High tPA activity but absent PAI-1 and plasminogen function early in intrapleural lytic therapy

To the Editor

Intrapleural fibrinolytic therapy (IPFT) for complicated parapneumonic effusions and empyema (CPE/E) is inefficient resulting in average hospital lengths of stay of 14 days and costs the U.S. healthcare system more than 1 billion dollars annually.¹ The current dogma is that resistance to IPFT is related to elevated plasminogen activator inhibitor-1 (PAI-1), which inhibits tissue plasminogen activator (tPA).^2–4 However, only 5–15% of PAI-1 in the pleural space is active, and the relatively high dose of tPA administered for standard IPFT should overcome active PAI-1 in a single dose given its 1:1 stoichiometric inhibitory capacity.^3,5,6

Our study aimed to establish whether pleural fluid PAI-1 activity is suppressed by tPA administered during IPFT and to evaluate for free, active tPA (unbound to PAI-1) after IPFT.

Methods

26 adult inpatients with clinical diagnoses of CPE/E whose treating team had prescribed IPFT were included with Institutional Review Board approval (UNMC IRB #0101–24-FB, COMIRB #24–1369). One patient had bilateral empyema leading to inclusion of 27 CPE/E cases. Pleural fluid from the intrapleural catheter was collected into 3.2% sodium citrate anticoagulant, centrifuged, aliquoted, and stored at −80°C until assays. Samples were collected immediately before the patient’s first dose of IPFT (pre-IPFT), and within 1–180 minutes after unclamping of the catheter following completion of the one-hour dwell of their second (post-IPFTD1) and fourth dose (post-IPFTD2) of IPFT.

Pleural fluid PAI-1 activity, which specifically assesses the capacity of the pleural fluid to inhibit tPA, and tPA activity were measured using commercially available activity assays (Innovative Research, Novi, MI).

A functional fibrinolysis turbidity assay was performed by mixing pleural fluid 1:1 with plasma from a healthy human donor or plasminogen-depleted plasma with or without supplemental plasminogen. Thrombin and calcium were added to clot fibrinogen and absorbance (turbidity) measured at 405 nm wavelength as described previously.^5,7

Results

The median age was 64.5 years and 23.1% were female. All patients were treated at the discretion of the clinical treatment team, which included antibiotics, chest tube placement, and standard intrapleural dosing of tPA/DNAse (10 mg/5 mg). Median pleural fluid pH was 7.16 and nine patients (39.1%) had positive pleural

fluid cultures. Five patients (19.2%) required pleural surgery and four patients (15.4%) died while in the hospital. Additional pragmatic failure outcomes included 13(48%) additional ipsilateral chest tubes to treat residual effusion after initiation of IPFT, five (18.5%) with receipt of more than the standard six doses of IPFT, and seven (25.9%) with <30% improvement of effusion on chest imaging.

We performed PAI-1 activity assays on pre-IPFT and post-IPFTD1 pleural fluid. While pre-IPFT pleural fluid had modest PAI-1 activity, there was a complete loss of activity in all patients but one after a single day of treatment (pre-IPFT median 37.22 U/mL IQR [26.47–77.93], post-IPFTD1 median 0.13 U/mL [0.068–0.48], P <0.0001, Figure 1). The single sample with elevated PAI-activity after 1 day of therapy was reduced to <1 U/mL at post-IPFTD2. These findings confirm that the clinical doses of tPA used in IPFT overcome PAI-1 inhibition capacity within the first day of IPFT in nearly all patients.

Figure 1.

Pleural fluid PAI-1 activity and tPA activity from pre-lytic and post-lytic day 1 samples demonstrate PAI-1 activity is absent with an overwhelming excess of tPA activity after one day of standard intrapleural fibrinolytic therapy. PAI-1 activity of patients’ pleural fluid (N = 27) prior to receipt of intrapleural fibrinolytic therapy is present, but after a single day of IPFT the PAI-1 activity is absent in all but one patient. Conversely, endogenous tPA activity is low in pre-lytic pleural fluid, but tPA activity is markedly elevated after one day of IPFT that introduces exogenous tPA. Outcomes did not differ depending on pleural PAI-1 or tPA activity. Data are shown as individual values (dots) with the line and error bars at the median +/− interquartile range, with comparisons made between pre-lytic and post-lytic day 1 values using Wilcoxon sign-ranked test. Comparisons of outcomes were assessed using Kruskal-Wallis test with Dunn’s correction for multiple comparisons with all P-values >0.999 between outcomes at respective timepoints. IU = International Units, PAI-1 = plasminogen activator inhibitor-1, tPA = tissue plasminogen activator, IPFT = intrapleural fibrinolytic therapy, Failure = pleural surgery or in-hospital death, Failure*= additional chest tube placed, more than 6 doses of IPFT, or <30% improvement on chest imaging after IPFT.

Next, a functional tPA activity assay was performed on pleural fluid from pre-IPFT and post-IPFTD1 time points. While pre-IPFT pleural fluid had negligible endogenous tPA activity as expected (median 0.12 IU/mL [0.055–0.24]), post-IPFTD1 fluid had abundant tPA activity (median 8150 IU/mL [1503–12185]) (P <0.0001), which corresponded to an average 38000-fold ratio of tPA: PAI-1 activity (median 38315; [13631–113675]). For comparison, the peak plasma tPA activity for IV alteplase treatment of myocardial infarction or stroke is 1100–1900 IU/mL,⁸ indicating that pleural fluid active tPA concentrations markedly exceed therapeutic doses for other indications.

When stratified by outcomes, there were no differences in PAI-1 activity or tPA activity between groups of patients (Figure 1). This suggests that failure of standard dose IPFT is unrelated to intrapleural PAI-1 or tPA and may depend on other unmeasured factors.

To assess whether residual tPA in post-IPFTD1 pleural fluid could initiate fibrinolysis, turbidity assays were performed by mixing samples 1:1 with healthy plasma (a fresh source of fibrinogen and plasminogen) and initiating clotting via addition of thrombin and calcium. Pre-IPFT specimens exhibited no fibrinolytic activity, as expected, whereas post-IPFTD1 samples promoted rapid clot dissolution (Figure 2). When post-IPFTD1 fluid was mixed with plasminogen-deficient plasma, there was no lysis observed, but fibrinolysis was readily restored with addition of plasminogen. These findings confirm that residual active tPA in the pleural space after 2 doses of IPFT was sufficient to drive fibrinolysis in vitro under conditions where plasminogen was present, and with sufficient excess to overcome the additional fresh PAI-1 present in healthy plasma.

Figure 2.

Functional fibrinolytic potential measurements of pleural fluid demonstrate sufficient tPA is present in pleural fluid on post-IPFT day 1 to generate rapid fibrinolysis when plasminogen is present on turbidity plasma clot lysis assays. Pre-IPFT and post-IPFT day 1 pleural fluid samples from 27 total CPE/E sample sets were each mixed 1:1 with healthy 3.2% citrate anticoagulated plasma, which served a source of clottable fibrinogen and fresh plasminogen (and also provided additional PAI-1 for tPA to overcome). In additional experiments, plasminogen-depleted plasma with or without supplemental plasminogen was mixed 1:1 post-IPFT day 1 pleural fluid. These mixtures were then clotted with alpha-thrombin and calcium, and absorption (turbidity) was measured over time at 405 nm wavelength. While pre-lytic samples did not generate a fibrinolytic response, as expected, the post-lytic day 1 pleural fluid was able to generate a rapid fibrinolytic response, indicating abundant tPA activity was present in pleural fluid after a single day of IPFT that was able to overcome all PAI-1 present in the pleural fluid and additionally the PAI-1 present in healthy plasma. Post-IPFT day 1 fluid lost its ability to stimulate lysis when mixed with plasminogen-depleted plasma, which was restored with plasminogen supplementation. Absorbance was normalized to corresponding unclotted controls for each sample. Data are shown as a composite (mean +/− standard error of the mean) of all N = 27 samples from each respective timepoint. Abs, absorption; tPA, tissue plasminogen activator; IPFT, intrapleural fibrinolytic therapy.

While previous studies have reported elevated pleural PAI-1 at the onset of CPE/E,^2,3,8 our data unequivocally demonstrate that standard intrapleural dosing of tPA rapidly eliminates PAI-1 activity. This results in an abundance of free, active tPA in the pleural space after just one day of IPFT, yet fibrinolysis still remains inefficient. These findings raise a critical question.

Why does intrapleural fibrinolysis fail in CPE/E despite the presence of active tPA?

We previously identified a deficiency of intrapleural plasminogen in CPE/E in a pilot study, driven in part by neutrophil-mediated degradation,⁵ which depletes the essential substrate for tPA-mediated fibrinolysis. Our current turbidity assay using plasminogen-deficient plasma affirms this finding. Neutrophil elastase, which is elevated in CPE/E and is an independent predictor of poor outcome,⁹ degrades multiple fibrinolytic proteins and regulators including plasminogen and PAI-1.^7,10 This dual degradation not only impairs plasmin generation but also removes PAI-1’s inhibitory function, allowing free tPA to accumulate without driving effective fibrinolysis.

In summary, our findings fundamentally shift the understanding of the regulation of intrapleural fibrinolysis in CPE/E, demonstrating that intrapleural tPA remains abundant and active while PAI-1 activity is completely exhausted after just one day of IPFT. Notably, mixing pleural fluid with a plasminogen source restored pleural fluid fibrinolysis. These insights highlight that current standard tPA dosing may be unnecessarily high, and suggest that development of plasminogen-based therapeutics may enhance IPFT efficacy to improve outcomes for patients with CPE/E.

Supplementary Material

Supplementary material is available at American Journal of Respiratory and Critical Care Medicine online.