Treatment of Loculated Pleural Effusions with Transcatheter Intracavitary Urokinase

Transcription

1 941 Treatment of Loculated Pleural Effusions with Transcatheter Intracavitary Urokinase Jeffrey S. Moulton1 Surgical thoracostomy tube placement and radiologically guided catheter drainage Patrick Timothy Moore are standard therapy for loculated pleural fluid collections. Treatment may fail if the Raymond A. Mencini catheter is not placed optimally within the loculation or if the fluid is hemorrhagic or fibrinous. We studied the value of transcatheter urokinase instillation in facilitating Received May : accepted after revision June 22, drainage of hemorrhagic or fibrinous nonhemorrhagic loculated pleural collections in 11 patients with 13 loculated pleural collections. Eight of the fluid collections were hemorrhagic, five were nonhemorrhagic. Five patients had had a thoracostomy tube placed surgically and all had had radiologically guided placement of single lumen drainage catheters managed with suction, saline irrigation, and mechanical guidewire manipulation. This therapy had failed to drain the loculations completely over an average of 10 days (range, 1-22 days). Urokinase (1000 units/mi) was instilled into the drainage catheters in 80- to 150-mI aliquots. After 1-2 hr, suction was reinstituted and the procedure was repeated. Twelve (92%) of the 13 collections were drained completely after an average of 4.3 instillations (range, three to eight instillations). Successful urokinase therapy required an average of 28 hr (range, 8-75 hr). In one case, therapy was discontinued after partial resolution for unrelated clinical reasons. There were no complications. These results suggest that transcatheter intracavitary urokinase therapy is a safe and effective method to facilitate drainage of loculated hemorrhagic or fibrinous nonhemorrhagic pleural fluid collections. AJR 153: , November All authors: Department of Radiology, St. An- Subjects and Methods thony Hospital, 4231 W. 16th Ave., Denver, Co Loculated pleural fluid collections may be treated by thoracentesis, closed thoracostomy tube drainage, rib resection and open drainage, or thoracotomy and decortication. Recent reports have advocated the use of image-guided placement of 1 0- to 1 4-French single lumen drainage catheters as the initial therapy [1-4]. Success rates have ranged from 72% to 92%, with failures usually due to the viscous or clotted nature of the fluid. Such drainage procedures tend to be prolonged, often requiring 1 week or more of drainage [1, 2]. In the early 1 950s, several reports described the use of streptokinase/streptodornase to facilitate treatment of loculated hemothorax and empyema [5, 6]. Enthusiasm for this technique waned, with only sporadic reports in the Swedish literature in the late 1 970s [7, 8]. In 1 987, Vogelzang et al. [9] reported the first use of urokinase in the treatment of infected extravascular hematomas. We evaluated the use of urokinase to facilitate transcatheter drainage in 1 1 patients with 13 loculated pleural collections Address reprint requests to J. S. Moulton. From July 1987 to April 1989, 13 pleural fluid collections in 11 patients were treated with x/89/ transcatheter intracavitary urokinase therapy. There were nine men and two women, American Roentgen Ray Society years old (mean, 48 yr). Indications for initial drainage included infection or compromise of

2 942 MOULTON ET AL. AJA:153, November 1989 pulmonary function. Infected collections (1 1 cases) were defined by positive pleural fluid culture (seven cases), organisms seen on Gram stain (one case), or elevated pleural fluid neutrophil count with a fluid ph less than 7.0 (three cases). Sterile collections (two cases) were drained because of significant atelectasis and hypoxemia. There were five infected hemothoraces. Two occurred after surgical drainage of a subphrenic abscess, and two were iatrogenic results of laceration of intercostal vessels (one during chest tube placement for a spontaneous tension pneumothorax and one after percutaneous transpleural drainage of a liver abscess). The fifth was due to a self-inflicted gunshot wound (Fig. 1). There were three hemorrhagic parapneumonic empyemas (all unilateral), three nonhemorrhagic parapneumonic empyemas (one bilateral and one unilateral) A B C (Fig. 2), and one case of bilateral sterile loculated exudative effusions after nephrectomy for a perinephric abscess. Nine collections involved the right pleural space, four were on the left. None of the patients had associated malignancy. All infected patients were receiving appropriate antibiotic therapy at the time of drainage. Estimated duration of the fluid collections before initial drainage ranged from 1 to 21 days (mean, 7 days). Five patients initially underwent blind 30-French thoracostomy tube drainage. These failed to drain the fluid collections because of tube malposition (two cases) or because of the nondrainable consistency of the collections despite good tube position (three cases). These patients were subsequently referred for image-guided drainage. Six patients were referred primarily for image-guided drainage. All patients underwent CT exami- Fig year-old man with a loculated infected right hemothorax after a self-inflicted gunshot wound. A, CT scan made after 13 days of dual 30-French chest tube drainage (tubes have been removed) shows clotted posteromedial hemothorax. B, CT scan made after 3 days of 14-French single lumen catheter drainage shows no significant change in size of collection. Catheter (arrow) is in good position. C, CT scan made 2 days after final treatment shows minimal residual pleural thickening after seven instillations of urokinase over 48 hr. Tube was removed immediately after scanning. Fig. 2.-Left nonhemorrhagic parapneumonic empyema in a 60-year-old man. A, CT scan made 3 days after admission shows a large loculated left pleural fluid collection with left lower lobe and lingular consolidation. B, CT scan made 9 days after placement of two 14-French drainage catheters shows persistent fluid. Catheters (arrows) are in good position. Patient remained febrile. C, CT scan made 1 week after three instillations of urokinase over 10 hr shows minimal pleural thickening. Tubes were removed 1 day after therapy was completed.

3 AJA:153, November 1989 UROKINASE AND LOCULATED PLEURAL EFFUSIONS 943 Fig year-old man with a multiloculated right hemothorax. A, CT scan made 8 days after dual 30-French chest tube placement for a spontaneous tension pneumothorax. Anterior chest tube (solid arrow) is well positioned in an anterolateral loculation. Lateral chest tube (open arrow) is outside posterior loculation. B, CT scan obtained immediately after final treatment shows resolution of both loculations after CT.guided placement of two 14-French single lumen drainage catheters (arrows) and urokinase therapy (three instillations in each tube over 20 hr). No fluid could be aspirated from the 14-French catheters before urokinase therapy. Posterior loculation was drained effectively even though catheter was positioned posteromedially. nation before radiologic drainage to determine the presence and site of loculations. Fluid collections were considered loculated if CT showed pleural adhesions, a convex inner margin, a nondependent collection, or a collection that did not conform to the normal crescentic shape of the pleural space. Eight collections were unilocular, five were multilocular. Each patient subsequently underwent CT or fluoroscopically guided placement of 1 2- to 1 4-French single lumen vansonnenberg chest drainage catheters (Medi-tech, Watertown, MA). The technique of percutaneous image-guided pleural drainage has been described [1, 4]. Catheter insertion was preceded by 22-gauge needle aspiration with specimens tested for protein and glucose levels, cell count, Gram stain, and culture. If large separate loculations were present, catheters were placed in each. Nine patients were treated with a single drainage catheter, two patients required two catheters, and one patient required three catheters. Catheters were connected to 20-cm water suction via a closed underwater seal system (Thora- KIex, Davol, Cranston, Al). Irrigation with 15-mI saline aliquots was performed every 8 hr to prevent occlusion, with guidewire manipulation performed if occlusion occurred. Catheter manipulation was performed if the catheter did not remain optimally positioned within the loculation. Patients were considered candidates for transcatheter intracavitary urokinase therapy if standard tube drainage failed to drain the collection completely. Failed standard therapy was defined as persistent loculated fluid despite good tube position in a patient with persistent symptoms (fever and leukocytosis in infected collections, hypoxemia in sterile collections). We did not define a specific trial period for nonfibrinolytic therapy in terms of time, residual volume, or rate of drainage. In our series, standard therapy failed to drain the loculations adequately over an average of 1 0 days (range, 1-22 days). Aadiologically placed catheters had been in place an average of 7 days (range, 1-17 days) before urokinase therapy. Transcatheter intracavitary urokinase therapy was instituted in an attempt to decrease the duration of tube drainage. Instillation was performed via the percutaneously placed drainage catheter. If more than one catheter was required for large separate loculations, each was treated separately (Fig. 3). A total of 250,000 units of urokinase was dissolved in 250 ml of normal saline and divided into 80-mI aliquots. In very large collections (>500 ml estimated volume, two cases in this series) 100- to 1 50-mI aliquots were used. After instillation of 80 ml via the chest drain, the suction tube was clamped for 1-2 hr. Fluid was then aspirated, net output was recorded, and suction was reinstituted. After 1 hr of suction, a second instillation was performed. A third instillation was performed after an additional hour of suction or the next morning after overnight suction. Tube position and results were monitored with CT after the third treatment. If loculated fluid remained and tube position was satisfactory, the procedure was repeated. If no residual fluid was seen on CT scans, tube irrigations were stopped and suction was reinstituted. If the patient remained asymptomatic with less than 1 0 ml drainage over 24 hr, the tubes were removed in one step. Coagulation studies were performed on both serum and pleural fluid before and after treatment in three cases (one hemorrhagic and two nonhemorrhagic). Vital signs were measured every hour during treatment, and hematocrit was measured every 6 hr. Patients were encouraged to walk and change position during treatment to aid in mixing of the urokinase with the Ioculated fluid. The presence of a bronchopleural fistula was considered a contraindication of transcatheter intracavitary urokinase therapy in order to prevent intrapulmonary lavage. We encountered no patients with bronchopleural fistula during the study period. Recent intrapleural hemorrhage was considered a relative contraindication because of the risk of recurrent bleeding. In known hemorrhagic collections, we did not perform intracavitary urokinase instillation until at least 3 days after the initial bleeding. Results In eight cases, a single procedure (three instillations) resuited in complete resolution of the fluid collection (average procedure time, 1 7 hr; range, 8-20 hr). In four patients, the initial treatment resulted in incomplete resolution and the procedure was repeated. In one of these patients, the collection resolved after six instillations of 80 ml of urokinase given over 40 hr. The second collection required seven instillations of 80 ml of urokinase given over 44 hr. The third required seven instillations given over 48 hr; in the last four instillations, 1 50 ml of urokinase was used. The fourth collection required eight instillations given over 75 hr; the first five instillations were of 1 50 ml of urokinase and the last three were of 80 ml. In one patient, therapy was discontinued after two instillations of 80 ml of urokinase given over 8 hr, with a moderate amount of loculated fluid remaining. This patient was critically ill, with end-stage chronic obstructive pulmonary disease, and all therapy was discontinued at the family s request.

4 944 MOULTON ET AL. AJA:153, November 1989 The fluid aspirated from hemorrhagic collections (eight cases) initially consisted of intact clots with some dark lysed blood. During urokinase treatment, the fluid became progressively thinner with few clots and was considerably easier to aspirate. No fresh blood was aspirated in any of the cases. Fluid initially aspirated from the nonhemorrhagic exudates (five cases) contained large clumps and strands offibrin, which often occluded the catheter. This fibrinous component gradually disappeared during treatment. The total volume of pleural fluid removed in the 1 2 successful cases averaged 850 ml (range, ml). Standard therapy before urokinase therapy had removed an average of 480 ml (range, ml). Transcatheter urokinase therapy removed an average of 370 ml (range, ml). In all 1 2 successful cases, the CT examination performed after the final instillation showed minimal residual pleural thickening without fluid. All patients remained afebrile, and the catheters were removed on either the first or second day after the final treatment. Twelve of the 13 fluid collections completely resolved with no further therapy needed, for a cure rate of 92%. If we exclude the one patient who did not undergo an adequate trial of drainage, the cure rate was 1 00%. The average duration of successful urokinase therapy was 28 hr, with patients requiring an average of 4.3 instillations of urokinase. There were no bleeding complications, febrile reactions, or allergic manifestations. Each patient had a stable hematocrit during the procedure. Serum coagulation studies were normal before and after treatment. No fibrin-split products were detected in the serum. The level of fibrin-split products was greater than 640 g/dl in the pleural fluid before and after treatment. Measurable plasminogen was present in the pleural fluid before treatment; this decreased to undetectable amounts after treatment. Discussion Blind placement of large-bore closed thoracostomy tubes has been the standard therapy for hemothorax and empyema [1 0, 1 1]. Although this usually suffices if the fluid is free flowing, loculated collections may not be drained completely if the tube is not positioned properly within the loculation or if clotted or fibrinous debris occludes the chest tube [1, 3]. Success rates for closed thoracostomy tube drainage of empyemas range from 16% to 91 % [1 11 3] with higher success rates for primary parapneumonic empyemas than for postoperative or posttraumatic cases. In surgical series, failed tube drainage is usually followed by open surgical drainage [11, 12]. Several recent reports have advocated the use of imageguided catheter placement to avoid the problem of malpositioned blind thoracostomy tubes in the treatment of loculated collections, with success rates ranging from 72% to 92% [1-4]. Failures have been due to the viscous or clotted nature of the fluid or the presence of a pleural peel [2]. Image-guided tube failures are usually managed successfully by large-bore closed thoracostomy tube drainage, although up to 1 2% will require open surgical drainage [1, 3]. In addition, percutaneous transcatheter drainage procedures tend to be prolonged. Silverman et al. [1 ] reported duration of drainage from 7 to 45 days, whereas Merriam et al. [2] reported a range of 1-20 days (mean, 9 days). This correlates well with the average 7-day duration of image-guided tube drainage before urokinase therapy in our series. The use of streptokinase/streptodornase to facilitate drainage of loculated hemorrhagic and nonhemorrhagic pleural collections was first reported in [5]. Although there were no hemorrhagic complications, allergic reactions were common. In 1 977, investigators in Sweden used more purified streptokinase in an attempt to decrease immunologic reactions [7]. Although allergic manifestations were less common, production of antibodies in response to intrapleural streptokinase was documented [7, 8]. Urokinase has now supplanted streptokinase as the agent of choice for lysis of intravascular thrombi. Urokinase produces no antibody response, and it is this feature that accounts for the absence of febrile reactions. Efficacy is also increased because there is no antibody-mediated deactivation of urokinase. In 1 987, Vogelzang et al. [9] reported the successful use of urokinase in the treatment of infected intraabdominal hematomas. No febrile or allergic reactions occurred. We first attempted transcatheter intracavitary urokinase in an attempt to speed drainage of a hemorrhagic parapneumonic empyema that was incompletely drained after 1 5 days of 30-French thoracostomy tube drainage and 9 days of 14- French single lumen catheter drainage (Fig. 4). As we became more familiar with the technique, we tended to use fibrinolytic therapy earlier in the course of drainage if resolution was slow. In three of our most recent cases, we began fibrinolytic therapy 1 day after image-guided catheter placement. During this time, we encountered no cases in which image-guided catheter drainage led to immediate complete evacuation of the pleural collection. In all cases, the water-seal suction drainage was minimal despite good catheter position. We did not define a specific trial period for nonfibrinolytic therapy in terms of time or rate of drainage. This lack of specific criteria makes it difficult to speculate on the exact role of urokinase therapy in the spectrum of image-guided pleural drainage. Given the success rate of image-guided drainage quoted in the literature, most of our cases probably would have been drained successfully had we continued nonfibrinolytic therapy. However, this is a pilot study and the 28-hr average duration of successful urokinase therapy is striking when compared with the average 1 0-day duration of prior incomplete standard drainage. This is significant when one considers the economic pressures that current reimbursement strategies put on hospitals to both shorten the length of patient stays and minimize consumption of resources. In future cases, more rigid criteria need to be defined regarding when to begin fibrinolytic therapy. Despite the promising results, the series was uncontrolled, and it is possible, although not likely, that saline irrigation

5 AJA:153, November 1989 UROKINASE AND LOCULATED PLEURAL EFFUSIONS 945 Fig. 4.-Right hemorrhagic parapneumonic empyema in a 62-year-old man. A, CT scan made after 6 days of 30-French chest tube drainage shows persistent loculated fluid. Chest tube (arrow) is positioned anterior to loculation. B, CT scan made 7 days after CT-guided 14-French catheter drainage shows persistent fluid. Drainage catheter (arrow) is in good position. C, CT scan made 2 weeks after three instillations of urokinase over 20 hr shows minimal residual pleural thickening. Tubes were removed I day after therapy was completed. would have produced similar results. We used small-volume saline irrigation to maintain catheter patency, but we did not duplicate the urokinase protocol with saline. However, the presence of measurable plasminogen and fibrin-split products in the pleural fluid lends support to the theory that an enzymatic fibrinolytic process was occurring, a process that urokinase should accelerate. Our measurements, as well as those of Vogelzang et al. [9] and Berglin et al. [8], have shown no systemic fibrinolytic effect after intracavitary instillation of fibrinolytic agents. No hemorrhagic complications have occurred in all series reported to date. However, Coselli et al. [1 0] anecdotally mention that the use of streptokinase in the acute stage after traumatic hemothorax occasionally reinstituted the bleeding. Although we did not treat hemorrhagic collections until several days after the initial bleeding, others have begun therapy immediately with no adverse effects [6]. In addition, there is evidence that transcatheter drainage will be more successful if begun before the fibrotic stage of empyema/ hemothorax is reached [1]. This is estimated to begin approximately 7 days after pleural hemorrhage [1 0]. A compromise is faced between the theoretical risk of early fibrinolytic therapy inducing bleeding and the increased efficacy of early complete drainage. More experience with this technique is needed to determine the best timing of fibrinolytic therapy. Treatment protocols in all series to date have been empirical. It appears to be safe to instill up to 150,000 units of urokinase per treatment without adverse effect, although it is not known whether a dose-response relationship exists. Instillations have been left in place for 1 to 4 hr each and have been repeated at 1- to 24-hr intervals. Further studies are needed to determine the best dose, volume, timing, and duration of treatments. ACKNOWLEDGMENTS The authors thank Laura Herrmann for help in management of patients and data collection and Sharon Green for help in preparing the manuscript. REFERENCES 1. Silverman SG, Mueller PR, Saini 5, et al. Thoracic empyema: management with image-guided catheter drainage. Radiology 1988:169: Merriam MA, Cronan JJ, Dorfman GS, Lambiase RE, Haas AA. Radiographically guided percutaneous catheter drainage of pleural fluid collections.ajr 1988:151: vansonnenberg E, Nakamoto 5K, Mueller PA, et al. CT and ultrasoundguided catheter drainage of empyemas after chest-tube failure. Radiology 1984:151: Westcott JL. Percutaneous catheter drainage of pleural effusion and empyema. AJR 1985:144: Tillett WS, Sherry S. The effect in patients of streptococcal fibrinolysin (streptokinase) and streptococcal desoxyribonuclease on fibrinous, purulent and sanguinous pleural exudations. J Clin Invest 1949:23: Sherry 5, Tillett WS, Read CT. The use of streptokinase-streptodomase in the treatment of hemothorax. J Thorac Surg 1950;20: Bergh NP, Ekroth A, Larsson 5, Nagy P. Intrapleural streptokinase in the treatment of hemothorax and empyema. Scand J Thorac Cardiovasc Surg 1977;1 1: Berglin E, Ekroth A, Teger-Nilsson AC, William-Olsson G. Intrapleural instillation of streptokinase: effects on systemic fibrinolysis. Thorac Cardiovasc Surg 1981:29: Vogelzang AL, Tobin AS, Burstein 5, Anscheutz SL, Marzano M, Kozlowski JM. Transcatheter intracavity fibrinolysis of infected extravascular hematomas. AJR 1987:148: Coselli JS, Mattox KL, Beall AC. Reevaluation of early evacuation of clotted hemothorax. Am J Surg 1984:148: Wehr CJ, Adkins RB. Empyema thoracis: a ten-year experience. South MedJ 1986;79: Lemmer JH, Botham MJ, Orringer MB. Modern management of adult thoracic empyema. J Thorac Cardiovasc Surg 1989;90: Mandal AK, Thadepalli H. Treatment of spontaneous bacterial empyema thoracis. J Thorac Cardiovasc Surg 1987;94: