Are isolated ocular motor nerve palsies in the elderly truly ‘low risk’ for abnormal neuro-imaging outcomes?

Abstract

Background

Neurologically isolated ocular motor nerve palsies often present a management dilemma. Neuroimaging is more likely to be offered to patients <50 years without coexisting ischaemic risk factors as their risk of sinister underlying causes is thought to be higher. However, populations are rapidly ageing and advanced neuroimaging is now more widely available. We thus investigated the incidence of abnormal neuroimaging outcomes in the traditionally low-risk older patient group.

Methods

This is a retrospective cohort study of 353 patients presenting with isolated ocular motor nerve palsies to a tertiary neuro-ophthalmology service in Singapore over a four-year (2015 to 2019) period. Clinical data was obtained through manual review of case records. Common aetiologies, age-based differences in prevalence of causes and abnormal neuroimaging outcomes were statistically analysed.

Results

Abnormal neuroimaging outcomes were significantly greater in the younger cohort only when age segregation was performed at 60 years of age. In a multivariate analysis, acute onset rather than ischaemic risk factors were independently predictive of normal neuroimaging outcomes. After adjusting for prior cancer risk and clinical bias from presumed ischaemic palsies, abnormal neuroimaging outcomes were seen in 14.1% ≥ 50 yrs, 10.9% ≥ 60 yrs and 15.1% ≥ 70 yrs.

Conclusions

In patients presenting with isolated ocular motor nerve palsies, acute onset may be a more reliable indicator of an ischaemic palsy rather than advanced age or presence of ischaemic risk factors. If onset is not acute, neuroimaging should be considered irrespective of age and coexisting ischaemic risk factors.

Introduction

Neurologically isolated ocular motor nerve palsies involving cranial nerves (CN) III, IV or VI are a common clinical presentation. As the underlying pathology can range from microvascular ischaemia to sinister intracranial pathology such as compressive tumours and stroke, deciding whether or not to neuroimage is often a difficult judgement call [1].

Traditionally, isolated ocular motor nerve palsies in older patients with co-existing ischaemic risk factors such as diabetes mellitus, hypertension and hyperlipidaemia are thought to be largely ischaemic in nature. A ‘watchful waiting’ approach without neuroimaging has therefore been advocated for these patients, given that ischaemic palsies are expected to self-resolve within 8–12 weeks [2]. An arbitrary age cut-off of 50 years is most often used to define the ‘low-risk’ older population group [3–7]. This has been supported by historical studies suggesting that those aged ≥50 are less likely to have underlying non-ischaemic aetiologies [8]. However, more recent prospective studies have indicated that a significant proportion (10–30%) of patients ≥50 years may have underlying non-ischaemic aetiologies which can be diagnosed on neuroimaging [3, 4, 7]. Other studies uphold that neuroimaging yield is low in patients ≥50 years, with a prospective study of 93 such patients finding only one with a non-ischaemic aetiology [5].

With changes in population epidemiology, especially with increasing life spans and the associated higher incidence of cancer, many clinical practices are seeing an increase in the average age of their isolated ocular motor nerve palsies patient pool. Clinical practice has further changed with advances in the sensitivity and acceptability of neuroimaging techniques. However, neuroimaging for older patients with isolated ocular motor nerve palsies, particularly with coexistent ischaemic risk factors, remains a point of debate [4, 9]. It has been suggested that a higher age cut-off should be used for delineating the low-risk older age group [9]. This study was performed with the aim of further understanding the epidemiology and abnormal neuroimaging outcomes in these palsies, especially in the older ‘lower-risk‘ age group, to better inform subsequent management decisions.

Methods

Study design and data collection

This was a retrospective cohort study of all patients presenting to our Neuro-ophthalmology service at the National Eye Centre, diagnosed with isolated ocular motor nerve palsies from January 2015–December 2018. Patients were identified based on diagnostic codes for CN III, CN IV and CN VI palsies. Clinical data was obtained through manual review of case records. Ischaemic risk factors were defined as presence of diabetes mellitus, hypertension, hyperlipidaemia, ischaemic heart disease, cerebrovascular accident or smoking. Patients were excluded if they had multiple ocular motor nerve palsies or if clinical details were unavailable. The study was performed in accordance with guidelines by the SingHealth Central Institutional Review Board.

Cases were defined as acute if they presented within 7 days of onset, and chronic if they presented after 7 days of onset. Aetiology was determined based on clinical course and outcome of investigations. Ischaemic aetiology was ascribed if the patient had ischaemic risk factors, the palsy was of acute onset, resolved spontaneously within 6 months and neuroimaging was not abnormal. In patients without neuroimaging, ischaemic aetiology was also ascribed if there was spontaneous recovery of the palsy. All other causes were diagnosed based on neuroimaging findings. In post-viral and herpetic cases, pre-existing viral illness or concomitant herpetic lesions informed the diagnosis despite normal neuroimaging. Isolated ocular motor nerve palsies caused by radiation or intracranial bleeds were categorised as ‘other intra-cranial’ aetiology. Incidental neuroimaging findings of questionable clinical significance which did not explain the palsy anatomically were not classified as abnormal neuroimaging outcomes. Cause was ascribed as ‘unknown’ for patients lost to follow-up without full investigation or resolution.

Neuroimaging protocol

As per standard practice in our centre, neuroimaging was offered to all patients who presented with isolated incomplete CN III palsy. CN VI palsies were offered neuroimaging at presentation if they had a past cancer history or lack of any ischaemic risk factors. CN IV palsies were likewise not always offered neuroimaging at presentation as they are, if not ischaemic, often traumatic or congenital [10, 11]. Patients who did not undergo neuroimaging at presentation (patient choice or lack of clinical indication initially) were offered neuroimaging at follow-up if the palsy worsened or did not resolve. MRI orbit and brain with contrast, with additional MR angiography in select cases, was the standard neuroimaging protocol used. ‘Abnormal neuroimaging’ refers to cases where abnormal findings on neuroimaging explain the isolated ocular motor nerve palsies.

Statistical analysis

Prevalence of ischaemic vs non-ischaemic aetiologies between specific subgroups of patients were compared using Fisher’s exact test (two-tailed). To compare the effect of age, sex, race, type of palsy, onset of palsy, presence of ischaemic risk factors, history of cancer or recent viral illness on the odds of abnormal neuroimaging, a multivariate logistic regression analysis was carried out and significance tested using Wald’s test. Fisher’s exact tests were performed using GraphPad Prism version 6.04 for Mac, GraphPad Software (La Jolla, California, USA). All other statistical analyses were conducted using R version 4.0.0 for Mac, R Foundation for Statistical Computing (Vienna, Austria) with the epiDisplay package [12]. Statistical significance was defined as p < 0.05.

Results

353 patients met inclusion criteria during the time period of the study. Patients ≥50 yrs of age accounted for 85.6% (302/353) of the population. The mean age of the study population was 63 years and the majority (74.8%, 264/353) were of Chinese ethnicity. CN III palsy was seen in 19.8% (70/353); CN VI accounted for almost half the palsies seen (46.7% 165/353) and a third of the patients had CN IV palsies (33.4%, 118/353). (Supplementary Table 1). 19 cases were lost to follow-up and excluded from further analysis on aetiologies.

Over 50% of all cases in our cohort were of ischaemic aetiology and this was true of each individual isolated ocular motor nerve palsy subgroup as well (56.7%, 38/67 of CN III; 61.9%, 96/155 of CN VI and 59.8%, 67/112 of CN IV). A detailed breakdown of the spread of aetiologies for individual palsies is depicted in Supplementary Fig. 1. If not ischaemic, CN III and CN VI palsies were most likely to be related to compressive tumours while CN IV palsies were likely to be congenital.

Neuroimaging was performed most frequently in isolated CN III palsies (90.0%, 63/70) and CN VI (73.9%, 122/165) palsies; less so in CN IV (55.1%, 65/118). Overall, 30.0% (75/250) of patients who underwent neuroimaging were found to have abnormalities. This was 31.7% (20/63) in CN III palsies, 26.9% (45/122) in CN VI palsies and 15.4% (10/65) in CN IV palsies. Univariate analysis showed that CN III (odds ratio [OR] 2.56, 95% confidence intervals [CI] 1.09–6.03) and CN VI (OR 2.89, 95% CI 1.34–6.24) palsies were correlated with significantly higher odds of abnormal neuroimaging.

Ischaemic aetiology was more common in patients aged ≥50 years (64.8%, 190/302) than those aged <50 years (26.8%, 11/51) (Fig. 1A, B). The proportion of patients neuroimaged in both age groups was similar (69.5% of ≥50 and 78.4% of <50 years, p > 0.05, Fisher’s exact test). Among the neuroimaged patients, frequency of intracranial abnormalities explaining the isolated ocular motor nerve palsies was similar in the ≥50 years and <50 years age groups (29.5%, 62/210 in ≥50 years compared to 32.5%, 13/40 in <50 age group; p > 0.05, Fisher’s exact test) (Fig. 1C, D).

Fig. 1. Aetiologies and neuroimaging outcomes of isolated ocular motor nerve palsies by age group.

A, B Aetiologies of isolated ocular motor nerve palsies from least to most common in (A) patients ≥50 years old and (B) patients <50 years old. C, D Abnormal neuroimaging rates and outcomes in (C) patients ≥50 years old and (D) patients <50 years old.

Even after excluding all patients with prior cancer history from the analysis, abnormal neuroimaging outcomes remained equivalent between the two groups (20.2%, 33/163 in ≥50 group and 27.3%, 9/33 in the <50 group) (p > 0.05, Fisher’s exact test).

To control for selection bias in the neuroimaged subgroup, when analysing the yield of abnormal neuroimaging in ‘low-risk’ patients (≥50 years old with ischaemic risk factors, but with no other risk factors for non-ischaemic aetiologies), we included in the total number of cases all patients who had been ascribed as having presumed ischaemic aetiology without any neuroimaging. With this adjustment, 13.1% (26/199) of ‘low-risk’ patients were found to have abnormal neuroimaging outcomes. These 26 patients, despite their older age group, exclusive presence of only ischaemic risk factors in their clinical history and lack of other red flags, had a range of sinister intracranial pathologies requiring urgent care. This included 8 cases with vascular anomalies (aneurysms, arteriovenous fistulas), 6 cases with intracranial tumours, and 9 cases of ischaemic or haemorrhagic stroke.

Patients ≥60 years had a significantly lower rate of abnormal neuroimaging than those <60 years (24.5%, 39/159 vs 39.6%, 36/91 respectively, p < 0.05). This difference remained significant after excluding patients with prior cancer history (16.1%, 20/124 in over-60s vs 30.6%, 22/72 in under-60s, p < 0.05) (Fig. 2). In the over-60 cohort with only ischaemic risk factors, including those with presumed ischaemic aetiology not confirmed by neuroimaging, the abnormal neuroimaging rate was 11.8% (19/161).

Fig. 2. Abnormal neuroimaging yield using different age cutoffs.

Abnormal neuroimaging yield in the older and younger age cohorts, using different age thresholds. Using an age cutoff of 60 resulted in a significantly higher neuroimaging yield in the younger age group (Fisher’s exact test, p < 0.05).

There was no significant difference in the rate of abnormal neuroimaging between old and young groups with age of 70 years as the cut off (31.3%, 25/80 vs 29.4%, 50/170 respectively, p = 0.76); and this was unaffected by removing patients with prior cancer history (23.3%, 14/60 vs 20.6%, 28/136 respectively, p = 0.33) (Fig. 2). In this over-70 cohort, among patients with only ischaemic risk factors and those with presumed ischemic aetiology, 16.0% (13/81) had abnormal neuroimaging outcomes.

In our cohort of 353 cases of isolated ocular motor nerve palsies, 160 were acute (onset ≤1 week prior to presentation), 190 were chronic and 3 had unknown onset. The acute subset had a significantly greater proportion of ischaemic aetiology compared to chronic (78.2%, 122/156 vs 44%, 77/175, p < 0.001) (Supplementary Fig. 2). Patients with an acute isolated ocular motor nerve palsy also had a significantly lower yield of abnormal neuroimaging compared to those with a chronic isolated ocular motor nerve palsy (16.4%, 18/110 vs 41.6%, 57/131, p < 0.001).

136 patients in the acute group had no prior history of trauma or cancer and their median age was 66 years (range 26–89) (Table 1). 65.4% (89/136) underwent neuroimaging and only 12.3% of those neuroimaged (11/89) had abnormal neuroimaging outcomes. These 11 patients had a median age of 70 (range 34–89) and neuroimaging abnormalities included compressive tumours [3], strokes [3], other intra-cranial pathology [2], compressive non-tumour [1] and inflammatory [1].

Table 1.

Characteristics of acute onset and chronic onset isolated ocular motor nerve palsies.

	Acute onset (≤1 week)	Acute onset (≤1 week) with no risk factorsa	Chronic onset (>1 week)
Number	n = 160	n = 136	n = 190
Median age (range)	66 (26–89)	66 (26–89)	64 (6–95)
% neuroimaged	68.8% (110/160)	65.4% (89/136)	72.1% (137/190)
% yield of abnormal neuroimaging	16.4% (18/110)	12.3% (11/89)	41.6% (57/137)
Number with known aetiology	n = 156	n = 132	n = 175
% ischaemic aetiology	78.2% (122/156)	87.9% (116/132)	44.0% (77/175)

^aNo previous trauma or cancer. One patient with decompensated congenital CNIV palsy was also excluded from this sub-analysis.

A multivariate regression analysis controlling for age, sex, race, palsy type, ischaemic risk factors, history of cancer, history of viral illness and onset of palsy showed that presence of ischaemic risk factors was not an independent predictor of abnormal neuroimaging outcomes (adj.OR 0.66, 95% CI 0.31–1.43); that prior cancer history was a strong independent predictor of abnormal neuroimaging (adj.OR 5.50, 95% CI 2.71–11.16); and that acute onset (≤1 week) was independently a strong negative predictor of abnormal neuroimaging (adj.OR 0.26, 95% CI 0.13–0.53) (Table 2).

Table 2.

Risk factors associated with abnormal neuroimaging.

Factors	Crude OR (95% CI)	Multivariatea OR (95% CI)	p-value
Age ≥50 years old	0.81 (0.39–1.68)	0.79 (0.31–2.03)	0.629
Male sex	0.963 (0.55–1.68)	1.01 (0.54–1.89)	0.983
Race (ref. = Chinese)
Indian	0.36 (0.08–1.65)	0.42 (0.08–2.17)	0.301
Malay	0.63 (0.20–1.96)	0.72 (0.21–2.40)	0.588
Others	1.34 (0.53–3.37)	1.58 (0.51–4.91)	0.433
Palsy (ref. = CNIV)
CNIII	2.56 (1.09–6.03)	2.86 (1.12–7.30)	0.028
CNVI	2.89 (1.34–6.24)	3.29 (1.40–7.72)	0.006
Ischaemic risk factors present	0.52 (0.29–0.92)	0.66 (0.31–1.43)	0.294
Cancer history present	6.33 (3.30–12.12)	5.50 (2.71–11.16)	<0.001
Viral illness present	0 (not calculable)	0 (not calculable)	-
Acute onset	0.25 (0.13–0.46)	0.26 (0.13–0.53)	<0.001

^aMultivariate analysis controlled for the following variables: age, sex, race, palsy type, presence of ischaemic risk factors, presence of cancer history, presence of viral illness and acute onset.

Discussion

Outcomes from this large retrospective study challenge previous paradigms in our practice patterns when managing isolated ocular motor nerve palsies.

When analysing isolated ocular motor nerve palsies, most previous studies concentrate on the age based distinction between ischaemic and non-ischaemic palsies in their analysis [6–8]. While the distinction between ischaemic and non-ischaemic aetiologies is important given their diverse clinical course (70% of ischaemic palsies are thought to self-resolve within 3 months) [13], it is important to note that congenital, inflammatory or traumatic causes are also associated with normal neuroimaging and usually warrant little intervention. In many of these cases, the diagnosis is based on history; and neuroimaging is often used, if at all, as an adjunct to rule out other pathology. As seen here and in previous studies, these non-ischaemic aetiologies are more common in the younger age group. Therefore, while they contribute to higher numbers of non-ischaemic palsies, they may not form part of the abnormal neuroimaging group. On the other hand, non-ischaemic aetiologies such as malignancy, aneurysms or demyelinating disease require more urgent intervention and are invariably associated with abnormal neuroimaging. We believe that rather than the aetiological classification into ischaemic and non-ischaemic palsies, assessment of abnormal neuroimaging is therefore of greater utility when analysing a cohort of isolated ocular motor nerve palsies with the intention of guiding management principles.

Our data suggests that advanced age (as defined by age over 50 years) and presence of ischaemic risk factors alone does not reliably correlate with an ischaemic isolated ocular motor nerve palsies and normal neuroimaging. By specifically analysing neuroimaging outcomes in our case cohort as opposed to analysing aetiology, we show that while the likelihood of an ischaemic palsy was still greater in the older cohort, there was no difference in the rates of abnormal neuroimaging outcomes between the older and younger cohorts. Interestingly, while an age cutoff of 50 years showed no difference, using an age threshold of 60 years successfully separated out an older population with lower rates of abnormal neuroimaging. This may indicate a need to rethink age threshold for the ‘older population’.

The retrospective nature of our study led to unavoidable selection biases. The two key biases were (1) higher prevalence of previous cancer in older patients which in itself is a risk factor for non-ischaemic aetiology; and (2) bias introduced by clinician’s presumption of ischemic aetiology in select cases that resulted in lack of neuroimaging. We adjusted our analysis to take these into account, and demonstrated that despite controlling for all of these, 10–15% of the older cohort of patients (regardless of age cutoff used) presented with abnormal neuroimaging outcomes which explained their isolated ocular motor nerve palsies and required urgent care. These included compressive, malignant or vascular tumours and intra-cranial haemorrhages. There is a clear risk of missing sinister diagnoses in these patients if decision to neuroimage is based solely on age, presence of ischaemic risk factors and absence of red flags.

Previous studies looking at presumed ischaemic isolated ocular motor nerve palsies [3] also focused on better defining their cases for probable ischaemic aetiology based on acute presentation. Although the retrospective nature of this study limited an accurate clinical characterisation of ‘acute’ onset of the palsy, we used time to presentation as an indirect measure of time of symptom onset; and defined our cases as acute if they presented within 7 days of onset. Acute onset in cases thus defined in our study population showed excellent correlation with normal neuroimaging outcomes; and was found to be an independent predictor of normal neuroimaging in patients with isolated ocular motor nerve palsies regardless of patient age, presence of ischaemic risk factors and absence of red flags in history.

Overall, this study offers a detailed analysis of patients traditionally thought to be low risk for non-ischaemic isolated ocular motor nerve palsies and abnormal neuroimaging, and highlights the profile of cases that could be missed and inappropriately managed if guided by advanced age or ischaemic risk factors alone. We propose that incorporating presence of cancer history and a non-acute onset (>1 week) may better guide appropriate decisions on neuroimaging and subsequent management of patients with isolated ocular motor nerve palsies irrespective of age. Our proposed clinical protocol is summarised in Fig. 3.

Fig. 3. Proposed protocol for management of isolated ocular motor nerve palsises.

New proposed protocol for management of isolated ocular motor nerve palsies, incorporating acuity of presentation, previous cancer history and new age cutoff of 60.

Limitations

The limitations of this study primarily stem from its retrospective nature as detailed above. Time to presentation was used as a proxy indicator of onset and is less accurate than a focused history to elicit this particular detail. Non-neuroimaged patients and patients lost to follow-up affected accurate estimation of the true prevalence of neuroimaging abnormalities. Moreover, there was a likelihood of pre-test bias for neuroimaging. Although we tried to mitigate this in our data analysis, subtle clinical features that prompted the managing clinician to order neuroimaging may not have been captured. There may also have been selection bias of a tertiary service towards more serious structural pathologies not diagnosed in primary or secondary services.

Conclusions

In this large retrospective study of isolated ocular motor nerve palsies, neither age ≥50 nor presence of ischaemic risk factors were predictive of a ‘low risk’ population. Despite controlling for various confounders, 10–15% of patients in the older age group had abnormal neuroimaging outcomes. Acuteness of onset emerged as the most independent predictor of normal neuroimaging outcomes in our cohort. Our data supports consideration of neuroimaging despite older age and ischaemic risk factors if the palsy is not of acute onset. Raising the age threshold to 60 years may also improve risk stratification for patients with isolated ocular motor nerve palsies.

Summary

What was known before

• Isolated ocular motor nerve palsies may be microvascular or a sign of a more sinister underlying pathology.

• Older patients above 50 with ischaemic risk factors are thought to be relatively low risk for sinister underlying pathology.

• This may mean that they do not receive neuroimaging early on in the process of investigation.

What this study adds

• In this four-year cohort of patients in a tertiary eye referral centre, patients aged above 50 were just as likely to have abnormal neuroimaging outcomes as those aged under 50.

• A higher age cutoff of 60, after accounting for patients with a previous cancer history, was a better stratifying factor in this cohort.

• Conversely, an acute onset to presentation of less than 7 days was a more reliable independent predictor than age of normal neuroimaging outcomes.

Author contributions

SYC analysed the data and wrote the manuscript together with SS. JX collected the data. ST, DM, JL and SS conceptualised and helped to supervise the project with SS primarily supervising. All authors discussed the results and contributed to the final manuscript. We thank and acknowledge Prof Umapathi N Thirugnanam for critical comments on the manuscript.

Data availability

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41433-023-02775-7.