Efficacy of sirolimus for the prevention of recurrent pneumothorax in patients with lymphangioleiomyomatosis: a case series

Lymphangioleiomyomatosis (LAM) is a rare, progressive, cystic lung disease that mostly affects women of child-bearing age and is characterized by abnormal smooth muscle cell proliferation. LAM is associated with a series of clinical manifestations such as dyspnea, recurrent pneumothorax, hemoptysis, chylous effusion, renal angiomyolipoma (AML), retroperitoneal masses, and respiratory failure [1]. Pneumothorax is a common manifestation of LAM. Previous studies have demonstrated that approximately 66% of patients with LAM may exhibit pneumothorax; importantly, 70% of these patients may experience recurrent ipsilateral or contralateral pneumothoraces [2, 3]. However, management strategies for recurrent spontaneous pneumothorax in patients with LAM remain controversial and inadequate. Sirolimus, a common mammalian target of rapamycin (mTOR) inhibitor, is considered the first effective drug for patients with LAM. According to the guidelines published in 2016 [4], sirolimus is recommended for the following types of patients with LAM: patients with moderately impaired lung function [a forced expiratory volume in 1 s (FEV1) of less than 70% predicted] or progressively declining lung function (decline rate for FEV1, over 90 ml/year) and patients with chylous effusion. However, to date, sirolimus has not been recommended for patients with LAM presenting with pneumothorax. Here, we describe the clinical course of five women with LAM who presented with recurrent pneumothorax that was successfully prevented by sirolimus treatment. We also discuss the effectiveness of sirolimus therapy and other therapeutic options for the prevention of pneumothorax recurrence in patients with LAM, in an attempt to promote the development of better treatment strategies for this patient population.

A 33-year-old female nonsmoker with a 4-month history of intermittent chest pain and dyspnea at rest, which recurred every 2 weeks, was admitted to our hospital at 31 weeks of gestation. Four months ago, she had been admitted after experiencing these symptoms for the first time. A chest radiograph at that time revealed left hydropneumothorax with 90% lung compression. The patient received closed chest tube drainage (CTD). However, left pneumothorax recurred during rest or minimal activity in the 20th, 25th, 28th, and 30th weeks of gestation. For every recurrent episode, she was admitted to a local hospital, where she received CTD and was discharged only after radiographic confirmation that the pneumothorax had completely resolved. At the current admission, arterial blood gas analysis indicated type I respiratory failure with a partial pressure of oxygen (PaO₂) of 51 mmHg. The patient was treated with supplemental oxygen and continuous CTD. Between the 31st and 32nd weeks of gestation, abdominal ultrasound revealed that the umbilical cord was twisted around the neck of the fetus. At 33rd weeks, the patient underwent a cesarean section and successfully delivered a baby with a low birth weight of 1720 g and normal Apgar scores. High-resolution computed tomography (HRCT) revealed small, thin-walled cystic lesions diffused throughout all lung fields. The serum level of vascular endothelial growth factor-D (VEGF-D) was 6608 pg/ml. The patient was diagnosed with LAM, and she began treatment at a dose of 2 mg/day from 28 days after delivery. At 18 months after treatment initiation, the patient’s exercise capacity and quality of life exhibited considerable improvement, and she was able to resume work. She was followed up for 3 years and had not experienced recurrent pneumothorax at the time of writing this report. She could perform all daily activities, including jogging, housekeeping, and routine work. The only sirolimus-associated adverse effect was mucositis, which gradually ameliorated and resolved with time during the course of sirolimus treatment. A follow-up pulmonary function test (PFT) revealed a forced vital capacity (FVC) of 2.20 l (75.3% predicted), FEV1 of 1.85 l (66.3% predicted), and an FEV1/FVC ratio of 84.1% (100% predicted). Moreover, she could cover a distance of 480 m in a 6-min walk test (6MWT). Her baby showed normal growth and remained healthy without breast milk.

In this report, we described the successful prevention of recurrent pneumothorax by sirolimus treatment in five women with LAM. The clinical characteristics of the five LAM patients with recurrent pnuemothoraces are described in Table 1, and the time course of pneumothorax recurrences is shown in Fig. 1. Spontaneous recurrent ipsilateral or contralateral pneumothorax during rest or minimal activity is one of the most common manifestations, accounting for two-thirds of LAM patients [5]. In a previous study, the majority of patients with LAM initially presented with unilateral pneumothorax, and only 4% initially presented with simultaneous bilateral pneumothoraces [3]. After the first episode of pneumothorax, conditions such as Birt–Hogg–Dubé syndrome, pulmonary Langerhans cell histiocytosis, pulmonary bullae, lymphoid interstitial pneumonia, Sjögren syndrome, and amyloidosis, all of which are characterized by diffuse, thin-walled cystic lesions in the lungs on HRCT, should be ruled out [6]. All five patients reported here developed pneumothorax before LAM was diagnosed, and chest pain and dyspnea were the most frequent symptoms, and two of the patients experienced type I respiratory failure. In addition, all patients experienced recurrent homolateral or contralateral pneumothorax, which resulted in a poor quality of life and repeated hospitalizations. The management of recurrent pneumothorax in patients with LAM has been controversial. The latest official guideline from the American Thoracic Society/Japanese Respiratory Society recommends that ipsilateral pleurodesis should be performed when patients with LAM experience their first episode of pneumothorax (conditional recommendation, very low confidence in the estimated effects) [7]. An observational study of 395 patients registered with the LAM Foundation [3] revealed that two-thirds of patients presenting with pneumothorax subjected to conservative therapy for the first episode experienced recurrent pneumothorax, with recurrence rates of 32% and 27% for patients who underwent surgical management and chemical pleurodesis, respectively, for the first episode. However, approximately 62% of patients with LAM and pneumothorax select oxygen supplementation or CTD for the first episode, while 60% select pleurodesis for the second episode [2]. All of our patients received conservative treatment such as supplemental oxygen and small-bore chest drain insertion for re-expansion of their lungs after the first episode of pneumothorax, with three of them receiving chemical pleurodesis, surgical pleurodesis, and/or bullectomy after frequent recurrences. However, all five patients continued to develop recurrent ipsilateral, contralateral, or bilateral pneumothoraces despite conservative or aggressive surgical treatment. The efficacy of conservative treatment and pleurodesis for the prevention of pneumothorax recurrence in patients with LAM remains unsatisfactory. Repeated pneumothorax negatively influences the life quality of patients with LAM and significantly increases their healthcare burden. Therefore, it is necessary for physicians to identify an effective drug that can prevent the relapse of pneumothorax and improve the quality of life.

The findings from the five cases reported here suggest that sirolimus is a promising and effective drug for the prevention of recurrent pneumothorax in patients with LAM. Pneumothorax did not reappear in any of our patients as long as the plasma sirolimus level remained at 3–10 ng/ml. In addition, improvements in the subjective quality of life and exercise capacity were observed during sirolimus treatment for all five patients. All patients were able to resume their daily activities and work during sirolimus therapy. However, the pneumothorax recurred when sirolimus therapy was discontinued or the plasma sirolimus level was very low. The side effects of sirolimus experienced by these patients included the common ones such as mucositis, irregular menstruation, and delayed wound healing. On the basis of our experience, we would suggest that physicians use supplemental oxygen and CTD to facilitate gas discharge and re-expansion of the lungs in patients with LAM who present with pneumothorax. If these conservative methods are not effective, chemical pleurodesis or surgical intervention should be used. Alternatively, physicians can consider pleurodesis as the first choice of treatment for lung re-expansion. Once the pneumothorax has completely resolved and surgical wounds have healed, sirolimus therapy can be initiated as soon as possible to prevent relapse. It should be noted that sirolimus cannot enhance the absorption of pneumothorax and cannot be used to achieve the remission of existing pneumothorax. The potential benefits of sirolimus treatment for patients with LAM and recurrent pneumothorax include an improvement in lung function and the quality of life, an increase in the exercise capacity, and a decrease in the healthcare burden. On conducting a search of PubMed, we found only one case report where pneumothorax in a patient with LAM was successfully treated with sirolimus [8].

We observed that the plasma sirolimus level was remarkably low (0.01 ng/ml) during coadministration of sirolimus at 1 mg/day and anti-tuberculosis therapy in case 4, whereas it ranged between 1 and 3 ng/ml when the patient was receiving sirolimus at 2 mg/day and concomitant anti-tuberculosis therapy. When the patient discontinued anti-tuberculosis therapy, the level of sirolimus was > 15 ng/ml, and obvious adverse effects occurred in the form of pain and swelling of the ankles and fingers. We found that previous findings have suggested a potential interaction between anti-tuberculosis drugs and sirolimus [9, 10]. Therefore, when physicians coadminister sirolimus and anti-tuberculosis drugs, they should individually increase the dose of sirolimus under close monitoring of the plasma levels. In our case series, no patient received talc pleurodesis, considering that previous studies have demonstrated that exposure to talc may increase the risk of lung cancer [11, 12]. Therefore, autologous blood or hypertonic glucose is used as a sclerosing agent for chemical pleurodesis in many hospitals in China because of the low medical risk associated with these agents.

The present case series has some limitations. Because all patients presented to us with pneumothorax, PFT findings prior to sirolimus therapy were unavailable. Therefore, we could not evaluate the efficacy of sirolimus for improving pulmonary function in these patients. Second, pregnancy may have played a role in pneumothorax development in case 1. Therefore, the possibility that pneumothorax may have stopped recurring without sirolimus treatment after pregnancy cannot be completely excluded. Third, we cannot negate the effectiveness of surgical pleurodesis in minimizing recurrences on the basis of our case series. In future, we aim to determine whether sirolimus therapy is more effective than surgery, which is an invasive treatment. Finally, we cannot eliminate bias caused by individual differences in effect of sirolimus.

Thus far, no controlled clinical studies have been conducted for investigation of the possible efficacy of sirolimus for the prevention of pneumothorax recurrence in patients with LAM. In our case series, all patients received supplemental oxygen support, repeated CTD, or surgical treatment for lung re-expansion under case of recurrent pneumothorax before sirolimus treatment, although these strategies proved dissatisfactory in preventing the relapse of pneumothorax. During treatment with sirolimus, however, no patient developed pneumothorax. In addition, they exhibited a significantly improved quality of life. The findings from our case series suggest a potential therapeutic strategy for the management of recurrent pneumothorax in patients with LAM. However, further studies are necessary to clarify our findings.