Introduction and objectives Spontaneous Pneumothorax (SP) is a common pathology. Recurrence rates (RR) for Primary SP (PSP) are often quoted as approximately 30% (individual studies reporting anywhere between 17 and 49%), with less data available on Secondary SP (SSP) recurrence rates. Recurrence prevention at first episode remains controversial. International guidelines suggest pleurodesis for non-resolving air leak or recurrence prevention at second episode. There are numerous candidate agents for chemical pleurodesis.

This study aimed to comprehensively review the existing literature regarding chemical pleurodesis as a treatment modality.

Methods The systematic review methodology was based on the PRISMA approach and principles. Literature searches of multiple databases (PubMed, Embase, Medline, Web of Science, Cochrane Library) used combinations of terms including “spontaneous”, “pneumothor*”, “chemical”, “talc”, “tetracycline”, “minocycline”, “iodopovidine”, and “blood”. Abstracts were reviewed for relevance by two authors, who subsequently assessed and extracted data from the full articles.

Results Of 522 abstracts reviewed; 427 were excluded (e.g. case reports, letters, reviews, animal models or basic science articles); an additional 4 papers included via back-referencing. 99 full text papers were reviewed; 58 were excluded for the following reasons: foreign language only, not available, or inadequate data/follow-up information.

The remaining 41 papers’ data were extracted, showing variation in size, quality and type of studies. Eight randomised trials across differing patient groups (both medical and surgical) report markedly varying recurrence rates (Table 1). Six prospective series (n = 398) found thoracoscopic talc insufflation (RR: 3 to 7%) and tetracycline (9% via chest drain or poudrage) to be effective; with iodopovidone less so (13%). Of 27 retrospective case series (n = 4,990), seven were reasonable quality, finding good efficacy of adding talc or silver nitrate post-bullectomy (RR: 1 to 2%); better than minocycline or acromycin post-bullectomy (3 and 4%) or talc post-electrocoagulation (5%). The remaining 20 were poorer quality with high risk of bias, assessing 7 different agents.

Conclusions Numerous agents have been used for chemical pleurodesis for spontaneous pneumothorax. Chemical pleurodesis post-surgical treatment or via thoracoscopy appears most effective. Evidence for definitive success rates of each agent is limited by the small number of randomised and comparative trials.