Systematic Review of Lymphangioleiomyomatosis Outcomes in Pregnancy and a Proposed Management Guideline

Abstract

Objective

Lymphangioleiomyomatosis (LAM) is a rare, multisystem disease that primarily affects women of reproductive age. Disease progression has been linked to estrogen exposure, and as such many patients are advised to avoid pregnancy. Data are limited regarding the interaction between LAM and pregnancy, and as such we performed a systematic review to summarize available literature reporting outcomes of pregnancies complicated by maternal LAM.

Study Design

This was a systematic review including randomized controlled trials, observational studies, systematic reviews, case reports, clinical practice guidelines, and quality improvement studies with full-text manuscripts or abstracts in the English language with primary data on pregnant or postpartum patients with LAM. The primary outcome was maternal outcomes during pregnancy as well as pregnancy outcomes. Secondary outcomes were neonatal outcomes and long-term maternal outcomes. This search occurred in July 2020 and included MEDLINE, Scopus, clinicaltrials.gov, Embase, and Cochrane Central. Risk of bias was ascertained using the Newcastle–Ottawa Scale. Our systematic review was registered with PROSPERO as protocol number CRD 42020191402.

Results

A total of 175 publications were identified in our initial search; ultimately 31 studies were included. Six (19%) studies were retrospective cohort studies and 25 (81%) studies were case reports. Patients diagnosed during pregnancy had worse pregnancy outcomes compared to those diagnosed with LAM prior to pregnancy. Multiple studies reported a significant risk of pneumothoraces during pregnancy. Other significant risks included preterm delivery, chylothoraces, and pulmonary function deterioration. A proposed strategy for preconception counseling and antenatal management is provided.

Conclusion

Patients diagnosed with LAM during pregnancy generally experience worse outcomes including recurrent pneumothoraces and preterm delivery as compared to patients with a LAM diagnosis prior to pregnancy. Given that there are limited studies available, and that the majority are low-quality evidence and subject to bias, further investigation of the interaction between LAM and pregnancy is warranted to guide patient care and counseling.

Lymphangioleiomyomatosis (LAM) is a multisystem disease most commonly affecting women of childbearing age.^1–3 An estimated 3 to 5 per one million women are diagnosed with LAM at an average age of 35 years at diagnosis. Sporadic LAM is rare in the general population, but is more commonly seen in patients with tuberous sclerosis complex.⁴ Clinical manifestations of LAM vary based on affected organ systems. Patients most commonly experience pulmonary manifestations^5,6; however, LAM can affect other systems including lymphatic and renal systems.⁶

The pathophysiology of LAM is characterized by mutations in tumor suppression genes TSC1 and TSC2 leading to activation of the mTOR pathway and abnormal cellular proliferation.⁷ The location of abnormal cellular proliferation leads to the clinical sequelae of the disease including abdominal lymphatic masses (lymphangioleiomyomas), benign renal tumors (angiomyolipomas), or progressive destruction of the lung parenchyma with proliferation of thin-walled cystic lesions (pulmonary LAM).⁷ Pulmonary LAM is a progressive cystic lung disease associated with the infiltration of atypical smooth-muscle-like LAM cells.³ Proliferation of and enlargement of LAM cells leads to destruction of the pulmonary parenchyma, obstruction of airways and lymphatics, and ultimate loss of pulmonary function in patients affected by pulmonary LAM.⁸

LAM typically affects patients who are premenopausal and of childbearing age, and several hormonal factors are thought to play a role in the disease process.^9,10 LAM may be unmasked with estrogen exposure including pregnancy and estrogen replacement therapy.⁹ Many LAM cells express estrogen and progesterone receptors, supporting the theory that this disease is hormonally mediated.^11,12 Since disease progression or exacerbation has been associated with estrogen exposure, younger patients have generally been advised to avoid pregnancy.⁴ Patients diagnosed after menopause also demonstrate slower growth of pulmonary cysts and are known to have slower disease progression compared to premenopausal counterparts.

Data are limited regarding the effects of pregnancy on LAM, and the effects of LAM on pregnancy outcomes. In light of the limited available information, we performed a systematic review to summarize available literature reporting outcomes of pregnancies complicated by maternal LAM.

Materials and Methods

A protocol was developed following the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement¹³ and registered with PROSPERO as protocol number CRD 42020191402. Before initiation of study procedures, the protocol was submitted to the University of Alabama at Birmingham Institutional Review Board and it was deemed to be exempt from full board review (Fig. 1).

Fig. 1.

PRISMA flow diagram of included studies. PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-analyses.

The objective was to summarize the available literature reporting outcomes of pregnancies complicated by maternal LAM. The primary outcome included maternal outcomes during pregnancy and pregnancy outcomes. Secondary outcomes consisted of neonatal outcomes and long-term maternal outcomes. Eligibility criteria were tailored to include as much detail as possible given the paucity of data in the literature; therefore, we included randomized controlled trials, observational studies, systematic reviews, case reports, clinical practice guidelines, and quality improvement studies if there was a full-text manuscript or abstract in the English language that reported primary data on pregnant or postpartum patients with LAM. Studies were excluded if the full-text manuscript or abstract was unavailable in English.

With the assistance of an experienced health sciences librarian with expertise in systematic reviews (CHS), MED-LINE, Scopus, clinicaltrials.gov, Embase, and Cochrane Central were searched from inception through July 2020 using the terms pregnancy, pregnancy complications, pregnancy outcome, puerperal, maternal, or neonatal and lymphangioleiomyomatosis. All searches occurred on July 2, 2020 except the Embase search which occurred on July 6, 2020. The itemized search strategy is listed in its entirety in the Supplementary Appendix (available in the online version).

All sources initially identified by the exhaustive search of the databases were loaded into Covidence, an internet-based systematic review management system. All records were reviewed by two independent reviewers (M.L.C. and J.G.M.). Those appearing to meet inclusion criteria then moved to full-text review by three independent reviewers (M.L.C.,J.G.M.,and R.G.S.) who assessed their eligibility for inclusion in the final analysis. Any disputes during the study selection process were resolved through discussion with the senior author (R.G.S.).

Key elements of included studies were then abstracted by the lead author (M.L.C.) and confirmed by the senior author (R. G.S.). Items of interest included the study title, journal, year of publication, country where the majority of the data were collected, type of study, sample size, and the main results regarding LAM and pregnancy outcomes including maternal pulmonary status during pregnancy and the postpartum period, long-term maternal outcomes, pregnancy complications, gestational age at delivery, mode of delivery, neonatal outcomes, and neonatal intensive care unit admissions.

Risk of bias was evaluated by two separate reviewers (M.L. C., R.G.S.) who independently reviewed all included papers and assigned risk of bias according to the Newcastle–Ottawa Scale (Peterson, 2011 #176) for assessing the quality of manuscripts included in systematic reviews.

Results

A total of 175 publications were identified and screened on the basis of the predefined search strategy. Two of these were duplicates; therefore, 173 studies underwent title and abstract screening. Of all titles screened, 75 studies were excluded and ultimately 98 studies were assessed for full-text eligibility. Of these, 67 studies were excluded for multiple reasons. The majority were excluded as they did not report outcomes on pregnant patients with LAM. Thirty-one studies were included in the systematic review (Table 1).

Table 1.

Studies included in systematic review

Study (journal)	Year	Country	Publication details	Population/sample size (n =)	Main results
Clinical experience of lymphangioleiomyomatosis in the United Kingdom9 (Thorax)	2000	The United Kingdom	Retrospective cohort Risk of bias 5	Patients with LAM in the United Kingdom (n = 7)	The primary outcome was to determine the natural history of LAM in the UK compared to other centers. Pregnancy was uncommon (7 out of 50 patients) but associated with increased complications. One pregnancy was uncomplicated, one terminated, two with chylous pleural effusions, and three with pneumothoraces. Three required lung surgery during pregnancy. Overall, the incidence of complications was 11 times higher in pregnancy than those with no pregnancy (05% CI: 5.3–25.0, p < 0.001).
Pregnancy experiences among women with lymphangioleiomyomatosis4 (Respiratory Medicine)	2008	Canada	Retrospective cohort Risk of bias 4	Women registered with the U.S. LAM Foundation and U.K. LAM action (n = 328)	The primary outcome was to determine pregnancy and health outcomes in patients with LAM. Many women with LAM chose not to pursue pregnancy due to concerns about disease worsening. In total, 62% of women surveyed had successful pregnancies prior to diagnosis of LAM. Women diagnosed with LAM during pregnancy had the worst outcomes, with almost half of the pregnancies ending in preterm delivery.
Pregnancy and delivery after the establishment of the diagnosis of lymphangioleiomyomatosis15 (American Journal of Respiratory Critical Care Medicine)	2011	The United States	Retrospective cohort Risk of bias 5	Patients with LAM diagnosed prior to pregnancy (n = 8)	The primary outcome of this study was to evaluate the effect of pregnancy on the progression of LAM. Out of eight pregnancies there were seven episodes of pneumothorax during pregnancy and two postpartum (all had prior pneumothorax). One patient had progression of LAM at 1-year follow-up. Six had SVD. Two had CD.
Pregnancy does not always cause deterioration of lymphangioleiomyomatosis16 (Respirology)	2014	Japan	Retrospective cohort Risk of bias 6	Patients pregnant after diagnosis of LAM (n = 10)	The primary outcome of this study was to evaluate the effect of pregnancy on disease progression of LAM. No pregnancy loss. A total of 88% had pneumothorax recurrence. Pregnancy did not worsen FEV1. Serum VEGF-D levels decreased (768 vs. 1,850 pg/dL, p = 0.053).
Changes in lung function during pregnancy in patients with lymphangioleiomyomatosis14 (American Journal of Respiratory and Critical Care Medicine)	2017	The United States	Retrospective cohort Risk of bias 3	Patients with LAM (n = 6)	The primary outcome was to assess changes in lung function in LAM patients during pregnancy. Patients with normal FEV1 and DLCO may tolerate pregnancy without major complications. Patients with abnormal lung function may experience a profound decline in lung function.
Spontaneous pneumothoraces during pregnancy in patients with lymphangioleiomyomatosis17 (CHEST)	2019	The United States	Retrospective cohort Risk of bias 5	Adult women with LAM (n = 50)	The goal of this study was to characterize the risk and burden of spontaneous pneumothoraces during pregnancy in patients with LAM. Spontaneous pneumothorax occurs in 25–30% of patients with LAM during pregnancy, most commonly in the second trimester. This may be the presenting disease manifestation.
Pregnancy exacerbating unsuspected mediastinal lymphangioleiomyomatosis and chylothorax41 (International Journal of Gynecology and Obstetrics)	1996	Italy	Case report	26-year-old white female at 35 weeks of gestational age	Presented at 35 weeks of gestational age with dyspnea. Imaging consistent with bilateral chylothorax. Required chest tube placement and was delivered by cesarean. Diagnosed with LAM at postpartum follow-up.
Pulmonary lymphangioleiomyomatosis: prolonged survival despite multiple pregnancies and no hormonal intervention40 (Respirology)	1996	Israel	Case report	26-year-old P0	Patient presented at 26yoa with new pneumothorax which was treated with chest tube. Recurred 1 week later. Subsequently had four term deliveries without complication. She remained well until age 37 when she had recurrent left sided pneumothoraces.
A diagnosis of lymphangioleiomyomatosis in a pregnant woman presenting with a retroperitoneal mass24 (British journal of Obstetrics and Gynecology)	1999	The United Kingdom	Case report	30-year-old P0 at 10 weeks of gestational age	Presented at 10 weeks of gestational age with pain, fatigue, and malaise. Imaging showed a large retroperitoneal mass. Underwent XL which showed nonresectable mass. Histology consistent with lymphangioleiomyomatosis. Developed dyspnea postoperatively and underwent CT scan with pulmonary findings consistent with LAM. Required 2L O2 through pregnancy. Delivered via CS at 39 weeks of gestational age.
Pregnancy and lymphangioleiomyomatosis: anesthetic management25 (International journal of Obstetric Anesthesia)	2002	The United Kingdom	Case report	29-year-old P0 at 20 weeks of gestational age	Presented at 20 weeks of gestational age with known h/o LAM. Renal angiolipoma identified during pregnancy due to complaints of groin pain which was expectantly managed. Underwent uncomplicated VAVD at 39 weeks of gestational age and discharged on PPD2.
Pregnancy unmasking lymphangioleiomyomatosis39 (journal of the Association of Physicians in India)	2004	India	Case report	30-year-old 10 days postpartum	Patient initially developed respiratory symptoms prior to pregnancy which was presumed tuberculosis. Pregnancy was complicated by term fetal demise. Patient was readmitted postpartum with shortness of breath and was diagnosed with LAM based on CT scan findings confirmed with biopsy.
Conservative management of an unusual case of breathlessness during pregnancy18 (Respiratory Medicine Extra)	2006	The United Kingdom	Case report	37-year-old at 14 weeks of gestational age	Patient presented at 14 weeks of gestational age with worsening shortness of breath. Chest radiograph showed pneumothorax. She subsequently had another pneumothorax antepartum followed by a third pneumothorax postpartum and underwent VATS with biopsy consistent with lymphangioleiomyomatosis.
A pregnant woman with complications of lymphangioleiomyomatosis and idiopathic thrombocytopenia purpura29 (Internal Medicine)	2006	Japan	Case report	26-year-old P0 in second trimester	Presented in the second trimester with chest pain diagnosed with pneumothorax requiring chest tube. LAM diagnosed at this time based on CT findings. Pregnancy additionally complicated by ITP. Delivered at term via CD. Postpartum course complicated by recurrent pneumothorax.
Cesarean delivery under general anesthesia for a woman with undiagnosed tuberous sclerosis complex and lymphangioleiom-year-oldmatosis38 (Anaesthesia and Intensive Care)	2009	Australia	Case report	33-year-old P0 at 39 weeks of gestational age	Presented at 39 weeks of gestational age for scheduled CD due to placenta previa. SaO2 70% during induction of anesthesia. Extubation complicated by pulmonary hemorrhage. CT showed pulmonary findings consistent with LAM with bilateral renal angiolipomas. Diagnosed with tuberous sclerosis complex postpartum.
Conservative management of pulmonary lymphangioleiomyomatosis and tuberous sclerosis complicated by renal angiolipomas in pregnancy43 (The Journal of Maternal-Fetal and Neonatal Medicine)	2009	The United States	Case report	23-year-old with tuberous sclerosis at 26 weeks of gestational age	Presented at 26 weeks of gestational age with cough and progressive dyspnea. PMH remarkable for tuberous sclerosis and lymphangioleiomyomatosis. Treated for pneumonia. Remained inpatient until 34 weeks of gestational age at which time she underwent uncomplicated CD. Discharged home on POD4.
New diagnosis and subsequent postpartum improvement of lymphangioleiomyomatosis26 (Thorax)	2009	The United Kingdom	Case report	39-year-old P1 s/p primary CS	39-year-old P1 presented postpartum c/o dyspnea. CT showed cysts with thickened interlobar septa consistent with LAM. Follow-up scan 14 months later showed resolution.
Multiple pneumothoraces during second and third trimesters as first presentation of lymphangioleiomyomatosis37 (Obstetric Medicine)	2009	Scotland	Case report	30-year-old P1 at 19 weeks of gestational age	Presented with dyspnea at 19 weeks of gestational age found to have pneumothorax which was aspirated and she was discharged. Developed multiple recurrent pneumothoraces and underwent repeat CS at 34 weeks of gestational age. Overall needle thoracostomy 10 times and six intercostal drains throughout pregnancy.
Pregnancy complicated by lymphangioleiomyomatosis36 (Archives of Iranian Medicine)	2009	Iran	Case report	35-year-old P0100 presented at 8 weeks of gestational age	Presented to prenatal care at 8 weeks of gestational age with PMH LAM. Referred to pulmonology and followed by them throughout pregnancy. Uncomplicated primary CD at 37 weeks of gestational age.
Bilateral pneumothorax in pregnancy unmasking lymphangioleiomyomatosis35 (Irish Journal of Medical Science)	2011	Ireland	Case report	18-year-old P0 at 20 weeks of gestational age	Patient presented at 20 weeks of gestational age with worsening dyspnea. CXR showed left pneumothorax. She underwent drain placement complicated but persistent air leak. Bilateral thoracoscopy showed cystic lungs with histopathological exam consistent with LAM. She ultimately required a CD at 29.5 weeks of gestational age due to FGR with an uncomplicated PP course.
Successful pregnancy without disease progression in a LAM patient19 (CHEST)	2011	The United States	Case report	32-year-old P1	Presented prior to conception with dyspnea and diagnosed with LAM. Had a successful pregnancy with an uncomplicated antenatal course. Postpartum chest CT and PTFs overall unchanged.
A retroperitoneal mass, systemic lymphadenopathy, and pulmonary nodules in a pregnant woman30 (Skeletal Radiology)	2011	Japan	Case report	27-year-old at 18 weeks of gestational age	Presented at 18 weeks of gestational age with 1 week history of left supraclavicular lymphadenopathy. Biopsy consistent with malignancy. Underwent SVD of twins at 37 weeks of gestational age. Postpartum re-evaluation of mass showed lymphangioleiomyomatosis.
Lymphangioleiomyomatosis suspected to be a gynecologic disease31 (Journal of Obstetric and Gynaecology Research)	2011	Japan	Case report	30-year-old P0	Diagnosed with LAM at age 30 and became pregnant shortly after. Pregnancy complicated by pneumothorax at 36 weeks of gestational age. Delivered via cesarean section at 37 weeks of gestational age.
Successful pregnancy complicated by persistent pneumothorax in a patient with lymphangioleiomyomatosis on sirolimus34 (Sarcoidosis Vasculitis and Diffuse Lung Diseases)	2011	Germany	Case report	29-year-old P0	Patient presented at 6 weeks of gestational age after being diagnosed with LAM 10 months prior. Sirolimus was discontinued at the time of conformation of pregnancy. She had gradually increasing dyspnea and the decision was made to restart sirolimus at 23 weeks of gestational age when diffusion capacity declined. She was diagnosed with pneumothorax at 26.5 weeks of gestational age which was complicated by persistent leak s/p chest tube insertion. She ultimately underwent scheduled delivery at 32 weeks of gestational age in order to undergo thoracoscopy. Sirolimus was resumed postpartum and her lung function returned to prior to pregnancy state.
Exacerbation and remission of pulmonary micronodules with lymphangioleiomyomatosis around time of childbirth32 (Japanese journal of Radiology)	2013	Japan	Case report	33-year-old P0 at term	Patient diagnosed with TSC and LAM prior to pregnancy. Pregnancy was uncomplicated with term delivery. Postpartum repeat imaging showed improvement in imaging.
Renal angiomyolipomas and lymphangioleiom-year-oldmatosis in pregnancy27 (journal of Obstetrics and Gynaecology)	2013	The United Kingdom	Case report	23-year-old P0 at 31 weeks of gestational age	Patient presented at 34 weeks of gestational age with preeclampsia and worsening pruritis. Workup showed large renal angiolipoma. Underwent CD at 38 weeks of gestational age. Repeat CT scan was performed postoperatively which also showed pulmonary lymphangioleiomyomatosis. She ultimately required embolization of renal angiolipoma with nephrectomy.
Case of lymphangioleiomyomatosis discovered during cesarean section under spinal anesthesia33 (Masui)	2014	Japan	Case report	35-year-old P0 at 38 weeks of gestational age	Presented at 38 weeks of gestational age for scheduled primary CS 2/2 breech. Chest imaging obtained 2/2 hypoxia intraop showed findings consistent with LAM.
A curious case of acute respiratory distress syndrome20 (journal of Surgical Case Reports)	2015	The United States	Case report	37-year-old P0 at 21 weeks of gestational age	Presented at 21 weeks of gestational age with cough and progressive dyspnea. Subsequently decompensated requiring intubation. Initially found to have diffuse alveolar hemorrhage and influenza A infection. CT showed multiple cysts in the lungs concerning for LAM. Patient required ECMO and ultimately developed HELLP syndrome requiring cesarean delivery at 24 weeks of gestational age. Mother and neonate both discharged from the hospital in stable condition.
Lymphangioleiomyomatosis: an explosive presentation of a rare disease21 (Oxford Medical Case Reports)	2017	The United States	Case report	26-year-old in second trimester	Admitted during pregnancy for cough and hemoptysis. Workup revealed diagnosis of LAM. Underwent CD at 30 weeks of gestational age. Discharged home on supplemental O2.
No sirolimus? No problem: alternative management of lymphangioleiomyomatosis during pregnancy22 (CHEST)	2019	The United States	Case report	19-year-old P1 at 24 weeks of gestational age	Patient presented at 24 weeks of gestational age with known history of tuberous sclerosis complex presented with dyspnea. Chest radiograph showed pneumothorax requiring chest tube placement. Her antepartum course was complicated by two additional pneumothoraces and ultimately had a chest valve drain placed. She underwent an uncomplicated primary cesarean delivery at 38 weeks gestation.
Pregnancy unmasking symptoms of undiagnosed lymphangioleiomyomatosis: case report and review of literature23 (Respiratory Medicine Case Reports)	2019	The United States	Case report	32-year-old P2 3 months postpartum	Patient presented 12 weeks postpartum with worsening dyspnea. CT showed numerous small cystic spaces prompting VATS with histology resulting consistent with lymphangioleiomyomatosis. She was started on sirolimus with symptomatic improvement.
Postpartum bilateral tension pneumothoraces28 (Cureus)	2019	The United Kingdom	Case report	31-year-old P1 PPD5	Patient presented with obstructive shock and respiratory failure PPD5 s/p SVD. Found to have bilateral tension pneumothoraces and diagnosed with LAM on histopathology.

Abbreviations: CI, confidence interval; CD, caesarean delivery; CT, computed tomography; CXR, chest X-ray; DLCO, diffusing capacity of carbon monoxide; FEV1, forced expiratory volume in 1 second; FGR, fetal growth restriction; HELLP, hemolysis, elevated liver enzymes, low platelet count; ITP, Immune thrombocytopenic purpura; LAM, lymphangioleiomyomatosis; PMH, perinatal mental health; PP, postpartum; SVD, spontaneous vaginal delivery; TSC, tuberous sclerosis; VATS, video-assisted thoracoscopic surgery; VEGF, vascular endothelial growth factor.

Note: While overall our retrospective studies showed low risk of bias, the majority were case reports which inherently carry a risk of bias.

Of the 31 studies included, all were retrospective cohort studies or case reports. Six (19%) studies were retrospective cohort studies that included 6,¹⁴ 8,¹⁵ 10¹⁶, 50,¹⁷ and 328⁴ patients, respectively. The remaining 25 (81%) of studies were case reports that included one patient each. Of the 31 studies included, 9 (29%) were published in the United States,^{14,15,17–23} 7 (23%) in the United Kingdom,^9,18,24–28 6 (19%) in Japan,^16,29–33 and 1 from each of the following countries: Germany,³⁴ Ireland,³⁵ Iran,³⁶ Scotland,³⁷ Australia,³⁸ India,³⁹ Israel,⁴⁰ Italy,⁴¹ and Canada.⁴

Regarding risk of bias among the 7 cohort studies, 2 studies (33%) received <5 points, while 5 studies (67%) received ≥5 points, indicating the majority of cohort studies had a low risk of bias. However, given the majority of studies available were case reports, this information is certainly subject to bias.

The largest retrospective cohort study published by Cohen et al in 2008 included 328 women with LAM using a survey to evaluate their pregnancy outcomes.⁴ Of the 328 patients included, 123 (37.5%) had never been pregnant, 55% due to concerns that LAM may worsen their disease. Two hundred and five (62.5%) patients had been pregnant at least once. Most patients (n = 178, 54%) were diagnosed with LAM more than 1 year postpartum. Fifteen (5%) patients were diagnosed with LAM during pregnancy, and 12 (4%) conceived with a known diagnosis of LAM. Patients diagnosed with LAM during pregnancy had more premature births (53%, n = 8) compared to those diagnosed before pregnancy (20%, n = 3) or those diagnosed postpartum (15%, n = 50, p = 0.0001). LAM diagnosis antepartum was additionally associated with increased adverse maternal respiratory outcomes including pneumothorax or chylothorax.⁴

In a presentation at The American College of Chest Physicians (CHEST) annual meeting in 2019, Hyslop and Gupta evaluated the incidence of spontaneous pneumothoraces during pregnanciescomplicatedbyLAM.¹⁷ In survey data, 38 of 50 participants had been pregnant at least once with a reported 80 total pregnancies. Thirty-four percent of these patients were diagnosed with LAM during pregnancy. Eleven patients experienced at least one spontaneous pneumothorax (median two episodes per patient), with the majority of pneumothoraces occurring during the second trimester (59%).¹⁷

A retrospective cohort study published in the United States in 2011¹⁵ evaluated eight patients with a diagnosis of LAM prior to pregnancy and reported seven episodes of pneumothoraces out of eight patients. Each of these patients had a prior pneumothorax. A similar study from Japan in 2014¹⁶ evaluated pregnancy outcomes in 10 patients with LAM and reported that 88% experienced pneumothoraces; however, pregnancy was not associated with reduced forced expiratory volume (FEV₁) and the authors concluded that pregnancy did not worsen pulmonary function.¹⁶ Most recently, a retrospective cohort study published from the United States in 2017¹⁴ evaluated pulmonary function throughout pregnancy in six patients with known history of LAM. The authors concluded that patients with normal FEV₁ and diffusing capacity for carbon monoxide prior to pregnancy may tolerate pregnancy without major complications. However, patients with abnormal pulmonary function prior to pregnancy were more likely to experience adecline.¹⁴

Of the 28 case reports, 8 report cases of LAM were diagnosed prior to pregnancy, 9 cases were diagnosed antepartum, and 8 were diagnosed postpartum. Eight studies mention pneumothoraces, most of which were managed with chest tube placement, and 8 studies report preterm delivery. Of those that mentioned preterm birth, deliveries occurred between 24 and 34 weeks gestational age, with the majority occurring between 30 and 34 weeks gestational age. Differentiation between spontaneous versus indicated preterm birth was not apparent.

Discussion

In this systematic review, we summarize available literature evaluating outcomes of pregnancies complicated by LAM. All studies were retrospective cohort studies or case reports and subject to considerable bias. The largest study identified included 328 women with 205 pregnancies and observed that patients diagnosed during pregnancy had worse pregnancy outcomes than those diagnosed prior to pregnancy or postpartum, with a higher rate of preterm delivery as well as respiratory complications, primarily pneumothorax and chylothorax.⁴ Other studies reported a risk of further pulmonary function deterioration. Additionally, LAM appears to increase the risk of preterm delivery, though it is unclear if this is spontaneous preterm birth or an indicated preterm delivery due to maternal status.

The most common maternal complication described in the literature was pneumothorax. This was more commonly seen in patients with a history of pneumothorax, those diagnosed with LAM antepartum, specifically during the second trimester. Each pneumothorax carries significant immediate risks to the patient, including but not limited to pleural effusion, pulmonary hemorrhage, empyema, respiratory failure, cardiac arrhythmias, cardiac arrest, and sudden death. Pneumothoraces also individually carry a risk of recurrence, further lung damage, and sudden death.⁴² Entering pregnancy with abnormal lung function was also associated with a further decline in pulmonary status during pregnancy. The most common fetal complication was preterm delivery, ranging from 24 weeks gestationalageto34weeksgestationalage, with most preterm deliveries occurring between 30 and 34 weeks gestational age. Whether spontaneous preterm birth occurred or if the treating team deemed it safer to deliver a preterm fetus to preserve maternal status was not apparent in the available studies.

Our study has several strengths. Since LAM is a rare disease and data on pregnancy outcomes are limited, our study provides a unique opportunity to synthesize available data on this rare complication of pregnancy. Additionally, our database search was performed by an experienced medical librarian. Furthermore, our study was preregistered in PROS-PERO in order to reduce the risk of bias within our systematic review.

Our study is not without limitations. The majority of studies identified in our review are case reports which are low-level evidence. Although there were six retrospective cohort studies that included multiple patients, we did not identify any case-controlled prospective studies or trials available evaluating this condition. In addition, we only included papers in the English language. Studies published in other languages likely exist that would provide additional information on the effects of LAM and pregnancy. Additionally, there were not specific data available within our studies differentiating between spontaneous LAM, LAM as a complication of tuberous sclerosis complex, or other complications surrounding LAM (such as cardiac or renal comorbidities) which would affect pregnancy outcomes. Last, it was often unclear how the diagnosis of LAM was made, so it is possible that lung diseases other than LAM may be combined in the LAM cohorts which would affect outcomes.

Considering the available low-quality data subject to bias, a proposed management plan for patients with LAM considering pregnancy and for pregnant patients with LAM is presented in Table 2. Further, we propose a framework with which a provider may review common medications used in the treatment of LAM and management options during pregnancy (Table 3). We recommend that prior to conception patients discuss their pregnancy specific risks with specialists both from maternal–fetal medicine and pulmonology, specifically reviewing maternal risks of LAM in pregnancy including pneumothoraces and worsening pulmonary dysfunction. We recommend optimization of pulmonary function prior to pregnancy as entering pregnancy with reduced pulmonary function appears to be associated with poorer outcomes including pneumothoraces, chylothoraces, increased risk of further deterioration of pulmonary function, and preterm delivery. We additionally recommend optimization of other comorbidities that may exist concomitantly with LAM. If the patient and their pulmonologist desire to avoid or reduce estrogen exposure, the possibility of egg retrieval and use of a gestational carrier should be considered and discussed with reproductive endocrinology and infertility.

Table 2.

Management of lymphangioleiomyomatosis in pregnancy

Preconception counseling

• Assessment of pulmonary function  - Pulmonary function tests • Discussion of maternal risks Pneumothoraces  - Worsening pulmonary dysfunction  - Entering pregnancy with known LAM diagnosis and reduced lung function appears to be associated with poorer outcomes including pneumothoraces, chylothoraces, increased risk of further pulmonary function deterioration, and preterm delivery • Genetics  - Discussion regarding isolated LAM versus LAM as part of tuberous sclerosis complex. Low threshold for genetics referral since tuberous sclerosis complex is autosomal dominant. • Medication review  - In addition to reviewing safety of all mediations in pregnancy and recommendation for folic acid, medications the patient is on to manage LAM should be reviewed (Table 3) • Continued follow-up with pulmonology • Alternative family planning options  - Reproductive endocrinology referral if desired to discuss egg retrieval and gestational carrier to reduce estrogen exposure  - Adoption • Contraception counseling (avoidance of estrogen-containing contraceptives)

Pregnancy: co-management with pulmonology and maternal-fetal medicine

First trimester  Maternal • Baseline pulmonary function • Counsel on risks • Optimization of other medical comorbidities • Genetics discussion regarding possible tuberous sclerosis complex if not already performed Fetal • Early dating ultrasound Second trimester  Maternal • Pulmonary function assessment • Consultation with OB anesthesia  Fetal • Mid trimester targeted anatomy scan with cervical length screening • Offer prenatal diagnosis if maternal diagnosis of tuberous sclerosis complex Third trimester  Maternal • Timing of delivery per usual obstetric indications • Mode of delivery per usual obstetric indications (no outright indication for cesarean delivery)  Fetal • Growth and antenatal testing per usual indications

Obstetric triage evaluation

Patients with LAM who present with pulmonary complaints during pregnancy should undergo the following evaluation: • Pulse oximetry and vital sign assessment • Chest X-ray • Pulmonary consultation if worsening symptoms • Fetal assessment according to gestational age

Abbreviation: LAM, lymphangioleiomyomatosis; OB, obstetric.

Table 3.

Medications used in the management of lymphangioleiomyomatosis and management principles in pregnancy

Medication	Role in the management of LAM	Considerations for pregnancya
Sirolimus	Stabilizes lung function in patients with LAM by mTOR inhibition which inhibits T-lymphocytes44	Evidence is limited on use of sirolimus in pregnancy. It is not preferred during the first trimester due to increased risk of fetal loss and reduced weight in living offspring, though several reports of living children without birth defects have been reported.45–50 Resumption of sirolimus in the second trimester should be considered.
Everolimus	Stabilizes lung function in patients with LAM by mTOR inhibition which inhibits T and B cell activation51,52	Evidence is limited on use of everolimus in pregnancy. It is not preferred in the first trimester due to reports of cleft sternum and decreased fetal and neonatal viability in animal studies.53 Case reports in humans show no malformations that can be attributed to everolimus useage.54–59 Resumption of everolimus in the second trimester should be considered.
Albuterol	Provides symptomatic relief60,61	Albuterol is the preferred beta-agonist in pregnancy.62 There is no evidence of congenital anomalies or adverse perinatal outcomes attributable to albuterol therapy.63
Ipratropium (and other inhaled anticholinergics)	Provides symptomatic relief60,61	Ipratropium therapy during pregnancy does not increase the risk of congenital malformations based on animal studies.64,65

Abbreviation: LAM, lymphangioleiomyomatosis.

^aEach patient’s clinical course varies and management should be individualized. The benefits of some medications in pregnancy may outweigh risks depending on the patient’s disease severity.

Once the patient becomes pregnant, an early dating ultrasound is recommended as well as assessment of baseline pulmonary function in early pregnancy. Throughout pregnancy continued comanagement with maternal–fetal management and pulmonology is critical.

LAM is rarely encountered in pregnancy, but improvements in diagnosis and care in LAM foundation-based centers are leading to earlier diagnoses. Many women with LAM may choose to pursue pregnancy and parenthood after consultation with appropriate specialists. This systematic review revealed several areas for future research which are outlined in Table 4. Further, the heterogeneity of the available studies highlights an urgent need for consistent reporting of pregnancies complicated by LAM. Therefore, to adequately address key research questions, we include a list of recommended variables that should be reported when publishing LAM-related pregnancy manuscripts (Table 5). Including key variables will allow for more clinically meaningful synthesis of available data.

Table 4.

Suggested areas for future obstetric research in lymphangioleiomyomatosis

Preconception

• Does prepregnancy VEGF level offer prognostic information on pregnancy outcomes?

Assisted reproductive technology

• What are the effects of the estrogen concentrations in oocyte stimulation on LAM development and progression?

Pregnancy

• Does cervical length appropriately risk-stratify preterm birth risk in patients with LAM?

• Does LAM increase the risk of spontaneous preterm birth?

• Does a VEGF change from baseline predict pregnancy outcomes?

Long-term

• Does pregnancy reduce the lifespan of the patient with LAM compared to patients with LAM who do not carry pregnancies?

Abbreviations: LAM, lymphangioleiomyomatosis; VEGF, vascular endothelial growth factor.

Table 5.

Recommend variables to be included in obstetric-oriented LAM publications

LAM diagnosis

• Method of diagnosis (imaging, histopathologic, both)

• Timing of diagnosis (prior to pregnancy, during pregnancy, after pregnancy)

Pulmonary function

• Prior to pregnancy

• First, second, and third trimester lung function

• Postpartum lung function

• Lung function beyond 12 weeks postpartum

Conception

• Method of conception (spontaneous, artificial reproductive technology [ART])

• If ART, highest estradiol level during oocyte stimulation

• If in vitro fertilization, was gestational carrier utilized?

VEGF levels

• VEGF levels prepregnancy, each trimester of pregnancy, postpartum, and beyond 12 weeks postpartum

Pregnancy

• Mid-trimester cervical length

• Aspirin utilization

• Gestational age at delivery

• Pregnancy complications: spontaneous abortion, termination of pregnancy, hypertensive disorders of pregnancy, premature preterm rupture of membranes, spontaneous preterm birth, indicated preterm birth

• LAM complications: pneumothorax, pleural effusion, chest tube placement, decline in lung function respiratory failure

• Incidence of severe maternal morbidity and mortality (as defined by the Centers for Disease Control and Prevention; Obstetricians, 2016 no.: 179)

Neonatal outcome

Live birth, Apgars, neonatal intensive care unit admission, respiratory distress syndrome, neonatal death

Postpartum

• Postpartum contraception

Long term

• Lung function following pregnancy

• Need for lung transplantation

Abbreviations: LAM, lymphangioleiomyomatosis; VEGF, vascular endothelial growth factor.

In summary, this systematic review identified six cohort studies and multiple case studies reporting patients with LAM who were pregnant. The majority of these were low quality and were case report or retrospective cohort studies. Most relevant to clinicians are that data are limited. Some patients with LAM successfully carry a pregnancy, while others experience a high rate of maternal and perinatal morbidity. Overall, patients who were diagnosed with LAM during pregnancy had worse outcomes including recurrent pneumothoraces and preterm delivery. There are multiple case reports of successful pregnancies complicated by LAM without deterioration of pulmonary status; however, further investigation is warranted to aid in appropriate patient counseling and management. A multinational prospective registry would provide a unique opportunity to collect maternal and perinatal outcomes and to reduce the risk of bias associated with retrospective reviews.

Key Points.

• Data are limited on the effects of lymphangioleiomyomatosis on pregnancy outcomes.

• We performed a systematic review to summarize pregnancy outcomes complicated by LAM.

• Patients diagnosed with LAM during pregnancy experience worse outcomes.