Severe systemic Adverse Reactions to Ophthalmic Timolol in a CYP2D6 Homozygous *4 Allele Carrier: A Case Report

Abstract

A woman with ocular hypertension suffered from severe bradycardia, hypotension and syncope attacks in temporal relation with ophthalmic timolol application. Topically applied timolol is nasally absorbed and has been shown to reach potentially relevant systemic concentrations. Timolol is mainly metabolized by CYP2D6, which exhibits interindividual metabolic capacity due to genetic variations. A reactive pharmacogenetic panel test identified the patient as a CYP2D6 homozygous *4 allele carrier, which has been associated with a poor metabolizer phenotype and lacking enzyme activity. Thus, the adverse drug reactions possibly resulted from increased systemic timolol exposure. This case report highlights that pharmacogenetic panel testing can contribute to safe and effective pharmacotherapy, even for topically applied drugs.

Keywords::

• drug safety
• glaucoma
• pharmaceutical care
• pharmacogenetics
• pharmacy

Glaucoma is a collective term for a progressive and neurodegenerative disease of the optic nerve, with a loss of retinal ganglion cells, a reduction of the retinal nerve fiber layer and increasing excavation of the optic nerve papilla [1]. With a prevalence of around 70 million affected people worldwide, and being the second leading cause of irreversible blindness, glaucoma is one of the most relevant diseases in ophthalmology [2]. The main risk factor for the development of the disease is increased intraocular pressure (IOP) [3]. Increased IOP is a consequence of overproduction or inadequate outflow of aqueous humor and usually progresses with age [4,5]. Once the IOP has constantly reached a threshold of >21 mmHg, it is termed ocular hypertension [6]. For the treatment, different drug regimens are used that enhance the outflow or inhibit the production of aqueous humor, such as topical parasympathomimetics (e.g., pilocarpine), prostaglandin analogs (e.g., latanoprost), β-adrenergic antagonist (β-blocker, e.g., timolol), carbonic anhydrase inhibitors (e.g., brinzolamide) or α₂-adrenergic agonists (e.g., brimonidine) [7].

The application of ophthalmic timolol (as single agent or fixed-dose combination) is commonly initiated in the treatment of increased IOP [8]. Timolol is a potent nonselective β-blocker and was first approved to treat glaucoma in 1978 in the USA [9]. Approximately 70% of patients with glaucoma or ocular hypertension respond to ophthalmic timolol, which reduces the IOP by suppressing the formation of aqueous humor by 34% ± 9% (mean ± standard deviation [SD]) [10,11]. This effect can already be measured 20 min after application, reaches its maximum after 1–2 h and persists for 24 h [11,12].

When applying the topical solution to the eye, the active substance is diluted by the lacrimal fluid and then passes through the nasolacrimal duct to the nasal mucosa, where it enters the systemic circulation [8,9,13,14]. Korte et al. showed that the plasma C_max is already reached 15 min after application of ophthalmic timolol to the eye (1.14 ± 0.34 ng/ml [mean ± SD]) [15].

Since the corneal epithelium is considered as a tight barrier for drug absorption, it is expected that less than 10% of the active substance is available at the target structure in the ciliary body [16,17]. Due to its nasal absorption, ophthalmic timolol is considered to be subject to limited first-pass effects [8,9,13].

The mean area under the plasma concentration/time curve from zero to infinity (AUC_0-∞) was measured with 4.78 ± 1.90 ng/ml/h (mean ± SD) after topical administration to the eye and with 6.46 ± 2.49 ng/ml/h (mean ± SD) after intravenous administration, assessing a systemic bioavailability of approximately 80% [15].

Taken together, ophthalmic timolol is nasally absorbed, and due to its bioavailability of 80%, it is considered to reach relevant systemic concentrations that could potentially be high enough to cause systemic adverse drug reactions (ADRs) [8,9,15]. Mild-to-life-threatening ADRs associated with ophthalmic timolol including bradycardia, hypotension, dyspnea, cardiac arrhythmia and progression of congestive heart failure, are described in drug labels [12] and literature [14,18–21].

Volotinen et al. assessed the extensive metabolism of timolol in vitro [22,23]. In total nine metabolites were found in human liver microsomes (M1, M4, M5, M6) and cryopreserved hepatocytes (M1–M9). The major formed metabolite of timolol was identified as the hydroxyl metabolite M1, in both the microsomes (72% of all metabolites) and hepatocytes (30% of all metabolites). The CYP450 enzyme 2D6 (CYP2D6) was detected as the major metabolizing enzyme to be responsible for >90% of the formation of M1. Beyond that, the CYP450 enzyme 2C19 (CYP2C19) was also determined to contribute to the intrinsic microsomal clearance of M1, albeit only to a small extent (<10%). While the clinical impact of CYP2C19 in the metabolism of timolol is still unclear, the relevance of CYP2D6 is widely studied.

It is known that comedication of CYP2D6 inhibitors (e.g., cimetidine, quinidine) with ophthalmic timolol can lead to relevant drug–drug interaction (DDI) resulting in increased timolol exposure [24–26]. In 2011, the European Medicines Agency (EMA) even reevaluated the safety of β-blockers for ophthalmic use. Here, it was agreed that the summaries of product characteristics of β-blockers for ophthalmic use (single-agent and fixed-dose combination) should highlight the probability of systemic ADRs, contraindications and DDIs when CYP2D6 inhibitors are concomitantly used [27]. However, the consideration of an individual’s genetic predisposition potentially affecting timolol exposure, especially due to the main metabolism by CYP2D6, is nowhere implicated.

In general, the effects of genetic variation in CYP2D6 on drug exposure and response have been intensely studied. Referring to the pharmacogenomics knowledge database platform (PharmGKB), CYP2D6 genetic variation is mentioned in the drug label of 77 US FDA-approved drugs, 44 Swissmedic-approved drugs and 22 EMA-approved drugs [28]. CYP2D6 is a highly polymorphic gene that exhibits phenotypes with altered activity influencing the interindividual response to substrate drugs [29]. Four major phenotypes, with different frequencies across the world population, range from poor metabolizer (PM; 0.4–5.4% of individuals), intermediate metabolizer (0.4–11% of individuals), normal metabolizer (NM; 67–90% of individuals) to ultrarapid metabolizer (1–21% of individuals) [30]. The influence of CYP2D6 phenotype on the metabolism of ophthalmic timolol is also described in literature. Individuals classified as CYP2D6 PMs may experience intensified ADRs and even toxicities after ophthalmic timolol use, whereas individuals categorized as CYP2D6 ultrarapid metabolizers may fail to achieve the desired therapeutic effect [31,32]. One study conducted early after timolol approval showed that the AUC was four-times larger in PMs compared with NMs [33]. Further, associations between specific CYP2D6 diplotypes and exercise heart-rate reduction in patients using ophthalmic timolol are known [34,35] and clinically annotated by PharmGKB [36].

Although the involvement of CYP2D6 in the metabolism of ophthalmic timolol seems to be well known, up to now there are no clinical recommendations on how to consider genetic variation of CYP2D6 in the drug therapy of patients with ocular hypertension or glaucoma.

In summary, the patient’s genetic predisposition can influence the pharmacokinetics of timolol and may exacerbate systemic ADRs, which however, is not yet considered in clinical practice. Here, pharmacogenetic (PGx) tests could be performed preemptively to identify the genotype-predicted CYP2D6 phenotype to improve safety of timolol. To emphasize the potential impact of PGx on timolol tolerability and to demonstrate how PGx testing results are interpreted and applied for medication optimization in clinical practice, we report an exemplary patient case.

Case presentation

Clinical case & medication

We present the case of a 71-year-old woman with bradycardia, hypotensive crisis and syncope attacks temporally associated with the application of ophthalmic timolol. The patient was diagnosed with ocular hypertension, which was initially treated with ophthalmic brinzolamide 10 mg/ml (two drops in the morning). However, brinzolamide had to be discontinued after only 3 weeks due to allergic reactions with redness and pruritus. Afterward, the medication was switched to ophthalmic timolol 0.5% (1 drop in the morning). Within the first 6 months after starting timolol therapy, the patient experienced progressive bradycardia (40–60 bpm), hypotension (<100/60 mmHg) and increased dizziness. During physical exertion (e.g., climbing stairs, hiking) the symptoms worsened. In this period, the patient experienced two syncope attacks and was hospitalized three times.

During the third hospitalization, physicians decided to discontinue ophthalmic timolol. Thereafter, a therapy with the prostaglandin derivate latanoprost was initiated, which has been well tolerated by the patient.

Due to the serious ADRs associated with ophthalmic timolol application, the patient was referred to our out-patient study pharmacy by her general physician for a medication review including PGx testing. The evaluation of the case was part of an observational study approved by the local ethics committee (ClinicalTrials.gov identifier: NCT04154553). Written informed consent for genetic testing and health data retrieval was obtained from the patient.

When experiencing the afore-described ADRs, the patient was taking comedication, which are listed in Table 1. Potential DDIs of the comedication were considered in a comprehensive medication review.

Table 1. Substance, dosage and indication of the patient’s medication by the time of cardiovascular adverse drug reactions.

Substance	Dosage (q.d.)	Indication
Timolol, 0.5%	1 drop	Ocular hypertension
Candesartan, 16 mg	1 tablet	Arterial hypertension (stage II)
Atorvastatin, 20 mg	1 tablet	Hyperlipidemia
Acetylsalicyl acid, 100 mg	1 tablet	Secondary prophylaxis of cardiovascular events
Pantoprazole, 40 mg	1 tablet	Prophylaxis NSAID-related gastrointestinal ulcers
Ibuprofen, 600 mg	1 tablet	Chronic pain syndrome (state after accident); arthrosis

q.d.: Daily.

PGx analysis

PGx panel test was conducted with a buccal swab sample by the commercial provider Stratipharm^® by humatrix AG (Pfungstadt, Germany). They apply real-time PCR using Life Technologies QuantStudio 12 k flex (Thermo Fisher, MA, USA) with the respective optimized and commercially available chemistry to determine polymorphisms. The panel test includes 100 polymorphisms in 30 genes. However, in the evaluation of the present case, emphasis was placed on the genes related to the pharmacokinetics and pharmacodynamics of timolol (Table 2). Interpretation of the genotype predicted the patient’s phenotype as CYP2D6 PM (*4 homozygous carrier, predicted activity score: 0). No variation was detected in adrenoceptor β₁ (ADRB1) polymorphisms rs1801252 and rs1801253, but the patient exhibited a heterozygous variation in the guanine nucleotide-binding protein subunit β₃ (GNB3) polymorphism rs5443.

Table 2. Selected results of pharmacogenotyping and phenotype interpretation.

Gene	Variant also tested variants in gene locus	Genotype	Diplotype	Predicted phenotype
ADRB1	rs1801252, rs1801253	A/A (= WT), C/C (= WT)	Not applicable	Substance-specific function
CYP2D6	rs1065852 c.100 C>T (in *4,*10,*114), rs3892097 c.506-1 G>A (in *4) (rs35742686, rs5030655, rs5030867, rs5030865, rs5030656, rs201377835, rs28371706, rs59421388, rs28371725)	T/T, A/A (= WT)	*4/*4	No function (poor metabolizer)
GNB3	rs5443 c.825C>T	C/T	Not applicable	Substance-specific function

WT: Wild-type.

To evaluate the comedication, the genotype-predicted phenotype of CYP2C19 (NM, *1 homozygous carrier), CYP2C9 (intermediate metabolizer, *2 heterozygous carrier, predicted activity score: 1.5) and SLCO1B1 (normal function, *1 homozygous carrier) was further considered in the medication review.

Interpretation & discussion of results

In this case, the patient was identified as a homozygous carrier of the CYP2D6 *4 allele, which codes for nonfunction genes due to a splicing defect. Nearly no enzyme activity is expected resulting in the prediction of the patient’s CYP2D6 PM phenotype. Timolol is predominantly metabolized via CYP2D6, suggesting increased systemic exposure which increases the risk to experience ADRs. In addition to the patient’s genetic predisposition, we checked the patient’s medication for interactions involving CYP2D6. However, no comedication interacting with CYP2D6 was reported by the patient. After interpretation of PGx testing results and consideration of afore-presented literature [31–33], we evaluate the cardiac ADRs to potentially result from the patient’s genetic predisposition (CYP2D6 PM). Still, PharmGKB has annotated this association with only a low evidence level (level 3 of evidence) based on the findings of two studies (PharmGKB ID: 1183690112; 1183689924) [36]. One study showed that patients with CYP2D6 PM phenotype (two nonfunctional CYP2D6 alleles) exhibit increased exposure to aqueous timolol (higher AUC, longer elimination half-lives), which is associated with enhanced heart-rate reduction compared with patients with two or one functional CYP2D6 allele [35]. This was further confirmed in a study showing that CYP2D6 PM phenotype is associated with increased timolol plasma concentration and exercise heart-rate reduction [34].

PharmGKB has also listed a clinical annotation for the ADRB1 polymorphism rs1801252. ADRB1 encodes for the G-protein coupled receptor adrenoceptor β₁ which is mainly expressed in cardiac tissue [37]. The receptor is known to be the main target structure for the sympatholytic effects of β-blockers [38]. This clinical annotation of PharmGKB is again classified with level 3 of evidence (PharmGKB ID: 1444703187) [39] and is based on findings that homozygous and heterozygous variations in ADRB1 polymorphism rs1801252 are associated with a higher sensitivity for ophthalmic timolol with a higher risk for blood pressure reduction in treated patients [35]. However, our patient has no variation in this polymorphism. It seems noteworthy that interpretations based on the PharmGKB annotations need to be handled with caution. Clinical annotations listed by PharmGKB are generally categorized descending from level 1A-4. Level 3 classification is a low evidence score usually depending on preliminary data or single studies [40]. This indicates that there is a need to conduct more studies to provide clinically relevant recommendations.

Beyond that, PGx testing results identified the patient’s heterozygosity of the GNB3 polymorphism rs5443. Heterotrimeric G-proteins play a key role in transmembrane signaling for a huge variety of receptors (e.g., adrenoceptors) [41]. Variations in GNB3 have been shown to affect α- and β-adrenergic responses and the antihypertensive effect of β-blockers [42,43]. One study also associated the GNB3 polymorphism rs5443 with an increased risk of decreased heart rate in patients exposed to orally administered β-blockers [44]. Yet, we considered this finding in the genetic interpretation of our patient case, since there are studies indicating that one drop of 0.5% ophthalmic timolol is equivalent to a 10 mg oral dose [11,45].

At the time of study participation, the patient’s therapy was already changed to latanoprost. Thus, the PGx testing results were also checked for latanoprost and for brinzolamide, which was taken in the past. We found no evidence of drug–gene interactions (DGIs) in the PGx profile for neither of the two substances. We therefore support the decision of the physician to change the therapy to latanoprost.

This case demonstrates that without PGx information, ADR due to DGI might be overlooked, leading to delayed treatment adjustments. It required three hospitalizations within 6 months until the therapy was changed. Especially for the treatment of ocular hypertension, multiple topical drugs are available, which have a comparable or even higher efficacy in lowering the IOP compared with timolol [7].

Thus, therapy change should have been discussed much earlier, so that these severe ADRs would not have emerged at all. Since the ADRs started to progress insidiously, also a reduction of the concentration of aqueous timolol or a change of formulation could have been tried at the beginning of ADRs onset. The patient applied an aqueous solution of 0.5% timolol daily, which is reported to have more cardiac side effects compared with 0.1% timolol hydrogel [46,47]. However, since the patient is a CYP2D6 PM with minor-to-zero CYP2D6 activity, ADRs might even be expected with lower timolol concentrations, so that the complete change to another topical drug (i.e., latanoprost) seems to be more reasonable in this case.

Since a PGx panel test was conducted, it was also possible to analyze the impact of the patient’s genetic predisposition on the comedication (Supplementary Table 1) and on possible future drug therapies, providing preemptive PGx-based recommendations. Of note, the low blood pressure could also have been interpreted as being based on an inappropriate dosage of candesartan. Yet, the patient described a stabilization of the blood pressure after discontinuation of ophthalmic timolol, which contradicts an overtreatment with candesartan. In addition, the patient was diagnosed with arterial hypertension stage two (systolic blood pressure: 160–179 mmHg/diastolic blood pressure: 100–109 mmHg prior to drug therapy), so a single-agent therapy with an angiotensin-II-receptor-subtype-1 antagonist (e.g., candesartan), would rather be categorized as insufficient (undertreated), according to the current European guidelines [48].

In addition to the DGI with timolol, the patient’s CYP2D6 PM status may also have an impact on future medication choices: In particular, with regards to her chronic pain syndrome, certain opioids (e.g., codeine, tramadol) should be avoided. Likewise, antidepressants that are mainly metabolized via CYP2D6 should be avoided or only prescribed with a reduced starting dose to prevent ADRs. In summary, this case report highlights that PGx panel testing may not only identify DGI-causing ADRs, but also provide evidence to evaluate future drug therapies preemptively. This would not have been possible with a single gene test.

Conclusion

A reactive PGx panel test was used to explain cardiac ADRs during ophthalmic timolol treatment in a female patient with ocular hypertension. As aqueous timolol is a commonly used drug for the therapy of elevated IOP, the patient and the referring physician wanted to know whether the experienced ADRs were potentially caused by a DGI. Indeed, we identified the patient as a CYP2D6 PM, which may explain the experienced ADRs. Topical drugs are often assumed to be non-systemically available, which is wrong. Our case showed that topical drugs can not only act systemically, but that they can even be impacted by the genotype of a patient.

Also, we noticed in the interpretation process of the genetic results that there are various references in literature describing systemic ADRs in the context of ophthalmic timolol application, under consideration of the CYP2D6 phenotype. To date, there are no general recommendations on how to consider patients’ individual genetic information for ophthalmic timolol treatment. This supports the need to conduct more research in this area to replicate the already existing associations. However, a Clinical Pharmacogenomics Implementation Consortium recommendation guideline for this DGI is currently under consideration [49]. Such a guideline would facilitate treatment decisions in clinical practice and would raise awareness for potential ADRs at an early stage.

The patient suffered from bradycardia, hypotension and syncope attacks for 6 months. It required three hospitalizations until the physician considered the possibility that the systemic symptoms might have been caused by the ophthalmic timolol and decided to discontinue the drug therapy. It may be assumed that PGx testing at an earlier stage might have reduced the patient’s burden. As we witnessed in this case, it also required three attempts of different drug regimens until the patient experienced a safe and effective pharmacotherapy to treat increased IOP. This finding underlines the potential benefits of a pre-emptive genotyping approach for PGx-associated drugs in clinical practice. It may be hypothesized that such an approach would even be cost-effective in the long term. However, we are aware that we are reporting a single case, which does not allow for generalizability to the overall population. Therefore, the evaluation of cost–effectiveness for a preemptive PGx service in clinical practice should be conducted in prospective implementation–effectiveness studies.

Executive summary

Case report

• A 71-year-old women, diagnosed with ocular hypertension, suffered from bradycardia, hypotension and two syncope attacks in temporal relation with ophthalmic timolol application.

• The patient was hospitalized three times, as the adverse drug reactions intensified over 6 months.

• Pharmacogenetic panel test was conducted to determine if genetic variations might have influenced patient’s drug response.

Selected results of pharmacogenetic panel test

• CYP2D6 *4/*4 (reduced to no enzyme function, poor metabolizer).

Discussion

• Metabolism of topical applied timolol, mainly mediated by CYP2D6, is well described.

• There is evidence on potential clinical outcomes based on the drug–gene interaction between CYP2D6 and timolol, of which clinicians should be aware.

Conclusion

• Patient’s experienced adverse drug reaction were potentially caused by a drug–gene interaction between ophthalmic timolol and CYP2D6.

• Heartbeat rate and blood pressure stabilized after changing the therapy to latanoprost.

• Pharmacogenetic testing at an early stage might have reduced the patient’s burden over 6 months.

• We propose to consider patients’ genetic predisposition in ophthalmic pharmacotherapy, when the respective drug exhibits a pharmacogenetic background.

Author contributions

C Jeiziner, H Meyer zu Schwabedissen, K Hersberger and C Stäuble: conceptualization and study design. A Bollinger and C Jeiziner: investigation and interpretation of genotyping data. A Bollinger: writing original draft preparation. C Jeiziner, H Meyer zu Schwabedissen, K Hersberger, S Allemann and C Stäuble: critical review and editing. C Stäuble: supervision. All authors have read and agreed to the published version of the manuscript.

Ethical conduct of research

The case was collected within the observational study ‘Pharmacogenetic Testing of Patients with unwanted Adverse Drug Reactions or Therapy Failure’, which was conducted according to the guidelines of the Declaration of Helsinki and approved by the local ethics committee of ‘Ethikkommission Nordwest- und Zentralschweiz’ (2019-01452, 31.10.2019). The patient provided written informed consent to use the data for research purposes, as well as for publishing this case report.

Supplementary data

To view the supplementary data that accompany this paper please visit the journal website at: www.tandfonline.com/doi/suppl/10.2217/pgs-2023-0122

Financial disclosure

The authors have no financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Competing interests disclosure

The authors have no competing interests or relevant affiliations with any organization or entity with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

Writing disclosure

No writing assistance was utilized in the production of this manuscript.

Data sharing statement/data availability statement

The genetic data presented in this study are available on request from the corresponding author. The data are not publicly available for ethical and privacy reasons.