Seeing through the MIST: A Pathway Toward Optimal Intervention for Pleural Infection

Community-acquired pneumonia accounts for more than 1.5 million hospitalizations each year in the United States (1). It is estimated that approximately 40% of hospitalized patients with bacterial pneumonia will develop a parapneumonic effusion (PPE) (2), and 10–20% of these will evolve into empyema (defined as pus in the pleural space or a positive Gram stain or culture). The incidence of PPE and empyema (PPE and E) has increased over the last several decades (3, 4), and the presence of PPE/E has been shown to be an independent predictor for mortality (5). Although definitions vary, a parapneumonic effusion is considered complicated when pleural fluid pH is <7.2, pleural fluid lactate dehydrogenase is >900 IU/L, or pleural fluid glucose is <72 mg/dl (6).

Since the time of Hippocrates, the mainstay of treatment of patients with complicated PPE/E has been drainage of pleural fluid (7). This can be performed by tube thoracostomy or surgery. The landmark second Multicenter Intrapleural Sepsis Trial (MIST-2) (8) showed that a combination treatment with intrapleural t-PA (tissue plasminogen activator) and DNase (subsequently referred to as intrapleural enzymatic therapy [IET]) results in significant improvement in pleural fluid drainage as well as a reduction in the frequency of surgical referral and duration of hospitalization. IET has revolutionized the approach to patients with complicated PPE/E, and the most recent British Thoracic Society Guidelines recommend IET if tube thoracostomy alone does not completely drain the effusion (6).

The protocol for MIST-2 was quite labor intensive, with instillation of t-PA, followed by a dwell time of up to 1 hour, then instillation of DNase with another 1-hour dwell time, repeated twice a day for 3 days (8). There have been several subsequent publications that now suggest concomitant administration is safe and effective, and this approach was recently recommended in a consensus statement (9) and the most recent British Thoracic Society Guidelines (6).

To date, however, there have only been a few small trials comparing surgery to more conservative management strategies. Wait and colleagues randomized 20 patients with complicated PPE/E to video-assisted thoracoscopic surgery (VATS) or chest tube with intrapleural streptokinase (10). VATS was the definitive procedure in 91% versus 44% in the streptokinase group and was also associated with fewer chest tube days, less time in the ICU, less time in the hospital, and lower overall costs. Not only was this a small trial, but the MIST-1 randomized trial showed that streptokinase was no better than placebo (11), making the results of the Wait trial less impressive. More recently, Wilshire and colleagues randomized 20 patients to IET versus surgical decortication and found no difference in the median duration of chest tube and a trend toward a lower hospital length of stay in the surgical group compared with the IET group (5 d vs. 11 d; P = 0.08) (12).

In this issue of the Journal, Bedawi and colleagues (pp. 1305–1315) present the results of MIST-3, a feasibility randomized controlled trial of early VATS versus IET (13). Unlike prior studies and case series, which were troubled by selection bias (i.e., only enrolling patients “fit for surgery”), the goal of MIST-3 was to randomize all patients to either standard of care (SOC, defined as intravenous antibiotics and tube thoracostomy), early IET, or early VATS. Patients were enrolled if they had a clinical presentation consistent with pleural infection, a pleural effusion with a chest tube in place, empyema, or a pleural fluid pH < 7.2 and evidence of residual pleural fluid collection or ongoing sepsis and/or systemic inflammation. The study was well designed, with a priori definitions of noncompliance with the assigned treatment arm. The primary outcome measure was the feasibility of randomizing patients into the three arms of the study, with secondary outcome measures including length of hospital stay, frequency of readmission, need for reintervention, and a visual analog score of pain and quality of life. Despite the limitations imposed by the coronavirus disease (COVID-19) pandemic, the team was able to screen 178 patients from eight centers across the United Kingdom from November 1, 2019 through July 30, 2021. Of the 178 patients screened, 110 met eligibility criteria, and 13 of these (11.8%) had good response to initial tube thoracoscopy and were excluded from randomization. Sixty of the remaining 97 patients were then randomized to the three treatment arms.

The authors should be congratulated on successfully completing this feasibility trial during the COVID-19 pandemic. Participant retention rate to hospital discharge was 100%, and the 2-month follow-up completion rate was 87.5%. Intention-to-treat analysis revealed treatment compliance was 47.6% in the SOC arm, 73.6% in the IET arm, and 50% in the VATS arm. Noncompliance in the SOC arm was due to the treating clinician’s concern that the patient required either IET or VATS. Noncompliance in the IET arm was primarily from patient intolerance due to pain, and noncompliance in the VATS arm was primarily due to clinical improvement (7/10) or the anesthetic risk being deemed too great (2/10). Of note, the median time for VATS was 3.5 days after enrollment.

On analysis of the secondary outcome measures, there was no significant difference in the median length of stay between groups: 9 days overall, 10 days in SOC, and 7 days in the IET and VATS groups (P = 0.70). Although there was no difference in the need for further pleural intervention between groups, there was a trend toward higher mortality at 12 months in the VATS group (20%) compared with 4.8% in the SOC arm and 5.3% in the IET arm (P = 0.19). Patients in the IED arm had significant improvements.

The major strength of the trial is showing that enrolling patients in a study comparing more conservative versus more aggressive therapy, where there is clinical equipoise, is possible. This is the only way we can improve the care of our patients—by actually knowing which treatment is best. Clearly, however, one treatment may be best for one subgroup of patients, whereas another may be better for a different subgroup. For example, patients with coagulopathy, or those who are receiving antiplatelet agents, may be better served by saline irrigation of the pleural space (14). The biggest limitation, perhaps, lies in the fact that SOC for patients with residual effusion after tube thoracoscopy may actually be IET, because of the wide implementation of the MIST-2 protocol. Should this be a two-arm study comparing IET to early VATS? This is the plan for the authors’ phase 3 study.

This feasibility trial will certainly impact the future care of patients with complicated PPE/E. While we await the results of the full trial, clinicians should not interpret these early results as favoring one approach over the other. We should continue to treat each patient on an individual basis, accounting for local resources and expertise. I look forward to the final results of MIST-3.