Safety and Feasibility of Photodynamic Therapy for Percutaneous Image-guided Abdominopelvic Abscess Drainage: Phase 1 Trial

Abstract

Background

Standard-of-care abscess management includes image-guided percutaneous drainage and antibiotics; however, cure rates vary, and concern for antibiotic-resistant bacteria is growing. Photodynamic therapy (PDT), which uses light-activated dyes to generate cytotoxic reactive oxygen species, could complement the standard of care by sterilizing the abscess at the time of drainage.

Purpose

To evaluate safety and feasibility of PDT with methylene blue (hereafter, MB-PDT) at the time of percutaneous abscess drainage.

Materials and Methods

This prospective, open-label, dose-escalation, first-in-humans, registered phase 1 clinical study of MB-PDT included participants who underwent percutaneous abdominal or pelvic abscess drainage with CT or US guidance from January 2015 to March 2020 and September 2022 to September 2023. Following drainage, MB-PDT was performed with laser illumination at a fluence rate of 20 mW/cm², with fluence groups of 6, 12, 18, 24, 30, and 36 J/cm² (n = 3 each). The primary outcome was safety, indicated by absence of fat embolism, MB escape, abscess wall damage, and need for surgery to remove optical fibers. Preliminary efficacy end points included the time to drainage catheter removal, drainage catheter output volume, and clinical symptom and fever duration. Relationships between fluence and outcomes were analyzed with Spearman correlation and linear regression analyses, and ordinary one-way analysis of variance was used for group comparisons.

Results

MB-PDT was safe and feasible in all 18 participants (mean age, 60.1 years ± 18.3 [SD]; 10 female), with no negative safety outcomes observed for any participant. No study-related adverse events were encountered, and the procedure did not increase reported pain (P = .1). Clinical symptom and fever duration was shorter in participants receiving higher fluences (30 and 36 J/cm² vs 6 J/cm²) (P = .03). The presence of antibiotic-resistant bacteria was not predictive of clinical symptom and fever duration (β = 0.13, P = .37).

Conclusion

MB-PDT was a safe and feasible adjunct to image-guided percutaneous abscess drainage. Clinical measures indicated a dose-dependent response to PDT.

ClinicalTrials.gov registration no.: NCT02240498

© RSNA, 2024

Supplemental material is available for this article.

See also the editorial by Johnston and Goldberg in this issue.

Summary

Photodynamic therapy with methylene blue performed at the time of percutaneous CT- or US-guided abdominopelvic abscess drainage was safe and feasible, demonstrating a dose-dependent improvement in clinical response.

Key Results

• ■ In this phase 1 clinical trial of 18 participants, photodynamic therapy with methylene blue (MB-PDT) performed at the time of percutaneous CT- or US-guided abdominal or pelvic abscess drainage was safe and feasible in all participants, with no study-related adverse events.

• ■ Faster resolution of symptoms and fever in participants receiving higher fluences (30 and 36 J/cm² vs 6 J/cm²) (P = .03) is preliminary evidence of MB-PDT effectiveness in eradicating residual infection following drainage.

Introduction

Abscesses form through the interaction between microbes in acute infection and the host immune system, resulting in fever, nausea, and acute abdominal pain. For untreated abscesses, mortality ranges from 35% to 100% (1,2). Abscesses are routinely managed by image-guided percutaneous drainage and antibiotic administration. This is a common procedure in interventional radiology (3), with improved outcomes compared with open surgical drainage (4,5). Despite this, abscesses remain a major source of morbidity, mortality, and hospital stay (5). Treatment response can vary widely between patients, with reported response rates of 30%–90% (5–8) and a mean time to clinical improvement of 4.5–10.2 days (9,10). This can result in abscess recurrence and necessitate repeated drainage procedures and catheter replacement, increasing the hospital stay duration, costs, and patient discomfort, as well as prolonging antibiotic administration. This is of particular concern because antibiotic-resistant bacteria are present in abscesses (11,12), and incidence of these bacteria is expected to increase. Therefore, development of alternative treatment options that improve response rate and are effective against antibiotic-resistant bacteria is paramount for management of future patients.

Photodynamic therapy (PDT) is a promising localized treatment for infectious disease that is based on excitation of photosensitive drugs, known as photosensitizers, by visible light to photochemically generate reactive oxygen species via intersystem crossing and energy transfer to molecular oxygen (13). These reactive oxygen species lead to microbial cell death through oxidative damage to the cytoplasmic membrane. PDT is effective against antibiotic-resistant bacteria (14) and does not result in acquired resistance (13). PDT with the photosensitizer methylene blue (hereafter, MB-PDT) has been applied in clinical trials for endodontic infection (15), periodontal pockets (16), infected wounds (17), and onychomycosis and infected diabetic foot ulcers (18), although these reports have not led to regulatory approvals. Presurgical PDT significantly reduced surgical site infection in a large patient population (19), and MB-PDT was shown to be efficacious against bacteria cultured from human abscess aspirates (12) and strains typically found in human abscesses (20). Simulation studies have shown that a large proportion of patients with abscesses would be eligible for PDT (21).

On the basis of these encouraging results and the potential clinical benefit, a prospective open-label phase 1 clinical trial investigating MB-PDT at the time of image-guided percutaneous abscess drainage was initiated. We hypothesized that MB-PDT would be safe and not increase pain at all illumination durations, with some evidence for efficacy apparent for longer illumination. Thus, the aim of this study was to evaluate the safety and feasibility of MB-PDT at the time of percutaneous abscess drainage.

Materials and Methods

All study procedures were approved by the institutional review board, and participants provided written informed consent. Data collection and handling complied with the Health Insurance Portability and Accountability Act, and this study was registered with ClinicalTrials.gov (no. NCT02240498).

Study Design

For this prospective open-label phase 1 clinical trial, a dose escalation scheme was used, with the MB concentration and fluence rate held constant and the fluence escalated using a 3 + 3 study design (22). The illumination time was escalated from 5 to 30 minutes in 5-minute increments (n = 3 each). At the 20-mW/cm² fluence rate described herein, this corresponds to fluence groups of 6, 12, 18, 24, 30, and 36 J/cm². In the absence of dose-limiting toxicity, this would yield a total sample size of 18. A maximum of 21 participants was allowed to accommodate treatment of an additional group at a de-escalated dose if dose-limiting toxicities were encountered.

Participants

Participants were recruited from a convenience sample of patients scheduled for image-guided abdominal or pelvic percutaneous abscess drainage from January 2015 to March 2020 and September 2022 to September 2023. Full inclusion and exclusion criteria, including rationale, are included in Appendix S1. A case report from one of these participants was previously published (23).

Study Procedures

Preprocedural diagnostic CT was performed for all participants, with acquisition details included in Appendix S1. All participants received standard-of-care image-guided percutaneous drainage, under CT or US guidance, and concomitant antibiotics. Details on drainage, including image guidance and antibiotic administration, are included in Appendix S1. Aspirated abscess material was sent to the clinical microbiology laboratory for identification and antibiotic susceptibility testing. All study procedures were performed by a board-certified interventional radiologist (A.K.S., with 14 years of experience).

Using the same access used for drainage, 1 mg/mL of sterile MB (BPI Laboratories) was infused into the abscess to match the volume of material aspirated during drainage. Following 10 minutes of incubation, MB was aspirated and the cavity was flushed with sterile saline. The abscess was filled with sterile 1% lipid emulsion (Intralipid [Baxter Health Care] or Nutrilipid [B. Braun Medical]) to gently distend the cavity and homogenize the light dose through scattering.

A sterile optical fiber (Vari-Lase Platinum Bright Tip; Vascular Solutions) connected to a 665-nm laser (ML7710; Modulight) was advanced to the approximate abscess center. Laser power was set to achieve a 20-mW/cm² fluence rate based on abscess dimensions on preprocedural CT images. Light was delivered for 5–30 minutes, depending on the fluence group. The optical fiber was then withdrawn, lipid emulsion was aspirated, and the cavity was flushed with sterile saline. Administered and aspirated volumes of MB and lipid were recorded.

A complete blood count was collected on the day of PDT, before the procedure, and on days 1 and 2 after PDT. Participants rated their pain on a 10-point Likert scale before and after PDT. Optical spectroscopy measurements were performed before and after MB administration, with results reported elsewhere (24).

Outcome Evaluation

The primary outcome was safety, evaluated according to the absence of (a) fat embolism related to lipid emulsion, (b) MB escape with evidence of adverse reactions, (c) disruption of the abscess wall or damage to surrounding tissue, and (d) the need for surgery to remove broken optical fibers. The rationale for safety outcomes is described in Appendix S1.

Feasibility criteria for technical success included successful administration of MB and lipid emulsion, placement of the optical fiber, laser irradiation, and removal of the entire optical fiber.

Although the study was not designed or powered to evaluate efficacy, preliminary evaluation was inferred through secondary clinical end points. These included the time to drainage catheter removal, drainage catheter output volume, and time to clinical symptom and fever resolution. Resolution of clinical symptoms and fever was defined as the first day after intervention when the participant was afebrile, reported no pain or malaise, and was able to have a bowel movement.

Details on follow-up, adverse event monitoring, and independent safety monitoring are included in Appendix S1.

Statistical Analysis

Continuous and categorical variables are summarized as means ± SDs and proportions, respectively. Outcomes were compared using ordinary one-way analysis of variance with the Tukey test. White blood cell count was analyzed using the Friedman test with the Dunn test. Changes in pain were analyzed with the Wilcoxon test. Correlations were assessed with Spearman coefficients. Linear regression was used in analyses controlling for abscess volume. P < .05 was considered indicative of a statistically significant difference. All analyses were performed by one of the authors (T.M.B.) using SPSS Software (version 28.0; IBM) and Prism (version 6.07; GraphPad Software).

Because this was a phase 1 safety and feasibility study, hypothesized effect sizes for efficacy were not used to determine sample size. Instead, a 3 + 3 design was used as described earlier.

Results

Participant Characteristics

A total of 56 participants were screened (Fig 1). Participants were excluded for the following reasons: abscess diameter greater than 8 cm (n = 11); taking serotonergic psychiatric medications (n = 10); presence of multiple abscesses (n = 9); and being of a vulnerable population, patient refusal, poor kidney function, study team unavailability, and lack of safe pathway to abscess (n = 1 each) (Appendix S1). Three additional participants consented but were not treated. For two of these participants, the abscess resolved without intervention between diagnostic imaging and the scheduled drainage procedure. The other participant could not adhere to the standard-of-care procedure and thus did not undergo study procedures. These consented but untreated participants were replaced to maintain the desired sample size. Ultimately, 18 participants (mean age, 60.1 years ± 18.3 [SD]; 10 female) were treated, and their characteristics are summarized in Table 1.

Figure 1:

Flowchart shows inclusion and exclusion of participants. MB = methylene blue.

Table 1:

Participant Characteristics

The treated abscesses were in pelvic (n = 7); abdominal wall (n = 2); pericolonic (n = 2); and femoral, left lower quadrant, presacral, periappendiceal, perirectal, psoas muscle, and tubo-ovarian (n = 1 each) locations. The mean abscess volume was 36 mL ± 40 (range, 2–150 mL). Seven participants had reduced or absent bowel movement before intervention.

Primary Safety and Feasibility Outcome Evaluation

No study-related adverse or serious adverse events were encountered, and MB-PDT did not increase reported pain (mean pain score, 3.8 ± 3.2 before PDT and 2.7 ± 2.9 after PDT; P = .1). One participant developed gastric bleeding after the intervention; however, both the independent safety monitor and study team determined that this bleeding was due not to the procedure but to an unrelated perforated duodenal ulcer at a site remote from the treated abscess.

No negative safety outcomes related to MB or lipid emulsion were observed in any participant. The mean administered MB and lipid emulsion volumes across participants were 36 mL ± 40 (range, 2–150 mL), with the volumes of MB and lipid emulsion set equal for each individual participant. The mean amount of infused MB that was aspirated was 99% ± 3 (range, 90%–100%), and the mean amount of infused lipid emulsion aspirated was 97% ± 5 (range, 83%–100%). For the worst case, where 90% of MB was aspirated, the unaspirated MB corresponded to a maximum whole-body MB dose of 0.1 mg/kg. This is less than the 4 mg/kg threshold for potential adverse effects, so carbon monoxide oximetry or administration of ascorbic acid and/or riboflavin after PDT was not needed for any participant.

Criteria for technical success were met for all participants, indicating that MB-PDT is technically feasible. Depending on the treatment group, the addition of PDT to drainage increased the procedure time by approximately 20–45 minutes (10 minutes for MB incubation, 5–30 minutes for laser illumination, and approximately 5 minutes for laser calibration and MB and lipid dilution).

The treatment fiber was removed intact for all participants, and postprocedural imaging showed no evidence of abscess wall disruption (Fig 2). Because no study-related adverse events were encountered, no fluence values were repeated between groups and no dose de-escalation was needed under the 3 + 3 design.

Figure 2:

Representative images at image-guided percutaneous abscess drainage with photodynamic therapy plus methylene blue. (A–C) Axial CT images obtained before (A), during (B), and at the end of (C) the procedure in a 59-year-old female participant with a history of appendiceal cancer who had a pelvic abscess and received a fluence of 12 J/cm², and (D–F) axial CT images obtained before (D), during (E), and at the end of (F) the procedure in a 67-year-old male participant with a history of diverticulitis who had a pelvic abscess and received a fluence of 36 J/cm² show the position of the pelvic abscess (arrow in A and D), the distal tip of the needle used for access (arrow in B and E), the position of the placed pigtail catheter (dashed arrow in C and F), and the intact abscess wall (solid arrow in C and F). Preprocedural images (A, D) were obtained with iodinated contrast material, whereas the other images were not.

Secondary Clinical Outcome Evaluation

Of the 18 abscesses treated, an additional procedure (exchange of the initially placed drain) was needed for only one. This abscess was the first treated and received a fluence of 6 J/cm². Two abscesses, one in the 12 J/cm² group and one in the 24 J/cm² group, were not completely resolved at subsequent imaging 4 months and 1 month after PDT. However, these both resolved at the last follow-up without any additional procedures.

For the entire sample, the mean white blood cell count was reduced from before PDT to 1 day after PDT (before, 14 570 µL⁻¹ ± 5999; 1 day after, 11 750 µL⁻¹ ± 4777; P = .02) and from before PDT to 2 days after PDT (2 days after, 10 850 µL⁻¹ ± 4474; P < .001).

As shown in Figure 3A, the mean time to resolution of clinical symptoms and fever was negatively correlated with fluence (ρ = −0.66, P = .004). This difference among fluence groups was significant (6 J/cm², 17.0 days ± 2.6; 12 J/cm², 9.7 days ± 8.1; 18 J/cm², 8.0 days ± 5.6; 24 J/cm², 8.0 days ± 1.4; 30 J/cm², 2.3 days ± 0.6; 36 J/cm², 4.0 days ± 2.6; P = .03), with the 30 and 36 J/cm² groups showing pairwise improvement in symptom and fever resolution compared with the 6 J/cm² group. Controlling for abscess volume, this relationship between fluence and symptom and fever resolution persisted (β = −0.65, P = .002), whereas abscess volume was not a predictor (β = 0.31, P = .09), suggesting that fluence was the main determinant of symptom and fever resolution.

Figure 3:

Graphs show clinical outcomes after abscess drainage with photodynamic therapy according to fluence group, including the (A) time to resolution of clinical symptoms and fever, defined as the first day after intervention when the participant was afebrile, reported no pain or malaise, and was able to have a bowel movement; (B) time to drainage catheter removal; (C) daily catheter drainage volume; and (D) cumulative catheter drainage volume for time points greater than 1 day after the procedure. Horizontal lines indicate mean values across participants within each group, and error bars indicate SDs. * = P < .05, Tukey test.

As shown in Figure 3B, the mean time to drainage catheter removal after PDT showed a negative correlation with increasing fluence (ρ = −0.18), although there was no evidence of a difference among groups (6 J/cm², 13.3 days ± 10.4; 12 J/cm², 6.3 days ± 10.9; 18 J/cm², 4.0 days ± 3.5; 24 J/cm², 0 days ± 0; 30 J/cm², 5.0 days ± 2.6; 36 J/cm², 5.7 days ± 6.4; P = .37). Drainage catheter output volume was negatively correlated with the number of days after the drainage and PDT procedure (ρ = −0.13) (Fig 3C). The mean cumulative drainage catheter output volume from day 2 after PDT onward was negatively correlated with increasing fluence (ρ = −0.18) (Fig 3D), but there was no evidence of a difference among groups (6 J/cm², 255 mL ± 374; 12 J/cm², 20.0 mL ± 34.6; 18 J/cm², 31.7 mL ± 54.9; 24 J/cm², 0 mL ± 0; 30 J/cm², 21.7 days ± 18.9; 36 J/cm², 60.6 days ± 104; P = .43). Controlling for abscess volume, fluence was still not a predictor of the time to drain removal (β = −0.15, P = .36) or the cumulative drainage volume (β = −0.20, P = .38).

Microbial Species Isolated

Aspirated fluid was retained for identification and antibiotic susceptibility testing for all treated abscesses. The identified species are tabulated in Table 2. The most common species identified was Escherichia coli (five of 18 abscesses). Five abscesses contained antibiotic-resistant bacteria, with four of these containing multidrug-resistant organisms. Antibiotic resistance was not a predictor of clinical symptom and fever duration (β = 0.13, P = .37), cumulative drainage volume (β = −0.02, P = .91), or the time to drainage catheter removal (β = −0.11, P = .41). The dependence of clinical symptom and fever resolution on fluence was preserved (β = −0.92, P < .001), indicating that antibiotic resistance did not affect the apparent PDT dose response.

Table 2:

Identified Microbes, Number of Abscesses in Which Each Species Was Found, and Number Showing Resistance

Discussion

Standard-of-care abscess management includes image-guided percutaneous drainage and antibiotics; however, cure rates vary between patients and there is growing concern for antibiotic-resistant bacteria. Photodynamic therapy (PDT) could complement the standard of care by sterilizing the abscess at the time of drainage, but studies on the use of PDT to treat infections deep in the human body are lacking. In this phase 1 clinical trial including 18 participants, PDT with methylene blue (hereafter, MB-PDT) performed at the time of percutaneous CT- or US-guided abscess drainage was safe and feasible in all participants, with no study-related adverse events. A reduced mean duration of symptoms and fever in participants who received higher fluences (30 and 36 J/cm² vs 6 J/cm²) is preliminary evidence of MB-PDT effectiveness in eradicating residual infection following drainage (6 J/cm², 17.0 days ± 2.6; 12 J/cm², 9.7 days ± 8.1; 18 J/cm², 8.0 days ± 5.6; 24 J/cm², 8.0 days ± 1.4; 30 J/cm², 2.3 days ± 0.6; 36 J/cm², 4.0 days ± 2.6; P = .03).

The standard-of-care treatment for abscesses is image-guided percutaneous drainage. Although translation of this to routine practice has reduced morbidity and mortality (4,5), several problems persist. The main issue is the variable response rate. Reported clinical success rates range from 30% to 90% (5,7,8). Even for technically successful procedures, catheters can require frequent exchange and extended placement (7,8).

For these reasons, other adjuvant treatments have been investigated. The most widely investigated has been postprocedural flushing with fibrinolytic agents, such as urokinase. Clinical studies have shown apparent benefit using urokinase flushing (25,26), although a randomized clinical trial did not show significant differences between urokinase and saline flushing (27). Fibrinolytic agents have been associated with adverse reactions, including hypotension, hypersensitivity reaction, apnea, and bleeding (28,29). Urokinase treatment also requires flushing multiple times daily for days or weeks. In contrast, MB-PDT was performed at a single time point and no adverse events were observed.

The expected clinical benefit of MB-PDT is a reduction in or elimination of bacteria remaining in the cavity following drainage, whether as undrained fluid or bacterial biofilm on the abscess wall. We envision that the eventual use case for MB-PDT could be the elimination of drainage catheter placement, with MB-PDT and drainage performed as a single procedure without the need for drainage catheter placement. Moreover, MB-PDT could be particularly useful for abscesses containing antibiotic-resistant bacteria. Multiple groups have found antibiotic-resistant bacteria in abscesses (11,12), and the incidence of these bacteria is expected to increase. Even for bacteria that are responsive to antibiotics, drug penetration can prove challenging (30). Fortunately, MB-PDT is efficacious against antibiotic-resistant bacteria (31,32), including members of the ESKAPE (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species) class of pathogens (33) cultured from human abscess aspirates (12). These ESKAPE pathogens are of particular importance for development of alternative antimicrobial treatments because they display multiple drug resistance mechanisms and account for a large proportion of hospital- and community-acquired infections and deaths.

One of the main criteria excluding potential participants from enrollment was use of medications with potential MB interactions. Based on reports that interactions between MB and certain serotonergic psychiatric medications can be related to serotonin syndrome (34), the U.S. Food and Drug Administration issued a drug safety communication (35). Because this was, to our knowledge, the first human study investigating MB-PDT for this indication, we conservatively excluded potential participants taking these medications. However, these safety concerns are based on reports where high concentrations (1–8 mg/kg) of MB were delivered intravenously. In contrast, we delivered lower concentrations (0.02–1.5 mg/kg) directly to the cavity, followed by aspiration. A mean total of 99% of MB was recovered in all participants, with 83% of participants showing complete MB aspiration. The largest concentration of unaspirated MB was 0.1 mg/kg, well below concentrations at which drug interactions were reported. Exclusion criteria related to serotonergic psychiatric medications could potentially be eliminated in future clinical trials, greatly expanding the eligible patient population.

This study had several limitations. First, this was an open-label phase 1 study with a limited sample size, so the results may be biased by participant characteristics. Second, we excluded potential participants with abscess diameters greater than 8 cm or multiple abscesses; these participants may have responded differently. In future phase 2 studies, we will use a laser with higher maximum output power to allow for treatment of larger abscesses. Third, although we tracked surrogates for response to MB-PDT, direct efficacy data were not collected and the study was not adequately powered to determine efficacy. Future phase 2 studies will be designed to directly evaluate efficacy.

In summary, we demonstrated that photodynamic therapy with methylene blue (MB-PDT) at the time of image-guided percutaneous abscess drainage is safe and feasible. Furthermore, preliminary evidence indicates that MB-PDT may improve the rate of symptom and fever resolution. Therefore, we are planning a phase 2 clinical trial to assess the efficacy and quality-of-life benefits of adding MB-PDT at the time of drainage compared with drainage alone.