Background

Multiple pulmonary nodules remain a challenging prospect with regards to management. The differential diagnosis varies from benign entities to disseminated metastasis. Benign metastasising leiomyoma (BML), first described by Steiner et al. in 1939 [1], is a smooth muscle tumour of primary uterine origin, metastasising to a distant location within the body. With approximately 150 cases reported in the literature, there is limited knowledge surrounding this condition. The largest contribution of patients comes from two single-centre case series, published from continental Europe and North America [2, 3].

Extra-uterine metastases are commonly pulmonary; however, there have been numerous reports of BML in other organs [4]. The mode of dissemination remains uncertain; BML is more common in post-menopausal females who have undergone uterine surgery, suggesting surgery creates a haematogenous route for tumour spread [5, 6].

Microscopically, the tumour is formed of spindle cells showing no atypia, necrosis, mitosis or invasion [7]. However, uterine tissue located distant to the original location suggests BML is a benign tumour with malignant potential, a low-grade slow-growing leiomyosarcoma [7, 8]. It has also been suggested that BML may display similarities with endometriosis [7].

Diagnosis is through tissue biopsy, wherein microscopic histopathology shows smooth muscle proliferation within lung parenchyma, and immunohistochemistry identifies smooth muscle actin, desmin and caldesmin. The presence of oestrogen and progesterone receptors and the ability for the nodules to change in size confirm hormone-dependent growth [4, 9].

Presentation is often an incidental finding of ‘pulmonary nodule(s)’ on radiological imaging (Fig. 1), but patients can present with a wide variety of respiratory symptoms, prompting radiological imaging of the chest. Although rare, there are few known cases of a direct relationship between the nodules and subsequent symptoms due to close proximity with vital structures or respiratory failure [4].

Fig. 1
figure 1

Cross-sectional imaging demonstrating bilateral pulmonary nodules, later identified on histology as pulmonary benign metastasising leiomyoma

Full size image

Hysterectomy removes the primary source of uterine leiomyoma, with bilateral oophorectomy eliminating endogenous hormone source promoting nodule growth by surgical castration. Medical therapies such as gonadotropin-releasing hormone (GnRH) analogues, luteinizing hormone-releasing hormone (LHRH) agonists or aromatase inhibitors offer hormonal manipulation to reduce tumour size. For asymptomatic patients with no evidence of progression, treatment often involves ‘watchful waiting’ [4].

BML is an uncommon and often overlooked diagnosis. Our study aims to help clinicians recognise BML in the differential diagnosis of pulmonary nodules and describe the implications for these patients across specialties.

Methods

Six patients with a diagnosis of BML were identified using histopathological data from electronic records.

Retrospective data collection was undertaken at a single tertiary university hospital over a 17-year period (2003–2019). In line with local and national governance in the UK, the study was registered as a clinical audit. Additional clinical data was collected and analysed from patient case notes and imaging.

Results

Patient demographics

Our study identified six patients, all post-menopausal females who were diagnosed with pulmonary BML at a median age of 45 years (range 38 to 56). Four patients were Caucasian (67%) and two of Asian origin (33%).

Three patients had undergone hysterectomy for symptomatic uterine leiomyoma prior to diagnosis (50%), with the diagnosis of pulmonary BML prompting subsequent hysterectomy and oophorectomy in one patient (17%). Patient demographics and presentations are summarised in Table 1.

Table 1 Patient demographics and presentation
Full size table
Table 2 Description of pulmonary nodules on imaging
Full size table

Presentation

BML was an incidental finding in asymptomatic patients during routine imaging for two patients and in one patient with ureteric colic. Another three patients presented with respiratory symptoms, prompting plain imaging of the chest.

Radiological imaging

Five patients had multiple lesions, with bilateral distribution of spread (83%). The median number of lesions was 9 (range 2–12), with a median lesion size of 11mm (range 7 to 27). Description of the lung nodules on radiological imaging can be found in Table 2.

Pathway

Following the radiological diagnosis of pulmonary nodules, patients were placed on a pathway through an established multidisciplinary team meeting formed of respiratory physicians, thoracic surgeons, pathologists, radiologists and medical oncologists. Following individual case discussion, all patients underwent a pulmonary wedge resection as a diagnostic procedure.

Surgical procedure

Five procedures were right sided (83%), and one procedure was left sided (17%).

A posterolateral thoracotomy in the left 5th intercostal space was performed for the first patient, as the lesion in the left lower lobe was significantly larger than any of the other pulmonary nodules and the timing of the procedure (2005) was prior to an established VATS programme at our centre.

Video-assisted thoracoscopic surgery (VATS) on the right side was used as standard approach for the remaining five patients (83%).

Histopathology

Haematoxylin and eosin (H&E) stained positive for spindle cells in all six samples (100%) (Figs. 2 and 3). All six samples were positive for desmin and smooth muscle actin (100%), with three samples also positive for caldesmin (50%).

Fig. 2
figure 2

Low-power image (magnification ×12.5). Staining with haematoxylin and eosin. Well-circumscribed pulmonary BML nodule. Green ink marking the surgical resection margin

Full size image
Fig. 3
figure 3

Higher power image (magnification ×100). Staining with haematoxylin and eosin. Cellular spindle cell nodule with entrapped air spaces. Normal lung tissue to the left of the image

Full size image

Biopsies confirmed the presence of progesterone receptors in six patients (100%), oestrogen receptors in five patients (87%).

Follow-up

Median length of stay was 2.5 days (range 1 to 8). No patients experienced any postoperative complications with 100% survival at median follow-up at 59.5 months (range 29 to 166).

Follow-up imaging was performed for all of our cohort. The choice of imaging modality and duration was largely subjective and dependant on the patient’s lead clinician. For five patients, the nodules remained unchanged. One individual, who underwent a hysterectomy without oophorectomy, required adjuvant hormonal therapy with anastrozole to stabilise growth of pulmonary nodules in close proximity to hilar structures.

All patients remained under follow-up with interval imaging. As no established management protocol for patients with BML exists, three patients remain under the care of thoracic surgery, one patient under clinical oncology, and two under gynaecology oncology.

Discussion

Steiner et al. first described benign metastasising leiomyoma as a rare smooth muscle tumour of uterine origin found at a distant location in the body [1]. A fundamental pathological difference between a benign and malignant tumour is the ability to invade surrounding tissue [10], creating much uncertainty around the term ‘benign metastasising leiomyoma’.

The growth of BML lesions is driven by hormonal influence due to the presence of oestrogen and progesterone receptors on the tumour surface. This can influence the clinical course and presentation. Most commonly, BML affects post-menopausal females, and presentation varies from asymptomatic to life threatening compression of vital structures within the thorax. Immunohistochemistry of the biopsied tissue confirms smooth muscle tumour, with no malignant features and positive staining of oestrogen and progesterone receptors [4].

Once BML is diagnosed in asymptomatic patients, management is individualised and often involves watchful waiting and interval imaging. In symptomatic patients or in those with nodules increasing in size, surgical oophorectomy or medications such as GnRH analogues and anti-oestrogens can slow or halt growth completely [4].

Similar to the literature [6], all our study participants were post-menopausal females through natural ageing or induced surgically by hysterectomy. Four (66%) patients had a hysterectomy due to symptomatic uterine leiomyoma, with 50% of those having an additional oophorectomy. Whilst only 50% of our cohort were asymptomatic, those presenting with respiratory symptoms were due to co-existing pathology prompting imaging and not a causal relationship.

BML typically demonstrates variable uptake of fluorodeoxyglucose (FDG) intake on positron emission tomography (PET) [11, 12]; however, none of our patients underwent a PET scan prior to procedure, due to low suspicion of malignancy in the preoperative multidisciplinary team meeting. None of our patients had a PET scan postoperatively as suspicion of malignant transformation remained low, both clinically and radiologically at follow-up.

All of our patients had a wedge resection for diagnosis; the presence of oestrogen and progesterone receptors were found in 83% (5 patients) and 100% (6 patients) respectively. The spindle-shaped cells were all found to be positive for desmin and smooth muscle actin. Previous reports have noted low levels of Ki67 activity, a marker for cellular proliferation; however, this was not measured at our centre [13].

Follow-up imaging showed no morphological change or growth in follow-up of five patients; however, one patient required adjuvant hormonal treatment for nodule regression, due to persistent growth and proximity to structures. There was no consistent follow-up specialty or duration for our patients. Follow-up imaging and interval of imaging were dependant on the clinician in charge of the patient’s care.

In our sample of six patients, following successful tissue diagnosis, our patients with BML have a favourable clinical course and outcome with conservative, medical or surgical management. BML caused no complications in any of our patients at follow-up.

With limited knowledge surrounding BML in combination with an ambiguous definition, pathologists require knowledge to recognise a potential BML diagnosis as well as facilities and resources to initiate appropriate analysis of it. The prevalence of BML is unknown; it is questionable if BML is as rare as the literature suggests or if it is a case of under diagnosis.

Conclusions

BML is a rare entity usually found in women with a history of uterine leiomyoma and should be considered in the wide variety of differentials for pulmonary nodules.

There are currently no algorithms for the management of BML; however, collating multi-institutional data in prospective centralised databases may help to optimise these strategies.