Primary Spontaneous Pneumothorax in Menstruating Women Has High Recurrence

During the study period from May 2009 - June 2015, 33 female patients met the inclusion criteria for the study. Their clinical and demographic profiles are presented in Table 1 . The median age was 33.4 years (range 19 – 50) with a median follow up time of 14 months. Nine women (27%) had a left-sided spontaneous pneumothorax while the majority, 24 (73%) had right-sided pneumothoraces. All patients were initially managed with tube thoracostomy. There were no treatment failures after tube thoracostomy, as defined by lack of resolution of pneumothorax or development of recurrent pneumothorax within two weeks. Overall twenty-one patients (63.6%) had a recurrence of pneumothorax following management by tube thoracostomy during the study period. Right-sided pneumothorax was associated with ipsilateral recurrence in 16/24 (67%) patients, whereas left sided pneumothorax was associated with ipsilateral recurrence in 5/9 (56%) patients. Two patients refused any surgical treatment after developing ipsilateral recurrence and were treated with tube thoracostomy. The remaining 19 patients (58% of the original cohort) underwent surgical intervention as described in Table 2 . In contrast, 183 male patients were treated with PSP during this study period with the same height as well as inclusion/ exclusion criteria. The median age was 31.6 years (range 20 – 57) with a median follow up time of 16 months. The mean height and weight were 176 ± 6.1 cms and 173.5 ± 26.5 lbs, respectively. Seventy-five (41%) had a left-sided spontaneous pneumothorax while 108 (51%) had right-sided pneumothorax. Of these only 8 (4.4%, p<0.001 compared to female patients) recurred with a median follow up of 16 months and were successfully treated with surgical pleurodesis. The surgical procedures included apical wedge resection in all patients and mechanical pleurodiesis (12/19, 63%) or subtotal pleurectomy (7/19, 37%). During thoracoscopy, nodular deposits were noted over the diaphragm in three (16%) patients, and four (21%) patients had diaphragmatic fenestrations. When diaphragmatic fenestrations were noted, plication of the diaphragm was performed to cover the defects. Histology of the apical wedge resections confirmed blebs/subpleural fibrosis in 11 (58%) of operated patients while analysis of the diaphragmatic nodules as well as the fenestrations revealed mesothelial hyperplasia and reactive changes but no endometrial glands or stroma. Overall there was a recurrence rate of 64% in the study cohort at a median follow up of 14 months from the initial episode of pneumothorax ( Figure 1A ). Right-sided pneumothorax had a higher recurrence (70%) compared to left (56%, p=0.02, Figure 1B ). Of the study cohort, four patients (12%) presented with large recurrent pneumothoraces with signs of tension pneumothorax within the first six months of the initial episode. One patient developed a second ipsilateral recurrence following apical resection with subtotal pleurectomy and she was successfully treated with talc poudrage. None of the male patients treated with surgical pleurodesis developed recurrence. Of the study cohort, 8 (24%) patients reported that their symptoms began within 72 hours of the onset of menstruation. Of these, three patients had a documented history of pelvic endometriosis. All eight patients were either already receiving, or were placed on oral contraception after consultation with a gynecologist. Despite this, all patients developed ipsilateral recurrence of pneumothorax within the first year of the initial episode ( Figure 2 ). Of these, six (75%) were right and two (25%) left sided. All eight patients with CP underwent surgical therapy at the time of the second episode. Thoracoscopy revealed diaphragmatic fenestrations in four (50%) and nodular deposits in three patients (38%). However, upon careful evaluation, neither endometrial glands nor stroma were found in the tissues resected from any of the patients. We evaluated whether the presence of hemosiderin would be associated with thoracic endometriosis, as previously reported [ 6 ]. Biopsy materials from the apical resection, pleural and diaphragm biopsies were evaluated and revealed presence of hemosiderin-laden macrophages in three (38%) patients. These macrophages were found at the site of diaphragmatic fenestrations. Tissues biopsies from all of these patients revealed mesenchymal fibrosis and chronic epithelial cell changes similar to those patients without association of PSP with their menstrual cycles ( Figure 3 ). In this study, we investigated the recurrence rates of PSP in menstruating females with no underlying lung disease and those who were not tall. Tall males or patients with marfanoid features can have apical blebs that lead to PSP [ 3 ], but this does not apply to normal-statured females who likely have a different pathogenesis of PSP. Tobacco [ 7 , 8 ] and cannabis [ 9 ] smoking are associated with PSP even in the absence of parenchymal lung disease. Deeper inhalation observed frequently in cannabis smokers and valsalva-like maneuvers can lead to PSP [ 9 ]. Hence, we excluded active tobacco or cannabis smokers from our study. Therefore, our patients consisted of a homogenous group of menstruating females with no underlying lung disease or obvious cause for PSP. We found a much higher recurrence of pneumothorax in our patient population compared to the published literature [ 1 ]. This might suggest that PSP in female patients should be treated more aggressively. This is further emphasized by the fact that about 12% of patients presented with a potentially life-threatening recurrence within 6 months. During the study period, the male patients with same height and inclusion/exclusion had a significantly lower recurrence rate. The overall male recurrence rate was lower in our series compared to published literature likely due to the exclusion of tall males [ 2 , 3 ]. The etiology of the high recurrence rates in the female patients remains unknown. Catamenial pneumothorax has been proposed as the etiology of PSP in these patients. Nevertheless, it is generally considered a rare phenomenon with a previously reported prevalence of 0.9 – 5.6% of all cases of PSP in women [ 10 , 11 ]. However, others have shown a higher prevalence up to 18-29% [ 12 - 14 ]. We evaluated clinical, morphological, and histological evidence of CP in female patients with PSP after eliminating obvious causes of PSP. Roughly 25% of our patients were clinically diagnosed with CP, defined by association with menses within 72 hours. The higher prevalence of CP in our study and others is likely related to the selection of study cohorts. Furthermore, 50% of our patients with PSP associated with menses revealed morphological features of CP such as nodular deposits and diaphragmatic fenestrations. Thoracoscopy provides improved exposure and magnification during chest exploration that may otherwise be missed at thoracotomy [ 13 ]. Therefore, use of video-assisted thoracoscopy (VATS) might also explain the increased detection of morphological features of CP observed in our series as well as others. Nonetheless, classical histological features of endometriosis such as endometrial glands and stroma were lacking in all patients. Instead, all our patients revealed non-specific histological features including mesothelial hyperplasia and reactive epithelial cell changes. Three patients had evidence of hemosiderin-laden macrophages that are suggestive of, but not definitive evidence for, CP. Hence, we postulate that standard histology has poor utility in defining the diagnosis of CP. To improve the histologic evidence of endometriosis, estrogen and progesterone receptor staining has been performed and shown to increase sensitivity [ 15 ]. Additionally, immunostaining with CD10 has been reported to be useful. However, the published experience with these markers has also been highly variable 16 . We conclude that the diagnosis of catamenial pneumothorax should be made clinically based on the temporal relation of PSP to menses and the identification of gross pathologic findings (nodular brown lesions, diaphragmatic defects, and implants) at the time of surgery. Approximately 95% of catamenial pneumothoraces occur on the right side and are associated with a high recurrence rate. In our cohort, there was right side predominance of PSP as well in patients clinically diagnosed with CP. Several hypotheses have been postulated to explain the pathogenesis of CP. The coelomic metaplasia theory suggests that a metaplastic transformation of the cells lining the pleural cavity accounts for development of endometrial tissue. This, however, does not account for the observed right-sided predominance. Other theories involving ectopic spread of endometrium via lymphatic or vascular embolization or retrograde menstruation may better explain the increased prevalence of right-sided PSP. Diaphragmatic defects found at thoracoscopy corroborate the transperitoneal-transdiaphragmatic theory of endometrial implant migration. These defects should be repaired by suture plication that eliminates a potential mechanism for endometrial implant migration. In our series the recurrence rate for primary spontaneous pneumothorax in females was high, particularly for right-sided pneumothoraces. This high early recurrence and potential for life-threatening recurrence may warrant management with surgery at the initial presentation of spontaneous pneumothorax in this patient population. Further prospective studies are necessary to validate our findings. The limitations of our study include its retrospective nature. Additionally, we excluded patients with smoking history or evidence of lung disease. CP can certainly occur in those patients. However, by eliminating those patients, we obtained a more homogenous group to study CP. Data from our study cannot be applied to tall females or those with traumatic and secondary causes of pneumothorax. We conclude that primary spontaneous pneumothorax in menstruating women with no underlying lung disease has a high recurrence rate. The high recurrence of PSP in this patient population may warrant earlier follow up and/or primary surgical intervention. Furthermore, the diagnosis of CP should be made clinically as pathologic findings of endometrial glands or stroma are rarely identified histologically.

We retrospectively identified all women treated at our center for PSP between May 2009 and June 2015. Inclusion criteria included menstruating females between the ages of 18-55 years, of height less than or equal to one standard deviation from average American female (<169.2 centimeters, www.cdc.gov ). Exclusion criteria included a history of trauma as a cause of pneumothorax, pregnancy, known lung disease or presence of parenchymal lung disease on chest imaging, current smokers or smoking exposure > 10 pack years, marijuana or other drug inhalation. During the same study period, we also included male patients less than one standard deviation from average American male height (<183.6 centimeters) with the same inclusion and exclusion criteria as outlined for the female subjects. The study was approved by the Institutional Review Board. Catamenial pneumothorax was defined as spontaneous pneumothorax occurring within 72 hours of menstruation, as previously reported [ 4 , 5 ]. At the time of initial evaluation all patients were queried regarding the timing of their last menstrual cycle and the onset of symptoms to assess potential correlation. Recurrence of pneumothorax was diagnosed by radiographic evidence of pneumothorax at least two weeks following resolution of initial episode. Those patients who never had radiographic resolution of first episode, or developed a pneumothorax within two weeks, were defined as failure of initial management. All patients with recurrence of primary pneumothorax underwent video-assisted thoracoscopic (VATS) evaluation of the chest cavity and diaphragm for signs of thoracic endometriosis, including nodules and fenestrations. Visceral and parietal pleura were examined for nodular lesions, and lung for blebs or nodules. All visible foci of implants concerning for endometriosis in lung parenchyma were resected and sent for pathological evaluation. Surgical resection of apical blebs, removal of pleural implants, and repair of diaphragmatic fenestrations was carried out as appropriate. In all patients either a mechanical pleurodesis or subtotal pleurectomy was performed. Clinical and demographic variables were identified using a prospectively maintained database. Statistical analysis was performed using GraphPad Prism software (GraphPad, San Diego, CA). Kaplan Meier curves were constructed using GraphPad Prism software and the recurrence rates between right and left pneumothorax compared using log-rank test. Patients were censored at the time of their last clinical follow up. Development of ipsilateral recurrence during the follow up was considered a positive event on the Kaplan Meier curves. The threshold for statistical significance was defined as α of 0.05.