Intraocular Pressure Increases After Intraarticular Knee Injection With Triamcinolone but Not Hyaluronic Acid

Abstract

Background

Intraarticular steroid injections are a common first-line therapy for severe osteoarthritis, which affects an estimated 27 million people in the United States. Although topical, oral, intranasal, and inhalational steroids are known to increase intraocular pressure in some patients, the effect of intraarticular steroid injections on intraocular pressure has not been investigated, to the best of our knowledge. If elevated intraocular pressure is sustained for long periods of time or is of sufficient magnitude acutely, permanent loss of the visual field can occur.

Questions/purposes

How does intraocular pressure change 1 week after an intraarticular knee injection either with triamcinolone acetonide or hyaluronic acid?

Methods

A nonrandomized, nonblinded prospective cohort study was conducted at an outpatient, ambulatory orthopaedic clinic. This study compared intraocular pressure elevation before and 1 week after intraarticular knee injection of triamcinolone acetonide versus hyaluronic acid for management of primary osteoarthritis of the knee. Patients self-selected to be injected in their knee with either triamcinolone acetonide or hyaluronic acid before being informed of the study. The primary endpoint was intraocular pressure elevation of ≥ 7 mm Hg 1 week after injection. This cutoff is determined as the minimum significant pressure change in the ophthalmology literature recognized as an intermediate responder to steroids. Intraocular pressure was measured using a handheld Tono-Pen® applanation device. This device is frequently used in intraocular pressure measurement in clinical and research settings; 10 sequential measurements are obtained and averaged with a confidence interval. Only measurements with a 95% confidence interval were used. Over a 6-month period, a total of 96 patients were approached to enroll in the study. Sixty-two patients out of 96 approached (65%) agreed. Thirty-one (50%) were injected with triamcinolone and 31 (50%) were injected with hyaluronic acid. Patients with osteoarthritis of the knee who were suitable candidates for either a steroid injection or hyaluronic acid injection were included in the study. Exclusion criteria included previous glaucoma surgery, previous corneal injury precluding use of a Tono-Pen, current acute or chronic steroid use, and diagnosis of glaucoma other than primary open-angle. Patients with elevated intraocular pressure at the 1-week timepoint were invited to return at 1 month for repeat measurement; however, only five of nine (55.6%) were able to do so. The mean age of the total population was 64.1 ± 11.65 years. There were 46 (74%) women and 16 men. Patient in the hyaluronic acid injection group were younger than the triamcinolone group, 59.5 ± 11.7 versus 68.7 ± 9.7 years of age (p < 0.003).

Results

The mean intraocular pressure increased by 2.79 mm Hg 1 week after treatment with triamcinolone, but it did not change among those patients treated with hyaluronic acid (2.79 ± 9.9 mm Hg versus -0.14 ± 2.96 mm Hg; mean difference 2.93 mm Hg; 95% confidence interval, -0.71 to 6.57 mm Hg; p = 0.12). More patients who received triamcinolone injections developed an increase in intraocular pressure > 7 mm than did those who received hyaluronic acid (29% [nine of 29] versus 0% [zero of 31]; p = 0.002). Of the nine patients who developed elevated intraocular pressure after a triamcinolone injection, five returned for reevaluation 1 month later, and four of them had pressures that remained elevated > 7 mm Hg from baseline.

Conclusions

There appears to be an associated intraocular pressure elevation found in patients who have undergone a triamcinolone injection of the knee. Further larger scale randomized investigations are warranted to determine the longevity of this pressure elevation as well as long-term clinical implications, including optic nerve damage and visual field loss.

Level of Evidence

Level II, therapeutic study.

Introduction

Osteoarthritis is the most common form of chronic arthritis worldwide. Intraarticular steroid injections are a frequent first-line therapy for inflammation associated with symptomatic osteoarthritis in patients for whom initial conservative therapy with nonsteroidal antiinflammatory drugs, activity modification, and weight loss has failed. With an aging population in America, and an estimated 27 million people currently diagnosed with osteoarthritis, a large number of patients have and/or will receive intraarticular steroid injections [12]. Although intraarticular steroid deposition is intended to increase local concentration, while minimizing systemic effects, Armstrong et al. [3] found an increased serum level of steroids within the first day. This elevation can lead to suppression of the hypothalamic-pituitary-adrenal axis for up to 1 week after injection [3, 10]. This can lead to measurable physiologic changes. For example, patients with diabetes who previously had their diabetes well controlled have been reported to experience hyperglycemia after these injections [11, 13, 14].

Intraocular pressure elevation can occur as an adverse effect of corticosteroid therapy, regardless of whether this medication is delivered intranasally, inhalationally, systemically, or through topical application on the skin or eye [8]. Becker and Mills [4] found that patients who had glaucoma, or a diagnosis for a high risk for glaucoma, had marked intraocular pressure increases in response to several weeks’ exposure to topical corticosteroids. Of concern, this occurred in 30% patients without any diagnosis of glaucoma [4]. Intraocular pressure returned to baseline in most patients in their study after cessation of steroid use; however, some continued to have elevated intraocular pressure. Although there is a clearly defined association between steroid use through numerous routes of administration and ocular hypertension and glaucoma, we could find no published studies to date investigating the change in intraocular pressure after an intraarticular steroid injection.

We therefore sought to investigate the effect of intraarticular steroid injections on intraocular pressure. We asked: How does intraocular pressure change 1 week after an intraarticular knee injection either with triamcinolone acetonide or hyaluronic acid?

Patients and Methods

A prospective, nonrandomized cohort study of consecutive patients in an orthopaedic clinic was performed at a tertiary care center from December 2015 to May 2016 and was approved by our hospital’s institutional review board. We performed an a priori power analysis and found that a minimum of 21 patients in each cohort would be needed to obtain statistical power at the 80% level to detect an intraocular pressure difference of > 7 mm Hg. We compared patients who received an intraarticular steroid injection and those who received an intraarticular hyaluronic acid injection. Patients were asked to return 1 week after injection for repeat intraocular pressure measurement. If intraocular pressure elevation (> 7 mm Hg) was found at 1 week, patients were asked to return 1 month after their initial injection. This cutoff was selected based on previous literature by Armaly [2] as the benchmark for an “intermediate steroid responder.” In addition, with normal intraocular pressures ranging from 12 to 18 mm Hg, this change would reflect a 39% to 58% increase from baseline and would place a patient in a category of “ocular hypertensive” and merit serial evaluation from an ophthalmologist [6, 7].

Between December 2015 and May 2016, we saw 643 patients in the orthopaedic clinic for degenerative joint disease of the knee. We offered injections of either corticosteroid or hyaluronic acid to 267 of these. Ninety-six agreed to have a knee injection and were offered enrollment in the study. We included patients diagnosed with osteoarthritis of the knee based on radiographic and clinical and who were determined to be a candidate for either a steroid injection or hyaluronic acid injection. We excluded patients with previous glaucoma surgery, previous corneal injury precluding use of a Tono-Pen® (Reichert Technologies, Depew, NY, USA), current acute or chronic steroid use, and a diagnosis of glaucoma other than primary open-angle. Of these, 62 agreed to participate. Patients self-selected steroid versus hyaluronic acid injection before learning about the study. Exclusion criteria included previous glaucoma surgery, previous corneal injury precluding use of an applanation device, current acute or chronic steroid use, and diagnosis of glaucoma other than primary open-angle. After informed consent, the patient was enrolled in the study and a demographic form was filled out.

All intraarticular knee injections were performed using the superior parapatellar approach. Patients were injected with either 40 mg triamcinolone acetonide or 16 mg hyaluronic acid. Intraocular pressure measurements were recorded using a Tono-Pen after topical anesthesia (tetracaine 0.5%) application to the eye. The Tono-Pen is a handheld applanation device that uses microstrain gauge technology and a 1.0-mm transducer tip to display the average of 10 independent readings along with a standard error. It has been considered a clinically accurate measure of intraocular pressure since at least 1987 [15, 24]. The patient was supine for all intraocular pressure measurements. Measurement was recorded only if the standard error was < 5%, its lowest possible value. If the standard error was > 5%, the measurement was repeated.

The primary dependent variable of our study was intraocular pressure, whereas the independent variable was steroid versus hyaluronic acid injection in the knee. A baseline intraocular pressure was measured in both eyes before intraarticular injection. A repeat intraocular pressure was measured 7 ± 1 days after the initial injection. The primary endpoint was change in intraocular pressure. Patients with increased intraocular pressure at 1 week were asked to return 1 month after their initial injection for final followup and repeat measurement. The hyaluronic acid group was designed as a control for the act of injecting the knee. There have been no studies showing intraocular pressure elevation after hyaluronic acid injection. Potential confounders such as physiologic stress associated with the intraarticular injection, the tetracaine used to numb the eye, and measurement of intraocular pressure were intentionally made consistent across both arms of the study. Thus, differences in intraocular pressure recorded in the two groups were thought to be isolated to the effect of the steroid or to hyaluronic acid.

Study patients were recruited from an outpatient, primary care orthopaedic clinic. Thirty-one patients were included in each of the triamcinolone acetonide and hyaluronic acid injection groups for a total of 62 patients. Each patient received a single injection of either the triamcinolone or hyaluronic acid. All 62 patients (100%) were seen at the 1-week visit. Nine patients had elevated intraocular pressure at the Week 1 visit and were requested to return at 1 month. Five patients were able to followup at 1 month. Four patients were not able to return at 1 month.

The mean age of the total population was 64.1 ± 11.65 years (range, 38-87 years). In the steroid group, 22 were women and nine were men. Self-identified race was stratified as 18 white, 10 black, two Hispanic, and one Middle Eastern. In the hyaluronic acid group, there were 24 women and seven men. Self-identified race was stratified in 20 white, seven black, one Hispanic, and three Middle Eastern. The steroid group contained one patient with a previous diagnosis of primary open-angle glaucoma and one who had been diagnosed with suspected glaucoma. The hyaluronic acid group contained two patients with a previous diagnosis of primary open-angle glaucoma and one who had been diagnosed with suspected glaucoma. Fifty-seven patients had no previous glaucoma diagnosis. Patient age was the only difference between the groups. Patients in the hyaluronic acid injection group were younger (59.5 ± 11.7 versus 68.7 ± 9.7 years; p < 0.003; Table 1).

Table 1.

Patient demographics

Ninety-six patients met the inclusion and exclusion criteria and were informed of the study. Sixty-two (65%) of these patients agreed to participate in our study. All of the 62 patients (31 in each cohort) returned for a 1-week appointment. Nine patients (29%) in the triamcinolone group were found at 1 week to have a pressure elevation > 7 mm Hg. No patients in the hyaluronic acid group were found to have elevated intraocular pressure. Of these nine patients receiving triamcinolone with elevated intraocular pressure at their 1-week appointment, five were followed up at 1 month. When contacted, the other four refused followup as a result of the inability to take time off from work, inability to commute to the clinic so soon after their initial visit, or lack of desire to return without remuneration. All four also cited lack of desire for repeat eye examinations as a reason.

Statistical Analysis

Statistical analysis was performed using descriptive statistics with means and SDs for the continuous variables and frequency distribution and percentages calculated for categorical variables. To look for significant differences at particular times for continuous variables, Student's t-test (unpaired) was used. A chi-square test was used to find associations between the categorical variables. Trend analyses in variables were assessed using two-way analysis of variance for both groups separately. A p value < 0.05 was considered statistically significant. A spreadsheet program and SAS® Version 9.4 software (SAS Institute Inc, Cary, NC, USA) were used for these statistical analyses.

Results

The mean intraocular pressure increased by 2.79 mm Hg 1 week after treatment with triamcinolone, but it did not change among those patients treated with hyaluronic acid (2.79 ± 9.9 mm Hg versus -0.14 ± 2.96 mm Hg; mean difference 2.93 mm Hg; 95% confidence interval [CI], -0.71 to 6.57 mm Hg; p = 0.12; Table 2). More patients who received triamcinolone injections developed an increase in intraocular pressure > 7 mm Hg than did those who received hyaluronic acid (29% [nine of 31] versus 0% [zero of 31]; p = 0.002). The mean intraocular pressure difference between those patients who experienced a > 7-mm Hg pressure spike and those who did not was 14.67 mm Hg (13.22 ± 11.05 mm Hg versus -1.48 ± 5.19 mm Hg; mean difference 14.67 mm Hg; 95% CI, 8.77-20.57 mm Hg; p < 0.001).

Table 2.

Intraocular pressures (mm Hg) at individual times for both groups of patients

Five of nine patients who developed elevated intraocular pressure after a triamcinolone injection returned for reevaluation 1 month later. Of these five patients, four had pressures that remained elevated > 7 mm Hg from baseline.

Discussion

Elevated intraocular pressure, if sustained for a significant amount of time or of sufficient magnitude, can lead to irreversible damage to the optic nerve and visual field [4-7]. Unfortunately, these glaucomatous changes are frequently asymptomatic, allowing for significant irreversible damage to occur without the patient ever being aware of it [4-7]. Fortuitously, once the steroid administration is discontinued, the intraocular pressure frequently returns to its baseline. Unfortunately, however, any damage to the optic nerve during the period of elevated pressure is irreversible [7]. Although topical, oral, intranasal, and inhalational steroids are known to increase intraocular pressure [8, 17, 18, 20, 25], the association between intraarticular steroid injections of the knee and intraocular pressure has not been investigated, to the best of our knowledge. Our study finds that there may be a physiologic response with elevation of intraocular pressure in patients who underwent triamcinolone injection of the knee. In addition, more patients who received triamcinolone injections developed a clinically significant increase in intraocular pressure of > 7 mm Hg than did those who received hyaluronic acid.

There are several limitations to this study. First, we did not have 100% followup of patients at 1 month in the triamcinolone group. Four of the nine patients receiving triamcinolone requested to return at 1 month were lost to followup. Because there were no intraocular pressure elevations at 1 week in the hyaluronic acid group, none was asked to return at 1-month followup. Second, the mean age of the groups was different, with the triamcinolone group being 9 years older. The potential different physiology of this older group may alter their susceptibility to the effects of steroid injections. Third, the timing of the intraocular pressure spike has not been uniform in previous studies [1, 5, 7]. Therefore, it is possible that we may have missed the pressure spike in some patients, thus lowering our mean difference between cohorts. Fourth, only one type of steroid was used in this study. It is possible that different types and/or different dosages may exert a different physiologic effect. Lastly, although there is a documented increase in eye pressures after the injection, it is unclear whether this physiologic response has any long-term detrimental effect on the eye. Although not evaluated in this study, substantial glaucomatous optic neuropathy with concomitant visual field loss has been documented owing to steroid response alone [6, 9, 22]. Future studies should be performed with ophthalmologic evaluation of the optic nerve and visual field preinjection and postinjection. In addition, although this study was appropriately powered, it was performed on a relatively small cohort of patients who self-selected their treatment. As such, some form of selection bias cannot be excluded. A randomized controlled trial may be indicated to definitively prove or disprove the findings presented here.

Steroids have been shown in vitro and in vivo to cause numerous cellular, biochemical, and molecular changes at the site of outflow in the eye, the trabecular meshwork [7, 9, 13, 16, 19, 21, 23]. Tissue culture experiments using isolated, human anterior segments in a perfusion chamber show that glucocorticoids can have direct effects on the trabecular meshwork [8, 13, 21]. Thirty percent of these eyes had ocular hypertension develop. This percentage is in line with the percentage reported in previous studies of intraocular pressure elevation in normal volunteers exposed to steroids [4, 8]. Armaly [2], in his seminal paper on steroid-induced glaucoma, found that, although not all subjects had an increase in intraocular pressure in response to steroid administration, a subgroup he termed “steroid responders” did experience an idiopathic spike in their intraocular pressure > 7 mm Hg. He found that this group encompassed 25% of his study population.

Studies, outside of the ophthalmology literature, evaluating steroid-induced glaucoma have been mainly limited to case reports in various subspecialty literature. In dermatology, Zugerman et al. [25] reported on a case of a patient using triamcinolone cream on the face, which led to intraocular pressures of 55 mm Hg leading to severe vision loss. In pulmonology, Opatowsky et al. [17] found a strong association between inhaled corticosteroids and the presence of ocular hypertension. This risk increased with higher doses and more puffs in patients with a family history of glaucoma [17, 18].

Although designed to place a local deposit of steroids in a specific region, intraarticular steroids have been shown to elicit system-wide effects. Habib and Safia [11] performed a randomized trial comparing methylprednisolone versus sodium hyaluronate intrarticular injections of the knee and their effect on the hypothalamus-pituitary-adrenal axis. Patients from both groups underwent a low-dose (1 μg) adrenocorticotropin hormone stimulation test. Twenty-five percent of the patients receiving methylprednisolone had secondary adrenal insufficiency at 2 to 3 weeks after injection. In a separate study, Habib et al. [10] found that an intraarticular injection of betamethasone into the knee led to an average blood glucose level of 322.5 mg/dL in patients with previously well-controlled type 2 diabetes.

In conclusion, our study found that patients undergoing a triamcinolone steroid injection of the knee did experience intraocular pressure elevation, whereas patients receiving hyaluronic acid did not. Twenty-nine percent of our patients qualify as intermediate to high steroid responders using the Armaly [2] cutoff of > 7 mm Hg. Although our pilot study is small and has several limitations, we believe our findings merit further evaluation in larger randomized and blinded studies to further elucidate this physiologic finding.