1 INTRODUCTION

Pleural mesothelioma (PM) is a rare but aggressive malignancy accounting for just 0.2% of all cancer diagnoses.¹ The causal link between asbestos exposure and PM is well established. Australia is known to have one of the highest rates of asbestos exposure and PM incidence in the world.² Despite introducing a ban on asbestos-containing materials from 2003 onwards, the incidence continues to rise due to the long latency period between exposure and development of PM, typically 30–50 years.^3-5

Historically, PM is classified into three broad histologic subtypes namely: epithelioid (60%–80%); sarcomatoid (10%), and biphasic (10%–15%).⁶ The epithelioid subtype is associated with a better prognosis compared to non-epithelioid histology. In Australia, age-adjusted PM survival rates have shown gradual improvements, particularly 1-year survival outcomes.⁵ Nevertheless, global overall survival (OS) outcomes remain poor with 6-, 12-, and 60-month survival rates of 55%, 33%, and 5%, respectively.⁷

The role of treatment modalities including surgery⁸ and radiotherapy remains controversial and should only be offered to a small subgroup of selected patients following multidisciplinary team consensus.⁹ In surgically resectable PM patients, the MARS2 trial showed that surgery (pleurectomy decortication) plus chemotherapy had inferior OS (hazard ratio [HR] 1.28, 95% confidence interval [CI] 1.02–1.60, p = 0.032), higher rates of adverse events, and poorer quality of life outcomes compared to chemotherapy alone.¹⁰ Thus, palliative systemic therapy is the treatment of choice in most patients. Historically, this was treatment with platinum chemotherapy (cisplatin or carboplatin) and pemetrexed, with or without bevacizumab.^{11, 12} However efficacy is limited, particularly for non-epithelioid subtypes. The landmark CheckMate-743 trial established the role of immunotherapy in treating PM.¹³ This trial compared dual immunotherapy with ipilimumab and nivolumab to platinum chemotherapy in the first-line setting. The study showed superior OS for patients treated with immunotherapy, with a median OS of 18.1 versus 14.1 months (HR 0.73, 95% CI: 0.61–0.80) for ipilimumab/nivolumab and chemotherapy respectively.¹³ The largest benefit was seen for those with non-epithelioid histology, who historically did poorly with chemotherapy. Following the publication of the trial in February 2021, the drug was approved for first-line use on the Australian pharmaceutical benefits scheme (PBS) in July 2021.

More recently, three ongoing phase III clinical trials are investigating first-line chemoimmunotherapy against chemotherapy alone in PM. These include the DREAM3R (NCT04334759) investigating the benefit of adding durvalumab to platinum/pemetrexed; BEAT-MESO (NCT03762018) adding atezolizumab to platinum/pemetrexed/bevacizumab and IND.227 (NCT02784171) with the addition of pembrolizumab to chemotherapy. Preliminary results from the IND.227 trial were presented at the American Society of Clinical Oncology 2023 annual meeting which demonstrated improvements in OS, progression-free survival (PFS), and overall response rate with the addition of pembrolizumab. Similar to immunotherapy doublet, the benefit of adding immunotherapy was greatest in the non-epithelioid subtype. Patient selection and biomarkers for benefit from chemo-immunotherapy combination versus chemotherapy or immunotherapy alone remains unclear.

With the recent evolution in the treatment landscape of PM, we sought to evaluate real-world outcomes in patients with PM in an Australian context and understand the impact of immunotherapy treatment regimens on outcomes.

2 MATERIALS AND METHODS

We conducted a multicentre retrospective cohort study of patients with pleural mesothelioma (PM) diagnosed within the Sunshine Coast and Metro North Hospital and Health Services in Queensland, Australia (including three tertiary hospitals). Patients with a histologically confirmed diagnosis between July 2017 and June 2023 were eligible. Those with mesothelioma of non-pleural origin, without adequate follow-up, or enrolled in clinical trials were excluded.

Cases were identified from the statewide cancer repository and systemic therapy prescribing platforms. Data were extracted through a retrospective review of medical records by investigators. Baseline characteristics were recorded including demographics (age, sex, and Eastern Group Cooperative Group [ECOG] performance status) and clinical characteristics (histology and staging as per the American Joint Committee on Cancer 8th Edition). Treatment details were collected including surgery, radiotherapy, and systemic treatment (type and number of treatment cycles). Clinical outcomes were recorded including dates of progression and date of death.

The outcomes of interest in this study were OS, PFS, and time to progression (TTP) as defined below. Treatment toxicity data was not recorded as it was unable to be reliably extracted from a retrospective review of medical records.

This study was exempted by the Human Research Ethics Committee (Approval number: LNR/2020/QPCH/52968) and was conducted in accordance with the National Health and Medical Research Council guidelines.

3 RESULTS

3.2 Treatment patterns

Of the 90 patients, 57 (63%) patients received active treatment while 33 (37%) received best supportive care (BSC) alone. Chemotherapy with platinum/pemetrexed was given as first-line treatment in 36/57 (63%) patients. Platinum backbone was carboplatin in 61% or cisplatin in 39% of chemotherapy-treated patients respectively. First-line immunotherapy was received by 21/57 (37%) treated patients, with all receiving ipilimumab/nivolumab except for one who received pembrolizumab.

Table 2 summarises the first-line treatment choice by histological subtype and the first-line treatment choice before and after July 2021, when ipilimumab/nivolumab was reimbursed for first-line use on the PBS scheme. Chemotherapy was prescribed more commonly than immunotherapy for epithelioid and unspecified histology. In contrast, immunotherapy was more commonly given for non-epithelioid subtypes of sarcomatoid and biphasic histology. Prior to July 2021, chemotherapy was more commonly prescribed compared to immunotherapy, whereas immunotherapy was more commonly used after. This was shown across all histological subtypes, shown in Figure 1.

TABLE 2. First-line treatment^a by histology and before and after July 2021.^b

Histology/Treatment received—N = 90	Epithelioid (n = 49)	Biphasic (n = 15)	Sarcomatoid (n = 11)	Unspecified (n = 15)
Platinum/pemetrexed—n (%)	25 (51)	4 (27)	2 (18)	5 (33)
Immunotherapyb—n (%)	9 (18)	7 (47)	4 (36)	1 (7)
BSC—n (%)	15 (31)	4 (26)	5 (46)	9 (60)

Treatment received—N = 57	Before July 2021	After July 2021
Platinum/pemetrexed—n	33	3
Immunotherapyb—n	5	16

• Note: BSC = best supportive care.
• ^a First-line treatment between 2017 and 2023.
• ^b Ipilimumab/Nivolumab immunotherapy reimbursed for first-line use on the Australian Pharmaceutical Benefits (PBS) scheme from July 2021 onward.

Figure 2 summarizes the lines of treatment received by all included patients. Of the 57 treated patients with a documented progression event, 23 (40%) received second-line treatment while 34 (60%) received no further systemic therapy. Second-line treatments included chemotherapy in 14 patients (61%) (platinum/pemetrexed in 12 and vinorelbine in two patients) and immunotherapy in nine patients (predominantly single agent PD1-inhibitor).

Of those who received second-line treatment, 7/23 (30%) received third-line treatment. Only four patients received fourth-line treatment.

3.3 PFS and TTP

Median follow-up was 55.8 months in the chemotherapy arm and 25.7 months in the immunotherapy arm. PFS was analyzed in 56 patients who received any first-line treatment, with one patient excluded due to incomplete data as the patient was treated outside of the health service. The median PFS for all patients was 8.3 months (95% CI: 7.1–10.2). Landmark PFS at 12 and 24 months were 25% (95% CI: 15–40) and 4% (95% CI: 1–16) respectively.

There was no statistically significant difference in PFS between the chemotherapy and immunotherapy groups, with a median of 8.4 months (95% CI: 6.7–10.6) with chemotherapy compared to 7.8 months (95% CI 6.2–NR) with immunotherapy (HR 1.31, 95% CI: 0.69–2.40, p = 0.58) (Figure 3).

The estimated 12-month PFS for chemotherapy and immunotherapy groups were 28% (95% CI: 16–47) and 17% (95% CI: 5–55) respectively. (Figure 3A)

An additional four patients were excluded from the final TTP analysis due to not having subsequent imaging (N = 52). Median TTP was 8.8 months (95% CI: 6.7–13.6) with first-line chemotherapy compared to 7.8 months (95% CI: 7.2–NR) for immunotherapy. Estimated 12-month TTP for chemotherapy and immunotherapy were 30% (95% CI: 18–51) and 17% (95% CI: 5–59) respectively. (Figure 3B) There was no statistical difference between the two groups (HR 1.32, 95% CI: 0.68–2.55, p = 0.66).

4 DISCUSSION

In this retrospective study, we highlight the real-world clinical outcomes and treatment patterns of PM patients across 3 large tertiary referral centers in Queensland, Australia. This is the first study to examine patterns of care for PM in Queensland, Australia. Whilst previous studies within Australia predominantly focused on the outcomes and biomarkers for immunotherapy,^{14, 15} our study is the first to report outcomes of all treatments including chemotherapy, immunotherapy, and BSC within an unselected real-world cohort.

There was a lower representation of patients with epithelioid (54%) and higher non-epithelioid subtypes (29%) in our cohort compared to published historical epidemiological data and other recent real-world studies (epithelioid ranging from 66% to 75%).^{2, 16-18} A significant proportion of our patients (37%) did not receive any active treatment or BSC alone, which was higher compared to several recent real-world PM studies ranging between 21% and 31%. This may be attributed to the other studies having a younger median age between 68 and 70.^16-18 In our study, among those who received BSC alone, 27/33 patients were over 70 years old.

At our institutions, we enroll patients into any available clinical trials as the first-line option. Outside of clinical trials, the standard of care was rapidly evolving as new clinical evidence emerged. During the study period, we used either first-line chemotherapy or ipilimumab/nivolumab for epithelioid PM, considering patient factors such as comorbidities, anticipated tolerance to chemotherapy, social circumstances, and autoimmune conditions. For non-epithelioid PM, chemotherapy was used initially due to the lack of other options but once immunotherapy was funded, most investigators offered that regimen if patients were suitable.

Overall, first-line treatment was largely consistent with international guideline recommendations.¹⁹ Carboplatin was used more commonly (61%) in combination with pemetrexed in our patient cohort compared to other countries,^16-18 likely also reflecting the older population in our study. Prescribing patterns in Australia reflect the availability of treatment options at that time as immunotherapy was funded in Australia from July 2021. Since then, first-line ipilimumab/nivolumab has been more commonly used in all subtypes compared to chemotherapy, reflecting the published data from Checkmate 743.²⁰

We demonstrated that both chemotherapy and immunotherapy are active treatment options with similar outcomes to published data in our study. Nevertheless, the median OS for the immunotherapy arm in our study (16.1 months) was shorter compared to Checkmate 743 (18.1 months).²⁰ The relatively small sample size in our study did not allow for comparisons between treatment groups for epithelioid and non-epithelioid histology separately. Our study had an older population (median age of 75 years compared to 69 years for Checkmate 743) and a larger proportion of non-epithelioid subtypes, which is known to carry a more guarded prognosis.²¹ With chemotherapy being the most used first-line treatment option, immunotherapy has only been available since July 2021 and a relatively short median follow-up time of 25.7 months, the benefit of immunotherapy may not have been seen in our study. A similar finding was seen in the RIOMeso Australian retrospective study, which demonstrated no significant difference in median OS between epithelioid and non-epithelioid histologies in patients who received ipilimumab and nivolumab.¹⁴

Only one patient within the BSC arm received pleurectomy/decortication, indicating that most PM present with unresectable disease. With the results of the MARS2 trial showing no significant benefits from the addition of surgery to chemotherapy,¹⁰ the future role and utility of surgery will likely diminish in Queensland. On the other hand, there is limited good-quality evidence for the OS benefit of pleurodesis. One retrospective analysis of 49 clinical trials compared survival rates between those receiving talc pleurodesis versus surgery (extrapleural pneumonectomy and radical pleurectomy decortication) reported a median OS of 14 months for talc pleurodesis and 17–24 months for surgery. However, the studies had significant heterogeneity and the number of studies with talc pleurodesis was low.²² In our study, the higher percentage of patients who received talc pleurodesis in the immunotherapy compared to the chemotherapy arm may not have impacted outcomes.

There is no clear standard of care for second or subsequent-line treatments in PM. In our population, chemotherapy was the most used treatment in the second-line setting. One meta-analysis of 49 trials consisting of various second-line treatments including immunotherapy, chemotherapy (gemcitabine, vinorelbine, platinum/pemetrexed) and targeted therapies showed a median pooled PFS of 3.4 months (95% CI 2.87–3.93) and OS of 7.9 months (95% CI 7.01–8.72). There was a trend towards improved OS for immunotherapy compared to targeted and chemotherapy.²³ The CONFIRM trial randomized patients who progressed on first- or second-line platinum-based chemotherapy to nivolumab or placebo, whereby nivolumab demonstrated a statistically significant improvement in PFS (median 3.0 months vs. 1.8 months, HR 0.67, 95% CI 0.53–0.85, p = 0.0012) and OS (median 10.2 vs. 6.9 months, HR 0.69, 95% CI 0.52–0.91, p = 0.009) compared to placebo.²⁴ However, the PROMISE-meso trial failed to show a benefit of pembrolizumab compared to institutional choice single-agent chemotherapy (gemcitabine and vinorelbine) with similar PFS (HR 1.06, 95% CI 0.73–1.53, p = 0.76) and OS (HR 1.12, 95% CI 0.74–1.69, p = 0.59) despite adjusting for 63% cross-over from the chemotherapy arm.²⁵ With the low number of patients receiving subsequent treatments in our study, we did not perform a time-to-event analysis of these later-line treatments. There is a substantial drop off from first to second-line treatments, highlighting the need to optimize first-line treatment as many patients in a real-world setting receive no subsequent therapy.

There were several limitations in our study. This was a retrospective analysis with a relatively small sample size. There were patients who died without subsequent imaging which may have led to an overestimation of PFS. However, with the exclusion of these patients, the calculation of TTP yielded similar survival estimates, suggesting the lack of radiographic surveillance in some patients did not have a large influence on outcome estimates. Given the retrospective nature of the study, toxicity data was unable to be reliably extracted as there were missing or conflicting data surrounding the documentation of these outcomes. This was also limited by inter-clinician variability of toxicity grading, and it was difficult to determine causation between the toxicities and treatment received. Programmed-death ligand 1 (PD-L1) expression was not routinely tested within biopsy samples, therefore the association between PD-L1 expression and the efficacy of immunotherapy is unknown in our study.

Both chemotherapy and immunotherapy have a role to play in the treatment paradigm of PM. Based on the findings of this study, we recommend treatment be individualized based on patient factors. Both chemotherapy and immunotherapy have comparable median OS in this cohort. Chemotherapy or immunotherapy should be considered for patients with epithelioid histology but this space is rapidly evolving and we await the result of the recently closed DREAM3R study which should answer this question definitively. The choice of first-line treatment should be the one that will likely be best tolerated, given that many patients will not be suitable for later-line treatments. As we were unable to compare the outcomes between epithelioid and non-epithelioid subtypes, we would recommend immunotherapy as a first-line treatment based on published literature.

5 CONCLUSION

This study showed that the overall prognosis of patients with PM remains poor in the real world. Our study demonstrated that both chemotherapy and immunotherapy are active treatments for PM, but the considerable drop-off between first- and second-line treatments showed that better treatments are needed especially in the first-line setting. With the rapidly changing landscape of PM treatment with ongoing phase 3 trials of combination chemoimmunotherapy, further real-world studies along with identifying prognostic and predictive biomarkers will be crucial for patient selection. We encourage participation and enrolment in clinical trials to improve the treatment paradigm for this debilitating disease.

CONFLICT OF INTEREST STATEMENT

Associate Professor Brett Hughes holds a consulting or Advisory Role for Merck and Co., Bristol-Myers Squibb, Roche, Pfizer, Astra Zeneca, Eisai, and Sanofi/Regeneron, and received grant funding (institutional) from Amgen. Associate Professor Zarnie Lwin holds an honoraria or advisory role for Bristol-Myers Squib, Astra Zeneca, Roche, Pfizer, Takeda, Merck and Co., and Johnson and Johnson. Associate Professor Bryan Chan holds an honoraria or advisory role for Astra Zeneca, Roche, and Merck and Co. The other authors declare no conflict of interest.