Serum creatinine as a predictor of functional and anatomical success in diabetic tractional retinal detachment

INTRODUCTION

Diabetic retinopathy (DR) is a leading cause of visual impairment in adults aged 20–74 years and is responsible for causing blindness in approximately 4% of individuals with Type 1 diabetes mellitus (DM) and in 1.6% of those with Type 2 DM.^[1-3] DR is divided into non-proliferative and proliferative depending on the absence or presence of neovascularization and can be further classified by severity.^[4] These stratifications have been useful in the analysis of indicators of disease progression and treatment strategies.

In the Wisconsin Epidemiologic Study of DR, higher glycosylated hemoglobin, higher blood pressure level, greater body mass index (BMI), severe DR, longer duration of diabetes, and gross proteinuria were associated with a higher incidence and/or progression of DR in patients with Type 1 DM.^[5,6] Comparably, microalbuminuria has been associated with proliferative diabetic retinopathy (PDR) in patients with Type 2 DM.^[7]

Diabetic macular edema (DME) and PDR are among the most common causes of persistent visual loss in DR.^[8-10] Tractional retinal detachment (TRD) represents an advanced form of PDR that results from the development of neovascularization and contractile fibrous tissue into the vitreomacular interface. Excessive anteroposterior and tangential traction on the fibrovascular complex and ischemic retina causes the retina to detach. Diabetic TRD is then repaired with pars plana vitrectomy (PPV).^[11,12]

Despite recent innovations in vitreoretinal surgery including high-speed and small-gauge vitrectomy systems, improved vitrectomy fluidics, bimanual surgery, and the development of anti-vascular endothelial growth factor (VEGF) intravitreal injections, surgical management of diabetic TRD continues to represent one of the most challenging surgeries for the vitreoretinal surgeon.^[12,13] The presence of extensive fibrovascular membranes and an ischemic retina contribute to the complexity of the surgical repair.^[12,14] Earlier studies have demonstrated that preoperative factors associated with poorer visual outcome include iris neovascularization, cataract, visual acuity <5/200, macular ischemia, and tractional or rhegmatogenous retinal detachment (RRD).^[15-18]

Visual recovery after surgical management depends on the functional and structural recovery of the macula which may be severely injured by chronic macular detachment and edema.^[19] Moreover, these eyes often have severe complications such as neovascular glaucoma and ischemic optic neuropathy which have been defined as predictors for low visual outcomes following surgery.^[19,20] The unique set of characteristics of each patient with DR demonstrates the importance of undergoing an individualized and timely approach to treatment to preserve vision.

Hence, the aim of this study was to analyze the effect of pre-operative, intraoperative, and post-operative variables on visual and anatomical success in diabetic TRD treated with PPV at a single ophthalmological reference center. The evaluation of predictors associated with successful visual and anatomical outcomes may serve as guidance for improved surgical decisions and patient counseling.

MATERIAL AND METHODS

This study was approved by the Institutional Review Board and followed the tenets of the Declaration of Helsinki.

We retrospectively reviewed the medical records of patients with diabetic TRD surgically repaired with PPV between March 2014 and February 2015 at the Instituto de Oftalmología Fundación de Asistencia Privada Conde de Valenciana, IAP in Mexico City. The surgical approach, including the instrument gauge and the tamponade agent, was based on the surgeons’ preference at the time of surgery.

Out of the 250 records reviewed, 85 met the inclusion criteria. All patients required a pre-operative internal medicine and anesthesiology assessment, a complete eye examination by a retina specialist, a pre-operative A-scan and B-scan ultrasonography, and a follow-up consultation 1 and 3 months after the surgery. Patients were excluded if they had a non-diabetic TRD or a previous surgically repaired retinal detachment. Pre-operative, intraoperative, and postoperative variables were obtained from all records and are summarized in [Table 1].

The pre-operative variables included: Age, type of DM (1 or 2), the period between the diagnosis of DM and the TRD in years, type of systemic treatment, microvascular or macrovascular systemic manifestations due to DM, BMI, creatinine level (mg/dl), best-corrected visual acuity (BCVA) measured on a Snellen chart, axial length (measured by an A-scan ultrasonography), the period between the onset of visual symptoms and the surgical intervention, lens status (phakic, aphakic, and pseudophakic), neovascular glaucoma, vitreous hemorrhage (VH), macula status assessed on the B-scan ultrasonography, and history of pre-operative panretinal laser photocoagulation (PRP) and/or anti-VEGF intravitreal injection.

The intraoperative variables included: Concurrent cataract extraction at the time of PPV, use of tamponade agent (silicone oil, air, octafluoro propane gas, or sulfur hexafluoride gas), duration of the surgery, presence of retinotomies and/ or other surgical complications, intraoperative observations such as submacular hemorrhage, macular hole or optic disc pallor, and the use of special surgical techniques such as bimanual vitrectomy, viscodissection, or dissection with perfluoro-N-octane.

The post-operative variables included: The presence of VH, recurrent retinal detachment (tractional, rhegmatogenous, or combined), neovascular glaucoma, and glaucoma due to vitreoretinal surgery.

The main outcome measures were anatomical success and BCVA at post-operative month (POM) 3. A successful visual outcome was defined as gaining at least one line of vision at POM 3. Anatomical success was considered if a completely attached retina was observed at the final follow-up examination.

Statistical procedures used for this analysis included Fisher’s exact test, Kruskal–Wallis test, and Mann–Whitney U test. P < 0.05 was considered statistically significant.

RESULTS

Pre-operative factors

A total of 88 eyes of 85 patients met the inclusion criteria for the study. The mean age at the time of the surgery was 51.53 ± 11.99 years. Eighty-two eyes were from patients with Type 2 DM and six eyes were from patients with Type 1 DM. Forty-eight patients (54.5%) had systemic microvascular complications, including chronic kidney disease in 47 patients and diabetic neuropathy in one patient. Six patients (6.8%) had macrovascular complications, including diabetic foot in five patients and upper limb arterial thromboembolism in one patient. The mean BMI was 26.29 ± 4.97, and 35.7% of the patients were categorized as obese (BMI >26.9). The median pre-operative creatinine value was 1.25 mg/dl (P25 = 1, P75 = 1.68) [Graph 1]. Most of the patients (44.3%) were being treated with only oral hypoglycemic medications, while 21.6% were being treated with both insulin and oral hypoglycemic medications. Six patients (6.8%) were not receiving systemic treatment.

Close

Graph 1:

Distribution of pre-operative creatinine value (mg/dl).

Export to PPT

On average, the period between the diagnosis of DM and the TRD was of 15.72 ± 7.49 years with a median of 16.5 years (P25 = 10.25, P75 = 21). The median period between the onset of visual symptoms and the surgical intervention was of 8 months (P25 = 5, P75 = 12). The mean axial length was 22.84 ± 0.75 mm. Regarding the presence of VH, 22.7% of the patients had Grade 3, 20.5% had Grade 2, 12.5% had Grade 4, and 3.4% had Grade 1 VH. The Previous PRP had been performed in 47.7% of the eyes and only five eyes received pre-operative intravitreal anti-VEGF. Two eyes had pre-operative iris neovascularization. A total of 55 eyes (62.5%) had a macula involving TRD.

Intraoperative factors

The median surgical time was of 2.5 h (P25 = 2, P75 = 4) [Graph 2]. A total of 78 eyes (88.6%) underwent concurrent cataract extraction at the time of the PPV. Of the 10 eyes that were treated with only PPV, six eyes had a clear lens, and four eyes were pseudophakic. Posterior capsule rupture occurred in five eyes, and in all cases, a three-piece lens was implanted in the sulcus. Incidental retinotomies occurred in 62.7% of the cases [Graph 3]. Regarding the tamponade agent, silicone oil was used in most cases (39 eyes) followed by C₃F₈ in 19 eyes (22%), SF₆ in 16 eyes, and air in 12 eyes (13.9%).

Close

Graph 2:

Surgical time in hours.

Export to PPT

Close

Graph 3:

Number of retinotomies.

Export to PPT

DISCUSSION

In this study, we report pre-operative, intraoperative, and post-operative factors associated with functional and anatomical success in patients with diabetic TRD treated with PPV at a single reference ophthalmological center.

A limited number of studies have analyzed the effect of renal function on the outcomes of vitrectomy for TRD and DME. An earlier study led by Kojima et al. did not find a correlation between the pre-operative serum creatinine level and a change in post-operative retinal morphology in patients who underwent PPV for DME.^[21] Similarly, a study led by Larrañaga-Fragoso et al. reported that a reduced renal function does not adversely affect visual outcomes of patients with TRD repaired with PPV.^[22] Comparably, Tokuyama et al. demonstrated that hemodialysis has little effect on macular leakage in patients with DME and DR. The authors described that a possible explanation for the lack of change in macular leakage after hemodialysis despite clinical expectations could be the presence of a sufficiently preserved inner blood-retinal barrier and retinal blood flow autoregulation capable of protecting the capillaries from the dynamic changes in fluid and plasma volumes.^[23] Contrary to these findings, in our study, patients with a greater pre-operative serum creatinine level achieved a worse functional and anatomical outcome 1 month after the PPV. A possible explanation for this observation could be that patients with greater serum creatinine levels have greater microvascular damage overall.^[24] In patients with PDR and renal disease, high blood pressures and low protein serum levels may affect the hydrostatic and plasmatic osmotic pressures, respectively, which increase vascular leakage.^[23]

The complexity of the surgical repair of diabetic TRD largely depends on the amount of vitreoretinal traction since the dissection of firmly adhered fibrovascular membranes may lead to incidental retinotomies and prolonged operative duration.^[15] Consistent with these observations, we found that fewer retinotomies and a shorter surgical duration were associated with a better visual outcome at the initial and final follow-up examinations. Overall, patients with worse functional and anatomical outcomes were those who underwent complex surgeries due to a friable and ischemic retina.

Supporting the findings of earlier studies, our results demonstrate that a pre-operative macula involving TRD is an important predictor for a worse post-operative outcome.^[15] Overall, eyes with pre-operative macula involving TRD achieved worse functional and anatomical outcomes at POMs 1 and 3.

Initially, patients with a pre-operative Grade 3 or 4 VH had a significantly lower visual acuity than those with Grade 1 or 2 VH. Unpredictably, patients with pre-operative Grade 3 or 4 VH presented better final visual and anatomical outcomes than those with a lower grade of VH. Finally, a greater number of post-operative complications was found to be associated with a worse overall final outcome.

CONCLUSION

Consistent with earlier studies, we demonstrate that timely PPV is an effective procedure for diabetic TRD repair and that due to the often-unpredictable results; it is of utmost importance to fully grasp all the variables that could affect the patient’s outcome. We conclude that greater preoperative serum creatinine levels correlate with a worse final functional and anatomical outcome.

The limitations of our study include the retrospective design, the lack of a long-term follow-up, and the variability of surgeons. Thus, further research is warranted to better elucidate the relationship between pre-operative serum creatinine levels and/or renal function and overall surgical success.