Abstract
Background
Previous studies suggest that recurrence of hemoptysis after arterial embolization is associated with the underlying pulmonary disease.
Purpose
To compare the baseline information and imaging findings in patients with hemoptysis due to either chronic pulmonary tuberculosis (PTB) or bronchiectasis and to identify predictors of rebleeding after embolization treatment.
Material and Methods
Clinical data of all consecutive chronic PTB and bronchiectasis patients who underwent arterial embolization for hemoptysis from January 2010 to January 2017 in a single center were reviewed. Baseline clinical information, radiological features, and rebleeding rates were compared between patients with chronic PTB and patients with bronchiectasis. Multivariate analysis was used to identify risk factors of recurrence in each patient group.
Results
Seventy-six patients with chronic PTB and 97 patients with bronchiectasis were included. Male sex, pleural thickening, multiple embolized arteries, and non-bronchial systemic arterial (NBSA) blood supply were more common in chronic PTB patients. The short-term and long-term recurrence-free rates were significantly lower in the chronic PTB group (P < 0.001). For the chronic PTB group, the presence of lung destruction and shunts were independent predictors of rebleeding during follow-up. Compared with patients who did not undergo computed tomography angiography (CTA) before the procedure, patients with CTA showed less recurrence in the first month after treatment (P = 0.019).
Conclusion
Chronic PTB patients had more extensive NBSA blood supply and experienced higher short- and long-term recurrence rates compared with bronchiectasis patients. The risk of rebleeding was high in chronic PTB patients with lung destruction and/or shunts.
Introduction
Hemoptysis—spitting of blood that originated in the lungs or bronchial tubes—is a common symptom in pulmonary disease and can be life-threatening due to asphyxia and acute blood loss (1,2). Arterial embolization is increasingly accepted as the therapy for hemoptysis in patients for whom conservative treatment is ineffective (3). Novel technical advances and improved embolic materials have increased success rates (4,5), but recurrence of hemoptysis after successful arterial embolization remains a significant concern, with rates in the range of 10–55% (2,6,7). Previous studies suggest that recurrence of hemoptysis after arterial embolization is associated with the underlying pulmonary disease (5,6,8–11).
Chronic pulmonary tuberculosis (PTB) and bronchiectasis are two common benign conditions in which hemoptysis occurs. In both scenarios, hemoptysis can be well controlled by arterial embolization (5,10). However, there are many differences between hemoptysis caused by chronic PTB and bronchiectasis, specifically in regard to chest computed tomography (CT) imaging features and angiographic characteristics (5,8,11,12), which conceivably might indicate the possible therapeutic effectiveness of arterial embolization (5,11).
The primary purpose of this retrospective study was to compare the baseline clinical data and imaging findings in patients with hemoptysis due to either chronic PTB or bronchiectasis and the second was to investigate if any risk factors could be defined in case of rebleeding after embolization treatment.
Material and Methods
Study design and patient selection
Review of medical records and radiographic data and telephone follow-up interviews were approved by the Institutional Review Board and informed consent was waived.
The records of 283 patients who underwent arterial embolization at our institution for hemoptysis (January 2010–January 2017) were reviewed. The focus of this study was on hemoptysis caused by chronic PTB or bronchiectasis. Chronic PTB was defined based on the previous history of tuberculosis and current negative acid-fast bacillus smears, with or without imaging data indicating bronchiectasis, calcified nodules, or fibrosis (8). Chronic PTB was distinguished from bronchiectasis on the basis of past history of PTB and presence of lung parenchymal destruction on the CT image. The exclusion criteria were as follows: hemoptysis caused by underlying diseases other than PTB or bronchiectasis (including lung cancer in six cases, pneumonia in 20 cases, chronic obstructive pulmonary disease in 14 cases, radiation pneumonitis in one case, pneumosilicosis in two cases, and cryptogenic hemoptysis in 12 cases); active tuberculosis in 39 cases or relapse of a previous tuberculosis infection in one case; and previous history of arterial embolization in nine cases. Moreover, six patients with aspergillosis were also excluded as they underwent lobectomy within 1–2 weeks after arterial embolization. Ultimately, 76 patients with chronic PTB and 97 patients with bronchiectasis were selected. Among these patients, 93 were referred from a local hospital where seven underwent chest CT and CT angiography (CTA), and the others underwent non-contrast chest CT. Among the patients who directly came to our institution (n = 80), 62 underwent chest CTA examination; the other 18 patients underwent non-contrast CT either because the patients could not follow instructions necessary for the CTA examination or because their situation was critical, necessitating immediate embolization. Chest CT or CTA were used to assess the underlying pulmonary etiology, extent of pulmonary disease (number of affected lobes), presence of lung destruction, and pleural thickening. Imaging results were assessed by two chest radiologists. All patients received standard medical treatment and antibiotics were provided for those with suspected bacterial infection.
Arterial embolization procedure
All embolization procedures were performed by one of four interventional radiologists with at least five years of experience in arterial embolization. Angiography was performed by percutaneous catheterization via the femoral artery under local anesthesia. A 5-F Cobra catheter (COOK, Bloomington, IN, USA) or left gastric artery catheter were used for selective angiograms of the bronchial arteries or non-bronchial systemic arteries (NBSAs) based on the findings of chest CT, CTA, or both. The angiographic criteria for arteries with pathologic changes were as follows: engorgement and tortuosity of the artery, hypervascularity, neovascularity, systemic arterial-pulmonary circulation shunts, extravasation of the contrast agent, or bronchial artery aneurysm. After confirmation of arteries with pathologic changes, a 2.7-F microcatheter (Terumo, Tokyo, Japan) was introduced coaxially for super-selective catheterization in order to avoid important side branches. Before 2011, both polyvinyl alcohol (PVA) particles (COOK) and gelfoam were equally used for embolization treatment. After 2011, PVA particles became the primary treatment for certain abnormal arteries. Micro-coils (COOK) were also used in some cases with obvious shunts or in patients with significantly enlarged abnormal arteries after they were treated with PVA particles. The details of embolized arteries, including number and source (bronchial arteries or NBSAs), were recorded.
Follow-up and data analysis
The amount of hemoptysis was divided into three levels according to the estimated blood volume: (i) massive hemoptysis (estimated volume > 200 mL/day or need blood transfusion); (ii) moderate hemoptysis (estimated volume < 200 mL/day); and (iii) slight hemoptysis (estimated volume < 100 mL/day) (13). The volume of hemoptysis was estimated through patient observation by hospital medical personnel or from the description of patients and/or their family. Generally, we considered the volume of one whole mouthful of hemoptysis as 10 mL. Chest CT identifying pleural thickening ≥ 3 mm was considered significant (14). Lung destruction was defined as a clear history of chronic PTB, coupled with radiological findings of lung parenchymal destruction or lung volume loss in at least one lobe as determined by the chest CT (15). Technical success was defined as the complete embolization of all bronchial or pathological systemic vessels in which embolization was attempted. Recurrence after embolization was defined as a single episode of hemoptysis with > 30 mL of fresh bleeding beyond 12 h following the procedure or death due to recurrence (16). Rebleeding that presented within one month after the procedure was defined as early recurrence. Upon discharge from the hospital, patients were encouraged to follow-up with the outpatient clinic within three months after the procedure to discuss current health condition and possible recurrent hemoptysis. For prognosis evaluation, the patients were monitored by our follow-up personnel every six months and during the data collection for the present study via telephone interviews.
The endpoint of this analysis was recurrence after embolization. Recurrence-free time was calculated from the date of embolization procedure to the date of recurrence, date of death by any cause, or the date of last follow-up (March 2017 for available patients).
In those patients who underwent repeated angiography after rebleeding, we classified the causes of recurrence based on angiographic findings as missed culprit arteries, recanalization of previously embolized vessels, or recruitment of new collateral circulation (5).
Statistical analysis
Parametric data are presented as mean ± standard deviation (SD) or with confidence intervals (CI) as appropriate. Statistical differences in baseline clinical information, imaging findings, and embolization indexes were analyzed using the chi-square testing or Student’s t-test. Recurrence-free rates were estimated using the Kaplan–Meier method; log-rank test was employed to analyze differences between recurrence-free rates. Factors including age (< 60 years vs. ≥ 60 years), sex, volume of hemoptysis (massive/moderate/slight), CTA examination, lung destruction (for chronic PTB patients only), pleural thickening, systemic arterial-pulmonary circulation shunts, NBSAs involvement, and embolic material (gelfoam/PVA/PVA + coils) were introduced into a Cox proportional hazards model using a backward-elevation method to identify risk factors of recurrence in each group (9). SPSS 24.0 software (Statistical Package for the Social Sciences, Inc., Chicago, IL, USA) was used for data analysis. P < 0.05 was considered statistically significant.
Results
The clinical features of 76 chronic PTB patients (56 men, 20 women; mean age = 59.6 ± 12.6 years; age range = 26–89 years) and 97 bronchiectasis patients (55 men, 42 women; mean age = 56.3 ± 13.0 years; age range = 18–80 years) are presented in Table 1. Compared with the bronchiectasis group, the chronic PTB group showed more male dominance (P = 0.021). Pleural thickening ≥3 mm on chest CT images was also more common in this group (P < 0.001).
Comparisons of clinical features in patients with hemoptysis resulting from chronic pulmonary tuberculosis (PTB) and in patients with bronchiectasis.
*Statistically significant with P < 0.05.
SD, standard deviation; CTA, computed tomography angiography.
Characteristics of arterial embolization and outcomes
Technical success was achieved in 97.4% (74 of 76) of chronic PTB patients and 99.0% (96 of 97) of bronchiectasis patients. Technical failure occurred in one bronchiectasis patient because of the target artery dissection. Lobectomy for the diseased area of this patient was then performed. Procedures of another two chronic PTB patients were aborted because of inability to achieve stable catheter position. One of them died due to hemoptysis while the other was managed with conservative medical management. Characteristics of arterial embolization in 170 technical success patients are summarized in Table 2. In total, 153 bronchial arteries (79 right and 74 left) and 57 NBSAs (45 intercostal arteries, six internal thoracic arteries, three phrenic arteries, two subclavian arteries, and one thyrocervical trunk) were embolized in the chronic PTB patient group, and 198 bronchial arteries (97 right and 101 left) and 17 NBSAs (12 intercostal arteries, two phrenic arteries, two subclavian arteries, and one internal thoracic artery) were embolized in the bronchiectasis patient group. Compared with the bronchiectasis group, the chronic PTB group had significantly (P < 0.001) more embolized arteries (2.8 ± 1.3 and 2.2 ± 0.9, respectively), and a significantly (P < 0.001) higher occurrence of NBSAs. The representative angiographic findings from patients with chronic PTB and patients with bronchiectasis are shown in Fig. 1. All these 170 patients were associated with a marked reduction in bleeding within 12 h following embolization. No major procedure-related complications, as defined by the Society of Interventional Radiology (SIR), were observed in either group of patients (17).
Characteristics of arterial embolization in two patient groups with technical success and the follow-up results.
*Statistically significant with P < 0.05.
SD, standard deviation; PTB, pulmonary tuberculosis; NBSA, non-bronchial systemic artery; PVA, polyvinyl alcohol; IQR, interquartile range.
Typical imaging of chronic PTB and bronchiectasis patients. (a, b) Images of hemoptysis caused by chronic PTB in a 76-year-old man. (a) Mediastinal window chest CT image shows consolidative lesions in both upper lung lobes. (b) Selective artery angiogram shows enlarged and circuitous intercostal arteries—small branches which feed the diseased area—combined with the formation of systemic-pulmonary vein shunt (black arrow). (c–f) A 30-year-old man with chronic pulmonary tuberculosis hemoptysis. (c) Mediastinal window chest CT image depicts pleural thickening (black arrow) in right upper lobe and calcification (white arrows) in both upper lobes. (d–f) Angiogram showing the tortuous enlargement of bronchial arteries (black arrows) and intercostal artery (white arrow) with hypervascular parenchymal staining in both upper lobes. (g, h) A 45-year-old man with bronchiectasis and hemoptysis. (g) Transverse CT scan demonstrating bronchiectasis in left lower lobe (white arrow). (h) Selective angiogram shows tortuous enlarged left bronchial artery with hypervascular parenchymal staining.
Mean and median follow-up times for all patients were 771 and 597 days, respectively. During follow-up, exactly 50% of the chronic PTB patients (37/74) and only 20.8% of bronchiectasis patients (20/96) experienced recurrence (P < 0.001). Cumulative recurrence-free rates for the chronic PTB and bronchiectasis groups after embolization are shown in Fig. 2. The recurrence-free rates in the bronchiectasis group were significantly higher than those in the chronic PTB group (P < 0.001).
Cumulative recurrence-free rates for chronic PTB and bronchiectasis groups after embolization. The 1-, 6-, 12-, 24-, and 36-month recurrence-free rates were 79.7%, 75.7%, 68.0%, 51.5%, and 42.8%, respectively, for chronic PTB patients and 92.7%, 89.5%, 88.3%, 83.8%, and 77.5%, respectively, for bronchiectasis patients (P < 0.0001, log-rank test).
In the chronic PTB group, repeated angiograms of 16 patients who experienced recurrence revealed that the causes of rebleeding were missed culprit arteries in eight cases (of which NBSAs were in six cases and bronchial arteries in two cases), recanalization of previously embolized vessels in three cases, and recruitment of new collateral circulation in five cases. In the bronchiectasis group, repeated angiograms of 10 patients who experienced recurrence revealed the causes to be missed culprit arteries (NBSAs) in two cases, recanalization of previously embolized vessels in six cases, and recruitment of new collateral circulation in two cases.
Risk factors influencing rebleeding in two groups
Multiple regression analysis of the 96 hemoptysis patients with bronchiectasis revealed that no independent factor was statistically associated with recurrence. However, in the chronic PTB group, lung destruction detected on the chest CT and systemic artery-pulmonary circulation shunts were identified as independent risk factors associated with rebleeding (Table 3). In consideration of the high recurrence rate in patients (19 of 74) with chronic PTB during the first month following procedure, we examined factors influencing early rebleeding in this group. For this condition, lung destruction was found to be an independent risk factor for early recurrence. CTA examination before the procedure was proved to be valuable in reducing early rebleeding (Table 3).
Factors influencing total and early recurrence in chronic pulmonary tuberculosis (PTB) patients after the procedure.
*Statistically significant with P < 0.05.
HR, hazard ratio; CI, confidence interval; CTA, computed tomography angiography.
Among the 31 patients who underwent CTA in the chronic PTB group, all 57 bronchial arteries, which were considered pathologically changed at angiography, were detected at CTA. A high majority (21 out of 27) of NBSAs considered pathologically changed at angiography were also identified as abnormal during CTA examination.
Discussion
In this retrospective study, we evaluated the success rate of arterial embolization for hemoptysis in patients with chronic PTB and in patients with bronchiectasis. Our results revealed that both the short- and long-term recurrence-free rates after arterial embolization were significantly lower in the chronic PTB group compared with those in the bronchiectasis group. These results are consistent with a previous retrospective study (12) comparing embolization in chronic PTB and in bronchiectasis in a cohort of only 49 patients.
The higher rebleeding rates in the chronic PTB group are attributed to the underlying pathology of this disease. Chronic PTB is commonly associated with continuous inflammation, proliferation, and extensive disease process, which infringes the bronchial wall and pulmonary interstitium, frequently leading to pleural adhesions (8,9,18). Angiography of patients with chronic PTB revealed a large number of arteries surrounding the area around the lesion, especially NBSAs (Fig. 1a–f) (5,19). With these characteristics, incomplete embolization during the initial procedure and evidence of recruitment of new collateral circulation during follow-up are likely to be present in chronic PTB patients. This was confirmed by repeated angiograms of patients with rebleeding in the present study. In contrast, bronchiectasis is distinguished from chronic PTB by repeated bronchial mucosal infections (20), and angiography of patients with bronchiectasis frequently reveals hypertrophy and tortuosity of the feeding artery to diseased bronchial mucosa (Fig. 1g and h), most of which are bronchial arteries.
In agreement with previous studies (2,21,22), our results also demonstrated that chest CTA was beneficial in identifying abnormal vessels. In this study, 100% of pathologic bronchial arteries and 77.8% of pathologic NBSAs were identified before the procedure in the chronic PTB patients who underwent CTA examination. However, instead of supporting reduction of rebleeding throughout the entire follow-up period, chest CTA was only found to do so in the early period after embolization. Specifically, it lowered the recurrence within the first month by 82.9% in patients with chronic PTB compared to patients who did not undergo chest CTA (Table 3). These results suggest that incomplete embolization, especially for NBSAs, is likely the main cause of early recurrence in hemoptysis patients with chronic PTB. Therefore, to improve the immediate control rate of hemoptysis and to reduce the probability of early recurrence in chronic PTB patients, thoracic CTA is recommended as a prerequisite for all PTB patients with indication for arterial embolization (23).
Multiple regression analyses in the present study identified that in the chronic PTB group, lung destruction visible on chest CT was an independent rebleeding risk factor within the first month and throughout follow-up after embolization. Extensive circulation from different sources supplies blood to the damaged lung segments, making complete devascularization challenging during the embolization procedure (15). Continuous inflammation in the damaged area also contributes to formation of new collateral circulation (9). The presence of systemic arterial-pulmonary circulation shunts detected in the angiogram was identified as another independent risk factor for recurrence in chronic PTB patients. Although similar results have been reported in several studies (6,7,16,24), no consensus has been reached on this matter. Furthermore, we did not find that the presence of shunts was associated with recurrence in the bronchiectasis group of patients, suggesting that additional studies examining larger patient populations are needed.
It is important to note that some patients in our study were only followed up for a few months, which may not be enough time to comprehensively evaluate long-term consequences of arterial embolization. Moreover, we did not examine the severity of the primary lung disease in all patients, which may likely influence recurrence rates (6,16). Only about 40% of our included patients underwent CTA examination during treatment and overall this number was relatively small. Though the results of the multivariate analysis showed that CTA examination before arterial embolization could reduce early recurrence in patients with chronic PTB, the risk of biased selection of patients for CTA in our study considerably reduced the value of this estimation.
Finally, the number of patients included in our analysis is relatively small, therefore further studies examining larger population cohorts are required to corroborate our results.
In conclusion, our results suggested that both the short-term and long-term recurrence-free rates of hemoptysis after arterial embolization were lower in patients with chronic PTB compared with patients with bronchiectasis. This result was found to be related to involvement of more NBSAs in blood supply. Additionally, CTA examination before the procedure was beneficial for finding abnormal vessels and appeared to support reduction of early recurrence in chronic PTB. Importantly, chronic PTB patients who present with lung destruction, shunts, or both, require special attention during follow-up after arterial embolization.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Jiangsu Province's Key Talents Program (QNRC2016559 to Qing-Quan Zu) and Construction Program of Jiangsu Province Clinical Research Center Support System (BL2014084).