Thoracoscopy for diagnosis and management of refractory hepatic hydrothorax

Dianjie Lin, Zhang Min, Gui-xin GAO, Bin Li, Mao-fen Wang, Ling Zhu, Li-fu XUE

Chinese Medical Journal · 2006 · 16 citations · 9 references

DOIFull text

Open access

Concepts

Abstract

The prevalence of pleural effusion caused by hepatocirrhosis ranges from 0.4% to 30%. Some patients with hepatic hydrothorax (HH) fail to respond to aggressive medical management such as albumin application, diuretics, and thoracentesis, and remain refractory massive pleural fluid.1 The repeated suctions to relieve the symptoms such as respiratory distress are associated with significant hyponatremia and hypoalbuminemia caused by the large volumes of fluid loss. Review of the literature has revealed that, to date, no therapy is ideal. This report describes our experiments in treating HH of patients with cirrhosis and ascites by thoracoscopy and talc poudrage, to offer our efforts to approach the pathogenesis of HH and to explore the new therapeutic strategy. METHODS A retrospective study was conducted, consisted of 20 men and 6 women ranging in age from 32 to 64 years [mean, (44.2±8.5) years], who were admitted to Shandong Provincial Hospital from March 1992 to April 2004. Twenty-one of them had post-hepatitis cirrhosis, 3 patients had alcoholic cirrhosis, 1 had biliary cirrhosis, and 1 had Budd-Chiari syndrome. Twenty-two cases remained refractory right-sided symptomatic HH and 4 with bilateral HH (large volume in right and small in left). Twenty-one cases presented massive ascites and 5 with moderate and modest ascites. All patients had HH history more than one month. Before pleurodesis, all patients called for repeated thoracentesis for rapid relief of syndromes and average volume of drainage ranged from 4500 to 24000 ml [mean, (9800±910) ml]. The diagnostic criteria of HH was as follows: (1) patients had the history of chronic liver disease, albumin<30 g/L, the proportion of albumin/globin was inverted; (2) portal vein pressure had stepped up, bypass circuit was performed with varices of esophagus, stomach fundus or abdominal wall; (3) spleen enlarged complicated with hypersplenia; (4) patients had massive pleural effusion and more or less ascites. The characteristic of pleural effusion and ascites was transudate. If the symptom of patients accorded with (1) and (4) plus (2) or (3), the diagnosis could be established.2 To estimate the patients' tolerance, clotting time, blood platelet count (BPC), liver and renal function and ECG were performed before operation. All procedures were processed on patients' permission. The HH was drained 600-1000 ml and 300-600 ml gas was injected into the pleural cavity instead to establish artificial pneumothorax before thoracoscopy. Then the extent of pleural symphysis was observed by chest radiography. The patients were positioned in the lateral decubitus position under local anesthesia. Twenty-four video-assisted thoracic surgery (VATS) procedures with two ports (one in the fifth and one in the eighth intercostals space) were performed using a rigid thoracoscope. Two procedures with one port were performed using a fiber thoracoscope. After complete evacuation of the HH, the pleural cavity was covered with 3-5 g of aerosolized medical sterile talc powder and drained by placing one or two chest tubes in the cavity. Postoperatively, medical therapy was maintained, suction (20 cmH20) was inserted, and the chest tubes were removed when the drainage volume remained <50 ml/d, or at least 2 to 3 days after sealing the air leaks. The chest radiography, chest ultrasound and liver function were performed every week after thoracoscopy for 2 or 3 weeks. Twenty-four cases of 26 were undergone pleurodesis via stereoplasm thoracoscopy. Complete response (CR) was defined as the complete absence of HH; partial response (PR) was defined as the obvious reduction of HH or encapsulated hydrothorax formation without respiratory symptoms; if the symptoms had no any change, it was defined as non-change (NC). The follow-up period ranged from 6 months to 3 years and any reaccumulation of pleural fluid was regarded as recurrence. RESULTS The pleural surfaces of 25 cases were smooth without any symphysis. The veins in parietal pleura and diaphragm became visible in 19 cases and azygos vein was observed in 6 cases. There were 16 cases with orifices located on the diaphragmatic muscle area, 14 cases with vesicles on the diaphragm tendinous area, 7 cases on costal part, and 5 cases with orifices on the diaphragmatic part. The vesicle size was different and changed with the respiration and abdominal pressure, in which the biggest one was 0.6 cm×0.6 cm×0.5 cm. In some patients we detected some fiber scars formed in the repair of the injured diaphragm. One week after thoracoscopy, 14 cases that performed ultrasound and X-ray examinations were found that the pleura adhered widely and the HH vanished. There were 8 cases formed encapsulated hydrothorax. One of them was drained twice after the large encapsulated hydrothorax located by ultrasound and injected medical sterile talc powder 1-2 g spontaneously, and then the HH disappeared. The average time to remove chest tubes ranged from 3 to 8 days [mean, (4.3±1.5) days]. Only two patients failed to respond to talc and the drainage volume >600 ml/d. Of 24 patients who received pleurodesis via thoracoscope, 14 cases (58%) acquired complete response, 8 cases (33%) acquired partial response and 2 cases (8%) had non-change. Intrapleural insufflation of talc was associated with minimal morbidity. We detected 6 cases with mild thoracic pain for several minutes and 17 cases with modest or moderate chest pain for 1-3 days. Fifteen cases had mild fever and 1 case had ardent fever for 2-3 days. Only 2 patients suffered fever for 1 week. One case presented intensive bellyache accompanied the taut abdominal muscle for 6 hours. As far as liver function test, we found 7 patients with aminopherase slightly elevated and 1 patient with aminopherase twice higher than normal level. The first liver function test performed in 23 patients one week after the operation. The second test administered in 17 patients two weeks after the operation. We observed that 8 patients' aminopherase slightly increased and 1 patient's aminopherase obviously raised accompanied by aurigo, however at the end of experiment, all patients' aminopherase came to pre-operation levels except 3 patients. There were 2 unsuccessful cases, one discharged and lost the contact and one died of hepatic encephalopathy 1 month after the operation. The other 22 cases were followed up from 6 months to 3 years. One died of hemorrhage and hepatic encephalopathy one month after the operation. Three died of hemorrhage in 6, 12 and 18 months respectively. One case presented recurrence 18 months after operation. DISCUSSION Morrow et al3 first described a rapid accumulation of massive right pleural effusion after a diagnosis of hepatic cirrhosis and defined it as HH in 1958. Although a wealth of literature on the treatment of symptomatic HH published, the optimal or standard therapy remained elucidated. Morbidity of hepatic hydrothorax was various because of different standards to choose patients. Usually it was found in the right hemoithorax (approximately 70%), but could also be seen on the left side (approximately 15%), bilaterally (approximately 15%), and even in the absence of ascites.4 The exact mechanism of HH remained unknown. A number of different mechanisms had been proposed to explain the development of HH, including the following: (1) Hypoalbuminemia and a decrease in colloid osmotic pressure. But this model could not explain why HH commonly appeared at right-sided hemothorax. Furthermore, some patients with a remarkable decrease in albumin did not produce hydrothorax, while others with normal albumin formed massive hydrothorax. (2) Portal hypertension and bypass circulation established. Varices of esophagus and stomach fundus lead to the hypertension of azygos, hemi-azygos vein, and then the body fluids extravasated into pleural cavity. However, this theory still could not interpret why hydrothorax was commonly found at right-sided hemothorax and why some patients with collateral circulation did not produce hydrothorax. One patient with Budd-Chiari syndrome had large ascites accompanied by hydrothorax. After the first thoracoscopy we found that the chest wall vein appeared and azygos vein varix formed. After caval vein bracket treatment the ascites disappeared, but the massive HH still remained. In second thoracoscopy, we detected that the pleura was commonly normal and the veins did not appear or distend, which suggested the hypertension of azygos vein could not account for the pathogenesis of hydrothorax. (3) The lymph stream of liver increased, and recirculation of hepatic sinus post-vein obstructed when cirrhosis performed. The lymph leakage could increase from 1 ml/min to 10 ml/min, which resulted in the pressure of lymph passage increased similar to that of portal vein (300 mmHg), and thus the lymph vessels in pleura distended and ruptured. (4)Ascites transferred from peritoneal cavity to the pleural space via lymphatic channels in the disphragm. When the pleural lymph circulation obstructed, the hydrothorax appeared. It had ever been reported that the formation of right-sided hydrothorax could be more closely related to the lymph distribution.5,6 (5) Defects of diaphragm. These defects probably resulted from anatomic thinning and separation of the taut collagenous fibers of the tendinous portion of the diaphragm. Congenital factors, high intra-abdominal pressure (cough, defecation, parturition), trauma or prolonged bed rest might contribute to the defects.7 The transdiaphragmatic defects could be divided into four types. Type I was big. Gas and fluid could go through between the thoracic cavity and abdominal cavity freely. One patient in this group who underwent artificial pneumothorax therapy and formed hydropneumothorax was belonged to this type. Type II was small. Fluid could go through the pore but gas could not. It played a role of one-way channel to prevent hydrothorax to flow into abdominal cavity. Type III was smaller. The negative pressure of pleural cavity induced ascites into pleural cavity. When the ascites volume that abdomen generated was less than or as much as that pleural cavity absorbed, the HH appeared without ascites. Type IV is the smallest. Gas and fluid could not go through the orifice into pleural cavity. But HH could go back to abdominal cavity via gravity. So the hydrothorax rapidly decreased after draining the ascites fluid.8 Type II and type III were commonly in clinical. Diagnostic modalities of diaphragmatic fenestration had included instillation of air, contrast agents, or nuclear scintigraphy (for example 1311, 99Tc) into abdomen.9 In our experiment, 16 cases presented diaphragmatic vesicles, which were often multiple and distributed in the various place of diaphragmatic muscle. It was also found on costal part and diaphragm part. Our observations were different from the previous views that diaphragmatic vesicle were commonly seen in weak point like chorda and tendinous portion of the diaphragm.10 HH was mainly attributed to hepatic cirrhosis. The chief characteristic of the composition of HH was transudate and as same as that of ascites. When infected, HH appeared biphasic changing. Some patients could produce sanguineous hydrothorax. Four cases in this group with sanguineous hydrothorax were caused by diaphragmatic vesicle rupture. The sanguineous hydrothorax also could be caused by short circuit of collateral circulation, rupture of communicating branch and injury of endotoxin to blood vessel.11 At the same time other diseases like tuberculosis, tumor, kidney disease and so on should be eliminated. Furthermore, the position of contracted lung altered was a vital evidence to diagnose HH in radiology. All the patients in this group underwent thoracoscopy. Only one patient with the history of tuberculous pleuritis appeared pleural symphysis and the others had smooth partial pleura, which were consistent with the appearance in radiology. To date the optimal or standard therapy for HH remained unclear. Using diuretic caused fluid and electrolyte imbalance and did little good to reduce hydrothorax. Drainage was also ineffective. Repeatedly thoracentesis could lead to substantial protein depletion, which in turn resulted in the deterioration of the clinical status.12 Hydrothorax retransefusion could avoid the loss of protein and body fluid, but it could not prevent hydrothorax from reaccumulation and increased the danger of infection, hepatic coma and disseminated intravascular coagulation (DIC). Multiple medical management should be adopted such as protecting liver function, supplying plasma products and branched chain amino acid, accompanied with water and salt restriction. This comprehensive treatment could make HH and ascites decreased or even disappeared in one third of the reported patients.13 The pleurodesis should be performed when the clinical management was ineffective. Talcum was proved to be one of the most effective sclerosing agents to induce pleurodesis by experiments and clinic research. It was a kind of magnesium trisilicate with chemical architecture Mg6(SiO5)4(OH)4, which could stimulate pleura fibrosis and granuloma degeneration. Pleurodesis using talc poudrage via thoracoscope was considered to be the most excellent method to treat malignant hydrothorax and refractory innocent hydrothorax.7 Thoracoscopy could be operated under local anesthesia, and therefore, it has wide availability and low cost. In this group, the pleural effusion could be drained completely and the pleura were covered by aerosolized drugs. This therapy was effective in treating malignant pleural effusion and refractory pneumothorax and was most useful in pleural adhesion. The success rate was 92%, CR was 58%, PR was 33%, and NC was 8%. This method was also proved to be effective to HH. The side effects about the treatment of insufflating talc in the pleural cavity under thoracoscope were slight such as chest pain, fever.14 Half of the patients have chest pain and fever because of the aseptic inflammatory reaction caused by the talc stimulating the pleura after operation. The majority of patients had modest and moderate fever for 2-3 days. A few patients had high fever with the temperature of 39°C. Patients who had modest and moderate chest pain and fever needed no special treatment generally. In this group, only one patient presented serious stomachache with the abdomen intension. The symptom was not relieved effectively after using spasmolysant. Abdominal X-ray and ultrasound examinations were performed to exclude acute abdomen such as gastrointestinal perforation. Whether the pain related to talc or talc leakage to the pleural cavity through the diaphragmatic defects remained elucidated. Using talc for pleurodesis we detected no adverse effects on aminopherase of patients who had normal liver function with refractory pneumothorax and malignant pleural effusion. But in this group, 8 patients had higher aminopherase, 1 patient whose aminopherase rose obviously accompanied by icterus. The aminopherase level of 7 patients came back to the level before operation 2 weeks later. The patients whose aminopherase increased obviously died of massive hemorrhage, hepatic encephalopathy one month after operation. When the patients with HH underwent the treatment of pleurdesis using talc, the liver function should be closely checked-up. The effect of talc on liver function was not definite. The possible mechanism might due to the inflammatory factors caused by aseptic inflammatory or the fine talc particle itself.15

References

9