- Email: [email protected]
Thoracoscopy in pediatric pleural empyema: a prospective study of prognostic factors
Journal of Pediatric Surgery (2006) 41, 1732 – 1737
www.elsevier.com/locate/jpedsurg
Thoracoscopy in pediatric pleural empyema: a prospective study of prognostic factors Nicolas Kalfaa, Hossein Allala,*, Manuel Lopeza, Magali Saguintaahb, Marie-Pierre Guibala, Edith Sabatier-Lavala, Dominique Forguesa, Franc¸ois Counilc, Rene´-Benoit Galifera a
Visceral Pediatric Surgery Department, Lapeyronie-Arnaud de Villeneuve Hospital, Montpellier Cedex 5 34295, France Pediatric Radiology Department, Lapeyronie-Arnaud de Villeneuve Hospital, Montpellier Cedex 5 34295, France c Pediatric Pneumonology Department, Lapeyronie-Arnaud de Villeneuve Hospital, Montpellier Cedex 5 34295, France b
Index words: Empyema; Child; Thoracoscopy; Prognosis; Complications; Pneumopathy
Abstract Purpose: The indications for thoracoscopy remain imprecise in cases of pleural empyema. This study aimed to identify preoperative prognostic factors to help in the surgical decision. Methods: From 1996 to 2004, 50 children with parapneumonic pleural empyema underwent thoracoscopy either as the initial procedure (n = 26) or after failure of medical treatment (n = 24). Using multivariate analysis, we tested the prognostic value of clinical and bacteriological data, the ultrasonographic staging of empyema, and the delay before surgery. Outcome measures were technical difficulties, postoperative complications, time to apyrexia, duration of drainage, and length of hospitalization. Results: The clinical and bacterial data did not significantly predict the postoperative course. Echogenicity and the presence of pleural loculations at ultrasonography were not independent significant prognostic factors. A delay between diagnosis and surgery of more than 4 days was significantly correlated ( P b .05) with more frequent surgical difficulties, longer operative time, more postoperative fever, longer drainage time, longer hospitalization, and more postoperative complications, such as bronchopleural fistula, empyema relapse, and persistent atelectasia. Conclusion: The main prognostic factor for thoracoscopic treatment of pleural empyema is the interval between diagnosis and surgery. A 4-day limit, corresponding to the natural process of empyema organization, is significant. The assessment of loculations by ultrasonography alone is not sufficient to predict the postoperative course. D 2006 Elsevier Inc. All rights reserved.
Parapneumonic effusion and empyema lie on a continuum from a nonspecific pleural reaction that does not require surgery to a potentially severe complication of
* Corresponding author. Tel.: +33 4 67 338 783; fax: +33 4 67 339 512. E-mail address: [email protected] (H. Allal). 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2006.05.066
bacterial pneumonia by an extension of the infection into the pleural space [1,2]. The risks of pleural and parenchymatous morbidity should therefore not be neglected, especially in young children. Failure to control the pleural process may lead to the development of multiple loculations and organized empyema, major pleural thickening, and restrictive ventilatory disorders [3].
Thoracoscopy in pediatric pleural empyema No clear consensus has been reached concerning the optimal therapeutic strategy to adopt for advanced empyema. The options include intrapleural fibrinolytics [4,5], thoracoscopy [6], and thoracotomy [7], and children are treated differently according to their clinical picture. Thoracoscopy has recently shown the advantages of minimal morbidity, hastened recovery, high safety, and efficacy [8-10]. But the precise prognostic factors of empyema treated by this technique are still debatable. A better definition of them is thus necessary both to provide the basis for the right therapeutic choice and to predict the postoperative disease progression. The aim of this study was to identify the preoperative clinical, bacteriological, and radiologic prognostic factors of thoracoscopy in this increasingly frequent pathology and to compare patients’ outcomes according to the timing of the surgical treatment.
1733
1.3. Surgical methods
1. Methods
The patient underwent a selective intubation and was placed in the lateral decubitus position on the opposite side of the empyema. Two ports were used (telescope and 1 instrument). A third one was necessary for empyema with extensive pleural peel. Pneumothorax with carbon dioxide (6-10 mm Hg) allowed the lung to collapse. An initial debridement with the telescope was often useful before placing the first working port. After a systematic sampling of liquid, abundant irrigation and aspiration with a doublecurrent system were performed. Extensive debridement and ablation of all septa allowed the whole pleural cavity to be unified. This decortication was as complete as possible in each case, with particular attention paid to the visceral pleura to avoid air leakage. Although large ports would have been advantageous, a 5-mm working port was used because of the limited space between ribs, especially in small children. Two chest tubes (minimum of 14F) were positioned anteriorly and posteriorly. A chest radiograph at the end of the procedure confirmed the lung reexpansion.
1.1. Patients
1.4. Statistical methods
From September 1996 to December 2004, 50 children with pleural empyema were admitted to our department. The diagnosis was confirmed in all cases by bacteriological or biochemical study of liquid sample [11] (lactate dehydrogenase N1000 IU/L, glucose b40 mg/dL) and surgical findings (fibrin organization, purulent pleural fluid). Seven patients were enrolled retrospectively in this study, and 43, prospectively. We recorded the following data: age, sex, and clinical findings; chronology of initial signs and diagnosis; oxygen saturation (pulse oximetry) at the time of admission and need for oxygen; delay before operative drainage; bacteriological and biochemical study of liquid; and radiologic and preoperative findings. Empyema was classified according to the American Thoracic Society staging [12]; stage I is exudative pleuresia, stage II is fibrinopurulent, and stage III is organizing. The outcome measures were technical difficulties, postoperative complications, time to stable apyrexia, duration of drainage, and length of hospitalization. The period of follow-up ranged from 6 months to 3 years.
The statistical analysis was performed by the Medical Information and Biostatistics Department of Montpellier University Hospital. The univariate analysis was performed with the nonparametric Mann-Whitney U test for quantitative data, and Pearson v 2 or Fisher’s Exact test for nominative data. Normal distribution of outcome measures was tested by a Shapiro-Wilk test. Each factor was then individually assessed by a multivariate analysis according to an analysis of variance test. Confusion factors were excluded.
1.2. Ultrasonographic methods Ultrasonographic (US) examinations were performed with an ATL UM9 HDI apparatus (ATL Ultrasound, Bothell, WA). A 10-MHz linear transducer with a low-wall filter (100 kHz) and low pulse-repetition frequency (1-2 Hz) was used. The patient was in a sitting position. The echogenicity of the liquid, the maximal thickness of the effusion, and the presence of loculations were studied. Examinations were performed every 2 days in the preoperative period until surgery only by senior pediatric radiologists.
Fig. 1 Increased incidence of bacterial pneumonia in children from 1996 to 2004.
1734
N. Kalfa et al.
2. Results A marked increase in the incidence of empyema has been observed in our institution since 1996 (Fig. 1).
2.3. Prognostic value of preoperative pleural ultrasonography
2.1. Prognostic value of clinical data The mean age was 4 1/2 years (ranging from 5 months to 16 years), and 27 patients (54%) were girls. All patients had previously been healthy. The clinical findings were principally fever (86%) and tachypnea (74%). The range of oxygen saturation was from 81% to 99%. Fifty-six percent of the patients needed oxygen (average of 1 L/min, maximum of 5 L/min). The clinical data did not allow us to predict the postoperative outcome. In particular, the severity of the initial clinical presentation could not be used to forecast the risk of surgical difficulties and postoperative complications (Table 1).
2.2. Prognostic value of bacteriological data Twenty-seven patients (54%) had a culture-proven bacterial pathogen in the pleural effusion. The most frequent pathogen was Streptococcus pneumoniae (18 cases). Staphylococcus aureus was found in 2 patients. Other pathogens were Streptococcus alpha haemolyticus, Peptostreptococcus, and Bacteroides. Streptococcus pneumoniae was penicillin-resistant in 17% of the cases, with antibiotic resistance that remained stable over the study period. In cases of multiresistant germs, the duration of postoperative fever was increased ( P b .05). The duration of drainage, the duration of postoperative fever, and the risk of complicaTable 1
b3 y (n = 18) N3 y (n = 32) Sex Male (n = 23) Female (n = 27) Respiratory Yes (n = 20) distress No (n = 30) Bacteriological Penicillin-sensitive datab pneumococcus (n = 15) Penicillin-resistant pneumococcus (n = 3) Pleural Yes (n = 35) loculations No (n = 15) at US Pleural fluid Yes (n = 39) echogenicity No (n = 11) Thickness of b30 mm (n = 28) empyema z30 mm (n = 22) a b c
Ultrasonographic examinations were performed at least twice in all patients. The delay between the last examination and surgery ranged from 12 to 48 hours. The last US before surgery showed an echogenic pleural liquid in 78% of the cases, which was suspected of being purulent or showed the beginnings of fibrin deposition. The mean thickness of the empyema was 37 mm (ranging from 15 to 81 mm). Loculations were present in 70%. When compared with the surgical exploration, the accuracy for detecting loculations was 80%, and US failed to detect organization of empyema in 9 cases. The multivariate analysis revealed that the echogenicity, the presence of pleural loculations, and the thickness of effusion at US were not independent significant prognostic factors. The durations of surgery, drainage, and hospitalization were not longer when the empyema appeared to be extensive or organizing on ultrasonography (Table 1).
2.4. Prognostic value of the interval between diagnosis and surgery During the interval between diagnosis and surgery, the initial nonsurgical treatment was started; this included a course of antibiotics, respiratory kinesitherapy, and a chest
Prognostic value of clinical, bacteriological, and radiologic data after thoracoscopy for empyema in children Operative time (min)
Agea
tions were also increased when compared with outcomes of cases with penicillin-sensitive pneumococcus, but the differences did not reach significance.
Postoperative fever (d)
89 90 81 96 85 93 84
P = NS 4.4 4 P = NS 5.9 4 P = NS 4.5 3.8 P = NS 3.
97
8.3
P = NS P = NS P = NS P b .05
Duration of drainage (d) 7.5 5.9 4.8 7.8 7 6.1 5.4
P = NS P = NS P = NS P = NSc
16.3
Hospitalization (d)
Surgical difficulties
15.8 13.5 14.6 14.1 14.2 14.4 12.6
7% 11% 12% 13% 6% 13% 6%
P = NS P = NS P = NS P = NSc
23
0%
Complications
P = NS 20% 19% P = NS 11% 26% P = NS 12% 26% P = NS 13%
P = NS P = NS P = NS P = NS
60%
87 P = NS 3.4 P = NS 83 4.3
5.8 P = NS 6.8
13.2 P = NS 15.3
14% P = NS 20% 0% 6%
P = NS
87 P = NS 4 P = NS 78 3.6 87 P = NS 4.1 P = NS 84 3.8
5.8 P = NS 8.1 6.5 P = NS 6
14 P = NS 13.7 14.6 P = NS 13.2
12% P = NS 15% 0% 18% 14% P = NS 10% 5% 22%
P = NS
Three years was chosen as the limit, since postnatal alveolar development mainly takes place during this period [29,30]. In 25 cases, no bacteria were found in either pleural samples or blood cultures. Nonsignificant because of high variances.
P = NS
Thoracoscopy in pediatric pleural empyema
1735
Table 2 Effect of the interval between diagnosis and surgery on the postoperative outcome Outcome measures
Delay V4 d Delay N4 d P (n = 26 children) (n = 24 children) Mean F SD
Operative time (min) Length of postoperative fever (d) Duration of drainage (d) Length of hospitalization (d) Surgical difficulties Complications
74 F 20 2.7 F 4.2
Mean F SD 98 F 26 5.3 F 3.2
b.05 b.05
8 F 6.8
b.01
17 F 8.4
b.01
0%
21%
b.05
3.8%
29%
b.05
4.5 F 4 10.9 F 4.3
tube when necessary. The length of treatment was decided by the referring pediatricians of the hospital and was based on clinical evolution and complementary examinations. Thus, this interval varied widely and ranged from 1 to 32 days (median, 5 days). A delay of more than 4 days (24 children) was correlated with more frequent surgical difficulties ( P b .05) (Table 2). In 5 cases, the decortication and release of the lung was a delicate procedure, especially in the lower parts of the pleural cavity, and surgery was never shorter than 2 hours. One conversion to thoracotomy was necessary because of an advanced pleural symphysis. Early thoracoscopy significantly reduced the duration of the surgical procedures (74 vs 98 minutes, P b .05) and the duration of postoperative fever by half (2.7 vs 5.3 days, P b .05). The drainage time and length of hospitalization were also notably shorter when surgery was decided on within the first 4 days after diagnosis ( P b .01) (Table 2). Seventy percent of the patients with late surgery had to stay in the hospital for at least 2 weeks because of persistent septic signs, need of intravenous antibiotic therapy, significant pleural fluid production, or postoperative complications. Only 19% of the patients operated on during the first 4 days were hospitalized for so long. More postoperative complications occurred in cases of postponed surgery ( P b .05), such as bronchopleural fistula (3 cases, 2 of them with necrotizing pneumonia), empyema relapse (3 cases), and pneumatocele (1 case). All of them occurred in patients operated on after a relatively long delay, except for 1 case of persistent atelectasis. Additional thoracoscopic procedures were required 3 times, only in cases of late surgery.
2.5. Effect of the surgeons’ learning curve To exclude the effect of the surgeons’ learning curve, we looked for an influence of the operative year on the outcome measures. We found no correlations between the year of surgery, the length of the procedure, the presence of technical difficulties, and the postoperative course.
3. Discussion An increase in the incidence of empyema has been reported since 1995, but the etiology is still largely unknown [13]. Although various therapeutic choices for pediatric empyema have been reported to be successful (chest tube drainage [14], thoracotomy [15], thoracoscopy [16], fibrinolytics [4,5]), the aim of this work was to highlight the earliness of surgical treatment as the most favorable prognostic factor. To ensure the homogeneity of the study group, all patients underwent ultrasonography as the reference imaging examination and thoracoscopy as the method of surgical treatment. The postoperative course was not influenced by the year of the surgery, and thus, the effect of the surgeons’ learning curve could be excluded as a bias. This study was not a randomized comparative trial of different therapeutic options, in particular, of thoracotomy and thoracoscopy. But an understanding of the prognostic factors is, nevertheless, critical to choosing the best therapeutic attitude. Unlike previous reports [17,18], we show that the surgical decision should not be based on clinical and bacteriological data alone. The clinical presentation was not a determinant of the pathologic outcome. Except for major respiratory distress, which is a rare occurrence in pleural empyema and may require urgent drainage, the potential severity of the empyema was not reflected by the clinical status at admission in our series. Even with moderate or minimal infectious or respiratory symptoms, pleural empyema exposes the patient to devastating morbidity if neglected or operated on too late. Similarly, bacteriological data alone were not sufficient to determine the need for aggressive treatment. Even if the postoperative course tends to be shorter in cases of nonresistant bacteria, the risk of complications does exist. As shown by Hardie et al [11], penicillin-sensitive streptococci do not result in a more favorable issue than resistant bacteria. Although the natural history of infection can be influenced by the bacterial virulence [19], bacteriological data did not appear as an independent and significant prognostic factor in our study. The timing of surgical treatment was expected to be a major prognostic factor [20], based on our preliminary report [21]. This result is supported by the pathophysiology of this disease. Stage I lasts a maximum of 3 to 4 days [22], and beyond this limit, the natural organizing process greatly reduces the likelihood that medical treatment will succeed [23]. The transthoracic chest drainage tube, without surgical removal of restrictive purulent debris, is inadequate and ineffective in complicated parapneumonic empyema because of tube clogging and multiple loculations. The present study demonstrated that whether surgery is performed within 4 days of diagnosis or later seems to be the best predictor of postoperative outcome. Early surgery facilitates the unification of the whole pleural space, allows complete decortication even in the pleural cul-de-sac, and significantly reduces the postoperative febrile period and the length of
1736
N. Kalfa et al.
Fig. 2
Algorithm for the surgical decision based on both clinical and US data from the first 4 days after diagnosis.
hospitalization [6]. In contrast, when performed after 4 days, thoracoscopy encounters more frequent technical difficulties and seems to be less effective. The risk of parenchymal injury is higher, and more complications occurred in our series, such as bronchopleural fistula, empyema relapse, and persistent atelectasis. In addition, some authors have emphasized the economic advantages of an early procedure that results in a shorter hospital stay with fewer additional procedures and a shorter course of antibiotics [24]. Nevertheless, the definition of an early thoracoscopy has not yet been clearly established. A 2-day limit was proposed by Schultz et al [25]. But recommending systematic thoracoscopy 2 days post diagnosis might expose too many children able to recover quickly without surgery to invasive treatment. A 4-day limit seems more acceptable because it would not lead to systematic surgery and would provide enough time for initial medical treatment with pleural drainage to succeed. At this point, only nonresponding patients could benefit from highly effective thoracoscopy
(Fig. 2). However, we do not recommend prolonging nonsurgical treatment beyond this 4-day limit. Surprisingly, our study did not confirm US findings as a significant prognostic factor of postoperative outcome. Computed tomographic scans and pleural US have been proposed to assess the empyema stage and provide surgical indications. Computed tomography is not specific enough to diagnose and stage empyema [26], however, and was not used in this study. Ultrasonography has been shown to be more reliable for evaluating the nature of the pleural fluid [27] and was chosen as the reference examination for all of our patients. The US findings were expected to be a valuable prognostic factor in this disease since Ramnath et al [28] reported that an early sonographic evaluation of parapneumonic effusion was useful in assessing disease severity. But the sensitivity of US for detecting loculations when compared with operative confirmation was only 77% in the present cohort. This discordance may be explained by 2 factors: first, the interval between the last US and surgery
Thoracoscopy in pediatric pleural empyema (12-48 hours) may be sufficient for the disease to progress; second, ultrasonography is very sensitive for detecting quite thin loculations because of the excellent contrast between anechogenic fluid and echogenic membranes, but this sensitivity may decrease in cases of very echogenic fluid. A low-grade organized empyema at US does not guarantee the absence of purulent debris and the success of medical treatment alone. Ultrasonography may thus underevaluate the disease progression, and the surgical decision may be inappropriately delayed. For these reasons, the US findings were not able to predict the outcome of our patients, the length of hospitalization, or the risk of complications. More closely spaced and regular exams— every 24 hours, for example—may counterbalance these limits and provide useful data. Despite these shortcomings for prognosis, US remains a central diagnostic tool for pleural effusion in children. This study highlights the delay between diagnosis and surgery as the main prognostic factor of the postoperative outcome of empyema in children. The 4-day limit is supported by pathophysiological considerations and corresponds to the natural course of pleural organization. Yet the efficacy of nonsurgical treatment should not be minimized, and a brief preliminary trial of medical treatment with chest tube in nonorganized empyema could reduce excessive surgical indications.
References [1] Light RW. A new classification of parapneumonic effusions and empyema. Chest 1995;108:299 - 301. [2] Hilliard TN, Henderson AJ, Langton Hewer SC. Management of parapneumonic effusion and empyema. Arch Dis Child 2003;88: 915 - 7. [3] Kohn GL, Walston C, Feldstein J, et al. Persistent abnormal lung function after childhood empyema. Am J Respir Med 2002;1:441 - 5. [4] Thomson AH, Hull J, Kumar MR, et al. Randomised trial of intrapleural urokinase in the treatment of childhood empyema. Thorax 2002;57:343 - 7. [5] Yao CT, Wu JM, Liu CC, et al. Treatment of complicated parapneumonic pleural effusion with intrapleural streptokinase in children. Chest 2004;125:566 - 71. [6] Kercher KW, Attorri RJ, Hoover JD, et al. Thoracoscopic decortication as first-line therapy for pediatric parapneumonic empyema. A case series. Chest 2000;118:24 - 7. [7] Karaman I, Erdogan D, Karaman A, et al. Comparison of closed-tube thoracostomy and open thoracotomy procedures in the management of thoracic empyema in childhood. Eur J Pediatr Surg 2004;14:250 - 4. [8] Rescorla FJ, West KW, Gingalewski CA, et al. Efficacy of primary and secondary video-assisted thoracic surgery in children. J Pediatr Surg 2000;35:134 - 8. [9] Cohen G, Hjortdal V, Ricci M, et al. Primary thoracoscopic treatment of empyema in children. J Thorac Cardiovasc Surg 2003;125:79 - 83.
1737 [10] Liu HP, Hsieh MJ, Lu HI, et al. Thoracoscopic-assisted management of postpneumonic empyema in children refractory to medical response. Surg Endosc 2002;16:1612 - 4. [11] Hardie WD, Roberts NE, Reising SF, et al. Complicated parapneumonic effusions in children caused by penicillin-nonsusceptible Streptococcus pneumoniae. Pediatrics 1998;101:388 - 92. [12] Andrews NC, Parker EF, Shaww RR, et al. American Thoracic Society Subcommittee on surgery management of nontuberculous empyema. Am Rev Respi Dis 1962;85:935 - 6. [13] Eastham KM, Freeman R, Kearns AM, et al. Clinical features, aetiology and outcome of empyema in children in the north east of England. Thorax 2004;59:522 - 5. [14] Shoseyov D, Bibi H, Shatzberg G, et al. Short-term course and outcome of treatments of pleural empyema in pediatric patients: repeated ultrasound-guided needle thoracocentesis vs chest tube drainage. Chest 2002;121:836 - 40. [15] Alexiou C, Goyal A, Firmin RK, et al. Is open thoracotomy still a good treatment option for the management of empyema in children? Ann Thorac Surg 2003;76:1854 - 8. [16] Grewal H, Jackson RJ, Wagner CW, et al. Early video-assisted thoracic surgery in the management of empyema. Pediatrics 1999;103:e63. [17] Satish B, Bunker M, Seddon P. Management of thoracic empyema in childhood: does the pleural thickening matter? Arch Dis Child 2003; 88:918 - 21. [18] Carey JA, Hamilton JR, Spencer DA, et al. Empyema thoracis: a role for open thoracotomy and decortication. Arch Dis Child 1998;79: 510 - 3. [19] Margenthaler JA, Weber TR, Keller MS. Predictors of surgical outcome for complicated pneumonia in children: impact of bacterial virulence. World J Surg 2004;28:87 - 91. [20] Gates RL, Caniano DA, Hayes JR, et al. Does VATS provide optimal treatment of empyema in children? A systematic review. J Pediatr Surg 2004;39:381 - 6. [21] Kalfa N, Allal H, Montes-Tapia F, et al. Ideal timing of thoracoscopic decortication and drainage for empyema in children. Surg Endosc 2004;18:472 - 7. [22] Bremont F, Baunin C, Juchet A, et al. Clinical course and treatment of pleural empyema in children. Arch Pediatr 1996;3:335 - 41. [23] Rodriguez JA, Hill CB, Loe Jr WA, et al. Video-assisted thoracoscopic surgery for children with stage II empyema. Am Surg 2000;66: 569 - 72. [24] Meier AH, Smith B, Raghavan A, et al. Rational treatment of empyema in children. Arch Surg 2000;135:907 - 12. [25] Schultz KD, Fan LL, Pinsky J, et al. The changing face of pleural empyemas in children: epidemiology and management. Pediatrics 2004;113:1735 - 40. [26] Donnelly LF, Klosterman LA. CT appearance of parapneumonic effusions in children: findings are not specific for empyema. AJR Am J Roentgenol 1997;169:179 - 82. [27] Park CS, Chung WM, Lim MK, et al. Transcatheter instillation of urokinase into loculated pleural effusion: analysis of treatment effect. AJR Am J Roentgenol 1996;167:649 - 52. [28] Ramnath RR, Heller RM, Ben-Ami T, et al. Implications of early sonographic evaluation of parapneumonic effusions in children with pneumonia. Pediatrics 1998;101:68 - 71. [29] Thurlbeck WM. Postnatal human lung growth. Thorax 1982;37: 564 - 71. [30] Zeltner TB, Burri PH. The postnatal development and growth of the human lung. II. Morphology. Respir Physiol 1987;67:269 - 82.
www.elsevier.com/locate/jpedsurg
Thoracoscopy in pediatric pleural empyema: a prospective study of prognostic factors Nicolas Kalfaa, Hossein Allala,*, Manuel Lopeza, Magali Saguintaahb, Marie-Pierre Guibala, Edith Sabatier-Lavala, Dominique Forguesa, Franc¸ois Counilc, Rene´-Benoit Galifera a
Visceral Pediatric Surgery Department, Lapeyronie-Arnaud de Villeneuve Hospital, Montpellier Cedex 5 34295, France Pediatric Radiology Department, Lapeyronie-Arnaud de Villeneuve Hospital, Montpellier Cedex 5 34295, France c Pediatric Pneumonology Department, Lapeyronie-Arnaud de Villeneuve Hospital, Montpellier Cedex 5 34295, France b
Index words: Empyema; Child; Thoracoscopy; Prognosis; Complications; Pneumopathy
Abstract Purpose: The indications for thoracoscopy remain imprecise in cases of pleural empyema. This study aimed to identify preoperative prognostic factors to help in the surgical decision. Methods: From 1996 to 2004, 50 children with parapneumonic pleural empyema underwent thoracoscopy either as the initial procedure (n = 26) or after failure of medical treatment (n = 24). Using multivariate analysis, we tested the prognostic value of clinical and bacteriological data, the ultrasonographic staging of empyema, and the delay before surgery. Outcome measures were technical difficulties, postoperative complications, time to apyrexia, duration of drainage, and length of hospitalization. Results: The clinical and bacterial data did not significantly predict the postoperative course. Echogenicity and the presence of pleural loculations at ultrasonography were not independent significant prognostic factors. A delay between diagnosis and surgery of more than 4 days was significantly correlated ( P b .05) with more frequent surgical difficulties, longer operative time, more postoperative fever, longer drainage time, longer hospitalization, and more postoperative complications, such as bronchopleural fistula, empyema relapse, and persistent atelectasia. Conclusion: The main prognostic factor for thoracoscopic treatment of pleural empyema is the interval between diagnosis and surgery. A 4-day limit, corresponding to the natural process of empyema organization, is significant. The assessment of loculations by ultrasonography alone is not sufficient to predict the postoperative course. D 2006 Elsevier Inc. All rights reserved.
Parapneumonic effusion and empyema lie on a continuum from a nonspecific pleural reaction that does not require surgery to a potentially severe complication of
* Corresponding author. Tel.: +33 4 67 338 783; fax: +33 4 67 339 512. E-mail address: [email protected] (H. Allal). 0022-3468/$ – see front matter D 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.jpedsurg.2006.05.066
bacterial pneumonia by an extension of the infection into the pleural space [1,2]. The risks of pleural and parenchymatous morbidity should therefore not be neglected, especially in young children. Failure to control the pleural process may lead to the development of multiple loculations and organized empyema, major pleural thickening, and restrictive ventilatory disorders [3].
Thoracoscopy in pediatric pleural empyema No clear consensus has been reached concerning the optimal therapeutic strategy to adopt for advanced empyema. The options include intrapleural fibrinolytics [4,5], thoracoscopy [6], and thoracotomy [7], and children are treated differently according to their clinical picture. Thoracoscopy has recently shown the advantages of minimal morbidity, hastened recovery, high safety, and efficacy [8-10]. But the precise prognostic factors of empyema treated by this technique are still debatable. A better definition of them is thus necessary both to provide the basis for the right therapeutic choice and to predict the postoperative disease progression. The aim of this study was to identify the preoperative clinical, bacteriological, and radiologic prognostic factors of thoracoscopy in this increasingly frequent pathology and to compare patients’ outcomes according to the timing of the surgical treatment.
1733
1.3. Surgical methods
1. Methods
The patient underwent a selective intubation and was placed in the lateral decubitus position on the opposite side of the empyema. Two ports were used (telescope and 1 instrument). A third one was necessary for empyema with extensive pleural peel. Pneumothorax with carbon dioxide (6-10 mm Hg) allowed the lung to collapse. An initial debridement with the telescope was often useful before placing the first working port. After a systematic sampling of liquid, abundant irrigation and aspiration with a doublecurrent system were performed. Extensive debridement and ablation of all septa allowed the whole pleural cavity to be unified. This decortication was as complete as possible in each case, with particular attention paid to the visceral pleura to avoid air leakage. Although large ports would have been advantageous, a 5-mm working port was used because of the limited space between ribs, especially in small children. Two chest tubes (minimum of 14F) were positioned anteriorly and posteriorly. A chest radiograph at the end of the procedure confirmed the lung reexpansion.
1.1. Patients
1.4. Statistical methods
From September 1996 to December 2004, 50 children with pleural empyema were admitted to our department. The diagnosis was confirmed in all cases by bacteriological or biochemical study of liquid sample [11] (lactate dehydrogenase N1000 IU/L, glucose b40 mg/dL) and surgical findings (fibrin organization, purulent pleural fluid). Seven patients were enrolled retrospectively in this study, and 43, prospectively. We recorded the following data: age, sex, and clinical findings; chronology of initial signs and diagnosis; oxygen saturation (pulse oximetry) at the time of admission and need for oxygen; delay before operative drainage; bacteriological and biochemical study of liquid; and radiologic and preoperative findings. Empyema was classified according to the American Thoracic Society staging [12]; stage I is exudative pleuresia, stage II is fibrinopurulent, and stage III is organizing. The outcome measures were technical difficulties, postoperative complications, time to stable apyrexia, duration of drainage, and length of hospitalization. The period of follow-up ranged from 6 months to 3 years.
The statistical analysis was performed by the Medical Information and Biostatistics Department of Montpellier University Hospital. The univariate analysis was performed with the nonparametric Mann-Whitney U test for quantitative data, and Pearson v 2 or Fisher’s Exact test for nominative data. Normal distribution of outcome measures was tested by a Shapiro-Wilk test. Each factor was then individually assessed by a multivariate analysis according to an analysis of variance test. Confusion factors were excluded.
1.2. Ultrasonographic methods Ultrasonographic (US) examinations were performed with an ATL UM9 HDI apparatus (ATL Ultrasound, Bothell, WA). A 10-MHz linear transducer with a low-wall filter (100 kHz) and low pulse-repetition frequency (1-2 Hz) was used. The patient was in a sitting position. The echogenicity of the liquid, the maximal thickness of the effusion, and the presence of loculations were studied. Examinations were performed every 2 days in the preoperative period until surgery only by senior pediatric radiologists.
Fig. 1 Increased incidence of bacterial pneumonia in children from 1996 to 2004.
1734
N. Kalfa et al.
2. Results A marked increase in the incidence of empyema has been observed in our institution since 1996 (Fig. 1).
2.3. Prognostic value of preoperative pleural ultrasonography
2.1. Prognostic value of clinical data The mean age was 4 1/2 years (ranging from 5 months to 16 years), and 27 patients (54%) were girls. All patients had previously been healthy. The clinical findings were principally fever (86%) and tachypnea (74%). The range of oxygen saturation was from 81% to 99%. Fifty-six percent of the patients needed oxygen (average of 1 L/min, maximum of 5 L/min). The clinical data did not allow us to predict the postoperative outcome. In particular, the severity of the initial clinical presentation could not be used to forecast the risk of surgical difficulties and postoperative complications (Table 1).
2.2. Prognostic value of bacteriological data Twenty-seven patients (54%) had a culture-proven bacterial pathogen in the pleural effusion. The most frequent pathogen was Streptococcus pneumoniae (18 cases). Staphylococcus aureus was found in 2 patients. Other pathogens were Streptococcus alpha haemolyticus, Peptostreptococcus, and Bacteroides. Streptococcus pneumoniae was penicillin-resistant in 17% of the cases, with antibiotic resistance that remained stable over the study period. In cases of multiresistant germs, the duration of postoperative fever was increased ( P b .05). The duration of drainage, the duration of postoperative fever, and the risk of complicaTable 1
b3 y (n = 18) N3 y (n = 32) Sex Male (n = 23) Female (n = 27) Respiratory Yes (n = 20) distress No (n = 30) Bacteriological Penicillin-sensitive datab pneumococcus (n = 15) Penicillin-resistant pneumococcus (n = 3) Pleural Yes (n = 35) loculations No (n = 15) at US Pleural fluid Yes (n = 39) echogenicity No (n = 11) Thickness of b30 mm (n = 28) empyema z30 mm (n = 22) a b c
Ultrasonographic examinations were performed at least twice in all patients. The delay between the last examination and surgery ranged from 12 to 48 hours. The last US before surgery showed an echogenic pleural liquid in 78% of the cases, which was suspected of being purulent or showed the beginnings of fibrin deposition. The mean thickness of the empyema was 37 mm (ranging from 15 to 81 mm). Loculations were present in 70%. When compared with the surgical exploration, the accuracy for detecting loculations was 80%, and US failed to detect organization of empyema in 9 cases. The multivariate analysis revealed that the echogenicity, the presence of pleural loculations, and the thickness of effusion at US were not independent significant prognostic factors. The durations of surgery, drainage, and hospitalization were not longer when the empyema appeared to be extensive or organizing on ultrasonography (Table 1).
2.4. Prognostic value of the interval between diagnosis and surgery During the interval between diagnosis and surgery, the initial nonsurgical treatment was started; this included a course of antibiotics, respiratory kinesitherapy, and a chest
Prognostic value of clinical, bacteriological, and radiologic data after thoracoscopy for empyema in children Operative time (min)
Agea
tions were also increased when compared with outcomes of cases with penicillin-sensitive pneumococcus, but the differences did not reach significance.
Postoperative fever (d)
89 90 81 96 85 93 84
P = NS 4.4 4 P = NS 5.9 4 P = NS 4.5 3.8 P = NS 3.
97
8.3
P = NS P = NS P = NS P b .05
Duration of drainage (d) 7.5 5.9 4.8 7.8 7 6.1 5.4
P = NS P = NS P = NS P = NSc
16.3
Hospitalization (d)
Surgical difficulties
15.8 13.5 14.6 14.1 14.2 14.4 12.6
7% 11% 12% 13% 6% 13% 6%
P = NS P = NS P = NS P = NSc
23
0%
Complications
P = NS 20% 19% P = NS 11% 26% P = NS 12% 26% P = NS 13%
P = NS P = NS P = NS P = NS
60%
87 P = NS 3.4 P = NS 83 4.3
5.8 P = NS 6.8
13.2 P = NS 15.3
14% P = NS 20% 0% 6%
P = NS
87 P = NS 4 P = NS 78 3.6 87 P = NS 4.1 P = NS 84 3.8
5.8 P = NS 8.1 6.5 P = NS 6
14 P = NS 13.7 14.6 P = NS 13.2
12% P = NS 15% 0% 18% 14% P = NS 10% 5% 22%
P = NS
Three years was chosen as the limit, since postnatal alveolar development mainly takes place during this period [29,30]. In 25 cases, no bacteria were found in either pleural samples or blood cultures. Nonsignificant because of high variances.
P = NS
Thoracoscopy in pediatric pleural empyema
1735
Table 2 Effect of the interval between diagnosis and surgery on the postoperative outcome Outcome measures
Delay V4 d Delay N4 d P (n = 26 children) (n = 24 children) Mean F SD
Operative time (min) Length of postoperative fever (d) Duration of drainage (d) Length of hospitalization (d) Surgical difficulties Complications
74 F 20 2.7 F 4.2
Mean F SD 98 F 26 5.3 F 3.2
b.05 b.05
8 F 6.8
b.01
17 F 8.4
b.01
0%
21%
b.05
3.8%
29%
b.05
4.5 F 4 10.9 F 4.3
tube when necessary. The length of treatment was decided by the referring pediatricians of the hospital and was based on clinical evolution and complementary examinations. Thus, this interval varied widely and ranged from 1 to 32 days (median, 5 days). A delay of more than 4 days (24 children) was correlated with more frequent surgical difficulties ( P b .05) (Table 2). In 5 cases, the decortication and release of the lung was a delicate procedure, especially in the lower parts of the pleural cavity, and surgery was never shorter than 2 hours. One conversion to thoracotomy was necessary because of an advanced pleural symphysis. Early thoracoscopy significantly reduced the duration of the surgical procedures (74 vs 98 minutes, P b .05) and the duration of postoperative fever by half (2.7 vs 5.3 days, P b .05). The drainage time and length of hospitalization were also notably shorter when surgery was decided on within the first 4 days after diagnosis ( P b .01) (Table 2). Seventy percent of the patients with late surgery had to stay in the hospital for at least 2 weeks because of persistent septic signs, need of intravenous antibiotic therapy, significant pleural fluid production, or postoperative complications. Only 19% of the patients operated on during the first 4 days were hospitalized for so long. More postoperative complications occurred in cases of postponed surgery ( P b .05), such as bronchopleural fistula (3 cases, 2 of them with necrotizing pneumonia), empyema relapse (3 cases), and pneumatocele (1 case). All of them occurred in patients operated on after a relatively long delay, except for 1 case of persistent atelectasis. Additional thoracoscopic procedures were required 3 times, only in cases of late surgery.
2.5. Effect of the surgeons’ learning curve To exclude the effect of the surgeons’ learning curve, we looked for an influence of the operative year on the outcome measures. We found no correlations between the year of surgery, the length of the procedure, the presence of technical difficulties, and the postoperative course.
3. Discussion An increase in the incidence of empyema has been reported since 1995, but the etiology is still largely unknown [13]. Although various therapeutic choices for pediatric empyema have been reported to be successful (chest tube drainage [14], thoracotomy [15], thoracoscopy [16], fibrinolytics [4,5]), the aim of this work was to highlight the earliness of surgical treatment as the most favorable prognostic factor. To ensure the homogeneity of the study group, all patients underwent ultrasonography as the reference imaging examination and thoracoscopy as the method of surgical treatment. The postoperative course was not influenced by the year of the surgery, and thus, the effect of the surgeons’ learning curve could be excluded as a bias. This study was not a randomized comparative trial of different therapeutic options, in particular, of thoracotomy and thoracoscopy. But an understanding of the prognostic factors is, nevertheless, critical to choosing the best therapeutic attitude. Unlike previous reports [17,18], we show that the surgical decision should not be based on clinical and bacteriological data alone. The clinical presentation was not a determinant of the pathologic outcome. Except for major respiratory distress, which is a rare occurrence in pleural empyema and may require urgent drainage, the potential severity of the empyema was not reflected by the clinical status at admission in our series. Even with moderate or minimal infectious or respiratory symptoms, pleural empyema exposes the patient to devastating morbidity if neglected or operated on too late. Similarly, bacteriological data alone were not sufficient to determine the need for aggressive treatment. Even if the postoperative course tends to be shorter in cases of nonresistant bacteria, the risk of complications does exist. As shown by Hardie et al [11], penicillin-sensitive streptococci do not result in a more favorable issue than resistant bacteria. Although the natural history of infection can be influenced by the bacterial virulence [19], bacteriological data did not appear as an independent and significant prognostic factor in our study. The timing of surgical treatment was expected to be a major prognostic factor [20], based on our preliminary report [21]. This result is supported by the pathophysiology of this disease. Stage I lasts a maximum of 3 to 4 days [22], and beyond this limit, the natural organizing process greatly reduces the likelihood that medical treatment will succeed [23]. The transthoracic chest drainage tube, without surgical removal of restrictive purulent debris, is inadequate and ineffective in complicated parapneumonic empyema because of tube clogging and multiple loculations. The present study demonstrated that whether surgery is performed within 4 days of diagnosis or later seems to be the best predictor of postoperative outcome. Early surgery facilitates the unification of the whole pleural space, allows complete decortication even in the pleural cul-de-sac, and significantly reduces the postoperative febrile period and the length of
1736
N. Kalfa et al.
Fig. 2
Algorithm for the surgical decision based on both clinical and US data from the first 4 days after diagnosis.
hospitalization [6]. In contrast, when performed after 4 days, thoracoscopy encounters more frequent technical difficulties and seems to be less effective. The risk of parenchymal injury is higher, and more complications occurred in our series, such as bronchopleural fistula, empyema relapse, and persistent atelectasis. In addition, some authors have emphasized the economic advantages of an early procedure that results in a shorter hospital stay with fewer additional procedures and a shorter course of antibiotics [24]. Nevertheless, the definition of an early thoracoscopy has not yet been clearly established. A 2-day limit was proposed by Schultz et al [25]. But recommending systematic thoracoscopy 2 days post diagnosis might expose too many children able to recover quickly without surgery to invasive treatment. A 4-day limit seems more acceptable because it would not lead to systematic surgery and would provide enough time for initial medical treatment with pleural drainage to succeed. At this point, only nonresponding patients could benefit from highly effective thoracoscopy
(Fig. 2). However, we do not recommend prolonging nonsurgical treatment beyond this 4-day limit. Surprisingly, our study did not confirm US findings as a significant prognostic factor of postoperative outcome. Computed tomographic scans and pleural US have been proposed to assess the empyema stage and provide surgical indications. Computed tomography is not specific enough to diagnose and stage empyema [26], however, and was not used in this study. Ultrasonography has been shown to be more reliable for evaluating the nature of the pleural fluid [27] and was chosen as the reference examination for all of our patients. The US findings were expected to be a valuable prognostic factor in this disease since Ramnath et al [28] reported that an early sonographic evaluation of parapneumonic effusion was useful in assessing disease severity. But the sensitivity of US for detecting loculations when compared with operative confirmation was only 77% in the present cohort. This discordance may be explained by 2 factors: first, the interval between the last US and surgery
Thoracoscopy in pediatric pleural empyema (12-48 hours) may be sufficient for the disease to progress; second, ultrasonography is very sensitive for detecting quite thin loculations because of the excellent contrast between anechogenic fluid and echogenic membranes, but this sensitivity may decrease in cases of very echogenic fluid. A low-grade organized empyema at US does not guarantee the absence of purulent debris and the success of medical treatment alone. Ultrasonography may thus underevaluate the disease progression, and the surgical decision may be inappropriately delayed. For these reasons, the US findings were not able to predict the outcome of our patients, the length of hospitalization, or the risk of complications. More closely spaced and regular exams— every 24 hours, for example—may counterbalance these limits and provide useful data. Despite these shortcomings for prognosis, US remains a central diagnostic tool for pleural effusion in children. This study highlights the delay between diagnosis and surgery as the main prognostic factor of the postoperative outcome of empyema in children. The 4-day limit is supported by pathophysiological considerations and corresponds to the natural course of pleural organization. Yet the efficacy of nonsurgical treatment should not be minimized, and a brief preliminary trial of medical treatment with chest tube in nonorganized empyema could reduce excessive surgical indications.
References [1] Light RW. A new classification of parapneumonic effusions and empyema. Chest 1995;108:299 - 301. [2] Hilliard TN, Henderson AJ, Langton Hewer SC. Management of parapneumonic effusion and empyema. Arch Dis Child 2003;88: 915 - 7. [3] Kohn GL, Walston C, Feldstein J, et al. Persistent abnormal lung function after childhood empyema. Am J Respir Med 2002;1:441 - 5. [4] Thomson AH, Hull J, Kumar MR, et al. Randomised trial of intrapleural urokinase in the treatment of childhood empyema. Thorax 2002;57:343 - 7. [5] Yao CT, Wu JM, Liu CC, et al. Treatment of complicated parapneumonic pleural effusion with intrapleural streptokinase in children. Chest 2004;125:566 - 71. [6] Kercher KW, Attorri RJ, Hoover JD, et al. Thoracoscopic decortication as first-line therapy for pediatric parapneumonic empyema. A case series. Chest 2000;118:24 - 7. [7] Karaman I, Erdogan D, Karaman A, et al. Comparison of closed-tube thoracostomy and open thoracotomy procedures in the management of thoracic empyema in childhood. Eur J Pediatr Surg 2004;14:250 - 4. [8] Rescorla FJ, West KW, Gingalewski CA, et al. Efficacy of primary and secondary video-assisted thoracic surgery in children. J Pediatr Surg 2000;35:134 - 8. [9] Cohen G, Hjortdal V, Ricci M, et al. Primary thoracoscopic treatment of empyema in children. J Thorac Cardiovasc Surg 2003;125:79 - 83.
1737 [10] Liu HP, Hsieh MJ, Lu HI, et al. Thoracoscopic-assisted management of postpneumonic empyema in children refractory to medical response. Surg Endosc 2002;16:1612 - 4. [11] Hardie WD, Roberts NE, Reising SF, et al. Complicated parapneumonic effusions in children caused by penicillin-nonsusceptible Streptococcus pneumoniae. Pediatrics 1998;101:388 - 92. [12] Andrews NC, Parker EF, Shaww RR, et al. American Thoracic Society Subcommittee on surgery management of nontuberculous empyema. Am Rev Respi Dis 1962;85:935 - 6. [13] Eastham KM, Freeman R, Kearns AM, et al. Clinical features, aetiology and outcome of empyema in children in the north east of England. Thorax 2004;59:522 - 5. [14] Shoseyov D, Bibi H, Shatzberg G, et al. Short-term course and outcome of treatments of pleural empyema in pediatric patients: repeated ultrasound-guided needle thoracocentesis vs chest tube drainage. Chest 2002;121:836 - 40. [15] Alexiou C, Goyal A, Firmin RK, et al. Is open thoracotomy still a good treatment option for the management of empyema in children? Ann Thorac Surg 2003;76:1854 - 8. [16] Grewal H, Jackson RJ, Wagner CW, et al. Early video-assisted thoracic surgery in the management of empyema. Pediatrics 1999;103:e63. [17] Satish B, Bunker M, Seddon P. Management of thoracic empyema in childhood: does the pleural thickening matter? Arch Dis Child 2003; 88:918 - 21. [18] Carey JA, Hamilton JR, Spencer DA, et al. Empyema thoracis: a role for open thoracotomy and decortication. Arch Dis Child 1998;79: 510 - 3. [19] Margenthaler JA, Weber TR, Keller MS. Predictors of surgical outcome for complicated pneumonia in children: impact of bacterial virulence. World J Surg 2004;28:87 - 91. [20] Gates RL, Caniano DA, Hayes JR, et al. Does VATS provide optimal treatment of empyema in children? A systematic review. J Pediatr Surg 2004;39:381 - 6. [21] Kalfa N, Allal H, Montes-Tapia F, et al. Ideal timing of thoracoscopic decortication and drainage for empyema in children. Surg Endosc 2004;18:472 - 7. [22] Bremont F, Baunin C, Juchet A, et al. Clinical course and treatment of pleural empyema in children. Arch Pediatr 1996;3:335 - 41. [23] Rodriguez JA, Hill CB, Loe Jr WA, et al. Video-assisted thoracoscopic surgery for children with stage II empyema. Am Surg 2000;66: 569 - 72. [24] Meier AH, Smith B, Raghavan A, et al. Rational treatment of empyema in children. Arch Surg 2000;135:907 - 12. [25] Schultz KD, Fan LL, Pinsky J, et al. The changing face of pleural empyemas in children: epidemiology and management. Pediatrics 2004;113:1735 - 40. [26] Donnelly LF, Klosterman LA. CT appearance of parapneumonic effusions in children: findings are not specific for empyema. AJR Am J Roentgenol 1997;169:179 - 82. [27] Park CS, Chung WM, Lim MK, et al. Transcatheter instillation of urokinase into loculated pleural effusion: analysis of treatment effect. AJR Am J Roentgenol 1996;167:649 - 52. [28] Ramnath RR, Heller RM, Ben-Ami T, et al. Implications of early sonographic evaluation of parapneumonic effusions in children with pneumonia. Pediatrics 1998;101:68 - 71. [29] Thurlbeck WM. Postnatal human lung growth. Thorax 1982;37: 564 - 71. [30] Zeltner TB, Burri PH. The postnatal development and growth of the human lung. II. Morphology. Respir Physiol 1987;67:269 - 82.