Contradictory Pleural based Mass Occurred in Patients with Pleural Tuberculosis during the Treatment: A Clinical Observational Study in China

Background: Recent years pleural based mass newly occurred during anti-tuberculosis (TB) treatment had been observed in patients with pleural TB. Its occurrence is contradictory and unfavorable for the outcome of patients. This study aims to explore its clinical, pathological and bacteriological characteristics. Methods: Patients newly diagnosed as pleural TB met included criterion were prospectively enrolled into the study. Patients were followed up throughout the treatment, clinical data were recorded. Percutaneous biopsy and surgical tissues from pleural based mass were pathologically and bacteriologically examined, related clinical factors were calculated by Fagan's nomogram and ROC curve. Results: A total of 122 patients with pleural TB were enrolled. 34.4% of them (42/122) were newly observed pleural based mass during the treatment, 12 cases received surgical resection at the 12±0.5 months of treatment course, under surgical observation only 58.3% of them (7 /12) were located on pleura, 41.7% (5/12) of them were located in lung parenchyma. Pathological observations showed that pleural based masses were divided into three types: granulomatous inammation, brous hyperplasia and necrosis type. Mycobacterium tuberculosis PCR was positive in 57.1% of cases (24/42), any rst-line anti-TB drug resistance gene mutation was positive in only 9.5% (4/42). Besides 12 cases receiving surgical operation, there were 86.7% of patients (26/30) in which pleural based mass still existed at the end of 12 months course. Young age, pleural thickness and adhesion, higher LDH, ADA, TGF-β and PAI-1, lower GLU and t-PA in pleural effusion were risk factors of occurring pleural based mass. Conclusions: Pleural based mass had been observed in about one third of patients with pleural TB, which were located in lung or pleura and divided into three pathological types. There are some clinical characteristics as its predictive risk factors.


Background
Tuberculosis (TB) remains one of the leading infectious diseases in the world. There were estimated 10.0 million new cases of TB worldwide in 2019 [1]. Pleural TB ranks the second most common form of extrapulmonary TB [2,3]. Pleural based mass in patient with pleural TB had been observed by imaging recent years [4], most of them were newly developed during anti-TB treatment, which was "contradictory changes" under chemotherapy. There were several literatures reported its occurrence [5][6][7][8]. We described it in the following case shown in gure 1, which presented radiological changes during the process of the occurrence of pleural based mass. Figure 1A showed the mass pleural effusion in the right chest; Figure   1B showed pleural effusion was clear after the drainage of pleural uid and starting the treatment; Figure  1C showed the occurrence of pleural based mass in the same site of the right chest after 4 months of chemotherapy; Figure 1D Pleural based mass resected by surgery was found no recurrence in a year's follow-up.
Patients with pleural based mass were frequently found in clinics in China. However, published studies only focus on case reports and mainly reported in Asian regions with high TB burden [5][6][7]. There had no uni ed treatment regimen for patients when pleural based mass was observed in imaging during patients received anti-TB treatment. Why does pleural based mass happen? How does pleural based mass show pathologically, which was located in pleural or lung? How should patients proceed to get treatment after pleural based mass occurred? Patients should maintain the origin regimen or change the regimen and how to change if changing the regimen was needed? However, all above questions still remain elusive.
In the present study, we conducted a clinical observational study to describe its clinical characteristics, explored the pathological ndings and associated risk factors of pleural based mass in order to clear the recognition about the occurrence of pleural based mass in the patients with pleural TB.
Materials And Methods

Study design and patients' information
The study was conducted in a national pulmonary disease specialized hospital, and the patients were mainly from eastern six provinces and one municipality in China. From January 1, 2017 to June 30, 2018, patients diagnosed as pleural TB met include criteria were prospectively enrolled into the study, which was approved by the ethics committee of Shanghai Pulmonary Hospital, Tongji University School of Medicine (No. K16-131). Written informed consent was obtained from all enrolled participants. Included patients were collected pleural effusion which were sent for biochemical and cytokines tests, then followed up throughout the treatment, observed the treatment outcome, all patients with pleural based mass were obtained biopsy tissue via surgical resection or aspiration biopsy. Biopsy tissues were tested by anti-TB drug resistant genes mutation and pathological observation.
Inclusioncriteria were as followed: newly diagnosed as pleural TB and pleural effusion could be punctured and obtained; and had no pleural based mass or nodules at the initiation of treatment; and negative in serum human immunode ciency virus (HIV) and had no previous history of anti-TB treatment.
Excluded criteria were as followed: Patients had malignant tumor; or with immunosuppressive status; had other pulmonary diseases; could not be drawn pleural effusion.
The standard of diagnosis of pleural TB was according to the WHO guideline: patients were diagnosed as pleural TB if there was at least one of the following criteria [9,10]: (1) identi cation of mycobacterium tuberculosis (MTB) in pleural effusion by acid-fast bacilli (AFB) smear; (2) detection of MTB by culture or DNA PCR in pleural effusion; (3) typical pathological ndings on a pleural biopsy; and (4) anti fast bacteria (AFB) stain or culture positive in sputum and chest radiology indicated abnormal lesions compatible with pulmonary TB; and clinical comprehensive standard in accordance with pleural TB (immunology, biochemical or histopathology and responsive to anti-TB treatment). All patients included were followed up throughout the treatment course. Outcome of pleural based mass was evaluated and included obviously absorbed (lesion absorbed by more than 50%); partly absorbed (lesion absorbed less than 50%), no change and deteriorated (lesion enlarged or diffused).

Data and sample collection
All clinical characteristics including age, symptoms, complications, diagnosis, pleural thickness, pleural nodule, bacteriological tests and treatment regimen were recorded completely. Pleura thickness > 2 mm was considered as plural thickness [11].
Patients with pleural TB were performed percutaneous transthoracic puncture to collect pleural effusion sent to clinical laboratory for biochemical test, part of which was frozen at -20 °C. Biomedical examinations included adenosine deaminase (ADA), lactate dehydrogenase (LDH), protein and glucose. Meanwhile, IFN-γ, PAI-I, t-PA and TGF-β in pleural effusion were tested by enzyme-linked immunosorbent assay (ELISA) (ELISA kit was from eBioscience Corp, San Diego, CA). ELISA were detected on the reader with 450nm wavelength within 30 minutes after the end of operation. All procedures were performed in strict accordance with the product instructions.

Surgery and biopsy
Part of patients with pleural based mass were received surgical resection decided by surgeons, physicians and patients. Surgeons decided the operative approach and recorded the precise location of pleural based mass during the course of operation. Other patients with pleural based mass were performed by CT guided percutaneous biopsy to obtain the tissue. Biopsy tissues from surgery and aspiration were sent for pathology and bacteriological tests.

Pathological observation and immunohistochemistry (IHC)
The biopsy tissues were pathologically tested. Formalin xation and para n embedding (FFPE) sections stained with hematoxylin-eosin (HE) were observed by two experienced pathologists, and another pathologist was invited to provide a diagnosis in cases of disagreement between two pathologists. Ziehl-Neelsen stain was used to examine AFB with 1000 times magni cation. IHC using an envision detection system (Leica Biosystems Melbourne Pty Ltd., Melbourne, Australia) was carried out to estimate expression of PAI-1(ab226946, 1:200, Abcam, USA) and t-PA (ab157469, 1:50, Abcam, USA) in pleural based mass lesions comparing with pleural TB cases without pleural based mass. As a classic methodology [12], H-score was used to access IHC results. The score calculated using the formula 1 × (% of 1 + cells) + 2 × (% of 2 + cells) + 3 × (% of 3 + cells). Zero for 'no staining', 1 + for 'light staining visible only at high magni cation', 2 + for 'intermediate staining' and 3 + for 'dark staining'.

Anti-TB drug resistance genes mutation detection in biopsy tissues
All biopsy tissues were sent for molecular pathological test, including real-time PCR and PCR-reverse dot blot to detect MTB and drug resistance gene mutations. Para n sections 4~5 µm thick, 8~10 pieces were put into Eppendorf tube. And then total DNA was extracted by using an E.Z.N.A. FFPE DNA Kit (Omega Bio-tek, Inc., USA) according to the manufacturer's protocol. Four µl of the extracted DNA was used for real-time PCR to detect speci c gene sequence IS6110 of MTB on a Stratagene Mx3000P QPCR System (Agilent Technologies, Santa Clara, CA, USA). After the ampli cation, the data was analyzed on Mx3000P. Subject to quality control (no S type ampli cation curve in negative control PCR batch; S type ampli cation curve in positive control PCR batch with the Ct value< 30; in internal standard channel, DNA sample with the Ct value< 45), the MTB is positive if there is an S type ampli cation curve with the Ct value< 37 in MTB detection channel. Another 4 µl of the extracted DNA was used for PCR-reverse dot blot to detect rifampicin, isoniazid, streptomycin and ethambutol resistance (rpoB, katG, inhA, embB and rpsL gene accordingly) by anti-TB drug resistance genes mutation detection kit (Yaneng Co, Shenzhen, China) according to the manufacturer's instructions. Clear blue dot indicates positive MTB drug resistance gene. Positive and negative controls were set up in all above test.

Statistics
The statistical analysis and data visualizing were performed using SPSS statistics version 25.0 (SPSS, Inc., Chicago, IL, USA), R 3.5.3 software (The R Foundation for Statistical Computing) and Stata/MP 14.0. Differences with P <0 .05 were considered signi cantly. The diagnostical value of each biomarker was calculated with the area under receiver operation curve (ROC) curve (AUC). The Fagan's nomogram was used to present pre-test probability and post-test probability which derived from Stata module (FAGAN: Stata module for Fagan's Bayesian nomogram Patients with occurring pleural based mass were younger (29 years old vs. 42 years old, p=0.001) and had higher rate of pleural thickness (78.6% vs. 26.3%, p=0.001) and pleural adhesion (47.6% vs. 18.8%, p=0.001) and higher proportion of packaged pleural effusion (47.6% vs. 18.8%, p=0.001) compared with those without occurring pleural based mass (P=0.001), however, MTB culture positive rates from any specimens were similar in both groups, P>0.05 (Table 2).

Biochemical and cytokines expression in pleural effusion
To further explore characteristics of patients with occurring pleural based mass, we tested biochemical and cytokines in pleural uid at the initiation of anti-TB treatment, LDH and ADA were signi cantly higher and glucose was lower in patients with pleural based mass patients than those without pleural based mass, p value was 0.001, 0.001, 0.008, respectively. TGF-β and PAI-1 were statistically higher and t-PA was signi cantly lower in patients with pleural based mass than those in patients without pleural based mass (20.021 μg/L vs. 16.386 μg/L, P=0.044, 241.014 μg/L vs 187.152 μg/ L, P=0.048, 1.733 μg/L vs 3.025 μg/L, P=0.024), Detailed data were showed in Table 3.

Pathological observations
Pathological diagnosis was made on 12 surgical and 30 biopsy specimens. According to pathological ndings, pleural based masses had three pathological types: (1) granulomatous in ammation type, in which lesions mainly included in ammatory cells, epithelioid histiocytes, Langhans giant cells with or without necrosis. Most pleural based mass lesions (34/42, 81.0%) presented this type; (2) brous hyperplasia type, in which lesions mainly consisted of broblasts and brous proliferation without any granulomatous changes, few cases (3/42, 7.1%) presented this type; (3) necrosis type, in which lesions showed large area of caseous necrosis with or without in ltration of in ammatory cells, brous hyperplasia and granuloma were not obviously observed. Necrosis seemed as dominant morphology including 2 cases of encapsulated empyema, 5 cases (5/42, 11.9%) presented type 3(See in gure 4).
AFB stain was positive in 52.4% of cases (22/42,). IHC expression of PAI-1 (50 vs. 20, p<0.05) was higher and t-PA (100 vs. 160, p<0.05) was lower in patients with pleural based mass compared to patients without pleural based mass, which were in accordance with the cytokines change trend in pleural effusions tested by ELISA.

Treatment and outcome:
We following all patients with pleural TB, all patients without pleural based mass had cured after full course of chemotherapy. Among 42 cases with pleural based mass, 12 cases received surgical resection got cure, 73.3% of cases (22/30) without received surgical resection had their pleural based mass no change after nished one year of chemotherapy, only 4 cases (4/30, 13.3%) got partly absorption and 4 cases (4/30, 13.3%) got obviously absorption, and no patient found their lesions deteriorated. All patients remained the anti-TB regimen no matter pleural based mass occurred or not.

Analysis of predictive risk factors of pleural based mass
We found nearly 30% of patients with pleural TB occurred pleural based mass during the treatment. To explore the risk factors of occurring pleural based mass on patients with pleural TB, ROC curve was used to choose the best appropriate clinical characteristics, biochemical and cytokines indexes for association of occurrence of pleural based mass. After screening and calculating the value of AUC, age, glucose, ADA, LDH, protein, TGF-β, PAI-1 and t-PA were associated with the occurrence of pleural based mass, data were shown in table 6.
We further used Fagan's nomogram to evaluate the prognostic possibility of pleural based mass occurrence. We found the probability of pleural TB to develop pleural based mass can reach 99.7% when patients have pleural thickness, packaged pleural effusion, LDH>461.5IU/L, ADA>62.9IU/L, GLU<4.75mmol/L, TGF-β>15.235μg/L, PAI-1>180.720μg/L, t-PA<2.875μg/L and age<32.5 years old. There was no probability (0.0%) to develop pleural based mass when patients do not have above nine characteristics (see in table 6 and gure 5).

Discussion
In recent years physicians found patients with pleural TB had likelihood to develop pleural based mass during anti-TB treatment (see in gure 1). However, there is little study regarding the reason and deep characteristics of occurring pleural based mass. We performed a prospective observational study focused on clinical, immunological, bacteriological and pathological ndings in order to get further knowledge on pleural based mass occurrence during the treatment.
During the follow-up of treatment on pleural TB, we found 34.4% of patients (42/122) developed pleural based mass. 12 cases were performed surgical resection due to failure absorption after months of anti-TB chemotherapy. Among the resting 30 cases, 73.3% of the patients (22/30) did not get pleural based mass any absorbed although nished enough course of anti-TB treatment; secondly, AFB positive was in 52.4% (22/42) of 42 cases within pleural based mass and MTB DNA positive was in 57.1% (24/42) of them; what is more, cases with pleural based mass in larger size (>3 cm ) had more possibility (90.9%, 20/22)of not being absorbed than those with pleural based mass in smaller size ≤3 cm (9.1%, 2/22), implying the poor outcome of cases once occurring the pleural based mass especially the larger lesions.
For further explore the characteristics of pleural base mass, we observed its location via surgical resection and pathological ndings, we observed only 58.3% of "pleural based mass" were located on pleura, 41.7% of "pleural based mass" were actually located at lung parenchyma , which suggested that pleural based mass was observed by imaging methods, which actually were partly located at pleura and partly located in lung, indicating pulmonary lesions closed to the pleura and involvement with pleural TB.
We further observed the pleural based mass through pathological ndings on biopsy specimens, they were three pathologically types: granulomatous in ammation type, brous hyperplasia type and necrosis type. The basic pathological change of TB was necrotizing granulomatous in ammation with varying numbers of accompanying non-necrotizing granulomas [13,14]. Our study showed granulomatous in ammation was the main type in pleural based mass (81.0%). However, the pathological morphology is various in different cases. MTB with strong virulence can cause massive necrosis without any granulomas in patients with hypo-immunity, and 11.95 of cases were as this type (necrosis type) in the present study. Despite two above pathological types, we found 7.1% of patients were exhibited as brous hyperplasia, which was thought to be more common in strong immunity environment.
We speculated pleural based mass was likely to be associated with treatment failure because MTB was positive within the lesions although patients had taken enough course of chemotherapy. How to treat the patients once occurring pleural based mass, our observation indicated that strengthen anti-TB regimens should be required because majority of pleural based masses were tuberculoma with positive MTB or AFB detected within lesions, majority of cases had treatment failure in absorbing the pleural based mass when they did not change their treatment regimen. Further studies should be designed to explore the optimal regimen for pleural based TB mass occurrence during anti-TB treatment.
In order to answer the question whether drug-resistance play an important role in the occurrence of pleural based mass, we performed the detection of anti-TB drug resistance gene mutation, the results showed that drug-resistance was not the main reason of occurring pleural based mass because only 9.5% of cases were detected having drug resistant gene mutation.
The occurrence of pleural based mass indicated the pleural TB contradictory change and make the further treatment complicated and bring dilemma. Therefore, exploring the predicted factors of pleural based mass is helpful for physicians to prevent its occurrence. We tested biochemical indexes and cytokines in pleural effusion before the treatment, we found that nine clinical and biochemical factors including LDH, plural thickness, ADA, glucose, t-PA, age, PAI-1, packaged pleural effusion and TGF-β were predictive risk factors of occurring the pleural based mass. Pleural thickness and packaged pleural effusion were indicator of local immune response [15,16]. Increased TGF-β was associated with the occurrence of pleural thickness. Increased PAI-1 and decreased t-PA was associated with pleural thickness in pleural TB [17,18]. Previous study showed pleural thickness may be associated with lower glucose and higher ADA in pleural effusion [19][20][21]. Therefore, the present results regarding to higher ADA, PAI-1, TGF-β and lower glucose and t-PA associated with pleural based mass should be reasonable. These nine factors were able to help physicians to prevent occurrence of pleural based mass by some measures such as thoroughly drainage of pleural effusion as early as possible, injecting urokinase into packaged pleural effusion to reduce the pleural thickness.
In summary, pleural based mass was common in patients with pleural TB and had one third of occurrence rate, half of them were located in lung, presented as granulomatous in ammation, necrosis or brous hyperplasia pathologically. Nearly one half of cases had AFB positive and had lower resistant rate within the lesion, majority of cases did not get any absorbed if did not changing the regimens. Combination of 9 clinical and biochemical factors could be useful for physicians to predict occurrence of pleural based mass. Tables   Table 1 Clinical    Surgical ndings of MSP. A Pleural based mass was located in lung, and one side of boundary was close to the pleura. B Pleural based mass was found on pleura adjacent to lung parenchyma. C Pathological morphology of case in the gure B, mass was based on pleura, and there was a clear boundary between pleural based mass and lung parenchyma (black circle). Fagan's nomogram for the 9 tests of 122 cohort. Which take in sequence of with LDH, plural thickness, ADA, GLU, t-PA age PAI-1, packaged pleural effusion, TGF-β.