A 55-year-old woman presented to the emergency department for a 2-day history of progressive right eye pain, redness, and swelling. She denied antecedent trauma, sinus congestion, rhinorrhea, diplopia, diminished vision, headache, fevers, or chills. Extraocular movements worsened her pain. She had recently been diagnosed with postmenopausal osteoporosis and received her first infusion of zoledronic acid by her rheumatologist 4 days prior to symptom onset. A review of systems was otherwise unremarkable.
Her medical history included well-controlled type 1 diabetes mellitus with proliferative diabetic retinopathy requiring panretinal photocoagulation laser treatment, primary hypothyroidism, osteoarthritis, gastroesophageal reflux disease, gastroparesis, and postmenopausal osteoporosis of the hip and lumbar spine. Due to her gastrointestinal comorbidities, her rheumatologist had treated her with an intravenous (IV) bisphosphonate rather than an oral bisphosphonate. Medications additionally included insulin, cetirizine, levothyroxine, probiotic, calcium, and vitamin D, as well as metoclopramide and antiemetics as needed.
Social and family history
She denied tobacco, alcohol, or illicit drug use and worked as a telephone dispatcher. Her mother and sister had osteoporosis.
Vital signs showed a temperature of 36.8°C, a blood pressure of 149/77 mm Hg, and a pulse of 75 beats/minute. She appeared alert, oriented, and in no distress. A head and neck examination revealed markedly swollen, tender, and erythematous right periorbital soft tissue, diffuse clear conjunctival chemosis, and restricted ocular motility (Figure 1). Eyelid fluctuance and crepitus were absent. Visual acuity was 20/20 bilaterally. Pupils were 3–4 mm and without afferent defect. The remainder of the ophthalmologic examination was normal, aside from bilateral panretinal photocoagulation scars. No lymphadenopathy was found. The remainder of her examination was normal.
Laboratory and radiologic evaluations
Admission laboratory values are listed in Table 1. A complete blood count and differential cell count were normal. Renal function was normal at baseline. Blood glucose was mildly elevated. Ionized calcium was expectedly low following an IV bisphosphonate infusion. Blood cultures were drawn. An orbital computed tomography (CT) scan with IV contrast (Figure 2) revealed diffuse inflammatory stranding within the right retrobulbar fat and thickening of the right posterior sclera and preseptal soft tissue inferiorly. The paranasal sinuses contained no fluid collection or evidence of subperiosteal abscess.
WBCs = white blood cells; BUN = blood urea nitrogen.
Ionized calcium, mmoles/liter
The patient is a 55-year-old woman with well-controlled type 1 diabetes mellitus, diabetic retinopathy, gastroesophageal reflux, gastroparesis, and postmenopausal osteoporosis who presented with a 2-day history of nontraumatic unilateral eye pain associated with periorbital edema, chemosis, and restricted extraocular movements. She had no associated fevers, leukocytosis, or clinical or radiographic evidence of sinusitis.
The clinical presentation of acute-onset, progressive unilateral periocular edema, chemosis, and ocular motility restriction raises concern for an orbital inflammatory process. CT or magnetic resonance imaging (MRI) is helpful for discerning whether orbital inflammation is diffuse or localized to a specific tissue type within the orbit, as in the case of dacryoadenitis (lacrimal gland) or myositis (extraocular muscle). Our patient exhibited radiographic findings of diffuse orbital inflammation, which in adults can arise from a number of conditions, including infectious, inflammatory, thyroid, neoplastic, traumatic, or idiopathic etiologies (1, 2) (Table 2).
Table 2. Differential diagnosis of orbital inflammatory disease*
Orbital cellulitis should first be excluded because it can cause irreversible visual compromise and life-threatening complications. It often presents acutely in the setting of antecedent periocular skin trauma or underlying sinus disease and is accompanied by fever and leukocytosis. Bacteria such as Staphylococcus or Streptococcus are the most common causes, although fungi such as Mucormycosis or Aspergillus should be considered, particularly in hosts with poorly controlled diabetes mellitus, renal transplantation, or chronic immunosuppression (3). Viral causes such as herpes zoster have been reported, but are rare (4). Although this patient lacked typical infectious features such as fever, leukocytosis, or sinus disease, the possibility of a bacterial orbital cellulitis could not be excluded entirely and initially prompted empirical broad-spectrum antibiotics. Fungal infection was less likely in the absence of underlying sinus disease or poor glycemic control.
Systemic inflammatory conditions such as sarcoidosis may present with acute diffuse orbital inflammation. Clinical features of sarcoidosis are variable and include pulmonary, hematologic, skin, musculoskeletal, neurologic, endocrine, and ophthalmic manifestations. Ophthalmic disease occurs in up to 50% of cases and is the presenting symptom in a subset of patients (5). Orbital involvement is characterized by insidious bilateral lacrimal gland enlargement and rarely by unilateral diffuse orbital inflammation. Hilar adenopathy supports the diagnosis, which is usually confirmed by lymph node or conjunctival biopsies (6).
Diffuse orbital inflammation can be a manifestation of a small- to medium-vessel vasculitis, particularly granulomatosis with polyangiitis (Wegener's) (GPA) (2). GPA is a necrotizing granulomatous vasculitis characterized by sinopulmonary disease and pauciimmune glomerulonephritis. Ocular involvement occurs in 25–50% of patients and manifests as scleritis, episcleritis, or uveitis. Diffuse orbital involvement is less common and often spreads from contiguous sinus granulomatous inflammation, although it may also originate in the orbits. Diagnosis is based on systemic clinical features and demonstration of vasculitis and granulomatous inflammation on tissue biopsy (7). Antibodies to proteinase 3 are highly specific for GPA, but a negative test should not exclude this diagnosis, especially in patients with limited disease that spares the lungs or kidneys (8). Systemic causes of orbital inflammation typically present with additional extraorbital findings that help elicit the diagnosis. Our patient presented with isolated orbital inflammation, a normal examination, and an unremarkable review of systems.
Thyroid orbitopathy, which is characteristically bilateral, is nonetheless the most common cause of unilateral proptosis in adults. Lid retraction is the most common early clinical finding, whereas more advanced disease can lead to proptosis, ocular motility restriction, exposure keratopathy, and compressive optic neuropathy (9, 10). CT or MRI typically shows orbital adipose changes and fusiform enlargement of the inferior, medial, and superior rectus muscles that spares the tendon insertions (11). Although acute exacerbations are possible, thyroid orbitopathy typically follows a chronic course (9, 10). Our patient's presentation was rapidly progressive and lacked the characteristic radiographic features.
Malignancy is an infrequent cause of orbital inflammation. Orbital lymphoma, uveal melanoma, and metastases have been reported and follow an insidious clinical course (12–14). CT or MRI and tissue biopsy are required for the diagnosis. Treatment is largely directed at the underlying neoplastic process.
Traumatic orbital fractures and subsequent edema and retrobulbar hemorrhage can cause significant proptosis, hemorrhagic chemosis, ocular motility restriction, and optic neuropathy, which mimic orbital inflammation. A history of trauma is essential and was not reported by our patient (1).
Finally, idiopathic orbital inflammation (also known as orbital pseudotumor) is a nonmalignant process that should be considered in the absence of the aforementioned etiologies. It is a diagnosis of exclusion based on clinical, radiographic and, if necessary, histopathologic data. The process can involve any orbital structure and usually presents with abrupt-onset eye pain, swelling, proptosis, ophthalmoplegia, and redness (2).
The patient was admitted to the hospital. Intravenous ampicillin/sulbactam and vancomycin were started due to concern for orbital cellulitis. Her temperature and leukocyte count remained normal. Serum blood glucose averaged at 183 mg/dl during the admission. Bacterial and fungal blood cultures revealed no growth.
Over the next 48 hours, her condition worsened. Her extraocular movements became restricted. Repeat orbital CT showed increased orbital fat stranding and inflammatory changes of the right extraocular muscles, posterior sclera, and optic nerve sheath.
In the absence of sinusitis or systemic symptoms, the diagnosis of idiopathic orbital inflammation or “orbital pseudotumor” was considered. A literature review revealed several cases of bisphosphonate-associated acute orbital inflammation whose features resembled those of our patient. IV methylprednisolone 1 gm was administered based on reported benefits in published similar cases. The patient's eye pain and swelling improved dramatically over the subsequent hours. A second dose was given the following day with further improvement (Figure 3). Antibiotics were discontinued 4 days after their initiation, and the patient was discharged on a 3-week prednisone taper (initial dosage 60 mg/day). Followup ophthalmologic examinations 1 and 5 weeks later (the latter after completion of the steroid taper) demonstrated no residual periorbital changes.
Ophthalmic complications of oral and IV bisphosphonates have been well described in clinical trials and postmarketing surveillance. They range from eye pain (in 2% of patients) to ocular inflammation, including conjunctivitis, scleritis, episcleritis, and uveitis (in 0.1–0.2% of patients) (15–17). Acute, diffuse orbital inflammation that mimics orbital cellulitis via proptosis, conjunctival chemosis, posterior scleritis, and ophthalmoplegia, however, has been rarely described.
We performed a PubMed search using the terms “orbital inflammation” and “bisphosphonate” together, reviewed all references of pertinent articles, and identified 17 cases of acute bisphosphonate-induced orbital inflammation published in the English literature (18–32) (Table 3). We also garnered additional cases from the National Registry of Drug-Induced Ocular Side Effects (www.eyedrugregistry.com) located at the Casey Eye Institute in Portland, Oregon. There were no other reports of bisphosphonate-induced diffuse orbital inflammation. Patients were ages 55–89 years and received a bisphosphonate for approved indications such as bony metastases, multiple myeloma, osteoporosis, and Paget's disease of bone, as well as for off-label uses, including osteonecrosis and Charcot joint. The majority of the patients received an IV aminobisphosphonate (either pamidronate or zoledronic acid) (19–32). Two patients received oral alendronate (18). In cases of IV bisphosphonate exposure, symptoms began within hours up to 6 days following drug administration (19–32); cases involving oral alendronate presented 10 days and 3 weeks after drug initiation (18). Systemic symptoms were present in one-third of patients (18–20, 23, 27, 29, 32). Ophthalmic symptoms primarily included unilateral eye pain (89%) (18–30, 32), diplopia (50%) (18–21, 23, 27, 31), and blurry vision (33%) (19, 24, 25, 27, 29, 32). Ophthalmic signs included lid edema (94%) (18, 20–32), conjunctival hyperemia (83%) (18, 20–32), conjunctival chemosis (78%) (19–29, 31, 32), motility defect (61%) (18–21, 23, 24, 27, 29, 31, 32), proptosis (50%) (19, 22–24, 26, 28, 31, 32), anterior uveitis (22%) (18, 23, 25, 32), fundus abnormalities (17%) (18, 22, 25), and an afferent pupillary defect (17%) (19, 24, 25). Orbital imaging, when available, revealed pre- and/or postseptal fat stranding, extraocular muscle inflammation, and scleral thickening. Antibiotics were administered to 6 of 18 patients without symptomatic improvement (19, 22, 24, 29, 30). All but 2 patients required treatment with systemic steroids (21, 29), which improved symptoms. In most cases, the bisphosphonate was discontinued. One patient who was continued on alendronate experienced recurrent symptoms upon discontinuation of systemic steroids (18). A second patient who had received zoledronic acid was switched to pamidronate and experienced mild conjunctivitis and hyperemia (22), but experienced no further symptoms after repeated dosing and steroid pretreatment. A third patient experienced recurrent eye symptoms on repeated pamidronate infusion; these resolved spontaneously (21). A fourth patient continued on monthly zoledronic acid without symptom recurrence (23).
Table 3. Summary of cases of bisphosphonate-induced diffuse orbital inflammation (including the current study)
Two cases with onset 10 days and 3 weeks post–alendronate initiation.
Three of 4 patients experienced recurrent orbital inflammation upon rechallenge.
Case characteristics (n = 17)
Age, range years
Skeletal metastases or multiple myeloma (7)
Paget's disease of bone (2)
Charcot joint (1)
Knee osteonecrosis (1)
Bisphosphonates given (no.)
Time from exposure to onset of eye manifestations, range days
The Naranjo algorithm is a validated tool designed to assess the likelihood of an adverse drug reaction and categorizes associations as doubtful, possible, probable, or definite (33). Bisphosphonate-induced orbital inflammation in our patient's case is possible or probable based on the temporal relationship to drug administration, lack of a clear alternate explanation, and similar case reports in the literature. Although a rechallenge with zoledronic acid and subsequent reoccurrence of orbital symptoms would have strengthened the association and would have supported causality, the severity of the patient's presentation precluded this test.
A recent case–control study of 69 patients presenting with their first episode of idiopathic orbital inflammation between 2000 and 2006 and compared to 296 adult controls with retinal detachment found an increased risk of orbital inflammation in users of an oral bisphosphonate (odds ratio 8.68, 95% confidence interval 1.16–65) (34), a finding that further adds support to this association.
Bisphosphonates, specifically IV aminobisphosphonates, have been associated with an acute-phase reaction (APR) characterized by fever, musculoskeletal pain, gastrointestinal symptoms, eye inflammation, and fatigue. A recent detailed analysis of the adverse events recorded in the Health Outcomes and Reduced Incidence with Zoledronic Acid Once Yearly pivotal fracture trial, which compared IV zoledronic acid to placebo for the treatment of postmenopausal osteoporosis, reported a 42% rate of APRs in the treatment arm compared to 11% in the placebo group. Ninety percent of the APRs were rated as mild to moderate, and 0.6% in the treatment arm had eye pain or inflammation compared to 0.1% in the placebo group (35). Fever following bisphosphonate infusion has been associated with a release of cytokines, specifically interleukin-6 and tumor necrosis factor α, that results from activation of γ/δ T cells via inhibition of farnesyl diphosphate synthase (36–38).
Aminobisphosphonates such as pamidronate and zoledronic acid inhibit the osteoclast cholesterol synthesis pathway and induce osteoclast apoptosis, thereby decreasing bone resorption. Specifically, these drugs inhibit farnesyl diphosphate synthase and the subsequent formation of lipid side chains required for GTP binding protein localization to the osteoclast cell membrane. This causes accumulation of the lipid side chain precursors, isopentyl pyrophosphate (IPP) and dimethylallyl diphosphate, which are potent activators of γ/δ T cells. In vitro treatment of peripheral blood cells with zoledronic acid results in the accumulation of these precursors in monocytes, which in turn have been shown to activate γ/δ T cells via direct cell contact (39).
Interestingly, in vitro studies have shown that hydroxymethylglutaryl-coenzyme A reductase inhibitors coadministered with bisphosphonates attenuate γ/δ T cell activation by blocking IPP formation (40). Subsequent clinical trials failed to demonstrate benefits of statins over placebo in the prevention of bisphosphonate-induced APR, however, and may be a consequence of the strong first pass hepatic metabolism of statins (41, 42). Fortunately, the incidence of APRs decreases with subsequent bisphosphonate infusions, an effect not due to patient dropout (41). The severity of APRs may also be reduced in patients pretreated with acetaminophen (35, 41).
We speculate that this and other similar cases of dramatic post–bisphosphonate orbital inflammation represent an extreme APR phenomenon. It is unclear why the orbital tissue is uniquely susceptible. While most APRs resolve after a few days, only 2 of 18 orbital inflammation cases were self-limited (21, 29). Systemic glucocorticoids were rapidly effective in all reported cases. A placebo- controlled study to determine the efficacy of glucocorticoids is not feasible, however, due to the extremely low incidence of this complication. Consequently, treatment will likely remain anecdotal. Route and dose of glucocorticoids should be individualized based on severity of orbital findings and concerns of optic nerve compromise. Finally, it is unknown if patients already taking glucocorticoids have a reduced risk of ophthalmic adverse events after bisphosphonate exposure.
Rheumatologists should be aware of the bisphosphonates' potential ophthalmic adverse effects and should educate their patients accordingly. Patients who develop otherwise unexplained recurrent or recalcitrant uveitis, episcleritis, or scleritis during oral bisphosphonate therapy should discontinue the drug. Early recognition of orbital inflammation after IV bisphosphonate exposure may prevent unnecessary invasive procedures or antibiotic exposure and may expedite treatment with glucocorticoids. Although several patients in the literature were rechallenged with an IV bisphosphonate and experienced no recurrence (23), mild recurrence with steroid pretreatment (22), or even self-limited recurrence (21), most cases were severe and a few were vision threatening (19, 24, 25). We therefore suggest avoidance of bisphosphonate rechallenge unless absolutely necessary, such as in the treatment of painful bony metastases. With osteoporosis treatment, clinicians should consider alternative therapies.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Schwab had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Schwab, Harmon, Bruno.
Acquisition of data. Schwab, Harmon, Bruno, Kim.
Analysis and interpretation of data. Schwab, Harmon, Fraunfelder.
The authors would like to thank Dr. Jim Rosenbaum for his helpful review of the manuscript.