Background
Post-cataract surgery endophthalmitis (PE) is a serious complication with guarded prognosis for useful visual acuity. The reported incidence rates after cataract surgery vary between countries, and range from 0.28 − 8 per 1000 individuals [
1‐
3] and 0.33 per 1000 individuals from 2006 to 2011 in China [
4]. The microorganisms involved in this disorder are usually the endogenous colonies that exist in the patients’ eyelids and conjunctiva. However, they could also have originated from unusual exogenous sources, e.g., contaminated surgical materials, ocular rinsing solutions, or a contaminated environment.
Staphylococcus epidermidis is the most common pathogen isolated in culture-proven endophthalmitis (60% of cases), with
Staphylococcus aureus,
Streptococcus spp., and
Enterococcus spp. each found in 5%–10% of the cases. Other Gram-negative bacteria account for approximately 6% of the cases [
5].
Stenotrophomonas maltophilia (
S. maltophilia) is an aerobic, nonfermentative, Gram-negative bacterium. Initially classified as
Pseudomonas maltophilia, it was once grouped in the genus
Xanthomonas[
6]. It is ubiquitous in aqueous environments, soil, and plants.
S. maltophilia has gained importance as a hospital pathogen because of its ability to colonize on plastic, glass, and Teflon [
7]. It can cause ocular infections such as keratitis, conjunctivitis, and scleritis. There have been a few sporadic cases of iatrogenic
S. maltophilia-induced PE since its first description in 1997 [
8‐
12]. Only two studies involved an outbreak in more than three patients [
13,
14], and only one study found the source of the infection, which was the rinsing solution [
14]. Currently, there have been no reports describing the symptoms of this bacterial infection. The purpose of this study was therefore to summarize the clinical features of PE caused by
S. maltophilia, and to investigate the risk factors of the outbreak. In addition, the potential source of contamination was also identified.
Methods
Patients
The patients were diagnosed and treated for endophthalmitis after receiving cataract surgery at the Eye & ENT Hospital, Fudan University, Shanghai, China, between December 10, 2010 and April 7, 2011. This study was designed to collect clinical data at the time of admission and during a minimum follow-up of 12 months. Patients without vitreous inflammation were excluded. Data collected included demographic information, eyes affected, date of surgery, surgical complications, type of intraocular lens implanted, operating room (OR) number and phaco machine, time between cataract surgery and diagnosis of endophthalmitis, site of culture, antibiotic sensitivity testing, treatment, and outcomes. All patients signed the informed consent form. The study protocol followed the tenets of the Declaration of Helsinki and was approved by the ethics committee of Eye & ENT Hospital, Fudan University (IRB number: KJ2010-41).
Preoperative details
Preoperatively, 0.3% ofloxacin eye drops were used three times a day for 3 days. Intraoperatively, povidone-iodine solutions were applied three times to conjunctiva, lids, eyelashes, and periorbital areas. Norvancomycin (4 mg) was used in the irrigation fluid (500 ml) to prevent infection by Gram-positive organisms. All surgical instruments and phaco tubes were sterilized using full-cycle steam sterilization. A new drainage bag or cassette of the phaco machine was used for multiple patients on the same day, but was discarded at the end of the day. All other instruments [e.g., syringes, viscoelastic agents, and IOL (intraocular lens) cartridges] were disposable. At the conclusion of surgery, Tobradex ointment was applied. Postoperatively, Tobradex and Pranoprofen eye drops were used to decrease inflammation.
Microbiological analysis
After noting endophthalmitis, sterilization procedures in the hospital were reviewed. To determine the origin of the S. maltophilia, multiple surveillance samples were collected from the air, from disinfectants, from the hands of surgeons, from the povidone-iodine solutions, from irrigation solutions, from viscoelastic agents, and from tubes and various instruments. At the beginning, only the irrigation tubes and other surveillance samples were tested and the results were negative. The fluids collected from aspiration tubes of phacoemulsifiers were not analyzed until a specialist in hospital hygiene was consulted on April 1, 2011. On April 7, a report showed that multidrug-resistant S. maltophilia was positive from the aspiration tube of one phacoemulsifier (Stellaris; Bausch & Lomb, Rochester, NY, USA). The planned surgeries using that machine were cancelled immediately on April 8, with no subsequent occurrence of endophthalmitis. Vitreous and aqueous fluids from tap or pars plana vitrectomy (PPV) were collected in the operating room using a sterile syringe before the administration of intravitreal antibiotics. Air contamination was measured with agar plates by the sedimentation method, using a 5-minute exposure of plates. Concentrations of airborne bacteria and fungi were expressed as colony-forming units per cubic meter (CFU/m3) by the formula CFU/m3 = 50000 N/(A*T), where N was the average CFU of plates, A was the plate area (cm2), and T was the exposure time (minutes). The collected fluids from different sources were inoculated on Columbia blood agar with 5% sheep erythrocytes (BioMerieux, Marcy l'Etoile, France), at 37°C, overnight. MicroScan autoScan-4 (Siemens Healthcare Diagnostics, Deerfield, IL, USA) was used for identification of all positive cultures. In accordance with the guidelines of the Clinical and Laboratory Standards Institute (CSLI), antibiotic susceptibility was tested by a broth dilution susceptibility test with NC 31 (Siemens Healthcare Diagnostics). The culture media were incubated for 7 days, and when negative was reported as “no growth”.
Treatments
The initial antibiotic treatment was topical, using fortified levofloxacin (5 mg/ml), tobramycin (3 mg/ml), intravenous ceftriaxone or ceftazidime (40 mg/kg), and intravitreal antibiotics (Norvancomycin 0.8 mg/0.1 ml and ceftazidime 2.25 mg/0.1 ml). Later, the antibiotics were changed on the basis of the antibiotic susceptibility test. PPV or IOL extraction was performed when the clinical course of infection led to a suspicion of impending blindness. Immediate PPV was first performed in patients who had light perception (LP) visual acuity, then later in patients with the visual acuity of hand motion (HM). IOLs were extracted from a diabetic patient with severe inflammation and a highly myopic patient with a negative power of the IOL.
Statistics
Statistical analysis included Fisher’s exact test and the Mann–Whitney U test. They were two-tailed, and significance was defined as P values less than 0.05. Statistical analysis was performed using the SPSS program for Windows (SPSS Institute, Chicago, IL, USA).
Discussion
The present study describes the S. maltophilia-induced PE of 14 patients who were infected within a period of 4 months at a single hospital because of a contaminated drainage cassette. Described in this study are the clinical characteristics, the effects of antibiotics and PPV, and the final visual acuities after treatments for S. maltophilia as the potential source of the infections.
In the culture-positive group, the median time from surgery to infection was 9.5 days (1–17 days). This time period was comparable with patients in a previous report [
13], whose symptoms began between postoperative days 1 and 19. However, in another report [
14], endophthalmitis by the same bacteria was fully developed in 26 patients within 2 days after surgery. Based upon the Endophthalmitis Vitrectomy Study (EVS), more virulent microorganisms (e.g., Gram-negative bacteria) were more likely to be diagnosed within 2 days of cataract surgery. However, in our study, as a Gram-negative bacterium,
S. maltophilia was less virulent, and the source of
S. maltophilia might affect the incubation time. In a previous report [
14], the continuous rinsing solution, which was contaminated, could have resulted in a higher level of bacterial inoculums, resulting in quicker and more pronounced inflammation, when compared with the contaminated aspiration solution that had been diluted by irrigation solution late in our study. In addition, in our study, the latent period was significantly longer for eyes that were culture-negative compared with those that were culture-positive. This was similar to another study [
17], where a shorter delayed onset was associated with a higher rate of bacterial infection.
The baseline clinical presentation of acute PE may be associated with the bacteria involved [
17,
18], which may assist clinicians to better medically and surgically manage patients before microbiologic identification is made. In our study, endophthalmitis induced by
S. maltophilia had special clinical characteristics. Pupil synechia was not seen in these patients. The liquidity of the aqueous humor was slightly influenced, and hypopyon was easy to deposit inferiorly. There was less fibrin formation and more cellular reaction in the anterior chamber. Fibrinolysin, one of the extracellular enzymes of
S. maltophilia, could have played a role in inhibiting the process of fibrin membrane formation. In addition, retinal periphlebitis was another early clinical sign of this infection. This was perhaps due to the toxins released by
S. maltophilia, such as lipases, hyaluronidase, protease, and elastase.
S. maltophilia shows intrinsic resistance to many currently used antibiotics, and therefore constitutes a special clinical challenge. Intravitreal injection of antibiotics combined with vitreous aspiration or PPV is considered standard treatment, and is usually initiated rapidly after diagnosis of PE.
S. maltophilia showed multidrug resistance, which included amino glycosides (e.g., gentamicin, tobramycin, and amikacin), most of the β-lactams (e.g., ampicillin, ticarcillin, cefazolin, cefotaxime, cefepime, ceftriaxone, and cefuroxime), and aztreonam, imipenem, and ciprofloxacin. Based upon our study, in the process of intravitreal injection of antibiotics, the following details should be considered. First, the sensitivity of the pathogen to the antibiotics (especially the intravitreal antibiotics) was the most important factor that determined the course of the infection. Patients did not respond well after initial intravitreal ceftazidime injection when the pathogen was resistant to this drug. Improvement was observed after a subsequent intravitreal injection of levofloxacin. Second,
S. maltophilia infections can be challenging because of contradictory findings between
in vitro and
in vivo antibiotic susceptibility studies. Two of the cases responded poorly to intravitreal ceftazidime, although the pathogens were reported to be sensitive to this drug. This finding is consistent with previous reports from other studies [
19]. Third, the antibiotic resistance was more severe than previously reported [
12]. In our study, the intravitreal-injected antibiotics such as amikacin and ceftazidime were not effective. In addition, the strain of
S. maltophilia was also resistant to SMZ/TMP and ciprofloxacin. However, it was sensitive to levofloxacin, another fluoroquinolone drug. Fourth, in our study, levofloxacin was applied systemically as well as locally. Although patients reported in the EVS derived no demonstrable benefit from these systemic antibiotics, the study made no recommendations regarding treatment with additional antimicrobial agents (e.g., systemic fluoroquinolones) [
20], so levofloxacin may ensure more predictable effects during the vulnerable period.
For the previous two patients with a visual acuity of HM, initial intravitreal injections were applied, but the infection could not be resolved. Therefore, we performed initial PPV on the remaining six patients, when their visual acuities were HM. This treatment differed from the EVS, which reported that there was no advantage to routinely performing immediate vitrectomy in patients who had better than LP visual acuity when first seen. This could be due to differences in the bacteria identified in these two studies. In the EVS, Gram-negative microorganisms were less commonly involved. Even if we performed earlier and near complete PPV, the infection was still difficult to control. The benefits of vitrectomy may be greater because of the mechanical removal of S. maltophilia and toxins from the eye.
In the present study, recurrence occurred in two patients (14.3%, patients 2 and 3). In another two reported outbreaks, two relapses were seen in 26 cases (7.7%), and one in six cases (16.7%) [
13,
14]. Both the previous reports and our study demonstrated that
S. maltophilia endophthalmitis may have a persistent and recurrent clinical course. For infections that were difficult to resolve, especially in patients with diabetes or high myopia with posterior capsule rupture, PPV with IOL extraction could be performed to complete the course [
10,
13]. In our study, a final visual acuity of ≥20/100 was achieved by 13 of 14 patients (92.9%), compared with 80% of patients with
S. maltophilia cultured from eye samples in a previous report [
14]. Despite recurrences, visual prognosis was good, except for the diabetic patient. The diabetic patient’s final vision was HM. His serious fundus lesions led to the worst vision of all the patients. The infection aggravated the proliferative diabetic retinopathy, and total tractional retinal detachment developed. PPV with silicone oil tamponade were then performed. Furthermore, diabetic patients in the EVS had worse outcomes than nondiabetic patients [
20].
S. maltophilia has been often associated with serious life-threatening, systemic, opportunistic infections, especially in immunosuppressed or debilitated patients [
21], but it could also be found in patients with no known predisposition. Nosocomial sources of
S. maltophilia have also been reported from the hands of health personnel, blood sampling tubes, contact lens solutions, disinfectants, nebulizers and inhalation systems, moistening water reservoirs, and ventilation tubes [
19].
S. maltophilia is not a normal commensal on the periocular skin or conjunctiva. Its exogenous origin was reported from contaminated ophthalmic solutions (e.g., balanced salt solution, BSS) [
11,
14]. In our study, we systematically eliminated these possibilities.
S. maltophilia was found in the resterilization aspiration tube for the drainage cassette of one phacoemulsifier. This finding was consistent with another study [
22], which showed that the most common microorganism cultured from the aspiration fluid of the internal vacuum control manifold (VCM) of the phaco and vitrectomy machines was
S. maltophilia. Bacteria often colonize on plastic, and it was possible that they survived within the resterilized tubes of a contaminated biofilm. In a previous study [
13], the phacoemulsifier (Alcon Legacy; Alcon Laboratories.) was possibly contaminated, but it was impossible to isolate any pathogen from either the aspiration fluid or the internal VCM [
13]. In some countries, for economic reasons, the tubes that should only be used once are sometimes resterilized for each patient. These tubes are at high risk for subsequent surgeries, when the aspiration fluids can contaminate the irrigation tubes. Furthermore, perioperative anterior chamber aspirates have demonstrated a high rate of microbial contamination after cataract surgery.
Finally, a possible limitation of our study was that the pathogen involved in the negative endophthalmitis cases was perhaps not S. maltophilia, with the inclusion criteria including the same ocular signs, and use of the same machine during the outbreak period.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
YJ contributed to the study design, conducted the data analysis, and drafted the manuscript. CJ, JJ, YL, and YJ were involved in data collection, data management, and interpretation of data. YL interpreted the data, ensured the accuracy and integrity of the data, and revised the manuscript. All authors read and approved the final manuscript.