FormalPara Key Summary Points

Why carry out this study?

There are no national or regional studies looking at pleural effusions related to patients with small cell lung cancer in the UK

Understanding this specific population could inform patient-centered care as well as inform future research

The study asked whether local patients with pleural effusions due to small cell lung cancer were similar to previous datasets in the literature

What was learned from this study?

Those presenting with a malignant pleural effusion had a poorer prognosis, probably signifying advanced disease and the presence of an effusion in our small cell lung cancer cohort seems higher

Larger prospective datasets are required to elucidate more about pleural effusions in small cell patients with lung cancer

Introduction

Lung cancer is the third commonest cancer in the UK (UK) and represents around 13% of all new cancer diagnoses [1]. Of these lung cancers, around 15% are estimated to be small cell lung cancer (SCLC) subtype [2]. Patients with SCLC have a poor prognosis: data from the National Lung Cancer Audit estimates a median survival of 6 months [3]. SCLC is part of the neuroendocrine group of lung cancer, which also comprises large cell neuroendocrine carcinomas and carcinoid tumours [4]. SCLC, with an incidence of 1–5 per 10,000, is classified as an orphan disease and has a strong correlation with smoking, with only 2% of SCLC occurring in never smokers [5, 6].

Patients with SCLC usually present with parenchymal lesions and extensive lymphadenopathy, with extra-pulmonary metastases seen in approximately 70% of cases at the time of diagnosis [6].

Malignant pleural effusions (MPE) arise when the resorptive capacity of the pleural space is overwhelmed. Pathogenesis may be directly attributable to tumor involvement and invasion of the pleural linings or infiltration of the lymphatic system, blocking drainage. MPE can also develop as a final pathway for inflammatory changes in response to tumor invasion with the resultant increase in vascular membrane permeability and pro-inflammatory cytokine production leading to influx of fluid in the pleural space [7, 8].

The prevalence and prognostic implications of MPE have previously been described in patients with non-small cell lung cancer compared to those with SCLC [9]. Shojaee et al. [10] found that prevalence of MPE was 11% in a population of patients with SCLC in the US. Furthermore, presence of MPE was an independent predictor of worse survival in this cohort (3 vs. 7 months). These findings were corroborated by Ryu et al. where the median survival of patients with SCLC with no, minimal or malignant pleural effusion was 11.2, 5.93 and 4.83 months.

We could not find a similar study for the UK and no local analysis has been conducted. Northumbria Healthcare NHS Foundation Trust is a large district general hospital in the North East of England with a well-established pleural center. The North East had the highest age-standardized cancer incidence rate for males and females at 717 and 603 per 100,000 people [11].

Our aim was thus to determine the characteristics of our local population of patients in a high volume center with SCLC including prevalence of MPE and survival.

Methods

Study Design

A retrospective cohort study design was used. Patients coded as having SCLC from January 2012 to September 2021 were identified from the Somerset Cancer register. Ethical approval was granted by the Northumbria Healthcare Information Governance department (Caldicott reference C3905).

Data Collection

Once identified, electronic patient records were searched to collate basic demographics including sex, age, performance status and date of death. All pathology reports were checked for confirmation as SCLC at the lung cancer multidisciplinary team meeting and patients with indeterminate pathology reports were excluded. Overall survival (OS) is defined as the time to death from date of first computed tomogram (CT) scan.

All CTs and chest x-rays were reviewed to identify patients with pleural effusions. Patients with radiologic pleural effusion on their first scan were categorized as “presentation with effusion.” Those who developed pleural effusions in subsequent images were labeled as “progressed with effusion” if there was evidence of disease progression (note that there was always evidence of disease progression with development of effusion). Patients with no radiologic evidence of effusion were labeled as “no effusion.” Patient records were also searched to identify which pleural effusions were sampled and characteristics including pH, lactate dehydrogenase (LDH), fluid and serum protein levels (to determine if effusions were transudates or exudates). Cytology reports were also reviewed to determine if effusions were cytology positive or negative. Notes were further reviewed in an attempt to elucidate effusion management by identifying chest drain and indwelling pleural catheter (IPC) insertion.

Data Analysis

Microsoft Excel 2019 was used for data analysis. Continuous variables were presented as mean (± range), median (± interquartile range [IQR]) when outliers were present and categorical variables as percentages where appropriate.

Results

Baseline Characteristics

Three hundred ninety patients with SCLC were identified (11% of all patients with lung cancer from January 2012 to September 2021) from the Somerset register. The mean age was 75 (range 13) years.

Baseline characteristics are presented in Table 1.

Table 1 Baseline characteristics of patients with SCLC, n = 390
Full size table

One hundred seventy-eight (46%) patients had an effusion at any point during their disease process, with 107 (28%) presenting with effusion. Overall median survival was 211 (IQR 296) days. Sixty-five (17%) patients had a World Health Organization Performance Status (PS) of 1 (median 0S 376 days), 160 (41%) a PS of 1 (median OS 258 days), 75 (19%) a PS of 2 (median OS 123 days), 42 (11%) a PS of 3 (median OS 47 days), and 14 (3%) a PS of 4 (median OS 12 days).

Pleural Effusion Characteristics

Of the 107 patients who presented with an effusion, 23 (21%) were sampled (Table 2); 71 patients progressed with an effusion, 5 of which were sampled.

Table 2 Characteristics of pleural effusion of patients presenting with MPE, n = 107
Full size table

Of the sampled effusions, ten (43%) were cytology positive, and all were exudative effusions (mean pH 7.51, LDH 399, fluid protein 37, serum protein 67). Most effusions were conservatively managed with eight (7%) patients treated with a chest drain and one (1%) with an IPC.

Pleural Effusion and Survival

Patients who presented with effusion had the worst OS with a median OS of 142 days (IQR 261) compared to 212 days (IQR 301) for patients who did not have an effusion and 327 days (IQR 368.5) for patients who progressed with effusion (Fig. 1).

Fig. 1
figure 1

Comparative histogram of overall survival (OS) of SCLC patients stratified by presence of MPE, n = 355

Full size image

Patients who presented with an effusion had worse levels of functioning than other SCLC patients, with a mean PS of 2 compared to 1 for those who either progressed with effusion or had no MPE (Fig. 2).

Fig. 2
figure 2

Comparative histogram of performance status (PS), stratified by presence of effusion

Full size image

Discussion

In our cohort, a pleural effusion was associated with SCLC, arising in 46% of patients during their disease process. Ten out of 23 patients with effusions and SCLC who had their effusions sampled had positive cytology and a true malignant pleural effusion. For the rest, we cannot be sure whether they have a para-malignant or malignant effusion; for the rest of this paper, we have considered the two together as malignant pleural effusion.

Of note, para-malignant pleural effusions are pleural effusions in patients with solid tumors without direct pleural involvement and no malignant cells in the pleural fluid. They are usually a result of lymphatic obstruction [12]. The prevalence of MPE in our cohort is greater than in previous estimates: Shojaee et al. found that 11% of patients with SCLC presented with MPE compared to 28% in our study [10].

Shojaee et al. looked at just under 70,000 patients in the US (mean age 68 years) and their study is the largest in the literature looking at pleural effusion in patients with SCLC. It was not possible for them to go into diagnostics and therapeutics because of how their data were collected. We surmise that our rates are higher as our population is slightly older, and our patients are presenting with more advanced disease and may also be a reflection of our use of CT scans to detect effusions. CT scans may identify minimal pleural effusions, those < 10 mm in thickness, which are not large enough to confer symptoms to patients but rather are identified on staging scans during the initial work-up of the disease [13]. These effusions are too small to be sampled, which explains why only 21% of patients who presented with effusions were sampled. Nevertheless, these minimal effusions have been demonstrated to be of prognostic significance: Ryu et al. found that presence of minimal pleural effusion was associated with a 45% increased risk of mortality compared to patients without effusion [14]. The increased prevalence of MPE within our cohort may also be reflective of high levels of socioeconomic deprivation in the North East of England, with 54.6% of households classified as deprived in the latest Census report (the highest proportion in England) [15]. Poor socioeconomic status has been linked to more advanced stage disease at presentation in patients with non-small cell lung cancer, as this group of patients is more likely to experience barriers to presentation [16, 17]. Delayed presentation could therefore account for a higher proportion of patients presenting with effusions. Another point to raise is that without knowing the exact cause of the effusions, which might have been due to any other concomitant co-morbidity (and thus treated, like heart failure for example), the effect of effusions on survival cannot be elucidated.

The effect of MPE on survival is unclear from these data. Although patients who presented with MPE had the worst OS (142 days), patients who progressed with effusions had better outcomes (OS 327 days) than those who never had an effusion during their disease process (OS 212 days). This may reflect the fact that patients who present with MPE are more likely to have more advanced stage disease and that those who progressed with an effusion are those who were well enough to receive anti-cancer treatment after their initial presentation without an MPE. The data are limited by the range of results and multiple outliers in the dataset as well as the fact that we did not look at treatment outcomes or staging as that data were not readily available on the current records. Previous studies have analyzed the effect of MPE on survival in patients with SCLC with Shojaee et al. finding MPE to be a predictor for worse survival [10].

Strengths and Limitations

Previous studies have assessed the prevalence of MPE in patients with SCLC; however, to our knowledge this is the first to examine this within the UK.

This study is limited by the descriptive analysis of survival data. This means that although correlations can be identified, we are precluded from drawing meaningful conclusions. Future work should focus on analysis after confounding for variables such as stage, age and treatment modalities. This would allow comment on whether MPE is an independent predictor of survival in this cohort. We have also purposefully not done any statistical analyses as the cohorts are small, multiple data points such as the exact cause of the effusions are not available, and it would thus not be correct to offer even inferential statistical values.

The study is also limited because of its retrospective nature, which poses issues such as incomplete or absent data. Future prospective studies should be conducted, potentially using the Lung Cancer Audit. Expanding the cohort outside of the North East would also improve the external validity of the study.

Conclusions

MPE is a common complication in patients with SCLC. Future prospective studies are imperative to further elucidate this trend.