Risk factors for the recurrence in pulmonary tuberculosis patients with massive hemoptysis

Abstract Objectives To evaluate the outcomes of bronchial artery embolization (BAE) for the treatment of massive hemoptysis in patients with pulmonary tuberculosis and identify risk factors that influence recurrence. Methods A total of 81 patients with massive hemoptysis who underwent BAE between January 2014 and December 2017 were retrospectively reviewed. All of the patients had either a history of pulmonary tuberculosis or a current diagnosis of pulmonary tuberculosis. Follow‐up ranged from 18 to 66 months. Results Hemoptysis was stopped or markedly decreased, with subsequent clinical improvement in 73 patients, while 11 patients experienced recurrence during the follow‐up period. Systemic‐pulmonary shunts and clinical failure showed a statistically significant correlation with the recurrence rate. The cumulative non‐recurrence rate was 95.3% for 3 months and 81.9% for more than 24 months. Complications were common (12.5%), but self‐limiting. Conclusions BAE is a safe and effective treatment option for the control of massive hemoptysis in pulmonary tuberculosis patients. Systemic‐pulmonary shunts and clinical failure are the risk factors for recurrence.

significant source of hemoptysis, which has a variety of sequelae and complications such as fibrosis, cavities, aspergilloma, end-stage lung destruction, airway lesions, and vascular lesions.
Generally, emergency treatment is performed to stop bleeding. However, the effects of conservative medical therapy are not quite satisfactory. The surgical excision has better effects, but the postoperative complications and mortality can be increased as well. 2 With the recent development of interventional therapy, bronchial artery embolization (BAE) has been clinically proven to be effective for pulmonary TB patients with massive hemoptysis. After BAE, the recurrent bleeding rate is 10%-45%. [3][4][5] Patients with chronic lung disease, including TB, have a particularly high recurrence rate. According to previous studies, recurrent bleeding may be caused by the recanalization of previously embolized arteries and/or the introduction of new bronchial and nonbronchial arteries through disease progression and can be treated by repeated embolization or surgery. 4,5 Although BAE in TB patients has been studied previously, the results on risk factors for hemoptysis recurrence are not consistent. Therefore, this study aimed to evaluate the short-term and long-term effects of BAE and identify the risk factors for the recurrence in pulmonary TB patients with massive hemoptysis.

| Study population
This was a retrospective clinical study. In total, 81 patients with massive hemoptysis of TB or post-TB sequelae in our hospital from January 2014 to December 2017 were reviewed. Massive hemoptysis was defined as (1) once expectoration of >100 mL blood or >500 mL within 24 h. 6 All patients were treated with BAE for the first time. Patients who were unavailable for follow-up were excluded. This study was approved by the Ethics Committee of Fujian Medical University Affiliated Fuzhou First Hospital, and all patients provided informed consent.

| Clinical definitions
All of the patients were classified as follows 7 : 1. Active TB: no known history of TB and (i) AFBpositive and/or images suggestive of active TB, or (ii) if AFB-negative or unknown, current TB medication as the first treatment episode or images strongly suggestive of active TB and clinical suspicion. 2. TB sequelae: previous history of TB and AFB-negative, with/without images suggestive of inactive TB. 3. TB reactivation: previous history of TB and AFBpositive, with images suggestive of active/inactive TB.
According to the guidelines for percutaneous transcatheter embolization established by the Society of Interventional Radiology (SIR) Standards of Practice Committee, 8 clinical success was defined as either complete stoppage of or reduced hemoptysis associated with a positive effect on clinical course. If hemoptysis continued without change, it was regarded as clinical failure. Recurrent hemoptysis was defined as the occurrence of frank hemoptysis with no blood-tinged sputum, which should be treated with hospitalization or BAE.

| Data collection
The general data on all patients were obtained from the standardized data sheet in the hospital's computer system, including demographic data, clinical manifestations, chest radiography, computed tomography (CT), and fiber optic bronchoscopy. The amount of bleeding and disease activity were measured. The outcomes of hemoptysis, postoperative complications, recurrent bleeding factors, and corresponding intervention measures were also recorded. For each patient, the end of follow-up was defined as the date of death or the last day of March 2019. All patients were followed up by inpatient and outpatient medical records and by contact with the patient through telephone calls.

| Management
All patients received conventional medical treatment after admission, including an unobstructed respiratory tract, oxygen uptake, hemostasis, symptomatic therapy, and anti-tuberculous or antibiotics if necessary. The standard procedures of BAE were performed as follows: The modified Seldinger technique was used for femoral artery puncture; a 4F/5F catheter was placed for selective angiography. After the diagnosis, a 2.4 F microcatheter was inserted into the bronchial artery branch, and the coronary artery and spinal artery should be avoided in the process. According to the artery diameter and arteriovenous fistula, the materials for embolization were chosen, which were composed of gelfoam or gelfoam + coil.

| Statistical analysis
Data were analyzed using SPSS v21.0 software (IBM Corp., Armonk, NY, USA). The Student's t-test, Mann-Whitney U test, or Kruskal-Wallis test was used for continuous variables; the χ 2 test or Fisher's exact test was used for categorical variables. The recurrent-free time was calculated using the Kaplan-Meier method. Univariate Cox regression analysis was used for the univariate analysis of recurrence, and p < 0.1 was used for the screening of variables. The Cox proportional hazards model was used for multivariate analysis to identify the independent factors of recurrence-free time. p < 0.05 was considered statistically significant.

| Demographic characteristics and outcomes
A total of 81 patients underwent BAE procedures for massive hemoptysis due to TB or its sequelae. Hemoptysis was immediately stopped or markedly reduced, with subsequent improvement of the clinical course, in 73 patients (90.1%). Hemoptysis continued without change in eight patients after BAE; one patient died because of hemoptysis on the eighth day after BAE; others were improved after medical treatment or blood transfusion. Eight patients were lost to follow-up after being discharged from the hospital. Eventually, 72 patients were involved in the analysis of recurrence outcomes.
The 72 patients (61 males, 11 females) had a median age of 53 years old (18-86 years old) and a median bleeding volume of 600 mL (100-2000 mL). A total of 32 patients (44.4%) had active TB, including 8 cases of TB reactivation and 40 cases (55.6%) of TB sequelae ( Table 1). The median follow-up period was 37 months (range 18-61 months). A total of four patients died during the follow-up period due to acute exacerbations of chronic bronchitis with respiratory failure (n = 1), pulmonary infection (n = 2), and lung cancer (n = 1), and none of them had hemoptysis recurrence. A total of 57 patients did not have recurrent hemoptysis, whereas 11 patients experienced recurrence. Among the 11 patients with recurrent hemoptysis, four patients improved upon conservative management, seven patients repeated BAE, and two cases underwent BAE for the third time (Tables 2 and 3). The cumulative hemoptysis control rate was depicted by the Kaplan-Meier curve ( Figure 1).

| Factors affecting outcome of BAE
Of the 72 patients, 11 (15.3%) experienced hemoptysis recurrence during the follow-up period. Neither of the groups showed statistically significant differences with regard to the following: the presence of TB sequelae such as 11.8% (6 of 51) for fibrotic scar change, 9.4% (3 of 32) for cavity, 23.1% (9 of 39) for bronchiectasis, and 20.0% (3 of 15) for TB-destroyed lung. Disease activity was not associated with the risk of rebleeding (OR, 4.288; 95% CI, 0.910-19.657; p = 0.066). Embolism materials, sex, amount of bleeding, or diabetes mellitus were not associated with the risk of recurrence. Table 5 shows the result of the multivariate Cox regression analysis. The outcome after BAE and systemicpulmonary shunts showed statistical significance. The cumulative hemoptysis control rates were depicted by the Kaplan-Meier curve for the outcome after BAE and systemic-pulmonary shunts, respectively (Figures 2 and 3).

| Complications
A total of 9 patients experienced procedure-related complications (12.5%), including gastrointestinal symptoms (n = 3), fever (n = 4), chest tightness (n = 1), and shortness of breath (n = 1). Only one patient died after intervention; the cause of his hemoptysis was secondary pulmonary TB, and the cause of his death was massive hemoptysis asphyxia. Most procedure-related complications were minor and could be improved spontaneously or by nominal therapy. No serious complications, such as a spinal artery embolism, occurred.

| DISCUSSION
Massive hemoptysis is one of the most common complications in pulmonary TB patients, which may lead to respiratory tract obstruction and asphyxia, hemorrhagic shock, or even death. Meanwhile, patients could suffer heavy psychological burdens due to aggravated illness states and the difficulties of medical treatment, which affect the final treatment outcomes. 9,10 Previous reports showed that the mortality rate for pulmonary TB patients with massive hemoptysis could be as high as 50% without standardized treatment. 11 Although the pulmonary lobectomy has shown good effects, it is limited due to severe trauma, which may affect the life quality, functional status, and complications of the patients. However, conservative medical therapy extends the treatment time and produces side effects, and patients are prone to have palpitations, emesis, and nausea while applying pituitrin for vasoconstriction in the long term. 12 With recent technological development, vascular interventional therapy has made great achievements. Interventional therapy has been primarily recommended for favorable therapeutic effects in pulmonary TB with massive hemoptysis. 13 Previous studies had shown that the treatment efficacy of bronchial arterial embolization in patients with massive hemoptysis of TB was as high as 90%. [14][15][16] Our immediate success rate of BAE was 90.1%, which was consistent with previous studies. During the follow-up period, the recurrence rate was 15.3% in this study, which was comparable to previous reports ranging from 10% to 45%. 5,17 F I G U R E 1 The cumulative hemoptysis control rate as depicted by the Kaplan-Meier method. BAE, bronchial artery embolization.
There are two peak periods of hemoptysis recurrence. The first peak for recurrence is from 1 to 2 months after BAE, which may be because of incomplete embolism. The second peak for recurrence is from 1 to 2 years after BAE. This seems to reflect the recruitment of blood supply and revascularization due to the progression of the underlying disease. No similar peak recurrence period was observed in our study. Consistent with previous studies, short-term bleeding recurrence reflects incomplete embolism, whereas long-term bleeding recurrence may  be probably because of recanalization after BAE, systemic-pulmonary shunts, and the progression of the underlying disease. 18 Overall, BAE has a high success rate in treating massive hemoptysis of TB, which can be recognized as the first-line treatment for replacement of high-risk emergency surgery or selective surgery. Moreover, for massive hemoptysis caused by TB, the outcome of repeated BAE procedures is generally good. High recurrence rates for patients with aspergilloma have been shown in previous reports. 19,20 Patients with aspergilloma have extensive parasite circulation from different sources, which may be associated with vasculitis in cases of thick vascular cavitary wall. In our study, aspergilloma was associated with the rate of recurrent hemoptysis in univariate Cox regression analysis. However, there was no significant difference in multivariate Cox regression. The sample size might be too small to detect the difference between the two groups.
Lee et al. 17 reported that active TB was related to the recurrence rate after BAE. Active TB patients have persistent active mucosal inflammation, which is regarded as a recurrence factor after BAE. However, other studies showed that active TB was not related to the recurrence rate after BAE. [21][22][23] Some studies have shown that active TB had more favorable results in terms of recurrence. In our study, we found no significant effects of TB sequelae and active TB on recurrence.
Our analysis showed that systemic-pulmonary shunts and clinical failure were risk factors for the recurrence of BAE in patients with tuberculous hemoptysis. Especially with the increased shunt, the large size of embolic materials should be chosen to prevent emboli from entering into the pulmonary circulation. As a consequence, the incomplete blood vessel embolism leads to recurrent hemoptysis. 9,[24][25][26] Clinical failure is an important factor for the risk of recurrence after BAE because patients who fail BAE for the first time will have a higher recurrence rate in the future. It might be postulated that surgical treatment should be recommended for these patients.
In this study, the complication rate was 11.1% (9/81). All of the symptoms were mild and could be managed conservatively. Other major complications, such as spinal cord ischemia or mediastinal structure necrosis, did not occur. Procedure-related complications are common, but F I G U R E 3 The cumulative hemoptysis control rate as depicted by the Kaplan-Meier method: p = 0.035; shunt (À) versus shunt (+). BAE, bronchial artery embolization. most are minor. 27 This is probably contributed to the use of digital subtraction angiography (DSA), super-selective catheterization of the abnormal vessel, and the use of a micro-catheter for the particle embolization.
Our study has some limitations. First, this is a singlecenter study, and the sample size might not be sufficient to identify minor factors that could be related to the recurrence of hemoptysis. Second, the study design is retrospective, which may cause high risk of biases.
In conclusion, BAE is a safe and effective procedure with minimal complications that can be performed routinely in pulmonary TB patients presenting with massive hemoptysis. Systemic-pulmonary shunts and clinical failure are the risk factors for recurrence.
AUTHOR CONTRIBUTIONS Qiong Lin, Jian Chen, Tianxing Yu, Bing Gao, Kaijin Kuang, Yong Fan, Junping Xu, Xiaohua Li, and Xin Lin collected and analyzed the data, Liyu Xu designed the study and wrote the manuscript. All authors read and approved the manuscript.