To determine the intravitreal anti–vascular endothelial growth factor (VEGF) injection techniques and preferences within the retinal community and to identify potential factors associated with the development of sustained intraocular pressure (IOP) elevation in patients treated with intravitreal anti-VEGF therapy for neovascular age-related macular degeneration (AMD).
Cross-sectional physician survey.
Five hundred and thirty retina specialists spanning both private and academic practices were surveyed regarding current anti-VEGF intravitreal injection protocols, including the anti-VEGF drug of choice, needle gauge, injection volume, injection technique, and self-reported prevalence of sustained IOP elevation. Multivariate logistic regressions were performed to assess the potential influence of these factors on long-term IOP.
Two hundred ninety-two specialists (55%) reported believing that intravitreal anti-VEGF therapy may cause sustained IOP elevation. Of these responses, the most common reported prevalence was 1%–2% (48%), followed by 3%–5% (34%). There was no relationship between the frequency of sustained IOP elevation and anti-VEGF drug of choice. Physicians who injected greater than 0.05 cc in less than 1 second were 5.56 times more likely to observe a high frequency of sustained IOP elevation ( P = .006, 95% CI 1.64–18.89).
Based on physician survey data, serial anti-VEGF injections using higher injection volumes with a rapid injection technique may potentially lead to sustained IOP elevation. The underlying mechanism for this complication may be injury to the trabecular meshwork resulting from rapid elevations in IOP. Further investigation of the relationship between injection techniques and sustained IOP elevation in the form of retrospective or prospective clinical studies is warranted.
Therapies targeting vascular endothelial growth factor (VEGF) have revolutionized the treatment of neovascular age-related macular degeneration (AMD) and other vision-threatening retinal diseases, including retinal vein occlusion and diabetic macular edema, where overexpression of VEGF, vascular leakage, and neovascularization have been identified as pathogenic mechanisms that may be addressed with therapy. In these diseases, anti-VEGF agents have been shown to be efficacious and to have a favorable safety profile. Given the need for frequent injections over the course of a disease, particularly in neovascular AMD, novel reports of long-term risks are surfacing. Intraocular pressure (IOP) elevation is a significant concern in chronic anti-VEGF therapy and is incompletely characterized.
Transient increases in IOP have been studied extensively in the setting of anti-VEGF therapy. This work has revealed that patients experience short-term pressure spikes after injection that normalize within 30 minutes without further intervention in the majority of cases. In contrast, sustained IOP elevation has been investigated, but its mechanism is poorly understood. Several associations have been described including male sex, short intervals between injections, a previous history of glaucoma, and the total number of injections administered. To the best of our knowledge, there are no prior studies in the literature on the influence of injection protocols on the risk of sustained IOP elevation.
This survey was performed primarily to determine the intravitreal anti-VEGF injection techniques and preferences within the retinal community. The second objective was to conduct statistical analysis to identify potential candidate risk factors for developing sustained elevations in IOP that could be examined further in future studies.
A database of 2638 retina specialists was created from existing membership data of the American Society of Retina Specialists, the Macula Society, and the Retina Society. Participants that spanned both private and academic practices were emailed an invitation to the survey hosted by SurveyMonkey (Palo Alto, California, USA). This cross-sectional survey consisted of 15 questions, including the number of monthly injections administered for neovascular AMD, the preferred anti-VEGF agent, the needle gauge used and volume of medication injected, the injection technique, the use of anterior chamber paracentesis, the use of a cotton-tipped applicator to prevent reflux, IOP monitoring practices both before and after injection, and the use of IOP-lowering drops prior to injection. Participants were also polled on the speed of administration, the practice of softening the globe with pressure during anesthetic preparation, the prevalence of sustained IOP elevation, and the proposed mechanism of pressure elevation. Sustained IOP elevation was defined as persistent elevation in preinjection IOP of >5 mm Hg above baseline (prior to the first anti-VEGF treatment) on at least 2 consecutive visits. The order of the answer choices to the suggested mechanism of sustained IOP elevation was randomized with each respondent so as not to create bias toward one selection. Since all surveys were anonymous and no patient identifiers were recorded in any manner, the Western Institutional Review Board ruled that the survey met the conditions for exemption.
Univariate relationships between the frequency of observing sustained IOP elevation and (1) drug type used, (2) needle gauge used, (3) volume of medication injected, (4) injection technique used, (5) speed of injection, (6) performance of anterior chamber paracentesis, (7) prevention of reflux with a cotton-tipped applicator, (8) administration of IOP-lowering drops, and (9) softening the globe with pressure prior to injection were analyzed. Specific variables based on the Likert-type scale (6–9) were dichotomized to never/sometimes and most of the time/always for univariate analysis. Compound variables for injection characteristics were also analyzed. Volume of medication injected, size of needle gauge, and speed of injection were dichotomized into high-volume (>0.05 cc) vs low-volume (≤0.05 cc), low gauge (<31 gauge) vs high gauge (≥31 gauge), and fast speed (<1 second) vs slow speed (≥1 second), respectively. All were calculated using the 2-sample t test or Wilcoxon rank sum test for continuous variables and the χ 2 test or Fisher exact test for categorical variables.
Specific clinical characteristics were further assessed for a potential independent effect on the frequency of observing sustained IOP elevation. Variables that resulted in a P value < .2 from the univariate analyses were further evaluated using a multivariable logistic regression model to adjust for confounding factors. An assessment of collinearity between predictors was performed prior to the specification of the final multivariate model. If 2 predictive factors were collinear (variance inflation factor value >10), then the one that was more closely related to the frequency of observing sustained IOP elevation in the univariate analysis was explored in the multivariate model. The number of subjects in the different outcome groups limited the number of variables that could be explored in the multivariate model. All P values are 2-sided with statistical significance evaluated at the .05 alpha level. Ninety-five percent confidence intervals (95% CI) for adjusted odds ratios (OR) were constructed to assess the precision of the obtained estimates. All analyses were performed in SPSS Version 22 (SPSS Inc, Chicago, Illinois, USA).
Five hundred thirty-nine of the 2638 retina specialists polled (20%) responded to the survey. Survey results are reported in Figure 1 . Of the 539 respondents, three hundred eighty-seven (72%) administered greater than 50 injections per month. Nine respondents (2%) reported that they do not routinely administer injections and were eliminated from subsequent analysis. Of the 530 specialists who administered injections, two hundred seventy-eight physicians (52%) most frequently used bevacizumab, 184 (35%) ranibizumab, and 68 (13%) aflibercept. Three hundred sixty-eight physicians (69%) used a 30 gauge needle, while the next most common needle sizes were 32, 31, and 27 gauge, respectively. A single physician reported also using 33 gauge needles. Three hundred sixty-one injectors (68%) elected to use a dosage volume of 0.05 cc, while the next most common volume was a little more than 0.05 cc.
Four hundred seventy-one respondents (89%) used a perpendicular injection technique, while the remaining 59 (11%) used an angled technique. One hundred thirty-nine respondents (26%) reported performing anterior chamber paracentesis immediately following injections at least sometimes, while 391 (74%) reported never performing this procedure. Two hundred ninety-two specialists (55%) reported covering the injection site with a cotton-tipped applicator after injecting to prevent reflux at least sometimes, while 238 (45%) elected not to use this technique. Four hundred seventy-nine physicians (90%) reported routinely checking IOP prior to injecting most of the time or more often, yet only 136 (26%) reported routinely checking IOP after injecting most of the time or more often.
Eighty-one specialists (15%) elected to use IOP-lowering drops prior to injections, while 449 (85%) reported never using IOP-lowering drops. Three hundred forty-six respondents (65%) reported injecting at a moderate speed (about a second), while the next most common technique was a fast injection speed (less than a second). One hundred and seven respondents (20%) reported softening the eye with pressure during anesthetic preparation of the globe.
Two hundred ninety-two specialists (55%) reported believing that intravitreal anti-VEGF therapy may cause sustained IOP elevation. Of the 292 physicians who reported believing in sustained IOP elevation, we asked the prevalence within their practice as well as the most likely mechanism. Eleven (4%) never observed it, 141 (48%) observed it rarely (in 1%–2% of patients), 99 (34%) observed it occasionally (in 3%–5% of patients), and 41 (14%) observed it in 6% or more of their patients. We also asked each physician the most likely mechanism of sustained IOP elevation. Mechanical obstruction to outflow by high-molecular-weight protein aggregates was the most common response, cited by 95 respondents (33%), followed by mechanical trauma to the trabecular meshwork from repeated pressure spikes with each injection, which was cited by 72 participants (25%).
Notably, 14 physicians (5%) commented that multiple mechanisms are likely to be responsible, depending on the individual patient. Several physicians proposed other mechanisms that were not included as choices in our survey, including the belief that sustained IOP elevation is related to the natural history of increasing IOP with aging, or that there may be mechanical injury to the trabecular meshwork by traction from insertion of dull needles.
Of those who observed sustained IOP elevation, the study population was dichotomized into those who reported frequent observation of sustained IOP elevation (>5%) and those who observed it infrequently (≤5%). The frequent observation group consisted of 35 respondents (12.5%), while the infrequent observation group consisted of 245 respondents (87.5%), as shown in Figure 2 . Clinical predictors of sustained IOP elevation with P < .2 from the univariate analysis are shown in Table 1 and included anti-VEGF drug type, needle gauge, high-volume and fast injectors (greater than 0.05 cc in less than 1 second), the decision to perform anterior chamber paracentesis, and the decision to administer IOP-lowering drops. The 2 groups were similar ( P > .2 from the univariate analysis) with respect to clinical characteristics that included angled vs perpendicular injection technique, high vs low-volume injection amounts, fast vs slow injection speed, prevention of reflux with a cotton-tipped applicator, and softening the globe with pressure prior to injection ( Table 2 ). Furthermore, the 2 groups were similar with respect to injectors who used high-volumes and high gauges (31 or 32 gauge); fast speed and high gauges; and high-volume, high gauges, and fast speed.
|High-Frequency Observers a (35, 12.5%)||Low-Frequency Observers b (245, 87.5%)||P Value c|
|Bevacizumab||23 (65.7%)||119 (48.6%)||.058|
|Aflibercept||2 (5.7%)||40 (16.3%)||.100|
|Ranibizumab||10 (28.6%)||86 (35.1%)||.446|
|27||1 (2.9%)||7 (2.9%)||1|
|30||22 (62.9%)||167 (68.1%)||.531|
|31||9 (25.7%)||27 (11.0%)||.027|
|32||3 (8.5%)||44 (18.0%)||.227|
|High-volume, fast injectors d||5 (14.3%)||9 (3.7%)||.020|
|Perform anterior chamber paracentesis||22 (62.9%)||176 (71.8%)||.119|
|Administration of intraocular pressure–lowering drops||3 (8.6%)||3 (1.2%)||.028|
|High Frequency Observers (35, 12.5%)||Low Frequency Observers (245, 87.5%)||P Value a|
|Angled injection technique||6 (17.2%)||30 (12.2%)||.420|
|Perpendicular injection technique||29 (82.8%)||215 (87.8%)||.420|
|High-volume b||12 (34.3%)||70 (28.6%)||.487|
|Fast injection speed c||11 (31.4%)||54 (22.0%)||.219|
|Prevention of reflux with a cotton-tipped applicator||13 (37.1%)||110 (44.9%)||.387|
|Softening the globe with pressure during anesthetic preparation||4 (11.4%)||30 (12.2%)||>.999|
|High-volume, high gauge d||4 (11.4%)||13 (5.3%)||.244|
|Fast speed, high gauge e||3 (8.6%)||20 (8.2%)||>.999|
|High-volume, high gauge, fast speed f||2 (5.7%)||12 (4.9%)||.690|
A multivariate logistic regression model for independent predictors of observing a high frequency of sustained IOP elevation was performed. A χ 2 test found a statistically significant association between the needle gauge and the drug type used (bevacizumab injectors were more likely to use 31 gauge needles than other drug type users), χ 2 (6) = 23.682; P < .001 (Cramer’s V = 0.206, P < .01). Therefore, the following variables were included, based on a priori considerations and the findings of the univariate analysis: drug type used, needle gauge used, and injectors who used a high-volume with a fast speed.
The results of the multivariable logistic regression are presented in Table 3 . There was no relationship between those who observed a high frequency of sustained IOP elevation and the drug type used. Respondents using 31 gauge needles were 2.83 times more likely to observe a high frequency of sustained IOP elevation than those who used 27, 30, or 32 gauge needles ( P = .028, 95% CI 1.12–7.17). Additional analysis showed that there was no statistically significant relationship between high frequency observers of sustained IOP elevation and injectors using high gauges (OR: 1.133, 95% CI 0.517–2.484).