Extrathoracic Metastases in Pleural Mesothelioma

Abstract

Introduction

Guidelines recommend obtaining a computed tomography scan of the chest for the staging of pleural mesothelioma and for assessing response to treatment. Consensus is lacking regarding the necessity of serial imaging of distant extrathoracic sites. In this study, we determined the prevalence of extrathoracic metastases in patients with pleural mesothelioma.

Methods

We conducted a retrospective review of patients with pleural mesothelioma treated at Massachusetts General Hospital between 1999 and 2022 who were referred for extrathoracic imaging during their disease course. Imaging reports were reviewed to determine sites of metastasis and calculate the time to development of extrathoracic metastasis. Overall survival and prevalence of extrathoracic metastasis were compared for patients with epithelioid versus nonepithelioid mesothelioma.

Results

The study included 148 patients, 69 (47%) of whom had undergone cytoreductive surgery. Histologic types included epithelioid (n = 82, 55%), biphasic (n = 49, 33%), and sarcomatoid (n = 10, 7%) mesothelioma. The median overall survival for the cohort was 24.0 months, specifically 34.7 months and 16.7 months for patients with epithelioid and nonepithelioid tumors, respectively (p < 0.001). There were 65 (44%) patients who developed extrathoracic metastases, with a median time to extrathoracic metastasis of 11.5 months. The most common sites of involvement were extrathoracic nodes (22%), peritoneum (20%), bone (11%), and liver (11%). Of the 76 patients referred for brain imaging, seven (9%) had brain metastases. The frequency of extrathoracic metastasis was identical for epithelioid and nonepithelioid mesothelioma (44%). Overall survival was shorter for patients who developed extrathoracic metastases (hazard ratio 5.9, p < 0.001).

Conclusions

Patients with pleural mesothelioma often develop extrathoracic metastases, providing a rationale for routinely obtaining imaging that encompasses sites outside of the thoracic cavity.

Introduction

Accurate imaging is critical for optimization of therapeutic strategies for pleural mesothelioma. To standardize imaging practices, several groups have developed guidelines that account for the unique morphology and biology of this disease.1, 2, 3 These guidelines recommend computed tomography (CT) of the chest extending inferiorly to the costophrenic sulci as the primary imaging modality. Magnetic resonance imaging (MRI) and positron emission tomography (PET) are advised to evaluate for chest wall and mediastinal invasion, transdiaphragmatic extension, or metastatic disease when cytoreductive surgery is being considered. In addition to the above, there is consensus across guidelines with respect to obtaining a CT of the upper abdomen at initial diagnosis that encompasses the most inferior extent of the pleura and enables evaluation of the diaphragm and the upper peritoneum irrespective of candidacy for cytoreductive surgery.

Specifically, recent guidelines from the European Society of Medical Oncology, a joint guideline from the National Cancer Institute Steering Committeee-International Association for the Study of Lung Cancer-Mesothelioma Applied Research Foundation, and a separate guideline from the International Mesothelioma Interest Group all recommend contrast-enhanced CT scans of the thorax and upper abdomen as standard imaging at initial diagnosis of pleural mesothelioma.^1,2,4 The American Society of Clinical Oncology guidelines also advise that staging imaging include the chest and upper abdomen for all patients while noting that providers should consider contrast-enhanced imaging of the lower abdomen and pelvis if there are abnormalities on the initial imaging that raise concern for distal metastases.⁵

The basis of these imaging recommendations, specifically the emphasis on the thorax and upper abdomen, originates in the prevailing understanding of pleural mesothelioma as a locally aggressive disease that seldom disseminates to distant organs.⁶ However, autopsy series suggest that extrathoracic involvement of pleural mesothelioma may be underrecognized.⁷ In the current guidelines, recommendations regarding longitudinal extrathoracic imaging for patients without baseline extrathoracic disease are lacking. Here, we conducted a retrospective, single-institution study to determine the prevalence of and timing of development of abdominopelvic and intracranial metastases in patients with pleural mesothelioma.

Materials and Methods

Study Population

Consecutive patients (n = 197) with pleural mesothelioma treated at Massachusetts General Hospital as part of routine care between 1999 and 2022 were identified. The study cohort was narrowed to 148 patients after excluding 49 patients who did not have abdominal imaging during follow-up (Fig. 1). Patients who only had abdominal imaging at initial diagnosis were excluded. The assessment of sites of metastatic involvement was on the basis of a retrospective review of imaging reports from CT scans, PET scans, and MRIs. Radiographs and ultrasounds were not considered for this analysis. In addition to determining sites of metastases, medical records (including operative reports) were retrospectively reviewed to capture demographics, pathology findings, stage, and treatment histories. Data were updated as of January 31, 2023. The studies supporting this analysis were performed under a protocol approved by an institutional review board.

Figure 1.

Study population. The study population is presented in the schema. ∗There were 27 patients who only had CT of the head (n = 8 without contrast). CT, computed tomography.

Statistical Analysis

The time to development of extrathoracic metastasis was measured from the date of mesothelioma diagnosis, as determined by an official pathology report, to the date of documentation of extrathoracic metastasis on an official imaging report interpreted by a radiologist. Overall survival was measured from the date of mesothelioma diagnosis (as defined above) until death. Patients without a known date of death were censored at the time of the last follow-up. Fisher’s exact test was used to compare the frequency of extrathoracic metastases among histologic types. Overall survival was estimated using the Kaplan-Meier method and compared using the log-rank test. The development of extracranial, extrathoracic metastases was defined as a time-dependent effect in a Cox proportional hazards model of overall survival used to estimate the hazard ratio (HR). p Values were based on a two-sided hypothesis. Analyses were performed using Statistical Analysis System version 9.4.

Results

Clinical Characteristics

Between June 1999 and July 2022, 148 patients with pleural mesothelioma were referred for extrathoracic imaging inclusive of abdominal imaging (Table 1). The stage at diagnosis is presented in Table 1, Supplementary Table 1, and Supplementary Figure 1. Most patients were of male sex (n = 110, 74%) and the median age was 69 years (range: 27–92 y). Epithelioid and biphasic histologic types were observed in 82 (55%) and 49 (33%) patients, respectively. A minority (n = 10, 7%) of patients had sarcomatoid mesothelioma. In the remaining cases, the histologic type was not documented in the medical record or pathology report. At initial diagnosis, 109 patients (74%) had a PET scan as part of staging imaging. For 57 patients (39%), chest MRI was included among staging studies. CT scan was the sole imaging modality used for staging for 26 patients (18%).

Table 1.

Clinicopathologic Characteristics of the Pleural Mesothelioma Cohort

Characteristics	N = 148
Age at diagnosis (y)
Median	69
Range	27–92
Sex, n (%)
Male	110 (74)
Female	38 (26)
Ethnicity, n (%)
White	134 (91)
Hispanic	7 (5)
Black	1 (<1)
Asian	3 (2)
Other/Unknown	3 (2)
Histologic type, n (%)
Epithelioid	82 (55)
Biphasic	49 (33)
Sarcomatoid	10 (7)
No subtyping performed	7 (5)
Prior Surgery, n (%)
None	79 (53)
Extrapleural Pneumonectomy	20 (14)
Pleurectomy/Decortication	33 (22)
Partial pleurectomy or palliative debulking	16 (11)
Stage (AJCC version 8.0), n (%)a
I	64 (43)
II	21(14)
III	45 (30)
IV	17 (11)
Radiation incorporated in initial treatment strategy, n (%)
Yes	33 (22)
No	115 (78)
Previous Chemotherapy, n (%)
Platinum + Pemetrexed	104 (70)
Platinum + pemetrexed + surgery	60 (41)
≥2 lines of chemotherapy	50 (34)
Previous immunotherapy, n (%)
Ipilimumab/nivolumab	15 (10)
Pembrolizumab	21 (14)
Chemotherapy + durvalumab	5 (3)
Multiple regimens	6 (4)

AJCC, American Joint Committee on Cancer.

^aThe pathologic stage is used for 69 patients who underwent cytoreductive surgery whereas the clinical stage is presented for the remaining patients. The initial stage was not known for one patient as they were originally diagnosed in another country and records were incomplete. Stage I includes 26 stage IA and 38 stage IB. Stage III includes 12 stage IIIA and three stage IIIB.

A total of 69 patients (47%) had undergone cytoreductive surgery, including 33 patients (22%) who had pleurectomy plus decortication and 20 patients (14%) for whom extrapleural pneumonectomy was performed (Supplementary Fig. 1). An additional 16 patients (11%) underwent partial debulking of pleural (n = 14) or paraspinous tumors (n = 2). There were 56 patients (81%) who had a preoperative PET scan. Of the 13 patients who did not have PET imaging before surgery, eight (62%) had a preoperative chest MRI. Eleven patients were upstaged at surgery (i.e., more extensive disease was observed than identified by preoperative imaging), including two patients who did not have a preoperative PET. Radiation was received by 33 (22%) patients as part of the initial treatment strategy, including 30 (43%) of patients who received “adjuvant” radiation after surgery. A total of 104 patients (70%) were treated with platinum-pemetrexed chemotherapy, including 60 patients (87%) who underwent cytoreductive surgery. A total of 47 patients (32%) were treated with immunotherapy, including 21 patients who received pembrolizumab and 15 patients who were given ipilimumab-nivolumab. Five patients received chemotherapy in combination with durvalumab. Six patients received multiple immunotherapy regimens.

Clinical and Disease Outcomes

The median duration of follow-up from diagnosis of mesothelioma was 4.8 years. The median overall survival was 24.0 months (95% confidence interval [CI]: 18.6–30.5). Consistent with the current understanding of disease biology,^6,8 survival was longer for patients with epithelioid mesothelioma compared with those with biphasic and sarcomatoid tumors (median 34.7 versus 16.7 mo, p < 0.001).

Development of Extrathoracic Disease

Imaging was reviewed to identify patients who developed extrathoracic disease during follow-up (Fig. 1 and Table 2). For this analysis, direct abdominal extension of pleural disease and chest wall bone or soft tissue metastases were not considered an extrathoracic disease. A total of 65 patients (44%) developed extracranial and extrathoracic metastases. Their median time to development of extracranial, extrathoracic metastasis was 11.5 months (95% CI: 7.3–16.9). Most (n = 37 of 65, 57%) patients with extracranial extrathoracic disease had multiorgan involvement. Representative examples are presented in Figures 2A-E and 3A-D. The most common sites of extracranial and extrathoracic disease were abdominal or retroperitoneal nodes (n = 32, 22%), peritoneum/omental (n = 30, 20%), bone (n = 17, 11%), and liver (n = 16, 11%). Other documented sites of extrathoracic involvement included soft tissue (n = 9), abdominal wall (n = 4), pancreas (n = 3), adrenal gland (n = 3), spleen (n = 2), adnexa (n = 1), and axillary nodes (n = 1).

Table 2.

Sites of Extrathoracic Involvement in Patients With Pleural Mesothelioma

Site	N = 148 n (%)
Any	65 (44)
Multiplea	37 (25)
Abdominal or retroperitoneal nodes	32 (22)
Peritoneum, omental, ascites	30 (20)
Bone	17 (11)
Liver	16 (11)
Extrathoracic soft tissue	9 (6)
Brainb	7 (9)
Abdominal wall	4 (3)
Pancreas	3 (2)
Adrenal	3 (2)
Spleen	2 (1)
Adnexa	1 (<1)
Axillary nodes	1 (<1)

^aThe proportion of patients with involvement of multiple extracranial extrathoracic sites, inclusive of patients captured in rows below.

^bThe denominator for brain imaging is 76.

Figure 2.

Multisite extrathoracic involvement in a patient with biphasic mesothelioma. At diagnosis, the CT scan reveals right pleural effusion and pleural thickening (arrow, A) associated with low-level uptake on fused positron emission tomography and CT scan (B). At 3.5 years after diagnosis, the patient developed new liver metastases (arrow, C) and bone metastases (arrows, D and E), as captured on abdominal CT and sagittal views of spine magnetic resonance imaging. CT, computed tomography.

Figure 3.

Abdominal and brain involvement of epithelioid mesothelioma. At diagnosis, coronal views of the computed tomography (CT) scan revealed pleural thickening and nodularity (arrow, A) with high-level uptake on fused positron emission tomography and CT scan (B). Scans 1 year after diagnosis reveal new lung metastases (arrows, C), brain metastases (arrows, D), and multiple new extrathoracic nodal metastases. CT, computed tomography.

Brain imaging was performed for 76 patients (51%), including 49 patients who had a brain MRI and 27 patients who only had a CT scan of the head. Of the 27 patients who had a CT scan of the head, eight had suboptimal imaging without contrast. Brain metastases were detected in seven patients in the overall cohort (Figs. 1, 3D, and 4A-D), all of whom also had extrathoracic metastases. All histologic types were represented among patients with brain metastases (n = 3 epithelioid, n = 2 biphasic, n = 1 sarcomatoid, n = 1 without histologic typing).

Figure 4.

Brain metastasis in a patient with biphasic pleural mesothelioma. A biphasic mesothelioma exhibiting an epithelioid component (A) characterized by nests or solid sheets of epithelioid cells with abundant eosinophilic cytoplasm and centrally located, large nuclei with nucleoli, and a sarcomatoid component (B) characterized by a proliferation of pleomorphic polygonal or spindled cells. Abundant tumor-infiltrating lymphocytes are also noted. The tumor cells are immunoreactive to pan-keratin, keratin MNF116, and calretinin (C) and negative for claudin 4 (D) consistent with a mesothelial lineage.

Extrathoracic metastases were equally common among patients with nonepithelioid histology (n = 26/59, 44%) as those with epithelioid mesothelioma (n = 36 of 82, 44%) (p = 1.000). Overall survival was significantly shorter for patients who developed extracranial, extrathoracic metastases (HR = 5.9, p < 0.001).

Discussion

In this retrospective analysis of nearly 150 patients with pleural mesothelioma treated at a single institution over two decades, we identified extrathoracic metastases in 44% of patients across histologic types. In addition to well-described sites of metastasis such as the peritoneum, involvement of retroperitoneal/abdominal nodes, liver, and bone were each observed in greater than 10% of patients. Furthermore, a subgroup of patients developed brain metastases.

Our findings are consistent with previous studies that primarily assessed distant metastases in patients participating in clinical trials. For example, in a retrospective analysis of patients with pleural mesothelioma referred to a phase 1 clinical trials center, metastatic disease was identified in two-thirds of patients.⁹ In addition, in a phase 2 study of 96 patients who received intensified local therapy with preoperative high-dose hemithoracic intensity-modulated radiotherapy followed by extrapleural pneumonectomy, more than 60% of patients developed distant recurrence at 5 years, including peritoneal cavity, liver, bone, retroperitoneal node, and kidney metastases.¹⁰ Collectively, these observations from both clinical trial–eligible populations and our unselected group of patients with pleural mesothelioma challenge the notion of mesothelioma as a largely locally confined disease and support routine comprehensive imaging that includes longitudinal assessment for distant disease. Future studies are indicated to explore whether particular disease features and molecular alterations are predictive of future development of extrathoracic metastases.

Our study has several limitations. First, our findings reflect the practice of providers from a single institution and were derived from a retrospective chart review. Second, extrathoracic imaging was not performed at fixed intervals or consistently performed with each CT chest for many patients. Given the reliance on medical records and notes, we cannot discount the possibility that imaging was specifically performed in some cases to assess symptoms or with the aim of enhanced vigilance in a patient with otherwise high-risk features. Third, the brain imaging (i.e., CT head without contrast) was suboptimal for many patients. Fourth, survival outcomes were more favorable than historical cohorts and a high proportion of patients underwent cytoreductive surgery and received radiation, potentially limiting generalizability.³ As radiation is not standard-of-care and many patients are not eligible for cytoreductive surgery because of the extent of the disease or aggressive disease characteristics at presentation and considering that controversy remains with respect to the use of surgery for the treatment of pleural mesothelioma, our study findings may not be applicable to all-comers with pleural mesothelioma. Finally, we did not have biopsy confirmation of metastatic involvement of mesothelioma for most patients. As such, it is possible that some metastases attributed to underlying mesothelioma may have had other origins.

In summary, consistent with a large autopsy series,⁷ our study suggests that pleural mesothelioma often disseminates to extrathoracic organs. Future guidelines should consider these findings when formulating imaging recommendations.

CRediT Authorship Contribution Statement

Ibiayi Dagogo-Jack: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing-original draft, Writing-review and editing.

Beow Y. Yeap: Data curation, Formal analysis, Investigation, Methodology, Writing-original draft, Writing-review and editing.

Mari Mino-Kenudson: Data curation, Formal analysis, Investigation, Methodology, Writing-original draft, Writing-review and editing.

Subba R. Digumarthy: Data curation, formal analysis, investigation, methodology, writing-original draft, writing-review and editing.