Catamenial pneumothorax: retrospective study of surgical treatment

GENERAL THORACIC

Catamenial Pneumothorax: Retrospective Study of Surgical Treatment Patrick Bagan, MD, Franc¸oise Le Pimpec Barthes, MD, Jalal Assouad, MD, Redha Souilamas, MD, and Marc Riquet, MD, PhD Service de Chirurgie Thoracique, Hoˆpital Europe´en Georges Pompidou, Paris, France

Background. Catamenial pneumothorax is a rare entity characterized by recurrent accumulation of air in the thoracic space during menstruation. Catamenial pneumothorax is also associated with a high rate of postoperative recurrence. The aim of this study was to discuss the etiology and to determine the optimal surgical treatment of this entity. Methods. From December 1991 to September 2000, 10 patients with catamenial pneumothorax were treated at our institution. Median age at time of operation was 37 years (range, 21 to 44 years). We retrospectively evaluated the pathologic findings, the operation performed, and the results in all patients. The mean follow-up was 55.7 months. Results. Pleurodesis alone was performed in 5 patients and an associated diaphragmatic procedure was performed in 5 patients. In 5 patients, no diaphragmatic anomaly was discovered: 3 experienced one or more recurrences and all still suffer from chronic catamenial

chest pain. Hormonal therapy temporarily improved outcome for 6 months in 2 patients. On the contrary, in 5 patients surgical pleurodesis was associated with the repair of diaphragmatic defects (simple closure or coverage by a polyglactin mesh): these patients experienced no recurrence (n ⴝ 0/5, p ⴝ 0.0016) and no subsequent catamenial chest pain. Conclusions. The postoperative outcome is influenced by the diagnosis of diaphragmatic defects with or without endometriosis. Surgical treatment should be accomplished during menstruation for an optimal visualization of pleurodiaphragmatic endometriosis. Because diaphragmatic lesion is frequent and may be occult, we propose the systematic coverage of the diaphragmatic surface by a polyglactin mesh to prevent catamenial pneumothorax recurrence even when the diaphragm appears normal. (Ann Thorac Surg 2003;75:378 – 81) © 2003 by The Society of Thoracic Surgeons

R

with diaphragmatic endometriosis confirmed histologically. Clinical characteristics are presented in Table 1. Mean age of patients was 37.2 years (range, 21 to 44 years). Pneumothoraces were unilateral and right-sided in 8 patients, left-sided in 1 patient, and bilateral in 1 patient. Symptoms consistent with pelvic endometriosis were observed in 5 of the 10 patients. One patient had a history of catamenial hemoptysis and 5 patients had been previously operated for right recurrent pneumothorax (2 in our institution, 3 at other institutions). We reviewed the different surgical treatment performed and their results using ␹2 analysis. All statistical analysis were performed using computerized software (StatView, Brain Power Inc, Calabasas, CA), with a p value of less than 0.05 considered as significant.

ecurrent spontaneous pneumothorax occurring in association with menstruation was first described by Maurer and colleagues [1] in 1958. Later, Lillington and colleagues [2] named this syndrome catamenial pneumothorax and presented 5 patients in 1972. The reports of catamenial pneumothorax (CPX) in the literature are rare and its physiologic mechanism remains unclear. Traditional therapy involving hormonal treatment or surgical pleurodesis seems to be associated with a high rate of recurrence [3]. The purpose of this retrospective study of 10 patient with CPX is to discuss the pathogenesis and to evaluate different methods of therapy for CPX.

Material and Methods From December 1991 to September 2000, 10 patients presenting with CPX underwent surgical treatment for CPX at our institution. The criteria used for diagnosis of CPX were as follows: recurrent pneumothorax documented by chest radiograph during or preceding menstruation and association of recurrent pneumothorax Accepted for publication Aug 24, 2002. Address reprint requests to Dr Riquet, Service de Chirurgie Thoracique, Hoˆpital Europe´en Georges Pompidou, 20-40 rue Leblanc, 75908 Paris, France; e-mail: [email protected].

© 2003 by The Society of Thoracic Surgeons Published by Elsevier Science Inc

Results Treatment Pathologic findings, treatment, and outcome are presented in Table 2. Of the 10 patients, 6 underwent thoracoscopy and 4 underwent posterolateral thoracotomy (as a second procedure in 3 patients). Surgical pleurodesis (pleurectomy n ⫽ 5, pleural abrasion n ⫽ 5) was performed in all patients. Isolated diaphragmatic defects were observed in 2 patients. Diaphragmatic de0003-4975/03/$30.00 PII S0003-4975(02)04320-5

379

BAGAN ET AL CATAMENIAL PNEUMOTHORAX

Table 1. Clinical Details of Patientsa Patient No.

Age (yr)

Pelvic Endometriosis

1 2 3 4 5 6 7

37 38 36 38 38 39 21

NO NO NO NO YES NO NO

Thoracoscopy Pl Ab Thoracotomy Pl Ab, Bl Res No No Thoracoscopy Pl Ab, Bl Res No No

8 9 10

44 43 38

YES YES YES

Thoracoscopy Pl Ab No Thoracoscopy Pl Ab, Bl Res

a

Previous Surgical Pleurodesis: Approach Technique

Indication (side) (No. of recurrence) Postoperative CPX (right) (2) Controlateral CPX (left) (1) Recurrent CPX (right) (3) Recurrent CPX (right) (3) Postoperative CPX (right) (3) Recurrent CPX (right) (3) Catamenial hemoptysis Recurrent CPX (Left) (2) Postoperative CPX (right) (2) Recurrent CPX (right) (3) Postoperative CPX (right) (3)

Patients are numbered chronologically.

CPX ⫽ catamenial pneumothorax;

Bl Res ⫽ blebs resection;

Pl Ab ⫽ pleural abrasion.

fects surrounded by endometrial implants were observed and confirmed by biopsy in 3 patients. The closure of diaphragmatic defects was obtained by interrupted sutures in all patients. During operation on patient 8 in 1997, we noticed an associated porous diaphragm. Because we feared to leave occult defects, we inserted a polyglactin mesh (Vicryl; Ethicon, Inc, Sommerville, NJ) through a thoracic approach to cover the tendinous part of the diaphragm. The mesh was held in place by stitches

at its periphery. The same technique was used for the remainder of the patients. There was no complication (except one wound infection of a trocar port). The mean duration of chest tube drainage was 3.9 days (range, 2 to 6 days) and the mean duration of hospital stay was 5.4 days (range, 4 to 8 days). Hormonal treatment was proposed to 7 patients: 1 patient refused and another was unable to tolerate the treatment. Five patients received gonadotropin-releasing hormone agonists during 6

Table 2. Pathological Findings, Treatment, and Outcomea Patient No.

Pathological Findings

Surgical Treatment (approach)

1

none

Pl Ab (VT)

2

Blebs

Pleur ⫹ Bl Res (VT)

3

Blebs

Pl Ab ⫹ Bl Res (VT)

4

Blebs

Pl Ab ⫹ Bl Res (VT)

5

Diaphragmatic defects

Pleur ⫹ Diaph R (Th)

6

Diaphragmatic defects

Pl Ab ⫹ Diaph R (VT)

7

Blebs

Pl Ab ⫹ Bl Res (VT)

8

Diaphragmatic defects and implants

9

10

Chest Tube Drainage/Hospital Stay (days)

Hormonal Treatment NO

Outcome Recurrence (n)/ Chest Pain (side)

Follow-up (mo)

Pleur ⫹ Pol M (Th)

2 4 2 8 4 5 5 6 4 5 4 5 4 5 4

YES

1 Right 0 Left 1 Right 1 Right 0 0 0 0 1 Bilateral 0

Diaphragmatic defects and implants

Pleur ⫹ Pol M (Th)

5 6

NO

0 0

30

Diaphragmatic defects and implants

Pleur ⫹ Pol M (Th)

7 4

NO

0 0

30

NO YES YES YES YES NO

120 110 66 60 48 12 36 45

5 a

Patients are numbered chronologically.

Bl Res ⫽ blebs resection; Diaph R ⫽ diaphragm repair; Pl Ab ⫽ mesh (vicryl); Th ⫽ thoracotomy; VT ⫽ videothoracoscopy.

pleural abrasion;

Pleur ⫽ apical pleurectomy;

Pol M ⫽ polyglactin

GENERAL THORACIC

Ann Thorac Surg 2003;75:378 – 81

GENERAL THORACIC

380

BAGAN ET AL CATAMENIAL PNEUMOTHORAX

months: 2 after a surgical pleurodesis alone, 3 after diaphragmatic repair.

Outcome The mean duration of follow-up was 55.7 ⫾ 34 months. We observed a high rate of recurrence after pleurodesis alone: 4 patients experienced one recurrence during menstruation and in 3 patients, the chest roentgenogram revealed a pneumothorax at the right base. Recurrences were all treated by placement of a chest tube. All patients (n ⫽ 5) treated by pleurodesis alone still suffer from chest pain during menstruation. Hormonal therapy during 6 months improved outcome in 2 of these 5 patients. On the contrary, diaphragmatic defect visualization and diaphragmatic repair procedures significantly improved outcome in 5 patients (p ⫽ 0.0016) with no recurrence and no pain after discontinuation of hormonal treatment.

Comment Pathogenesis The mechanism of CPX is unclear. Four proposed mechanisms for the cyclic occurrence of pneumothorax have been described in the literature. The first mechanism involves a congenital diaphragmatic fenestration or porosity. During menstruation, there is an open connection between the atmosphere and the peritoneal cavity because the cervical mucus plug is absent. Air can migrate through the fallopian tube into the abdominal cavity and through diaphragmatic fenestrations or porosities, mainly observed on the right side, causing a pneumothorax [4 – 6]. The second mechanism involves acquired diaphragmatic fenestrations caused by endometriosis. Endometrial tissue migrates to the diaphragm by retrograde flow from the uterus to the pelvis, hence in the peritoneal cavity to the subphrenic space [7]. When such endometrial tissue undergoes cyclical necrosis due to menstrual hormonal activity, the diaphragm perforates, producing a defect [8]. Endometrial tissue has been identified on the edges of the defect in many cases of CPX [9 –11] (and in 5 patients in our population). Pneumoperitoneum can migrate through these defects to cause a pneumothorax during menstruation [12]. The third mechanism involves the metastatic spread of endometrial tissue. Endometrial implants find their way through the uterine veins in the venous system [13]. In this model, endometrial implants reach the lung parenchyma and cause focal defects on the pleural surface during menstruation leading to air leaks [14]. The fourth mechanism cites the release of dinaprost tromethamine (prostaglandin F2). During menstruation prostaglandin F2 may be present in the plasma of some women. It is a potent constrictor of bronchioles and vascular structures. Alveolar tissue damaged by vasospasm may cause pneumothorax if bronchospasm impedes expiration [15]. In accord with the above described, our patient population can be divided into two groups. The first group is characterized by the presence of diaphragmatic defects (5 patients). Diaphragmatic de-

Ann Thorac Surg 2003;75:378 – 81

fects were identified during the first procedure in 2 patients and during the second procedure in 3 patients. All of these patients had a history of severe pelvic endometriosis (infertility, chronic pelvic pain). These observations lend support for the mechanism involving diaphragmatic perforation caused by retrograde peritoneal implantation of endometrial tissue. Shiraishi [16] reported that diaphragmatic defects were found in 29% of patients with CPX in the English literature and in 66% of cases in the Japanese literature. In our series, diaphragmatic defects were identified during the second procedure in 60% of the patients (n ⫽ 3). Pleurodiaphragmatic endometriosis undergoes cyclical changes due to hormonal activity. During menstruation, implants are congestive, intermittent bleeding causes necrosis, and subsequently the diaphragm perforates. Afterward, the lesions transform into fibrous tissue. Therefore, we recommend a complete exploration of the diaphragmatic surface, with the optimal time for the surgical procedure being the period of menses. In the second group, diaphragmatic defects were not visualized. One patient with bilateral recurrent CPX had a history of catamenial hemoptysis, which suggested the presence of parenchymal endometriosis. In this patient, lung metastasis of endometrial tissue through the venous system can explain the bilateral localization of parenchymal endometriosis. Four patients with unilateral recurrent CPX (rightsided in 3 patients and left-sided in 1 patient) had only blebs on the pleural surface. Surgical blebs resection combined with pleurodesis was inadequate as seen by postoperative recurrence in 3 patients and chest pain during menstruation in all patients. In these patients, the mechanism of CPX is still unclear but chest roentgenograms revealed a pneumothorax at the right base in 3 patients, which suggests that the recurrences were probably due to unobserved diaphragmatic lesions.

Treatment Current treatment of CPX combines the principles of treatment of spontaneous pneumothorax with that of hormonal treatment. Because a direct relation between menstruation and pneumothorax has been demonstrated, prophylactic hormonal therapy used to suppress ectopic endometrium activity by blocking hormonal support from the ovary is warranted. Gonadotropinreleasing hormone agonists induce an hypogonadotropic hypogonadism. The side effects are those of hypoestrogenism, especially osteoporosis, which limits duration of the treatment to 6 months [17]. In our patients, this treatment was not helpful when diaphragmatic repair was not performed. Apical pleurodesis can be achieved using pleural abrasion or pleurectomy, but this method does not prevent diaphragmatic perforation and catamenial chest pain that are, in the majority of the patients, both induced by cyclical proliferation of microscopical endometrial implants [3]. In 3 patients, we lined the diaphragm with a polyglactin mesh. This technique was performed for fear of leaving behind small defects. The goals of polyglactin mesh insertion was to reinforce the

diaphragmatic surface, to induce fibrotic adhesion with the lung [18], and to contain and prevent diaphragmatic perforation by endometrial implants. The good results obtained with this technique (no recurrence and no chest pain after the interruption of hormonal treatment for a mean follow-up of 35 months [range, 30 to 45 months]), suggest to use more widely this technique when no diaphragmatic defect is discovered during operation. In fact, recurrence may be due to diaphragmatic defects with or without endometriosis. This was probably the case in our 3 patients who experienced postoperative recurrences at the right base and who may have benefited from this procedure. To conclude, with respect to our series and reports in the literature, there is considerable evidence to support the involvement of endometriosis in the pathogenesis of CPX. It is important to differentiate between pulmonary endometriosis, which appears to be the result of vascular metastasis, and diaphragmatic endometriosis, which is presumed to be more frequent than reported in the literature. Conventional treatment of CPX appears to be insufficient, but when diaphragmatic lesions are visualized, appropriate treatment significantly improves patient outcome. This observation focuses on the importance of exploration of the diaphragm during menstruation. However, when diaphragmatic endometriosis is suspected during CPX but not observed during operation, we suggest lining of the diaphragm with a polyglactin mesh to help prevent recurrence.

References 1. Maurer ER, Schaal JA, Mendez FL Jr. Chronic recurring spontaneous pneumothorax due to endometriosis of the diaphragm. JAMA 1958;168:2013–4. 2. Lillington GA, Mitchell SP, Wood GA. Catamenial pneumothorax. JAMA 1972;219:1328 –32.

BAGAN ET AL CATAMENIAL PNEUMOTHORAX

381

3. Joseph J, Sahn SA. Thoracic endometriosis syndrome: new observations from a analysis of 110 cases. Am J Med 1996; 100:164 –9. 4. Crutcher RR, Waltuch TL, Blue ME. Recurrent spontaneous pneumothorax associated with menstruation. J Thorac Cardiovasc Surg 1967;54:599 –602. 5. Funatsu K, Tsuru M, Hayabuchi N. Catamenial pneumothorax and its relation to the peritoneal stomata of the diaphragm. Chest 1999;116:1843. 6. Brichon PY, Riquet M, Milongo R, et al. Epanchement pleural atypique par communication transdiaphragmatique. Ann Chir: Chir Thorac Cardio-vasc 1992;46:170 –3. 7. Foster DC, Stern JL, Buscema I, et al. Pleural and parenchymal endometriosis. Obstet Gynecol 1981;58:555. 8. Kirschner PA. Porous diaphragm syndromes. Chest Surg Clin North Am 1998;8:449 –72. 9. Soderberg CH, Dahlquist EH. Catamenial pneumothorax. Surgery 1976;79:236 –9. 10. Blanco S, Hernando F, Gomez A, et al. Catamenial pneumothorax caused by diaphragmatic endometriosis. J Thorac Cardiovasc Surg 1998;116:179 –80. 11. Slasky BS, Siewers RD, Lecky JW, et al. Catamenial pneumothorax: the role of diaphragmatic defects and endometriosis. AJR 1982;138:539 –42. 12. Downey DB, Towers MJ, Poom PY, et al. Pneumoperitoneum with catamenial pneumothorax. AJR 1990;155:29 –30. 13. Shearin RPN, Hepper NGG, Payne WS. Recurrent spontaneous pneumothorax concurrent with menses. Mayo Clinic Proc 1974;49:98 –101. 14. Van Schil PE, Vercauteren SR, Vermeire P, et al. Catamenial pneumothorax caused by thoracic endometriosis. Ann Thorac Surg 1996;62:585–6. 15. Rossi NP, Goplerud CP. Recurrent catamenial pneumothorax. Arch Surg 1974;109:173–6. 16. Shiraishi T. Catamenial pneumothorax: report of a case and review of the Japanese and non-Japanese literature. Thorac Cardiovasc Surgeon 1991;39:304 –7. 17. Olive DL, Pritts EA. Treatment of endometriosis. N Engl J Med 2001;345:266 –74. 18. Sugarmann WM, Widmann WD, Mysh D, et al. Mesh insertion as an aid for pleurodesis. J Cardiovasc Surg(Torino) 1996;37:173–5.

INVITED COMMENTARY Catamenial pneumothorax is an uncommon entity but will probably be seen at some point in most general thoracic surgeons’ careers. The article by Bagan and colleagues lends added evidence to the importance of diaphragmatic defects, with or without endometriosis, in the pathogenesis of this entity [1]. The search for diaphragmatic defects is essential, and thoracoscopy is the ideal tool. If no defect is seen, but blebs are identified, resection and pleural abrasion ⫾ pleurectomy versus talc poudrage can be performed. This first step of pleurodesis seems reasonable since some pneumothoraces may result from embolized endometrial tissue to peripheral lung parenchyma and not involve defects in the diaphragm, and surgical pleurodesis has documented success [2]. In this regard, I would not favor using talc poudrage as the pleurodesis agent, since recurrence in the right basilar location, as reported in this series, would suggest there will be cases of missed diaphragmatic defects. At this point, thoracotomy and © 2003 by The Society of Thoracic Surgeons Published by Elsevier Science Inc

placement of mesh over the tendinous portion of the diaphragm are reasonable. Tale poudrage at initial thoracoscopy could add technical difficulty if a second procedure is necessary. Carolyn E. Reed, MD Division of Cardiothoracic Surgery Medical University of South Carolina 171 Ashley Avenue Charleston, SC 29425-2279 e-mail: [email protected]

References 1. Bagan P, LePimpec-Barthes F, Assouad J, Souilamas R, Riquet M. Catamenial pneumothorax: retrospective study of surgical treatment. Ann Thorac Surg 2003;75:378 – 81. 2. Joseph J, Sahn SA. Thoracic endometriosis syndrome: new observations from an analysis of 110 cases. Am J Med 1996; 100:164 –70. 0003-4975/03/$30.00 PII S0003-4975(02)04628-3

GENERAL THORACIC

Ann Thorac Surg 2003;75:378 – 81