Comparison of outcome of surgery for tubercular and nontubercular empyema: An analysis of 285 consecutive cases.

BACKGROUND: Few studies have compared the surgical outcomes between tubercular empyema (TE) and nontubercular empyema (NTE), which were limited by a small sample size. We conducted this study with the objective of comparing the surgical outcomes of patients with tuberculous and nontuberculous empyema.
MATERIALS AND METHODS: This is a retrospective analysis of 285 consecutively operated cases of TE and NTE over 5 years conducted in a tertiary care center in New Delhi, India. A comparative analysis of demography, intraoperative, and postoperative variables including mortality between the two groups was carried out.
RESULTS: Out of 285 patients, 166 were tubercular and 119 were nontubercular. Nontubercular group had significantly higher age (45.4 ± 17.2 vs. 31.2 ± 13.6 in years), more comorbidities. Procedure was started by thoracotomy in 25.9% of tubercular group and 41.1% of nontubercular group. In patients where procedure started by video-assisted thoracoscopic surgery (VATS), complete decortication could be achieved by VATS in 91.1% of TE patients, whereas it was possible in 77.2% of nontubercular group. Need for postoperative ventilation (10% vs. 1.2%, P = 0.0011) and intensive care unit (ICU) stay (25.2% vs. 3%, P = 0.001) was significantly higher in nontubercular group. Nontubercular group was found to have significantly higher number of complications (13.4% vs. 5.4%, P = 0.02) and postoperative mortality (10% vs. 0, P = 0.001).
CONCLUSIONS: Higher percentage of TE cases were managed by VATS with reduced operative time, less blood loss, and lower conversions. Need for postoperative ventilation, ICU stay, and complications including mortality were more in NTE.

INTRODUCTION

Empyema thoracis continues to be a major public health problem across the world. Parapneumonic effusion and trauma are its major causes in developed countries,[1] whereas tuberculosis (TB) continues to be a major cause in developing countries like India.[23] No single procedure is ideal in the management of all stages of this complex clinical entity. Tubercular empyema (TE) and nontubercular empyema (NTE) not only have different etiology and demographic features, but their treatment outcomes have also been variously reported. Very few studies have compared the surgical outcomes of TE and NTE and these were limited by a small sample size.[34] The present study aims at comparing the surgical outcomes between TE and NTE and analyze the factors predicting perioperative mortality.

MATERIALS AND METHODS

It is a retrospective analysis of a prospectively maintained database of 285 consecutively operated cases of Stage III pleural empyema from March 2012 to March 2017. Out of 285 patients, 166 (58.2%) were of tubercular (TE) origin and the rest 119 (41.8%) were nontubercular (NTE). Various causes of NTE are enlisted in Table 1. The demographic data, details of present illness, and treatment received including antitubercular treatment were recorded in detail. All patients underwent a detailed clinical examination along with routine preoperative investigations. Computed tomography of the chest was performed in all to assess the disease state. Indications for surgery were trapped lung, multiloculated collection, and incomplete drainage of empyema even after tube thoracostomy.

Table 1

Nontubercular empyema: Causes (n=119)

Cause	Number of cases
Postpneumonic	72
CKD-associated empyema	20
Posttraumatic	8
Recurrent empyema	6
Postpneumothorax	4
Postesophageal perforation	2
Post-PCNL	2
Rheumatoid arthritis effusion	2
Subphrenic abscess	1
Postpancreatitis	2

CKD: Chronic kidney disease, PCNL: Percutaneous nephrolithotomy

Patients were taken up for surgery after thorough preoperative evaluation and adequate physical and nutritional preparation. All patients were counseled by physiotherapist for pre- and postoperative physical training. The patients were taken up for video-assisted thoracoscopic surgery (VATS) or open approach as per the described criteria [Figure 1]. However, VATS was converted to thoracotomy if there was a need for lung resection with unclear hilar anatomy not considered safe to proceed by VATS, presence of ruptured cavitary lesion(s) in the lobe(s) with bronchopleural fistula, and excessive oozing/bleeding from lung surface or pleura altering patient's hemodynamics.

Figure 1

Institutional protocol of management of empyema thoracis

Intraoperatively, pleural fluid/pus as well as pleural tissue was sent for Gram's stain, routine aerobic bacterial culture, direct fluorescent staining for acid fast bacilli, mycobacterial culture, KOH staining for fungus, and fungal culture in all patients. Procedure was chosen according to the disease stage and patient's fitness. Procedures performed included VATS or open debridement, decortication with or without lung resection, and window thoracostomy.

Surgical details

All the procedures were performed under general anesthesia with single lung ventilation. Patients were placed in lateral decubitus position with the diseased side up. In the VATS group, three ports were used with one being a camera port and two working ports. Standard thoracoscopic instruments were used during the procedure. Open group had a classical posterolateral thoracotomy through either fifth or sixth intercostal space. Initial debridement and breaking of loculations was done followed by decortication of visceral pleura. A combination of sharp and blunt dissection techniques was adapted for visceral decortication. In areas with inseparable peel, the same was criss-crossed with electrocautery till the soft parenchyma was reached, thereby helping in expansion of the lung. Lung was mobilized completely till the hilum including division of inferior pulmonary ligament. Fissures were also freed whenever feasible. Parietal peel was removed completely starting from apex till the diaphragm and any diaphragmatic adhesions were released, thereby achieving complete mobilization of diaphragm. Lung expansion was checked and any sites of major air leak assessed and solved. Hemostasis was ensured with a combination of pressure packing, electrocautery, and argon plasma coagulation. Two chest drains were placed and connected to Thopaz™ digital negative suction device (Medela, Switzerland) with 20 cm of H2O negative pressure. Patients were extubated on the table whenever possible, monitored in the recovery room overnight, and shifted to the floor the next morning. Patients who required ventilation were kept in intensive care unit (ICU) till they were extubated and then shifted to high dependency unit for further observation and transferred to the ward once stabilized.

Postoperative care and follow-up

Supervised, vigorous chest physiotherapy along with adequate nutritional support was continued to maintain good lung expansion. Antibiotics were modified if intraoperative cultures showed sensitivity different from the drugs already being given. If patients were not on Anti-tuberculosis therapy (ATT) at the time of surgery, the same was started postoperatively if there was evidence of TB on pleural fluid, pus, or tissue studies. The chest drains were removed when there was no air leak, the drainage was not purulent/hemorrhagic and was <100 ml in 24 h. Patients were discharged from the hospital either after drain removal or with drains if they had prolonged drainage or air leak. Duration of postoperative air leak and chest tube, hospital stay, wound infection, recurrence of disease, and mortality during hospital stay were monitored and recorded. After discharge, patients were monitored for status of lung expansion and any other complication. Follow-up was done in the outpatient clinic, first 1 week after discharge and every month thereafter for 6 months. All patients had a chest X-ray at 6 months to assess the status of lung expansion and for any recurrence of collection.

Statistical analysis

Statistical testing was conducted with the statistical package for the social science system version SPSS 22.0 (IBM SPSS Statistics for Windows, Armonk, NY: IBM Corp.). Continuous variables were presented as mean ± standard deviation or median (interquartile range). Categorical variables were expressed as frequencies and percentages. The comparison of normally distributed continuous variables between the groups was performed using Student's t-test. Nominal categorical data between the groups were compared using Chi-squared test or Fisher's exact test as appropriate. Nonnormal distribution continuous variables were compared using Mann–Whitney U-test. For all statistical tests, P < 0.05 was considered statistically significant.

RESULTS

Demography and preoperative variables

The study group consisted of 285 patients who were operated on for empyema thoracis during the period, i.e. March 2012 to March 2017. There were 166 patients in TE group and 119 in the NTE group. NTE patients had significantly higher age, and more comorbidities compared to TE group. However, patients in TE group had significantly lower body mass index, longer duration of symptoms, and also had undergone significantly more preoperative interventions, i.e., pleural tapping andIntercostal Drain (ICD) insertion compared to NTE group. Fever was the most frequent complaint in TE, whereas dyspnea on exertion was the most common presenting symptom in NTE. Detailed demographics are presented in Table 2.

Table 2

Demographic details tubercular versus nontubercular empyema

Characteristics	TE, n (%)	NTE, n (%)	P
Number of patients	166	119	-
Male:female	123:43 (74.1: 25.9)	91:28 (76.5: 23.5)	0.68
Age (years)	31.2±13.6	45.4±17.2	<0.001
Body mass index (kg/m2)	21.4±2.8	24.2±3.1	<0.001
Number of comorbidities
No comorbidity	133 (80.1)	57 (48)	<0.001
1 comorbidity	23 (13.9)	31 (26)	0.01
≥2 comorbidity	10 (6)	31 (26)	<0.001
Comorbidities
Hypertension	14 (8.4)	52 (43)
Diabetes mellitus	11 (6.6)	43 (36)
Coronary artery disease	9 (5.4)	20 (16.8)
CKD	10 (6)	20 (16.8)
COPD	8 (4.8)	9 (7.5)
Rheumatoid arthritis	1 (0.6)	0
Symptoms
Fever	102 (61.4)	102 (85.7)
Cough	63 (37.9)	56 (47)
Chest pain	51 (30.7)	90 (75.6)
Dyspnea on exertion	30 (18)	109 (91.5)
Mean duration of symptoms (months)	5.2±4.8	2.1±1.9	<0.001
Side of disease
Right	89 (53. 6)	62 (52.1)	-
Left	75 (45.2)	52 (43.6)
Bilateral	2 (1.2)	5 (4.3)
Preoperative interventions
Needle aspiration	149 (89.7)	92 (77.3)	0.005
Multiple (≥2) aspirations	38 (22.9)	36 (30.2)	0.17
ICD placement	85 (51.2)	34 (28.5)	0.005

TE: Tubercular empyema, NTE: Nontubercular empyema, COPD: Chronic obstructive pulmonary disease, CKD: Chronic kidney disease, ICD: Intercostal drain

Intraoperative variables

Procedure was started by thoracotomy in 25.9% of TE group and 41.1% of NTE group. In the rest of patients where procedure was started by VATS, complete decortication could be achieved by this minimally invasive approach in 91.1% of TE patients, whereas it was 77.2% in NTE group. This difference was statistically significant. The mean operative time along with mean intraoperative blood loss was also significantly higher in NTE group [Table 3].

Table 3

Comparison of intraoperative variables between tubercular empyema and nontubercular empyema group

Characteristics	TE (n=166), n (%)	NTE (n=119), n (%)	P
Method of surgery
Started by VATS	123 (74.1)	70 (58.8)	0.007
Successful	112/123 (91.1)	54/70 (77.2)	0.01
Converted to thoracotomy	11/123 (8.9)	16/70 (22.8)	<0.001
Started by thoracotomy	43 (25.9)	49 (41.2)	0.39
Surgery
Decortication only	157 (94.5)	109 (91.5)
Decortication wedge or lobectomy	6 (3.6)	7 (5.8)
Window thoracostomy	3 (1.8)	3 (2.5)
Operative time (min)	218±38.4	248.4±62.8	<0.001
Intra operative blood loss (ml)	288.7±37.3	301.7±42.3	0.006

TE: Tubercular empyema, NTE: Nontubercular empyema, VATS: Videoassisted thoracoscopic surgery

Postoperative variables

Need for postoperative ventilation andICU stay was observed to be significantly higher in NTE group. Nevertheless, there was no difference in the incidence of prolonged postoperative air leak and mean intercostal ICD duration between the two groups. However, the duration of hospital stay along with mean time to return to full work was significantly shorter in TE group. This may due to the younger age group of TE patients with lesser comorbidities [Table 4].

Table 4

Comparison of postoperative variables between tubercular empyema and nontubercular empyema group

	TE (n=166), n (%)	NTE (n=119), n (%)	P
Need for postoperative ventilation	2 (1.2)	12 (10.08)	0.001
Need for postoperative ICU stay	5 (3)	30 (25.2)	<0.001
Prolonged air leak (>7 days)	35 (21)	25 (21)	0.76
Postoperative ICD duration (days)	7.4±2.3	7.9±3.1	0.12
Hospital stay (days)	7.1±3.2	9.7±3.0	<0.001
Mean time to return to full work (days)	24.5±3.8	32.8±5.9	<0.001

TE: Tubercular empyema, NTE: Nontubercular empyema, ICU: Intensive care unit, ICD: Intercostal drain

Postsurgical complications and mortality

Nontubercular group was found to have a significantly higher number of complications compared to TE group. Detailed complications are discussed in Table 5. In total, 12 postoperative deaths were noted in the study population, all in the NTE group. Sepsis with septic shock was the major cause of mortality in 6 patients, respiratory failure in 3 patients, and acute renal failure, aspiration pneumonitis, and massive intraoperative bleeding in one patient each. On subgroup analysis, age >40 years, presence of comorbidities, and open surgery were the significant predictors of perioperative mortality [Table 6].

Table 5

Comparison of complications and mortality between tubercular empyema and nontubercular empyema group

	TE (n=166), n (%)	NTE (n=119), n (%)	P
Number of complications (overall)	9 (5.4)	16 (13.4)	0.02
Bleeding	0	2 (1.6)	0.17
Wound infection	4 (2.4)	9 (7.5)	0.04
Atelectasis and pneumonia	2 (1.2)	3 (2.5)	0.65
Cardiac arrhythmias	1 (0.6)	0	1.0
Renal complications	2 (1.2)	2 (1.6)	1.0
Recurrence	5 (3.01)	5 (4.2)	0.33
Perioperative mortality (<30 days)	0	12 (10)	<0.001

TE: Tubercular empyema, NTE: Nontubercular empyema

Table 6

Predictors of perioperative mortality

Variables	Characteristics	Mortality	P
Age (years)	≤40 (n=29)	0	0.03
	>40 (n=90)	12
Comorbidities	Present (n=62)	12	0.03
	Absent (n=57)	0
Duration of symptoms (months)	<2 (n=84)	6	0.18
	≥2 (n=35)	6
Method of surgery	VATS (n=70)	3	0.02
	Open (n=49)	9
Culture positivity	Yes (n=47)	7	0.21
	No (n=72)	5

VATS: Video-assisted thoracoscopic surgery

DISCUSSION

Empyema thoracis continues to be a menace worldwide, causing significant morbidity and mortality. In the western world, empyema secondary to pneumonia and trauma constitutes the major burden,[5] whereas TB is the leading cause in developing countries like India.[67] The treatment of this complex clinical condition includes pleural drainage and complete decortication to aid lung expansion, apart from antibiotics and supportive therapy. While traditionally it has been achieved by thoracotomy, VATS decortication is now being increasingly applied in all stages due to shorter hospital stay, less postoperative pain, and reduced morbidity and mortality.[8] There is a general belief that TE patients have poor surgical outcomes compared to NTE.[9] However, till date, there are no large series comparing the surgical outcomes among both the groups. In our study, tubercular versus nontubercular group constituted a ratio of approximately 3:2. As per the World Health Organization TB Statistics for 2016, the estimated incidence of TB burden in India is 2.79 million cases, making India the country with the highest burden of TB.[10] This figure explains the high proportion of tubercular cases in our series. A similar ratio of TE and NTE cases has been reported by other studies from India.[1112] In contrast, few Indian studies have also shown the predominance of NTE.[21314]

Young males were affected more in TE group which could be explained by the fact that TB affects a similar age group. A similar experience was reported in other studies also.[315] Lower body mass index observed in TE could be due to a prolonged clinical course as the process of inflammation in TB may go on for months without much clinical symptoms. Formation of thick peel practically isolates the tubercular bacilli from pleural space which leads to relatively long asymptomatic course in comparison to NTE.[1617] Similar findings, i.e. prolonged duration of symptoms with significant malnutrition were reflected in our study also. More than half of NTE group had comorbidities, out of which 50% had more than one comorbidity. Hypertension, diabetes mellitus, and chronic kidney disease were predominant comorbidities in descending order. Similar observations were reported by Tong et al.[18] in 420 patients of NTE. Patients in TE group had undergone frequent preoperative interventions, i.e. pleural aspiration and ICD placement. These observations represent the practice of physicians and pulmonologists, where patients were managed with repeat aspirations/ICD insertions without offering early surgical reference. Waller et al.[19] reported that the probability of success of VATS decreases in delayed referrals, thus indirectly altering postoperative outcomes.

A significantly higher proportion of patients in NTE group (41.2% vs. 25.9%) underwent open decortication per primum in view of comorbidities and altered preoperative coagulation profile. Cardillo et al.[20] also reported similar (40%) open decortication rates in their group. Efficacy of VATS was questioned in chronic Stage III empyema[21] in view of extremely narrowed intercostal spaces and thickened parietal peel which may make introduction and manipulation of the endoscopic instruments very difficult. By further studies, researchers proved the efficacy of VATS even in Stage III NTE[4192122] as well as TE.[923] In our series, more conversions were seen in NTE group than TE group (22.8% vs. 8.9%) which can be explained by higher coagulation abnormalities in old frail patients of NTE group. Few series also reported conversion rates as high as up to 40%–44% where the probability of conversion increases in Stage III empyema.[2425] We, in our own series, earlier reported a 10% conversion rate among one hundred chronic TE cases.[26]

No differences between these two groups were found in postoperative air leak and total ICD duration. Not surprisingly, the total duration of hospital stay along with time to return to normal activity was significantly shorter in TE group which could be explained by the mean age of patients and comorbidities. These observations were contrary to other studies. In a prospective comparative analysis of 75 cases of TE and NTE by Kundu et al.,[3] the authors reported that tubercular group required prolonged intercostal tube drainage and complicated surgery and had a poorer outcome. Another similar study by Malhotra et al.[2] also reported poorer outcomes with longer pleural drainage and longer duration for resolution of symptoms.

The complication rates were found comparable to previously published literature in both TE and NTE groups.[92223] On overall comparison, the postoperative complication rate was observed significantly more in NTE group compared to TE group (13.4% vs. 5.4%). This could be due to presence of significantly more comorbidities with relatively severe sepsis in NTE group. On analysis of individual complications, surgical site wound infection was found common in NTE group. This finding could be due to the 40% culture positivity in the pus.

We encountered 10% (n = 12) mortality all in NTE group. Various studies reported mortality ranging from 0% to 11.4% in NTE.[1927] In our study, severe sepsis with septic shock and postoperative respiratory failure were the predominant causes of mortality, whereas renal failure was the etiology in 12% of cases. Analysis of perioperative variables revealed that higher age (>40 years), presence of comorbidities, and open approach for surgery were associated with postoperative mortality. No mortality was present in TE group. Various other published series also demonstrated the mortality rates ranging from 0% to 3.1%.[2829] Late referral could be another factor to consider in the mortality group. A study by Cham et al.[30] showed that total hospital stay, postoperative morbidity, and mortality were higher in late referrals. In our own study of TE, we have reported that the probability of conversion and postoperative air leaks increases with an increase in the time before referral for surgical treatment.[28]

Retrospective nature is one of the major limitations of this study. The second most significant limitation is the heterogeneity of the cohort that is being studied. Third, the method/policy of surgical management was also heterogeneous, i.e. VATS/open and decortication with/without lung resection and window thoracostomy. Therefore, this analysis is prone to a number of biases, including selection, detection, and performance bias. However, it is the first and the largest study from a TB endemic country comparing the surgical outcomes in two major causes of empyema.

CONCLUSIONS

Our study reveals that a higher percentage of TE cases could be managed by VATS with reduced operative time, less blood loss, and lesser conversion rate as compared to NTE cases. Postoperative ventilation, ICU stay, and complications including mortality were more in NTE cases. Further prospective, multicenter, randomized studies with standardized treatment protocols are required to confirm these findings.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.