Masuno 1991.

Methods	Multicentre RCT of LC9018 plus doxorubicin vs doxorubicin alone in MPE secondary to lung cancer (Japan) LC9018 is a biologic response modifier prepared from heat‐killed, freeze‐dried Lactobacillus casei YIT 9018
Participants	Inclusion: positive histology for primary lung cancer; unilateral pleural effusion; expected survival > 8 weeks; no treatment within four weeks; performance score 0‐3; no concurrent cancer; no severe hepatic/renal/bone marrow failure; age ≤ 75 Exclusion: previous intrapleural (IP) treatment with a biologic response modifier; pregnant women and women of child‐bearing potential; history of allergy 95 participants randomised
Interventions	Effusion completely drained. Both treatment arms received a maximum of two intrapleural doses, 1 week apart Control group: doxorubicin 40 mg in 20‐50 ml saline LC9018 group: as control group, then LC9018 0.2 mg in 20‐50 ml saline
Outcomes	Efficacy of effusion control at four weeks (defined as 'complete response' (CR) (negative cytologic findings with no re‐accumulation of fluid), 'partial response' (PR) (negative cytologic findings with asymptomatic minimal fluid accumulation, not requiring additional aspiration) or 'failure' (detectable intrapleural fluid even after tube drainage with no improvement or exacerbation on radiology compared with before treatment, or failure to confirm conversion to negative cytology)) Side effects Change in performance status
Notes	People with trapped lung excluded post randomisation For this review, CR and PR counted as pleurodesis success Not included in network meta‐analysis NB: doxorubicin is the generic name for Adriamycin
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Central telephone randomisation system
Allocation concealment (selection bias)	Low risk	Central telephone randomisation system
Blinding of participants and personnel (performance bias) All outcomes	Unclear risk	Not clear
Blinding of outcome assessment (detection bias) All outcomes	Low risk	"Blinded committee assessed data regarding safety and efficacy"
Incomplete outcome data (attrition bias) All outcomes	High risk	19/95 participants excluded from final analysis, for a variety of reasons, including five participants with protocol violations
Selective reporting (reporting bias)	Low risk	All stated outcomes reported
Other bias	Unclear risk	Primary outcome measure included CXR resolution and conversion to cytology negative effusion. Not clear from methodology whether some participants who were asymptomatic had effusion drained to evaluate cytology status and were then classified as 'failures'