and kidney functions remained within normal limits. Neither the patient nor his family reported a history of tuberculosis, and the interferon-gamma release assay yielded negative results. Echocardiography revealed a preserved left ventricular ejection fraction of 62% with no regional wall motion abnormalities, confirming no significant cardiac dysfunction.

Figure.

Chest radiographic findings during the clinical course. (A) Initial visit: Chest radiography showing left pleural effusion. (B) Three months after the initial thoracentesis, the pleural effusion increased. (C) Six months after discontinuation of hachimijiogan, the pleural effusion had almost disappeared. (D) Two months after the initiation of goshajinkigan treatment, pleural effusion increased again. (E) Complete resolution of the pleural effusion was observed six months after discontinuation of goshajinkigan.

Table.

Laboratory Findings in Blood and Pleural Effusion.

Hematology				Tumor marker
WBC	5,700	/μL		CEA	1.3	ng/mL
Neutro	75.7	%		CYFRA	1.6	ng/mL
Eosino	0.9	%		Pro-GRP	58.6	pg/mL
Baso	0.5	%
Lymph	16.8	%		Infection marker
RBC	428×104	/μL		IGRA	negative
Hb	11.6	g/dL		C. neoformans-Ag	negative
Plt	19.7×104	/μL		β-D-glucan	5.1	pg/mL
Biochemistry/Serology				Autoantibodies
TP	6.7	g/dL		ANA	1:80 (Speckled)	titer
Alb	4.0	g/dL		Anti-dsDNA-Ab	2.0	IU/mL
BUN	21	mg/dL		Anti-SS-A-Ab	1.0	U/mL
Cre	0.92	mg/dL		Anti-CCP-Ab	0.6	U/mL
AST	25	IU/L		PR3-ANCA	1.2	U/mL
ALT	16	IU/L		MPO-ANCA	1.0	U/mL
ALP	107	IU/L
γ-GTP	53	IU/L		Pleural effusion
Na	142	mEq/L		Cell count	3,970	/μL
K	4.4	mEq/L		Neutro	3	%
Cl	106	mEq/L		Eosino	0	%
LDH	160	IU/L		Lymph	70	%
CRP	0.26	mg/dL		MΦ	27	%
BNP	157.9	pg/mL		TP	3.9	g/dL
IgG	1,302	mg/dL		Glu	119	mg/dL
IgG4	43.0	mg/dL		LDH	90	IU/L
IgA	297	mg/dL		CEA	0.7	ng/mL
IgM	118	mg/dL		ADA	19.2	IU/L
C3	118	mg/dL		Hyaluronic acid	13	ng/mL
C4	20	mg/dL		Bacterial culture	negative
CH50	54	U/mL		Mycobacterial culture	negative

WBC: white blood cell, Neutro: neutrophils, Eosino: eosinophils, Baso: basophils, Lymph: lymphocytes, RBC: red blood cells, Hb: hemoglobin, Plt: platelets, TP: total protein, Alb: albumin, BUN: blood urea nitrogen, Cre: creatinine, AST: aspartate aminotransferase, ALT: alanine aminotransferase, ALP: alkaline phosphatase, γ-GTP: γ-glutamyl transpeptidase, LDH: lactate dehydrogenase, CRP: C-reactive protein, BNP: brain natriuretic peptide, Ig: immunoglobulin, CEA: carcinoembryonic antigen, CYFRA: cytokeratin fragment, Pro-GRP: pro-gastrin-releasing peptide, IGRA: interferon-gamma release assay, C.neoformans-Ag: Cryptococcus neoformans antigen, ANA: antinuclear antibody, Anti-dsDNA-Ab: anti-double-stranded DNA antibody, Anti-SS-A-Ab: anti-Sjögren’s-syndrome-related antigen A antibody, Anti-CCP-Ab: anti-cyclic citrullinated peptide antibody, PR3-ANCA: proteinase 3-anti-neutrophil cytoplasmic antibody, MPO-ANCA: myeloperoxidase-anti-neutrophil cytoplasmic antibody, MΦ: macrophages, Glu: glucose, ADA: adenosine deaminase

A total of 300 mL of the left pleural effusion was aspirated and submitted for an analysis. The fluid, yellow in appearance, was classified as lymphocyte-dominant exudative pleural effusion. Pleural fluid adenosine deaminase levels did not increase, and bacterial cultures, including those for mycobacteria, returned negative results. Pleural fluid carcinoembryonic antigen and hyaluronic acid levels were within the normal ranges, with negative cytology findings. Following drainage, chest computed tomography (CT) revealed no abnormalities in the lung fields or pleura. Based on these test findings, malignancy and tuberculosis, which are common etiologies of lymphocyte-dominant exudative pleural effusion, were deemed unlikely.

As the patient's general condition was stable, he was followed-up as an outpatient. However, the left pleural effusion progressively increased despite initial thoracentesis (Figure B). The patient remained afebrile, and the CRP levels remained within the normal range throughout the course of the disease. Contrast-enhanced CT from the chest to the pelvis was performed to exclude abdominal tumors and pulmonary embolism, but no significant abnormalities were detected. Serum tumor markers, infection markers, and autoantibodies related to connective tissue diseases were also evaluated, with no notable findings (Table).

Upon reviewing the patient's medical history, it was noted that he had started taking hachimijiogan five months before the onset of pleural effusion. Given the potential for drug-induced pleuritis, hachimijiogan was discontinued. Following discontinuation, the left pleural effusion gradually decreased and nearly resolved after six months (Figure C). The resolution of the pleural effusion without additional intervention supported the diagnosis of drug-induced pleuritis secondary to hachimijiogan. A rechallenge was not performed because of safety concerns. The drug-induced lymphocyte stimulation test (DLST) for hachimijiogan yielded negative results.

However, three months after the diagnosis, pleural effusion recurred (Figure D). It was later discovered that the patient's family physician had prescribed goshajinkigan, a Kampo medicine that contains all eight components of hachimijiogan (Rehmanniae Radix, Corni Fructus, Dioscoreae Rhizoma, Alismatis Rhizoma, Poria, Moutan Cortex, Cinnamomi Cortex, and Aconiti Radix) plus two additional crude drugs (Achyranthis Radix and Plantaginis Semen), two months prior to this recurrence. After discontinuation of goshajinkigan, the pleural effusion gradually resolved without any further re-accumulation (Figure E). The accidental rechallenge further strengthened diagnostic certainty.

Discussion

Malignant tumors, tuberculosis, bacterial infections, pulmonary embolisms, and connective tissue diseases are the most common etiologies of exudative pleural effusion, whereas drug-induced pleuritis is rare (1). This case involved pleuritis, which was attributed to hachimijiogan. Blood tests and pleural fluid analyses failed to identify the cause, making the patient's medication history pivotal for the diagnosis.

Previous reports have documented drug-induced pleuritis associated with medications such as dasatinib, amiodarone, sodium valproate, imidapril, procainamide, and hydralazine (5-10). More than 30 additional drugs have also been implicated (2). Certain medications, including procainamide and hydralazine, induce pleural effusion as a manifestation of drug-induced lupus erythematosus (11); however, the mechanisms underlying the effusion onset remain poorly understood for most drugs. The hypothesized mechanisms include hypersensitivity or allergic reactions, direct toxic effects, increased production of oxygen free radicals, suppression of antioxidant defenses, and chemically induced inflammation (2). DLSTs are frequently employed to identify hypersensitivity reactions that cause drug-induced liver or lung injury (12). In this case, the DLST for hachimijiogan yielded a negative result. Notably, DLSTs are prone to both false positives and false negatives (12), precluding the complete exclusion of a hypersensitivity mechanism.

Drug-induced left-sided pleuritis was observed in the present case. Although systemic drug reactions indicate bilateral pleural effusions, unilateral presentation of drug-induced pleuritis is common (2,3). Potential mechanisms for unilateral presentation include anatomical independence of the left and right pleural cavities or localization of an immunological hypersensitivity reaction to one pleura due to factors such as uneven distribution of drug metabolites or differences in lymphatic flow. While the precise mechanism underlying the unilateral pleuritis in our patient remains unclear, the anatomical independence of the pleural cavities and localization of the immune response may be due to certain factors contributing to this presentation.

Currently, there are no standardized diagnostic criteria for drug-induced pleuritis. Diagnosis typically involves excluding alternative causes of pleural effusion followed by confirmation of symptom resolution after discontinuation of the suspected drug (2,3). The diagnosis becomes more plausible if symptoms recur after readministration of the causative drug. However, because of the high-risk method, it is performed infrequently, following guidelines for drug-induced interstitial lung disease (13). In the present case, blood tests, pleural fluid analysis, and contrast-enhanced CT excluded malignancy, tuberculosis, bacterial infection, pulmonary embolism, and connective tissue disease. Although the pleural effusion initially persisted, resolution occurred following hachimijiogan discontinuation, supporting the diagnosis of hachimijiogan-induced pleuritis. In addition, accidental re-administration further strengthened the diagnostic certainty.

Hachimijiogan, considered the causative agent in this case, is a Kampo medicine commonly used for patients with moderate or low physical strength who experience certain symptoms, such as fatigue, coldness of the limbs, frequent urination or residual urine, lower back pain, and leg pain. Goshajinkigan was formulated by adding Achyranthis Radix and Plantaginis Semen to the eight constituent crude hachimijiogan drugs. While also used for similar patterns of symptoms, goshajinkigan is often indicated for patients who present with more pronounced symptoms, such as a decreased urine output, lower back pain, and edema of the lower limbs. Kampo medicines are composed of multiple crude drugs; therefore, caution is warranted when prescribing alternative Kampo formulations that share common ingredients with preparations for which a patient has previously experienced adverse reactions.

The commonly reported adverse effects of Kampo medicine include liver injury, lung injury, pseudoaldosteronism, mesenteric phlebosclerosis, and drug eruptions (4). Although drug-induced lung injury from Kampo medicines is well-documented, pleuritis is extremely rare. To our knowledge, a case involving hochuekkito is the only instance previously reported (14). This is the first documented report of drug-induced pleuritis linked to hachimijiogan, highlighting the potential of hachimijiogan to induce this condition.

Thoracoscopy is recommended when blood tests and pleural fluid analyses fail to elucidate the etiology of pleural effusion (15). A pleural biopsy can provide a definitive diagnosis of malignancy or tuberculosis and is therefore an important procedure for evaluating pleural effusion of unknown etiology (16). Thoracoscopy has also been used in diagnosing drug-induced pleuritis (17-19), although its primary value lies in excluding malignancy and tuberculosis. For drug-induced cases, discontinuing the implicated drug is typically sufficient for both diagnosis and treatment. Thoracoscopy should only be considered if malignancy or tuberculosis cannot be excluded. Prior to invasive procedures, it was essential to ascertain whether new medications were initiated before effusion onset.

The latency of drug-induced pleuritis varies widely, with some cases emerging more than a year after drug initiation, although most manifest within a few days to six months (3). Therefore, it is important to review the medications introduced shortly before the onset of pleural effusion. In this case, the timing of hachimijiogan initiation was overlooked during the initial consultation. However, a subsequent review revealed its introduction five months before symptom onset, guiding the diagnosis.

In conclusion, clinicians should recognize that Kampo medicines, including hachimijiogan, can cause pleuritis, a rare adverse event. As hachimijiogan is a Kampo medicine widely used in the older population, drug-induced etiology should be considered if pleural effusion develops in patients during its administration. When evaluating pleural effusion of unknown etiology, drug-induced causes should be considered, and a comprehensive medication history should be obtained before pursuing invasive investigations.

The authors state that they have no Conflict of Interest (COI).