Bilothorax: A Case Report and Systematic Literature Review of the Rare Entity

Background Bilothorax is defined as the presence of bile in the pleural space. It is a rare condition, and diagnosis is confirmed with a pleural fluid-to-serum bilirubin ratio of >1. Methods The PubMed, Embase, Google Scholar, and CINAHL databases were searched using predetermined Boolean parameters. The systematic literature review was done per PRISMA guidelines. Retrospective studies, case series, case reports, and conference abstracts were included. The patients with reported pleural fluid analyses were pooled for fluid parameter data analysis. Results Of 838 articles identified through the inclusion criteria and removing 105 duplicates, 732 articles were screened with abstracts, and 285 were screened for full article review. After this, 123 studies qualified for further detailed review, and of these, 115 were pooled for data analysis. The mean pleural fluid and serum bilirubin levels were 72 mg/dL and 61 mg/dL, respectively, with a mean pleural fluid-to-serum bilirubin ratio of 3.47. In most cases, the bilothorax was reported as a subacute or remote complication of hepatobiliary surgery or procedure, and traumatic injury to the chest or abdomen was the second most common cause. Tube thoracostomy was the main treatment modality (73.83%), followed by serial thoracentesis. Fifty-two patients (51.30%) had associated bronchopleural fistulas. The mortality was considerable, with 18/115 (15.65%) reported death. Most of the patients with mortality had advanced hepatobiliary cancer and were noted to die of complications not related to bilothorax. Conclusion Bilothorax should be suspected in patients presenting with pleural effusion following surgical manipulation of hepatobiliary structures or a traumatic injury to the chest. This review is registered with CRD42023438426.


Introduction
Bilothorax, cholethorax, or thoracobilia is defined as the presence of bilirubin in the pleural cavity.The pleural cavity has no anatomical connection to the abdominal compartment, so the presence of bilirubin in the pleural space should always be considered pathological.The pathophysiology is poorly understood and likely due to the negative pressure generated by the thoracic cavity causing bile translocation through diaphragmatic pores or defects.Bilothorax is diagnosed when the pleural fluid-to-serum bilirubin ratio is equal to or greater than one [1].This entity appears underdiagnosed as it cannot be distinguished from other causes of hydrothorax with radiological imaging or physical examination, and its diagnosis requires a high index of suspicion.Delayed diagnosis can lead to inflammation, causing scarring and fibrosis of the pleura.We present a case of bilothorax and a review of the literature with a focus on incidence, etiology, and management.We also conducted a systematic literature review of the reported cases per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.
1.1.Case.A 72-year-old woman with a history of atrial fibrillation, left atrial thrombus on apixaban, heart failure with preserved ejection fraction on furosemide, iron deficiency anemia (hemoglobin 9-10 mg/dL), and morbid obesity (BMI 40 kg/m 2 ) underwent laparoscopic cholecystectomy for gangrenous cholecystitis a month prior to her presentation to the emergency department, reporting shortness of breath for three days.She denied fever, cough, chills, lower extremity edema, and orthopnea.The physical exam was unremarkable except for decreased breath sounds on the right lung base and a noted rapid ventricular response.A chest X-ray (CXR) revealed new-onset right-sided pleural effusion and bilateral pulmonary vascular congestion.Pro-BNP was 1815 pg/mL (it was 1120 pg/mL the month before).SARS-CoV-2 RT-PCR was positive on a nasal swab.She had leukocytosis of 16 k/μL, neutrophils of 74%, hemoglobin of 10.5 g/dL, and platelets of 552 k/μL.The basic metabolic profile and liver profile were grossly unremarkable.A point-ofcare ultrasound revealed a moderate-sized simple pleural effusion.She was started on intravenous ceftriaxone, azithromycin, and dexamethasone.Intravenous heparin was initiated, and her apixaban was stopped for anticipated thoracentesis.She was started on 80 mg/day intravenous furosemide, and a CXR three days later demonstrated an unchanged effusion.She underwent bedside thoracentesis with the removal of 600 mL of orange-colored pleural fluid.Considering her recent cholecystectomy and pleural effusion not responsive to diuresis, we suspected a bilothorax.The pleural fluid bilirubin level was 1.4 mg/dL, and the serum bilirubin level was 0.3 mg/dL with a pleural fluid-to-serum bilirubin (PB/SB) ratio of 4.67.Pleural fluid pH was 7.21, total protein (TP) was 4.2 g/dL, lactate dehydrogenase (LDH) was 223 IU/L, white blood cell (WBC) was 1556/ mm 3 , 85% lymphocytes, and glucose level was 96 mg/dL.The gram stain and culture had no growth, and cytology was negative for malignant cells.A subsequent computerized tomographic scan (CT scan) of the chest did not show diaphragm defects.A surgical consultation was obtained, and they recommended no further intervention in view of the chest CT results.After the thoracentesis, her oxygen was weaned off and her dyspnea improved.She was discharged in her usual state of health.A follow-up CXR after one month demonstrated no recurrent effusion.1).We searched the PubMed/MEDLINE, Google Scholar, Embase, and CINAHL databases.The Boolean parameters that were used to search were "bilothorax" OR "cholethorax" OR "biliary pleural effusion" OR "bilious pleural effusion" OR "thoracobilia".Table 2 in the Supplementary Material section provides a detailed description of the search strategy.We included studies published before January 31, 2023.The study's protocol was registered in PROSPERO (CRD42023438426).

Review of the Literature
2.1.2.Eligibility Criteria.Retrospective studies, case series, case reports, and conference abstracts were included.The comprehensive reviews were included if they had patients' information that could be pooled for description.Editorials, comments, viewpoints, and articles lacking full text were excluded.The studies in languages other than English were translated using the Google Translate website (Google LLC).
Inclusion criteria.The studies meeting the above eligibility criteria were included in this review if the study had patients with (1) measured bilirubin level in pleural fluid (2) and/or pleural fluid defined as green color or clearly mentioned bilious fluid/bile during thoracentesis or chest tube insertion.
Exclusion criteria.The studies were excluded if no pleural fluid bilirubin measurement or color description was available.
2.1.3.Data Extraction.Studies were identified and screened for eligibility by two authors (R.A. and S.K.) independently based on inclusion criteria.EndNote software was used to maintain the records of identified and screened studies and remove duplicated studies.Discrepancies were resolved by mutual consent obtained from another author (D.B.S.).A Microsoft Excel sheet (Microsoft Corp.) was used to extract the study characteristics, such as type of study, year of publication, age, sex, background, pleural and serum fluid bilirubin measurements in mg/dL, pleural fluid analyses, presence of bilopleural fistula (BPF), organisms isolated, management, and outcome.The patients with reported pleural fluid and/or serum bilirubin, age, presence of BPF, and other relevant data were pooled for quantitative data analysis.If a study had multiple patients reported, then all the patients from the study were pooled.
2.1.4.Outcome Measures.Our primary objective of the study was to identify the total number of reported cases of bilothorax.The secondary objectives were to calculate the ratio of pleural fluid-to-serum bilirubin and the total and differential cell counts in the pleural fluid and identify etiologies causing bilothorax, isolated organisms, radiological imaging used, treatment, and outcome.

Statistical Analysis.
The continuous variables were reported in mean with standard deviation (SD).The categorical variables were reported in frequency with percentages and 95% confidence intervals (CI).The statistical analysis was done in STATA 17.0 software (Stata Corp. LLC).
2.1.6.Risk of Bias Assessment.Joanna Briggs Institute's (JBI) critical appraisal checklists for case reports, case series, and cohort studies were used for risk of bias assessment.The checklist included 8 to 11 items.If the answer to an item was yes, it was scored 1; otherwise, it was scored zero.For case reports, quality scores of 2 or less, 3 to 5, and 6 or 2 Pulmonary Medicine greater were considered low, moderate, and high quality, respectively.For case series and cohort studies, quality scores of 4 or less, 5 to 7, and 8 or greater were considered low, moderate, and high quality, respectively (Supplementary Tables 3, 4, and 5).

Results
3.1.Literature Search.A total of 837 studies were identified through the databases, and one additional record was obtained from other sources.One hundred five studies were duplicated and, hence, were omitted.A total of 732 studies were screened with titles and abstracts, of which 285 qualified for full-text review.After applying inclusion and exclusion criteria, 123 studies were qualified for the qualitative analysis.From those 123 studies, 115 patients were pooled for quantitative analysis (Figure 1).

Primary Outcome.
Of the 838 studies identified through a database search, 123 qualified for the review.Of the 123 studies, 80 were case reports, 25 were abstracts (Table 1), and 18 were observational studies (Table 2).Eighty from case reports, 25 from abstracts, and ten from observational studies-a total of 115 patients-were pooled for quantitative analysis (Tables 3 and 4).

Pulmonary Medicine
Magnetic resonance imaging of the biliary tree was reported in seven patients and detected a diaphragmatic defect in 85.71% of cases (n = 6) (Table 4).Eighteen patients had an infected bilothorax, and the most common organism isolated was Escherichia coli (n = 5), followed by Klebsiella (n = 4) (Table 4).
Only two of the 115 patients had a left-sided bilothorax.
A chest tube was placed in 79 out of 107 patients (73.83%, 95% CI 64.59-81.35).In addition to chest tube thoracostomy or thoracentesis, 53 surgical interventions were reported in 52 patients with ERCP and/or biliary drain being the most common procedure, which was reported at 47.16% (n = 25).Similarly, VATS was reported in 24.52% of cases (n = 13), fistula repair in 22.64% of cases (n = 12), and open thoracotomy in one patient (Table 4).

Discussion
In this review, we found 123 studies that reported bilothorax, of which 115 met the criteria for quantitative analysis.The most common etiology was PTBD, followed by injury-related, and the prognosis was overall favorable with the institution of pleural fluid drainage.Chest CT was the most commonly used radiological investigation, and chest tube thoracostomy was the prevalent treatment modality.
Bilothorax seems to be underdiagnosed, requiring a high index of suspicion for an adequate diagnosis.A careful history with particular attention to any surgical manipulation, radiation, or infection of hepatobiliary structures can be the first clue to the diagnosis.The latency from the initial insult to the development of bilothorax varies from days to years [41,93,99].In our case, the bilothorax occurred about a month after laparoscopic cholecystectomy.Fortunately, the mortality from bilothorax remains low.In our review, around 84% of cases had favorable outcomes.Those with associated mortality had hepatobiliary or gastric carcinoma and succumbed to complications other than the bilious pleural effusion per se.The mainstay of the treatment was the drainage of the bilothorax, mostly through a chest tube.Surgical interventions were mentioned in 52 (45%) patients, and the most common procedure needed was biliary decompression.Surgical intervention was indicated if the bilothorax failed to resolve after placing the chest tube.Of those 14 positive CT scans, three patients had biliary stents that transverse through the diaphragm resulting in bilothorax.HIDA scan, or MR study of hepatopancreatobiliary structure, was more sensitive in detecting a BPF.Our patient was investigated with a chest CT scan, which showed no diaphragmatic defect.The mechanism of bilothorax is poorly understood.A bilopleural fistula was present in 51% of the patients, but in the remaining 49% of the patients, no diaphragmatic defects were present.It is possible that bile might have been sucked into the pleural cavity through congenital microdefects in the diaphragm during negative intrathoracic pressure, similar to that of hepatic hydrothorax.These defects are usually not detected with a CT scan or during the VATS procedure.
The mean pleural fluid LDH level was 2650 IU/L, and TP was 4.25 g/dL, consistent with an exudative process.Some studies suggested that the pleural fluid to serum fluid bilirubin ratio could be used to differentiate exudative pleural effusion from transudative effusion, especially in resourcelimited settings.The cutoff ratio suggested was less than 1 [124,125].The presence of bilirubin in the pleural space causes a cascade of inflammatory responses.This can lead to potential loculated pleural effusions [3] and also respiratory compromises like hypoxic respiratory failure or acute respiratory distress syndrome (ARDS) [62].In our study, the mean PB/SB was 3.47.The mean WBC count was 4540/mm 3 , with mostly neutrophils predominant.Only 18 cases reported organisms grown from the pleural fluid, which suggests that leukocytosis is likely a result of inflammation induced by bile in the pleural fluid and not necessarily related to infection.
Our study had some limitations.The quality of the evidence was low, as the identified studies were case reports and case series.Most of the case reports lacked additional pleural fluid studies and information on cultures and cytology.Despite the low level of quality of evidence, this is the first systematic literature review involving four databases and is expected to help clinicians diagnose and treat bilothorax.Secondly, we included the abstracts presented at reputed societal conferences, which is again low-quality evidence, but this was done to minimize publication bias.Another limitation was the lack of SARS-CoV-2 PCR testing in the pleural fluid of the patient we reported.However, pleural effusion due to SARS-CoV-2 infection is relatively rare, with an incidence of around 2-11%, and is mostly bilateral.It is a late complication that appears around three to four weeks, is seen with severe parenchymal involvement, and carries a worse outcome [126].Our patient had unilateral pleural effusion and did not have the parenchymal involvement that is commonly seen with COVID-19 pneumonia.We strongly believe this pleural effusion was unrelated to her concurrent SARS-CoV-2 infection.
There are no guidelines or consensus on how to treat bilothoraces, and based on the results of our review, we suggest that chest tube drainage should be the first line of treatment, with testing for the presence of infection with pleural fluid culture.If this is inadequate, nuclear studies should be done to investigate the presence of diaphragmatic defects.
One should not rely on a CT scan of the chest or abdomen for the diagnosis of the BPF, as the yield seemed low.Then, surgical consultation to correct the existing BPF should be obtained for persistent bilious pleural effusion or large diaphragmatic defects seen on the radiological scans.

Conclusion
Bilothorax should be considered in new-onset pleural effusions, particularly of the right side, in patients with a history of surgery, trauma, radiation, or infection of the hepatobiliary structure.The measurement of pleural fluid and serum bilirubin level usually confirms the diagnosis.Treatment is generally done with drainage of bilious pleural effusion, preferably with a chest tube.The presence of a bilopleural fistula plays a role in determining the need for surgical correction.

Table 1 :
Descriptive analysis of abstracts and case reports.

Table 2 :
Descriptive analysis of observational studies.was reported in 51% the patients.It was diagnosed either with radiological investigation, during the VATS procedure, ERCP, or through dye injected in the pleural or abdominal cavity.Only 12 patients needed repair of the fistula which accounted for only 10% of the patients.The most common radiological investigation used

Table 3 :
Quantitative analysis of the pooled patients-part I.

Table 4 :
Quantitative analysis of the pooled patients-part II.