Surgical interventions for infantile nystagmus syndrome

Abstract

Background

Infantile nystagmus syndrome (INS) is a type of eye movement disorder that can negatively impact vision. Currently, INS cannot be cured, but its effects can potentially be treated pharmacologically, optically, or surgically. This review focuses on the surgical interventions for INS.

Despite the range of surgical interventions available, and currently applied in practice for the management of INS, there is no clear consensus, and no accepted clinical guidelines regarding the relative efficacy and safety of the various treatment options. A better understanding of these surgical options, along with their associated side effects, will assist clinicians in evidence‐based decision‐making in relation to the management of INS.

Objectives

To assess the efficacy and safety of surgical interventions for INS.

Search methods

We searched CENTRAL, MEDLINE Ovid, Embase Ovid, ISRCTN registry, ClinicalTrials.gov, and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) to 3 July 2020, with no language restrictions.

Selection criteria

We included randomised controlled trials (RCTs) studying the efficacy and safety of surgical options for treating INS.

Data collection and analysis

Our prespecified outcome measures were the change from baseline in: binocular best‐corrected distance visual acuity; head posture; amplitude, frequency, intensity, and foveation period durations of the nystagmus waveform; visual recognition times; quality of life and self‐reported outcome measures; incidence of adverse effects with a probable causal link to treatment; and permanent adverse effects after surgery.

Two review authors independently screened titles and abstracts and full‐text articles, extracted data from eligible RCTs, and judged the risk of bias using the Cochrane tool. We reached consensus on any disagreements by discussion. We summarised the overall certainty of the evidence using the GRADE approach.

Main results

We only identified one eligible RCT (N = 10 participants), undertaken in India. This trial randomised participants to receive either a large retro‐equatorial recession of the horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus, or a simple tenotomy and resuturing of the four horizontal rectus muscles. We did not identify any RCTs comparing a surgical intervention for INS relative to no treatment.

In the single eligible RCT, both eyes of each participant received the same intervention. The participants’ age and gender were not reported, nor was information on whether participants were idiopathic or had sensory disorders. The study only included participants with null in primary position and did not explicitly exclude those with congenital periodic alternating nystagmus. The study did not report funding source(s) or author declaration of interests. The evaluation period was six months.

We judged this study at low risk for sequence generation and other sources of bias, but at high risk of bias for performance and detection bias. The risk of bias was unclear for selection bias, attrition bias, and reporting bias.

There is very uncertain evidence about the effect of the interventions on visual acuity and change in amplitude, frequency, and intensity of the nystagmus waveform. We were unable to calculate relative effects due to lack of data. None of the participants in either intervention group reported adverse effects at six‐month follow‐up (very low‐certainty evidence). There was no quantitative data reported for quality of life, although the study reported an improvement in quality of life after surgery in both intervention groups (very low‐certainty evidence).

Change in head posture, foveation period durations of the nystagmus waveform, visual recognition times, and permanent adverse effects after surgery were not reported in the included study.

We judged the certainty of the evidence, for both the primary and secondary efficacy outcomes, to be very low. Due to a lack of comprehensive reporting of adverse events, there was also very low‐certainty of the safety profile of the evaluated surgical interventions in this population. As such, we are very uncertain about the relative efficacy and safety of these interventions for the surgical management of INS.

Authors' conclusions

This systematic review identified minimal high‐quality evidence relating to the efficacy and safety of surgical interventions for INS. The limited availability of evidence must be considered by clinicians when treating INS, particularly given these procedures are irreversible and often performed on children. More high‐quality RCTs are needed to better understand the efficacy and safety profile of surgical interventions for INS. This will assist clinicians, people with INS, and their parents or caregivers to make evidence‐based treatment decisions.

Plain language summary

What are the benefits and risks of different surgical procedures for infantile nystagmus syndrome (an eye disorder that develops shortly after birth)?

Why is this question important?
Infantile nystagmus syndrome (INS) is an eye disorder that causes involuntary movement of the eyes from side to side, up and down, or in circles. INS typically develops shortly after birth, and persists throughout life. It is often associated with visual problems, such as:
‐ long‐sightedness (when people can see distant objects clearly, but near objects appear blurred);
‐ or short‐sightedness (when people can see near objects clearly, but distant objects appear blurred).

There is currently no cure for INS. However, it is possible to reduce eye movement and improve people’s vision. One of the main options is eye surgery, which involves operating on the muscles that control eye movement. Several different surgical procedures can be used: some procedures involve detaching and reattaching the eye muscles, whereas others require complete removal of these muscles.  

It is unclear whether some surgical procedures for INS have more benefits or risks than others. To find out which surgical procedures work best, we reviewed the evidence from research studies. We were particularly interested in whether different surgical procedures could improve vision and quality of life in people with INS. We also wanted to know about any adverse (unwanted) effects.

How did we identify and evaluate the evidence?
First, we searched for randomised controlled studies, in which people were randomly divided into two or more treatment groups. This makes it less likely that any differences between treatment effects were actually due to differences in the people who received them (rather than due to the treatments themselves, which is what we wanted to find out).

We then compared the results, and summarised the evidence from all the studies. Finally, we rated our confidence in the evidence, based on factors such as study methods and sizes, and the consistency of findings across studies.

What did we find?
We found one study, set in India, that involved a total of 10 people with INS who were followed for six months after surgery. The study authors did not report any information about: 
‐ the age and gender of the people who took part in it;
‐ or the source of funding for the study.

The study compared two different surgical procedures: 
‐ one procedure in which two horizontal rectus muscles (eye muscles that control side‐to‐side eye movements) were moved from their original position to a position further back on the eyeball; and
‐ another procedure in which four horizontal rectus muscles were detached and reattached, in their original position.
The procedures were conducted in both eyes.

The study compared the effects of the treatments on:
‐ clarity of vision six months after surgery;
‐ intensity of eye movement six months after surgery;
‐ adverse events six months after surgery; and
‐ quality of life.  

The study did not investigate the effects of treatments on:
‐ head posture;
‐ the amount of time when the eye is still; or
‐ the time it takes to recognise objects.

We have very little confidence in the evidence from the study we found, because:
‐ it is based on a very small number of people; and
‐ the patients and researchers in the study knew which type of surgery each patient received. This knowledge may have influenced the study results.

We therefore cannot determine whether the two procedures investigated in the study had different benefits and risks.

What does this mean?
There is insufficient evidence to determine whether some surgical procedures are better than others for INS. We need researchers to conduct robust randomised controlled trials in future, so that we can compare different procedures. This would help clinicians, and people with INS, to make treatment decisions based on evidence from research.

How‐up‐to date is this review?
The evidence in this Cochrane Review is current to July 2020.

Summary of findings

Summary of findings 1. Summary of Findings Table ‐ Large retro‐equatorial recession of horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus compared to simple tenotomy and resuturing of the 4 horizontal rectus muscles for the surgical treatment of INS.

Large retro‐equatorial recession of horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus compared to simple tenotomy and resuturing of the 4 horizontal rectus muscles for the surgical treatment of INS
Patient or population: infantile nystagmus syndrome Setting: eye hospital Intervention: large retro‐equatorial recession of horizontal rectus muscle Comparison: simple tenotomy and resuturing
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with simple tenotomy and resuturing	Risk with large retro‐equatorial recession of horizontal rectus muscle
Change in binocular best‐corrected distance visual acuity follow up: mean 6 months	The mean change in binocular best‐corrected visual acuity was 0.1 logMAR lower in the intervention group and 0 logMAR in the comparator group. No measures of variability (e.g., standard deviation) were provided. It is thus not possible to draw conclusions surrounding the relative effect of each intervention on this outcome.			10 (1 RCT)	⊕⊝⊝⊝ VERY LOW a,b
Change in head posture ‐ not reported	‐	‐	‐	‐	‐
Change in intensity of the nystagmus waveform follow up: mean 6 months	The mean change in intensity of the nystagmus waveform was 0.5 degrees cycles/second in both the intervention and comparator groups. No measures of variability (e.g., standard deviation) were provided. It is thus not possible to draw conclusions surrounding the relative effect of each intervention on this outcome.			10 (1 RCT)	⊕⊝⊝⊝ VERY LOW a,b
Change in foveation period durations of the waveform ‐ not reported	‐	‐	‐	‐	‐
Change in visual recognition times ‐ not reported	‐	‐	‐	‐	‐
Change in proportion of participants with adverse effects with a probable causal link to the treatment follow up: mean 6 months	None of the participants in either intervention group had restricted extraocular movements or significant complications as a result of their assigned surgery.			10 (1 RCT)	⊕⊝⊝⊝ VERY LOW a,b
Change in quality of life follow up: mean 6 months	No quantitative data were provided. The authors stated "Based on a questionnaire of subjective postoperative improvement, most patients noticed an improvement in their quality of life after the surgery in both the groups.			10 (1 RCT)	⊕⊝⊝⊝ VERY LOW a,b
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval
GRADE Working Group grades of evidence High certainty: We are very confident that the true effect lies close to that of the estimate of the effect Moderate certainty: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low certainty: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low certainty: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_419696919795401866.

a. Downgraded one level for risk of bias (as neither participants nor outcome assessors were masked)
b. Downgraded two levels for imprecision (as there was only study, with 10 participants, which only provided data related to the effect estimate but without associated measures of precision)

Background

Appendix 1 contains a glossary of medical terms used in the review.

Description of the condition

Nystagmus refers to a group of eye disorders that are characterised by rhythmic eye oscillations, in which the primary abnormality is in the regulation of slow eye movements. A smooth drift of the eyes away from the target (or desired gaze position), is then followed by a corrective saccade, a return to slow eye movement, or both. It can be further characterised as pendular nystagmus, if the corrective eye movement is slow, and as jerk nystagmus, if the corrective eye movement is fast. After strabismus, nystagmus is the second most common paediatric ocular motor disorder (Tychsen 1991). Nystagmus affects from one in 2000 to one in 10,000 people worldwide; congenital forms comprise 80% of cases, whilst acquired forms comprise 20% (Dell’Osso 1975; Sarvananthan 2009). Nystagmus can be either physiological or pathological (Daroff 1978; Dell’Osso 1997). The physiological forms of nystagmus include vestibular, optokinetic, and eccentric gaze (endpoint). Pathological nystagmus is divided into two categories: infantile‐onset (e.g. spasmus nutans syndrome, infantile nystagmus syndrome (INS), fusion maldevelopment nystagmus syndrome; CEMAS 2001), or acquired (e.g. gaze‐evoked, see‐saw; Maybodi 2003).

INS is an involuntary ocular motor oscillation that manifests at, or most often shortly after, birth (Gottlob 1997), and persists throughout life (Dell’Osso 1985). Rarely, it begins in later life (Gresty 1991). People with INS are usually diagnosed in the first few months of life. Estimates of the prevalence of INS range from one in 1000 to one in 6000 births; the disparity is attributed to different populations and ethnicities, different classification schemes, misclassifications, or a combination of these (Forssman 1971; Heemes 1924; Norn 1964; Stayte 1993). Many people with INS have an underlying disorder of the visual system (Abadi 2002; Gelbart 1988; Weiss 1989); approximately 30% of people with INS also have strabismus (Hertle 1999). The negative impact of nystagmus on vision has been described to be more profound than that associated with age‐related macular degeneration in adults (Pilling 2005).

Parameters used to describe INS oscillations include the type of waveform(s), and the amplitude, frequency, intensity, and foveation periods of the waveform (Abadi 1986; Leigh 2006). Foveation periods are those portions of the waveform where the image of the point of regard is on or near the fovea, and the eyes are stationary, or are only moving at a velocity less than some predefined limit. INS is characteristically bilateral and conjugate, with similar amplitude in both eyes. The intensity and foveation duration of the nystagmus, as well as its waveform, may vary in different gaze directions. The waveforms may also change with age (Hertle 1999; Reinecke 1988). An important diagnostic indicator is that this oscillation remains horizontal regardless of gaze position (Abel 2006; Dell’Osso 1997). In some cases, INS may have a latent component, whereby the waveform changes direction when covering one eye (Kestenbaum 1961). INS is reported to become worse with physiological factors, such as stress and motivated behaviours (Cham 2008; Salehi 2018). It usually decreases with voluntary lid closure (Dell’Osso 1997), and often with convergence (Khanna 2006). INS does not occur during sleep and is diminished when the gaze is directed to the null zone (i.e. the field of gaze in which the nystagmus intensity is minimal and the best foveation occurs).

People with INS can exhibit a null zone laterally, and so may adopt an abnormal head posture to shift the eyes into this null position (Abadi 2002; Dell’Osso 1974; Dell’Osso 1975; Hertle 2000a; Stevens 2003). A vertical head posture or tilt may also be present in cases of horizontal nystagmus when the null zone is perpendicular or tilted (Abadi 2002). This compensatory mechanism serves to minimise the nystagmus intensity and maximise visual acuity (Abadi 1991; Stevens 2003). The null region is usually stable, but it may shift in cases of periodic alternating nystagmus, smooth pursuit, or optokinetic stimulation (Khanna 2006). Occasionally, head oscillations, which can increase with visual intent, may also be seen in people with INS (Gresty 1981).

The ability to effectively treat INS has been hindered by the failure to ascertain the exact mechanism(s) underpinning the condition. Currently, INS cannot be cured, but its effects can potentially be treated. Treatment is indicated to improve the following INS characteristics: visual acuity, abnormal head posture, the appearance of the oscillations, effective field of best vision, and oscillopsia. Treatments that lessen the oscillations without affecting normal eye movements are preferred. Several treatment options exist that aim to maximise visual acuity by extending or lowering slow‐phase velocities during foveation periods and minimising oscillopsia, although the latter is rarely problematic (Leigh 1988). The intensity of the nystagmus can be reduced pharmacologically, optically, or surgically (Khanna 2006; Wang 2006a). Visual acuity is first improved by correcting any significant refractive error. Contact lenses can improve visual acuity by resolving ametropia and improving INS waveforms over a range of gaze positions (Biousse 2004; Taibbi 2008). Prisms can be incorporated into the prescription; in cases of people with INS whose nystagmus decreases with convergence, base‐out prisms are used to induce vergence (Spielmann 2000). Several medications have been found to be effective in reducing the intensity of certain types of INS; they can either decrease the nystagmus centrally or peripherally (Comer 2006; Dell’Osso 2011; Thurtell 2010). However, surgery remains the mainstay of INS treatment (Hertle 2010; Hertle 2011; Hertle 2017). Eye muscle surgery has been available since the 1950s, as an option to correct for abnormal head positioning by moving the eyes into the null zone, where the nystagmus is reduced (Anderson 1953; Kestenbaum 1954).

Description of the intervention

Surgical treatment for people with INS is primarily aimed at shifting the null zone to the primary gaze position, or correcting abnormal head tilts or turns, or both. Additional reported benefits include broadening of the null region, minimising nystagmus intensity at all gaze angles, and increasing foveation times and visual acuity (Dell’Osso 1973; Wang 2006a). The age at which these procedures are performed varies from a few months of life, up to adolescence and adulthood. Six main surgical interventions have been used in INS management:

• Anderson‐Kestenbaum procedures involve the recession of the pair of rectus muscles responsible for the direction of the face turn, with or without resection of the antagonist muscles (Anderson 1953; Kestenbaum 1954).

• Vertical recti surgeries (including recession‐resection of the vertical recti, the combined recession of the vertical recti muscles and weakening of the oblique muscles, and full tendon‐width horizontal transposition of the vertical recti muscles) are less common, but are proposed to have a similar mechanism of action as the Anderson‐Kestenbaum procedures. Bilateral superior recti recession, with inferior oblique myectomy, are typically performed for people with INS with chin depression (Hertle 2009). To correct head tilt, the insertion of the vertical recti muscles can be offset horizontally to overcome the cyclo‐torsional effect; e.g. for right head tilt, the following procedures should be performed: right superior rectus nasal horizontal transposition, right inferior rectus temporal transposition, left superior rectus temporal transposition, and left inferior rectus nasal transposition (von Noorden 1993). Bilateral superior oblique weakening combined with inferior rectus recession is proposed for a chin up head posture (Greenberg 2007; Hertle 2010).

• Maximal retro‐equatorial recession of the horizontal recti muscles involves recession of all four horizontal recti around the equator of the globe (medials 10 mm, and laterals 12 mm from insertion).

• Artificial divergence surgery is used when INS decreases with convergence, and this is proven to be an effective null in normal viewing conditions by wearing base out prism glasses, i.e. any eccentric null is markedly reduced or abolished in favour of using the convergence null (Hertle 2000b). This surgery aims to reduce nystagmus by encouraging the person to exert fusional convergence reserves to overcome the induced exophoria (Spielmann 2000).

• Tenotomy‐resuture describes the detachment and reattachment of all horizontal recti to their original insertion, without resection or recession.

• Myectomy involves complete or partial removal of all four horizontal recti muscles, without reattachment.

How the intervention might work

Surgical treatment of INS is most commonly performed in individuals with anomalous head postures secondary to eccentric null regions (Neely 1999). The Anderson‐Kestenbaum resection and recession procedure aims to relocate the nystagmus null zone to the primary position to eliminate any head turn, along with the benefit of reducing the nystagmus itself (Anderson 1953; Kestenbaum 1954; Parks 1973). The artificial divergence procedure is performed in individuals who can decrease their nystagmus with convergence (Cuppers 1971; Spielmann 2000). In people with INS without an abnormal head posture, large recessions of the horizontal recti muscles are undertaken with the intent of reducing the nystagmus amplitude, improving foveation times, and secondarily improving visual acuity and the ability to recognise objects more quickly (Alió 2003; Atilla 1999; Boyle 2006; Erbagci 2004; Helveston 1991; Sprunger 1997; von Noorden 1991).

Tenotomy with reattachment is performed with the aim of reducing the nystagmus, broadening the null region and increasing foveation periods (Hertle 2006; Wang 2006a). Though the mechanism of benefit(s) remains unclear, it has been suggested that disruption of feedback from the enthesis — the junction between the eye muscles and the eyeball — leads to reduction in the oscillation of the eyes (Hertle 2009). In recent years, myectomy of the extraocular muscles, by either completely or largely removing them (Lingua 2016; Sinskey 2002), has also been described in relation to treating INS; however, some raise concerns regarding the procedure’s detrimental and irreversible effects on other eye movements, the ability to maintain single binocular vision, and undesirable changes in the person's appearance (Dell’Osso 2018). 

Vertical and torsional head posture surgeries include the recession‐resection of vertical rectus muscles, combined recession of vertical recti, and weakening of the oblique muscles (Greenberg 2007; Pierse 1959; Roberts 1996; Yang 2004), and full tendon‐width horizontal transposition of the vertical recti (von Noorden 1993). These procedures may be performed when the abnormal head posture is in the vertical plane (chin elevation or depression) or torsional (head tilt or multiplanar positions). When correcting a chin down head posture, the aim, for example, would be to weaken the elevators of the eyes with a combined recession of the superior recti and myectomy of the inferior obliques. For a chin up position, the aim would be a weakening procedure of the ocular depressors, and tenectomy of the superior obliques with a recession of the inferior recti (Hertle 2010).

Why it is important to do this review

Although INS cannot be cured, its effects can potentially be reduced with intervention(s). Several potential options exist, including surgical, pharmacological, and optical treatments. As previously outlined, broadly, these procedures aim to reduce the intensity and the effect on vision of the involuntary ocular oscillations, correct abnormal head posture, and maximise visual performance over as wide a range of gaze angles as possible.

Despite the range of interventions available, and currently applied in practice for the management of INS, there is no clear consensus and no accepted clinical guidelines regarding the relative efficacy and safety of the various treatment options, relative to each other or to no treatment. Whether a specific type of surgery is best for treating a given case of INS remains debatable. Furthermore, heated debate has arisen in the literature between advocates and opponents of certain surgical procedures (Dell’Osso 2018; Hertle 2017). Surgical interventions vary widely in their level of invasiveness, ranging from detachment and reattachment of the extraocular muscles in situ, to complete removal of these muscles. A better understanding of the surgical options, along with their associated side effects, will assist clinicians in evidence‐based decision‐making in relation to the management of INS.

Understanding the evidence surrounding the efficacy and safety of surgical interventions for INS is especially important, given that many surgical strategies are non‐reversible, and are often performed on children. Therefore, a systematic review is essential to inform surgeons, healthcare practitioners, people with INS and their families and caregivers, and the wider community of the best available research evidence, as a basis for informing clinical decision‐making when considering surgical options for INS. This review will also identify gaps in the evidence, in order to inform future research in the field.

Objectives

To assess the efficacy and safety of surgical interventions for infantile nystagmus syndrome (INS).

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) only. As anticipated, there were limited RCT data in the literature. In view of this, per our a priori protocol (Cham 2019), we also included a narrative synthesis of non‐randomised clinical intervention studies, in the Discussion section of the review. We did not summarise the results from lower levels of evidence (e.g., case series or case studies), in view of the inherent limitations and risk of bias of these study designs.

Types of participants

We included studies in which people were diagnosed with infantile nystagmus syndrome (INS, which may also be termed idiopathic infantile nystagmus, congenital motor nystagmus or congenital idiopathic nystagmus, sensory nystagmus or sensory defect nystagmus), as defined by the study authors. We also included studies in which the participants with INS had underlying disorders of the visual system (e.g. optic atrophy, macular hypoplasia, albinism, retinal disorders, etc.).

We excluded studies that included other congenital types of nystagmus (e.g. fusion maldevelopment nystagmus and spasmus nutans) in the study population, and those that included participants with acquired forms of nystagmus.

Types of interventions

We included studies that considered all types of surgical interventions for INS, including:

• Artificial divergence;

• Anderson‐Kestenbaum procedure (and its variations);

• Retro‐equatorial recession of the horizontal recti;

• Four horizontal rectus muscle tenotomy with reattachment;

• Recession‐resection of vertical recti;

• Combined recession of the vertical rectus muscles and weakening of the oblique muscles;

• Full tendon‐width horizontal transposition of the vertical recti;

• Myectomy of the extraocular muscles without reattachment.

We considered a range of comparators, including:

• No intervention;

• Other surgical interventions;

• Non‐surgical interventions;

• Pharmacological interventions;

• Non‐pharmacological interventions (e.g. optical, acupuncture, biofeedback treatment, and afferent stimulation of the neck or forehead).

Types of outcome measures

For both primary and secondary outcomes, if data were not reported at six months' follow‐up, we accepted measures taken between three and 12 months follow‐up (Alió 2003; Dell’Osso 1973; Hertle 2006; Neely 1999; Sprunger 1997; Wang 2006a). We did not exclude studies from the review solely on the basis of lack of outcome data.

Primary outcomes

The primary outcome was the change in binocular best‐corrected distance visual acuity, measured using any acuity measure (e.g. LogMAR, Snellen, Teller acuity, grating).

Secondary outcomes

The secondary outcomes were:

• Head posture, measured in degrees;

• Amplitude of the nystagmus waveform, quantified using any of the following techniques from bilateral eye movement recordings: video‐recording, electro‐oculography, and the use of magnetic search coils;

• Frequency of the waveform, quantified as described above;

• Intensity (amplitude multiplied by frequency) of the waveform, quantified as described above;

• Foveation period durations of the waveform, quantified as described above; these may be measured directly by defining them in terms of thresholds for eye velocity and distance from the intended target. Alternatively, the Nystagmus Expanded Acuity Function (NAFX) or similar methods may be used (these are functions that extrapolate potential visual acuity based on measurements of the foveation characteristics of nystagmus waveforms);

• Visual recognition times, measured as search times (which may include number of saccades, duration and number of fixations);

• Quality of life and self‐reported outcome measures (including participant satisfaction, pain from the procedure, length of hospital stay, repeated or further surgeries, or both), and functional measurements obtained using validated questionnaires, such as the National Eye Institute Visual Function Questionnaire or the VF‐14 measurement of visual impairment.

Adverse effects

We considered any adverse effects related to the surgical intervention (including loss of vision, development or worsening of strabismus, surgical injury to the eye, postoperative infection, and loss of ocular rotation).

Although minimal clinically important differences are not established for these outcome measures, the key differences of interest (as determined by clinical consensus) were as follows:

• A 2‐line difference in binocular best‐corrected distance visual acuity;

• At least 5 degrees of change in head posture;

• A decrease in intensity (amplitude multiplied by frequency) of the waveform;

• An increase in foveation period durations of the waveform;

• Decreased visual recognition times;

• No permanent adverse effects after surgery (e.g. strabismus, double vision, loss or significant limitation of eye movement).

Search methods for identification of studies

Electronic searches

The Cochrane Eyes and Vision Information Specialist searched the following electronic databases for randomised controlled trials and controlled clinical trials. There were no restrictions to language or year of publication. 

• Cochrane Central Register of Controlled Trials (CENTRAL; 2020, Issue 7; which contains the Cochrane Eyes and Vision Trials Register) in the Cochrane Library (searched 3 July 2020; Appendix 2);

• MEDLINE Ovid (1946 to 3 July 2020; Appendix 3);

• Embase Ovid (1980 to 3 July 2020; Appendix 4);

• ISRCTN registry (www.isrctn.com/editAdvancedSearch: searched 3 July 2020; Appendix 5);

• US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov: searched 3 July 2020; Appendix 6);

• World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP; www.who.int/ictrp: searched 3 July 2020; Appendix 7).

Searching other resources

We did not handsearch conference abstracts for this review, as Cochrane Eyes and Vision routinely handsearches for RCTs from major ophthalmology meetings and incorporates these results into the CENTRAL database.

Data collection and analysis

Selection of studies

We adopted a two‐stage process to identify relevant RCTs. Two review authors (KMC and LAA) each assessed titles and abstracts returned from the electronic searches, to identify trials potentially relevant to the review. KMC retrieved the full‐text articles of studies considered to be relevant, or potentially relevant. KMC and LAA independently evaluated the full‐text articles to judge their suitability for inclusion in the review, according to the characteristics outlined in Criteria for considering studies for this review. The two review authors resolved disagreements in judgement by discussion. The same process was also adopted for identifying relevant non‐randomised clinical interventional studies.

We did not contact any study authors for additional information to determine the eligibility of a trial. We included details relating to the reason for excluding studies that progressed to the full‐text review stage in the ‘Characteristics of excluded studies’ section.

Data extraction and management

KMC and LED independently extracted study data (detailed in Appendix 8) using the online systematic review platform Covidence (Covidence). Each author used the Covidence data extraction form that was piloted for data extraction.  

Extracted information included: details relating to the design of the study, country, setting, participant characteristics, interventions and comparators, outcomes, results, and any other relevant information (e.g. funding source). We extracted quantitative data for our nominated primary and secondary outcomes, wherever possible. The two review authors resolved discrepancies in data extraction by discussion to reach consensus. After reaching consensus, LED exported data into Cochrane's RevMan Web and KMC verified the exported data (RevMan Web 2020).

Assessment of risk of bias in included studies

KMC and LED independently assessed the risk of bias in eligible studies using the guidelines in Chapter 8 of the Cochrane Handbook for Systematic Review of Interventions (Higgins 2017).

Specifically, they evaluated the risk of bias in each of the following domains:

• Selection bias (random sequence generation and allocation concealment);

• Performance bias (masking of participants and personnel);

• Detection bias (masking of outcome assessment);

• Attrition bias (incomplete outcome data);

• Reporting bias (selective reporting of outcomes);

• Other bias (funding source, other conflicts of interest).

They judged the risk of each bias as low risk, unclear risk (due to either lack of information or uncertainty over the potential for bias), or high risk. The two review authors resolved disagreements in assessment by discussion.

Measures of treatment effect

Where outcomes were continuous, we extracted data on the change from baseline, in means (and standard deviations of changes) of the outcome measures for the intervention and comparison group(s). If a measure of change from baseline was not reported, we extracted the means (and standard deviation) of the outcome for each group. We were unable to quantitatively report on treatment effects due to the limited availability of quantitative data.

Unit of analysis issues

The unit of analysis for this review was the study participant. Given the nature of surgical interventions for nystagmus, as might be expected, trials allocated individual participants (rather than individual eyes) to intervention groups. We did not identify any paired‐eye or multiple‐armed studies.

In future updates of this review, we will include paired‐eye studies, if they are identified.

If we include multiple‐armed studies, we will first combine all experimental groups of the study into a single intervention group, and then combine all control intervention groups into a single control group. If this is not possible because the experimental intervention groups are not sufficiently similar, we will select one pair of interventions and exclude the others, to best permit the pooling of data.

Dealing with missing data

We emailed the corresponding author of the only eligible RCT on 25 August 2020 to ask for:

(i) details regarding the variance parameters for both treatment groups for the following outcome measures: change in binocular best‐corrected visual acuity, change in amplitude of the nystagmus waveform, change in intensity of the nystagmus waveform, and change in frequency of the nystagmus waveform; and

(ii) quantitative data relating to quality of life outcomes.

As of  8 February 2021, we had yet to receive a response, and thus, we did not use any additional information for the review (Singh 2016).

Assessment of heterogeneity

As we only identified one RCT, we did not assess heterogeneity.

Assessment of reporting biases

As we only identified one RCT, we did not assess the potential for reporting bias.

Data synthesis

This systematic review identified a single, small RCT. As such, we were unable to conduct meta‐analyses for any of the prespecified outcome measures.

Subgroup analysis and investigation of heterogeneity

As we only identified one RCT, we were unable to conduct subgroup analyses for factors considered to potentially affect outcomes.

Sensitivity analysis

As we only identified one RCT, we could not perform a sensitivity analysis for the primary outcome measure.

Summary of findings and assessment of the certainty of the evidence

We planned to summarise the results of the analyses in a ‘Summary of findings’ table, using the approach recommended in Chapter 11 of the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2017).

In our protocol, we planned separate ‘Summary of findings’ tables for the following clinical comparisons:

• Any surgical intervention versus no intervention;

• Any surgical intervention versus any non‐surgical intervention;

• Four horizontal rectus muscle tenotomy with reattachment versus no intervention;

• Myectomy of the extraocular muscles without reattachment versus no intervention;

• Myectomy of the extraocular muscles without reattachment versus any other surgical intervention.

Although the single included RCT did not consider one of these prespecified comparisons, we had defined ‘retro‐equatorial recession of the horizontal recti’ as an intervention of interest, and ‘other surgical interventions’ as a possible comparator. Hence, we have generated Table 1, acknowledging that this was not prespecified in our published protocol (Cham 2019; see Differences between protocol and review).

We used the GRADE approach to grade the overall certainty of evidence. Outcome measures, measured between the treatment groups at six‐months follow‐up (with an acceptable follow‐up range between three and 12 months), included:

• Change in binocular best‐corrected distance visual acuity;

• Change in head posture;

• Change in intensity (amplitude multiplied by frequency) of the waveform;

• Change in foveation period durations of the waveform;

• Change in visual recognition times;

• Proportion of participants with adverse effects with a probable causal link to the treatment (this outcome will be reported at any follow‐up period);

• Change in quality of life (measured using a validated questionnaire).

Results

Description of studies

Results of the search

The electronic searches yielded a total of 673 records as of 3 July 2020 (Figure 1). After removing 241 duplicates, two review authors independently screened 432 studies and abstracts for potential inclusion. We classified 43 reports as potentially eligible, and these articles proceeded to full‐text screening. We excluded 42 studies (see Characteristics of excluded studies).

1.

Study flow diagram

No relevant ongoing or unpublished RCTs were identified in the register searches.

Included studies

We did not identify any RCTs comparing a surgical intervention for INS relative to no treatment.

We identified one randomised controlled trial (RCT), which was not prospectively registered, as eligible for inclusion in this review (Singh 2016). As only one RCT was included, we also produced a narrative summary of three non‐randomised intervention studies (Akbari 2013; Bagheri 2008; Kumar 2011), as outlined in the protocol (Cham 2019). These studies were identified for the narrative synthesis based on their study design, as they represented the studies providing the next highest quality of clinical evidence available that was relevant to the review question after the RCT.

We provided a detailed description of the included RCT in the Characteristics of included studies table.

In brief, this trial was conducted in India, and evaluated outcomes in 10 participants; the report did not provide details relating to participants’ age and gender, or whether the participants had congenital idiopathic infantile nystagmus syndrome (INS) or INS with sensory disorders. Furthermore, the study only included participants with null in primary position and did not explicitly exclude congenital periodic alternating nystagmus.

Both eyes of each participant received the same intervention. Participants were randomly assigned to undergo a large retro‐equatorial recession of the horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus, or a simple tenotomy and resuturing of the four horizontal rectus muscles. The participants were followed up at one week, four weeks, three months, and six months after surgery.

The outcome measures evaluated in this study were: binocular visual acuity, contrast sensitivity, colour vision, stereopsis, and electronystagmography (ENG) for nystagmus waveforms, amplitude, frequency, and intensity. Adverse events were considered by the intervention complication rate. A questionnaire was used to assess quality of life after the surgical intervention in both treatment arms, however, no details were provided about the survey tool that was used. Head posture, foveation period durations of the nystagmus waveform, visual recognition times, and permanent adverse effects after surgery were not evaluated. The study funding source(s) and author declaration of interests were not reported.

Excluded studies

Following full‐text evaluation, we excluded 42 studies from the review. These trials are listed in the Characteristics of excluded studies section, with the primary reason for exclusion. The two main reasons were non‐RCT study design (35 studies, either a case study, case series, or non‐randomised intervention study), or evaluation of an alternative intervention (seven studies).

We did not identify any studies as ongoing or awaiting classification.

Risk of bias in included studies

The risk of bias assessment for Singh 2016 is summarised in Figure 2 and Figure 3. Information is also provided in the Characteristics of included studies table.

2.

'Risk of bias' graph: review authors' judgements about each risk of bias item presented as percentages across included study

3.

'Risk of bias' summary: review authors' judgements about each 'Risk of bias' item across included study

Allocation

We judged Singh 2016 at low risk of bias for sequence generation, and an unclear risk for allocation concealment. The randomisation code was generated using a random number table; however, the study authors did not report how the randomisation list was implemented.

Blinding

We judged the domains of 'blinding of participants and personnel' and 'blinding of outcome assessors' both at high risk of bias. No information was provided in relation to masking. We assumed that in the absence of reporting, participants, personnel, and outcome assessors were not masked, which corresponds to a high risk of bias in these domains.

Incomplete outcome data

The extent of participant follow‐up was not explicitly reported, thus, we considered there was an unclear risk of attrition bias.

Selective reporting

We had no access to a clinical trial registry entry or trial protocol, therefore, we considered there was an unclear risk of reporting bias.

Other potential sources of bias

We judged the study to be at low risk of other bias, as we did not identify other potential sources of bias.

Effects of interventions

See: Table 1

Table 1 summarises information from Singh 2016 relating to the effect of large retro‐equatorial recession of horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus compared to simple tenotomy and resuturing of the four horizontal rectus muscles for the surgical treatment of INS.

Using the GRADE approach, we judged the certainty of the evidence as very low for all reported outcomes; we downgraded the certainty for risk of bias (one level, as neither participants nor outcome assessors were masked), and imprecision (two levels, as data were derived from a single study of small sample size (N = 10), with no reported measures of precision around the effect estimate).

Primary outcome

Our prespecified primary outcome was the change in binocular best‐corrected distance visual acuity at six‐month follow‐up, with an acceptable follow‐up range of three and 12 months. Singh 2016 reported no significant change from baseline (preoperative) binocular distance visual acuity in either intervention arm at six‐month follow‐up. They did not report between‐group comparison. Furthermore, the presented preoperative and postoperative data did not include a measure of variability (e.g. standard deviation) around the effect estimate, nor details regarding the number of participants used to derive the group summary data. We contacted the corresponding author via email to obtain this information, but did not receive a response in more than four weeks.

Secondary outcomes

Singh 2016 did not report on the following secondary outcomes:

• Head posture, measured in degrees;

• Foveation period durations of the nystagmus waveform;

• Visual recognition time, measured as search times;

• Permanent adverse effects after surgery.

Amplitude of the waveform

Singh 2016 measured the amplitude of the nystagmus waveform using electronystagmography (ENG). The authors reported that postoperatively, the nystagmus “amplitude decreased (relative to preoperative values) in both (intervention) groups in all gazes”, but there was no significant between‐group difference. As for the nystagmus amplitude data, the presented pre‐ and postoperative data did not include a measure of variability (e.g. standard deviation) around the effect estimate, or details regarding the number of participants used to derive the group summary data. We contacted the corresponding author via email to obtain this information, but did not receive a response in more than four weeks.

Frequency of the waveform

Singh 2016 reported “the frequency of nystagmus did not show any significant change in either (intervention) group (post‐surgery)” and there was no significant between‐group difference post‐operatively. As for the nystagmus frequency data, a measure of variability (e.g., standard deviation) on the effect estimate was not provided, and details regarding the number of participants used to derive the group summary data were not reported. We contacted the corresponding author via email to obtain this information, but did not receive a response in more than four weeks.

Intensity (amplitude multiplied by frequency) of the waveform (decrease in intensity)

In the abstract of the article, Singh 2016 reported that “the intensity of nystagmus showed a tendency to decrease in both the groups in all gazes”, however the inter‐group comparison was not reported.

Proportion of participants with adverse effects with a probable causal link to the treatment

The authors reported that no participants had restricted extraocular movements or significant complications as a result of their assigned surgery.

Quality of life and self‐reported outcome measures

Based on a questionnaire of subjective postoperative improvement, it was reported that most participants noticed an improvement in their quality of life after their surgery, in both intervention groups. No quantitative data were provided in the study report. We contacted the corresponding author via email to obtain this information, but did not receive a response in more than four weeks.

Discussion

Summary of main results

The aim of this systematic review was to examine the efficacy and safety of surgical interventions for infantile nystagmus syndrome (INS).

We did not identify any RCTs comparing a surgical intervention for INS relative to no treatment.

We identified one eligible randomised controlled trial (RCT)(Singh 2016), which evaluated outcomes in a total of 10 participants with INS with null in primary position, and excluded participants with an eccentric null or a convergence null. This limited the study outcomes since surgical treatment of INS is most commonly performed in individuals with anomalous head postures secondary to eccentric null regions (Neely 1999).The study randomised participants to receive either a large retro‐equatorial recession of the horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus, or a simple tenotomy and resuturing of the four horizontal rectus muscles, to both eyes, rather than any of our planned comparisons.

The authors found no significant change from baseline (preoperative) binocular distance visual acuity in either intervention arm at six‐month follow‐up; they did not report between‐group differences. We judged the certainty of the evidence to be very low.

Singh 2016 descriptively summarised their study findings, without providing complete quantitative data, for the following secondary outcomes: amplitude, frequency and intensity of the nystagmus waveform; the proportion of participants with adverse effects having a probable causal link to the treatment; and self‐reported quality of life. They reported no significant between‐group differences for all of these outcome measures at the study endpoint. We judged the certainty of the evidence for each of these secondary outcomes to be very low.

Singh 2016 reported no participants had adverse effects with a probable causal link to the intervention, but did not specifically report permanent post‐surgical adverse effects. Due to a lack of comprehensive reporting of adverse events, there was also very low‐certainty about the safety profile of the evaluated surgical interventions in this patient population.

Overall, we identified a scarcity of high‐quality evidence relating to the efficacy and safety of surgical interventions for INS.

Overall completeness and applicability of evidence

We only identified a single, small RCT investigating two types of surgical interventions for INS, retro‐equatorial recession of the horizontal rectus muscles and simple tenotomy and resuturing of the four horizontal rectus muscles. As such, we were unable to conduct meta‐analyses for any of the prespecified outcomes, which impacted our ability to draw definitive conclusions surrounding the efficacy or safety of surgical interventions for INS (see Table 1).

Since we only included one RCT (Singh 2016), we included a narrative summary of three non‐randomised interventional studies (Akbari 2013; Bagheri 2008; Kumar 2011), as outlined in our methods.

In all three studies, both eyes received the same intervention. The studies had different review schedules and follow‐up periods, ranging from three months to three years. Neither primary nor secondary outcomes were explicitly defined in any of the studies. Information relevant to the review prespecified primary outcome (change in binocular visual acuity) was reported in each study.

As summarised in Table 2, the three non‐randomised intervention studies included different numbers of participants, of different ages. Participant numbers ranged from 14 to 58, and the mean ages of participants ranged from 10.9 to 26.8 years. Two studies were conducted in Iran, one in an eye hospital (Akbari 2013), the other in a medical centre (Bagheri 2008); the remaining study did not report either the country or setting (Kumar 2011). Funding sources were declared in one study (Akbari 2013), and author declarations of interests in another (Kumar 2011).

1. Summary of the characteristics of the non‐randomised intervention studies.

	Akbari 2013	Bagheri 2008	Kumar 2011
Study design (as described in the study report)	Prospective interventional case series	Prospective comparative case series	Prospective, non‐randomised, interventional study
Country	Iran	Iran	not reported
Setting	eye hospital	medical centre	not reported
Total participant numbers (male/female)	14 (7/7)	58 (35/23)	28 (13/15)
Participant population	Participants > 18 years with a clinical diagnosis of INS, with binocular distance BCVA of 20/400 to 20/30, and no history of EOM surgery or other treatments	Participants with a clinical diagnosis of congenital nystagmus and no history of EOM surgery	Participants with a clinical diagnosis of INS and no history of EOM surgery
Age (mean + SD), years	26.8 ± 7.5	18.7 ± 9.1	10.9 ± 4.2
Intervention(s) and comparator(s)	Combined recession with tenotomy (Group 1); four‐muscle tenotomy (Group 2)	Large recession of all four horizontal rectus muscles (4‐rectus group); large recession of two horizontal rectus (2‐rectus group); Anderson‐Kestenbaum	Anderson‐Kestenbaum; modified Anderson; modified Anderson + tenotomy
Adjunctive therapies	Topical steroids and antibiotics, four times daily for 7 to 10 days	not reported	not reported
Follow‐up period, including review schedule	Next day, one week, one month, three months, and every three months postoperatively	Next day and one week postoperatively, then one, three, and six months, and every six months thereafter	One week and one month postoperatively
Trial duration	not reported	3 years	6 to 18 months
Losses to follow‐up	not reported	not reported	not reported
Group differences	not reported	not reported	not reported
Reported outcomes. Note: in all studies, comparisons were reported as change from baseline rather than as between‐group comparisons	Binocular BCVA: no change in Group 1; 'borderline significance' change in Group 2 Nystagmus waveform amplitude: 'reduced postoperatively' (presumably across both groups) Nystagmus waveform frequency: 'no clinically significant change' (presumably across both groups) Amount or duration of abnormal head posture: 'no remarkable change' (presumably across both groups) Adverse effects: no participants lost their fusion or stereopsis; no complications or reported discomfort in either group Participant satisfaction: 6/6 (Group 1) and 7/8 (Group 2) participants were satisfied with the cosmetic results of the procedures, would recommend the surgery, and would take it themselves if they were at the beginning of the study	Binocular BCVA: improved in participants who underwent large recession of two horizontal rectus recession Nystagmus waveform amplitude and frequency: 'a decrease in the amplitude and frequency of the nystagmus waveform only in the large recession of all four horizontal rectus muscles intervention' Adverse effects: none of the participants in the large recession of two horizontal rectus group developed motility restrictions postoperatively Quality of life and self‐reported outcome measures: 'the quality of vision improved in the majority of cases' across all intervention arms	Binocular BCVA: improved overall (across all participants); there was no difference in acuity outcomes between intervention arms Head posture: 'mean abnormal head posture (20.89°) improved significantly to a mean of 3.21°' (across all participants) Adverse effects: no participant in the trial lost fusion after surgery Quality of life and self‐reported outcome measures: 'a varying degree of improvement in binocularity in 3 participants'
Study funding sources and declarations of interest	Funding sources: Research and Thesis Committee at Eye Research Centre, Farabi Hospital, Tehran University of Medical Sciences, Iran Declarations of interest: not reported	Funding sources: not reported Declarations of interest: not reported	Funding sources: not reported Declarations of interest: “The authors have no financial or proprietary interest in the materials presented herein.”

Abbreviations: BCVA: best‐corrected visual acuity; EOM: extraocular muscle; INS: infantile nystagmus syndrome

The INS surgical interventions (and their variations) considered in the studies were four horizontal rectus muscle tenotomy with or without associated recession (Akbari 2013), large recession of all or two of the horizontal rectus muscles relative to the Anderson‐Kestenbaum procedure (Bagheri 2008), and the Anderson‐Kestenbaum procedure relative to a modified procedure, with or without tenotomy (Kumar 2011).

In summary, the only RCT deemed eligible for inclusion in this review considered the relative efficacy and safety of two types of surgical procedures: large retro‐equatorial recession of the horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus, and simple tenotomy and resuturing of the four horizontal rectus muscles to both eyes. We did not identify any RCTs evaluating the clinical effects of artificial divergence, the Anderson‐Kestenbaum procedure (and its variations), recession‐resection of vertical recti, combined recession of the vertical rectus muscles and weakening of the oblique muscles, full tendon‐width horizontal transposition of the vertical recti, or myectomy of the extraocular muscles without reattachment. High‐quality RCT evidence is not currently available for the majority of surgical procedures for INS, including the relative safety and efficacy of surgical interventions relative to no treatment.

Quality of the evidence

For all efficacy and safety outcomes reported in Singh 2016 (i.e. change in binocular best‐corrected distance visual acuity; amplitude, frequency, and intensity of the nystagmus waveform; the proportion of participants with adverse effects having a probable causal link to the treatment, and self‐reported quality of life outcome measures), we judged the certainty of the body of evidence to be very low. We downgraded the certainty of the findings due to risks of bias (i.e. the absence of participant or outcome assessor masking) and imprecision (data derived from a single study involving only 10 participants). Therefore, there is currently a dearth of high‐quality clinical evidence relating to the efficacy or safety of surgical interventions for INS.

Potential biases in the review process

We used the standard methodological procedures recommended in Chapter 8 of the Cochrane Handbook for Systematic Review of Interventions to minimise any potential source of bias during the review process (Higgins 2017).

The review protocol was prospectively published, and as such, all outcome measures were specified in advance of undertaking the review (Cham 2019). An a priori search strategy was developed that was comprehensive and included grey and non‐English literature, minimising selection bias. Two review authors independently oversaw each stage of the review process.

Agreements and disagreements with other studies or reviews

With only one small eligible RCT identified (Singh 2016), we included a narrative synthesis of three non‐randomised interventional studies (Akbari 2013; Bagheri 2008; Kumar 2011). However, being non‐RCTs, we were unable to make meaningful comparisons with Singh 2016.

Given that the main aim of surgical interventions for treating INS is to correct abnormal head posture, with additional aims including minimising nystagmus intensity and increasing foveation times (Dell’Osso 1973; Wang 2006b), it is surprising that these secondary outcomes were not uniformly reported, or quantified, in current studies. Singh 2016 reported on the intensity of the waveform, and Akbari 2013 and Kumar 2011 both reported outcomes for head posture; however, the outcomes were mainly descriptive. Only Akbari 2013 reported that no participants had any complication or reported discomfort during their follow‐up period across both interventions. Due to a lack of comprehensive reporting of adverse events, we were unable to definitely conclude about the safety profile of the INS surgical interventions considered in these studies.

To our knowledge, there are no other systematic reviews on the current topic.

Studies that have considered INS treatment options are non‐systematic and not specific to INS. They are often compilations of congenital and acquired forms of nystagmus treatment in general, which is irrelevant to our current review (e.g. Hobson 2009; Thurtell 2012). In summary, no definitive conclusions could be made regarding agreements and disagreements with other studies or reviews.

Authors' conclusions

Implications for practice.

We did not find any definitive evidence to support one infantile nystagmus syndrome (INS) surgical procedure over another. We also did not identify any randomised controlled trial (RCT) evidence comparing a surgical intervention for INS relative to no treatment. The safety profile of surgical treatments also remains unclear, as is the relative efficacy of different interventions. These uncertainties, and the possibility of adverse effects, must be clearly outlined to people with INS and their families or caregivers to ensure an appropriate level of informed consent before undergoing a surgical intervention for INS.

Despite the range of interventions currently available, and used in practice for managing infantile nystagmus syndrome (INS), there are no accepted guidelines to inform clinical decision‐making. Given the present review’s findings, this lack of clarity regarding the relative efficacy and safety of the various surgical options is not surprising. Small sample sizes and incomplete reporting of the study outcomes were evident in both the single included RCT(Singh 2016) and three non‐randomised intervention studies (Akbari 2013; Bagheri 2008; Kumar 2011) considered in this review.

Despite the lack of RCT evidence, the gestalt of the surgical treatments for INS is that individuals are generally satisfied by having their abnormal head postures routinely improved by ‘sensible surgery’ to improve their clinical signs and symptoms. In this systematic review, our primary outcome was the change in binocular best‐corrected distance visual acuity, and in this regard the evidence is limited.

Currently, the decision to perform a particular type of surgery, and the outcomes that are assessed when evaluating the treatment, is largely dependent on a surgeon’s own clinical judgement. Considering the evidence identified in the current review, distinguishing between INS surgical procedures lacks evidence to support a choice. Furthermore, while non‐RCT studies suggest that surgical procedures could be beneficial for reducing the intensity of the nystagmus and/or improving abnormal head posture in people with INS relative to no treatment, these claims lack robust evaluation in a RCT. The safety profile of surgical treatments also remains unclear, as is the relative efficacy of different interventions. These uncertainties, and the possibility of adverse effects, need to be clearly outlined to patients and their families/caregivers to ensure an appropriate level of informed consent before undergoing a surgical intervention for INS.

There is thus a paucity of high‐quality evidence relating to the efficacy or safety of surgical interventions for INS. More high‐quality RCTs are needed, to assist clinicians and patients in evidence‐based decision‐making in relation to the appropriateness of surgical treatments for INS.

Implications for research.

This is the first systematic review to appraise and synthesise evidence from RCTs relating to the efficacy and safety of surgical interventions for infantile nystagmus syndrome (INS). We only identified one relevant RCT, with noted limitations in the study design that may confound the reported findings (Singh 2016). In addition, we did not identify any ongoing RCTs or trials awaiting classification, highlighting a lack of ongoing research in the field. This may be due to the perception that INS is rare and less sight‐threatening than other eye conditions, or this may be a niche area of research, or both. In practical terms, the relative rarity of INS (one in 1000 to one in 6000 births) may also create challenges with participant recruitment for clinical trials (Forssman 1971; Heemes 1924; Norn 1964; Stayte 1993). In addition, surgical RCTs are hard to design and perform, due to difficulties in participant consent and randomisation to a surgical treatment. It is challenging to randomly assign people with INS to a particular type of surgery that is equally effective for all. Treatment needs to be tailored according to each person's clinical presentation, taking into account the presence of a head posture (null region, and its extent, position, and eccentricity), baseline sensory status, foveation period durations of the nystagmus waveform, and excluding other congenital forms of nystagmus via proper eye movement recordings.

Also, there is currently no consensus regarding appropriate outcome measures for clinical trials evaluating the efficacy of surgical interventions for INS. In general, these procedures aim to correct abnormal head posture and maximise visual performance, which can be indicated by minimising the nystagmus intensity, increasing foveation period durations of the waveform, and decreasing visual recognition times (Dell’Osso 1973; Wang 2006a). However, these clinically relevant parameters were not uniformly assessed in current studies. Only the change in binocular best‐corrected distance visual acuity, the primary outcome in this review, was consistently reported by the RCT (Singh 2016), and the three non‐randomised interventional studies (Akbari 2013; Bagheri 2008; Kumar 2011).

In order to obtain a more complete understanding of the efficacy and safety of surgical interventions for INS, more robust clinical studies are required. Such trials would need to be adequately powered and clearly define the population of interest, i.e. people with congenital idiopathic INS or INS with sensory disorders. Eye movement recordings must also be documented to ensure the participants have INS, and other congenital forms of nystagmus, such as periodic alternating nystagmus and fusion maldevelopment nystagmus syndrome, are excluded.

Development of a core outcome set would facilitate improved selection of standardised outcome measure in different trials, and enhance future capacity to synthesise relevant evidence and draw more definitive conclusions about the efficacy and safety of INS surgical interventions. To better understand the safety profile surgical options for INS, future trials should include longer follow‐up periods (minimum of 12 months), more comprehensive reporting of quality of life and self‐reported outcome measures using validated questionnaires, and outline any permanent adverse events after surgery. For transparency, reports should comprehensively report study funding sources and author declarations of interests.

History

Protocol first published: Issue 8, 2019
Review first published: Issue 2, 2021

Appendices

Appendix 1. Glossary of terms

Acquired – not inherited, but developing after birth

Age‐related macular degeneration – a disease that blurs central vision (required for activities such as reading and driving). This condition is caused by a degeneration of light‐sensitive cells in the region of the retina known as the macula, which allows us to see fine details.

Ametropia – vision disorders (e.g. near‐sightedness, far‐sightedness, or astigmatism) that result in the eyes being unable to focus clearly on objects without the aid of a refractive correction (e.g. glasses or contact lenses)

Amplitude – in nystagmus, the distance in visual angle (measured in degrees) between the start of the drift away from fixation (point of focus) and the start of the corrective eye movement in the opposite direction

Base‐out prism – a triangular glass prism with the flat base positioned towards the person’s ear (facing out), which is used to reposition an image to improve vision with both eyes open

Binocular – using both eyes together

Congenital – present at, or before, birth

Conjugate – both eyes behaving in almost identical ways

Convergence – the co‐ordinated movement of both eyes inwards when looking at an object up close

Eccentric gaze or endpoint nystagmus – a drift of the eyes towards the centre when held in extreme positions of gaze (e.g. far right or far left)

Extirpating – complete removal

Fixation – point of focus

Frequency – number of cycles of an oscillation (in this case, nystagmus) per second, measured in Hertz (Hz)

Infantile‐onset – an individual acquires, develops, or first experiences a condition during infancy

Intensity – in nystagmus, the product of amplitude and frequency

Monocular – with, for, or in one eye

Nystagmus – involuntary eye movements, during which an initial slow movement away from fixation is corrected by either a fast or slow movement back

Oculomotor – defined strictly, this refers to a small brain region that controls four of the six eye muscles, but sometimes, the term is used more generally to refer all the brain structures that control eye movements, thereby enabling normal perception of the world.

Optokinetic nystagmus – an involuntary eye movement in response to the rotational movement of a full field visual image

Oscillopsia – a visual disturbance, during which objects that are stationary appear to oscillate (move)

Periodic alternating nystagmus – horizontal nystagmus that reverses its direction every few minutes

Recession – a surgical weakening procedure, whereby an eye muscle is removed from its original position and repositioned further back on the eyeball

Resection – surgical removal of part, or all, of a tissue. In strabismus and nystagmus, this is a surgical strengthening procedure, whereby part of an eye muscle is cut off and the muscle length is shortened.

Saccade – the rapid eye movement when looking at two objects

Smooth pursuit – an eye movement that keeps an image of a slowly moving object on the fovea (a tiny area in the macula that allows the clearest vision)

Strabismus – an abnormality in the alignment of an eye, such that only one eye actually views the intended target

Vergence – the simultaneous movement of both eyes, in opposite directions, in order to maintain a single, clear vision of an object simultaneously in both eyes

Vestibular nystagmus – nystagmus resulting from a disturbance in the inner ear and to the system in the brain that is responsible for keeping an image steady on the eye during rapid head movements

Visual acuity – the measurement of the smallest target (e.g. a letter) that can be accurately identified

Appendix 2. CENTRAL search strategy

#1 MeSH descriptor: [Nystagmus, Congenital] explode all trees
#2 (infantile or idiopath* or congenital* or motor or sensory) near/4 nystagm*
#3 INS near/4 nystagm*
#4 #1 or #2 or #3
#5 MeSH descriptor: [Ophthalmologic Surgical Procedures] explode all trees
#6 surg*
#7 MeSH descriptor: [Oculomotor Muscles] explode all trees and with qualifier(s): [surgery ‐ SU]
#8 MeSH descriptor: [Tenotomy] this term only
#9 Kestenbaum:ab
#10 (resect* or recession or surg* or transpos*) near/3 (rectus next muscle*)
#11 (tenotomy or myectomy)
#12 (artificial near/2 divergence)
#13 (oblique next muscle*) near/3 weak*
#14 #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13
#15 #4 and #14

Appendix 3. MEDLINE Ovid search strategy

1. exp nystagmus, congenital/
2. ((infantile or idiopath$ or congenital$ or motor or sensory) adj4 nystagm$).tw.
3. (INS adj4 nystagm$).tw.
4. or/1‐3
5. exp Ophthalmologic Surgical Procedures/
6. surg$.tw.
7. Oculomotor Muscles/su [Surgery]
8. Tenotomy/
9. Kestenbaum.ab.
10. ((resect$ or recession or surg$ or transpos$) adj3 rectus muscle$).tw.
11. (tenotomy or myectomy).tw.
12. (artificial adj2 divergence).tw.
13. (oblique muscle$ adj3 weak$).tw.
14. or/5‐13
15. 4 and 14

Appendix 4. Embase Ovid search strategy

1. congenital nystagmus/
2. ((infantile or idiopath$ or congenital$ or motor or sensory) adj4 nystagm$).tw.
3. (INS adj4 nystagm$).tw.
4. or/1‐3
5. eye surgery/
6. surg$.tw.
7. surgical technique/
8. extraocular muscle/
9. muscle resection/
10. tenotomy/
11. Kestenbaum.ab.
12. ((resect$ or recession or surg$ or transpos$) adj3 rectus muscle$).tw.
13. (tenotomy or myectomy).tw.
14. (artificial adj2 divergence).tw.
15. (oblique muscle$ adj3 weak$).tw.
16. or/5‐15
17. 4 and 16

Appendix 5. ISRCTN search strategy

(infantile OR idiopathic OR congenital OR motor OR sensory) AND nystagmus

Appendix 6. ClinicalTrials.gov search strategy

(infantile OR idiopathic OR congenital OR motor OR sensory) AND nystagmus

Appendix 7. WHO ICTRP search strategy

infantile nystagmus OR idiopathic nystagmus OR congenital nystagmus OR motor nystagmus OR sensory nystagmus

Appendix 8. Data on study characteristics

Mandatory items		Optional items
Methods
Study design	· Parallel group RCT (i.e. participants randomised to treatment) · Within‐person RCT(i.e. eyes randomised to treatment) · Cluster‐RCT(i.e. communities randomised to treatment) · Cross‐over RCT · Other, specify	Exclusions after randomisation Losses to follow up Number randomised/analysed How missing data were handled (e.g. available case analysis, imputation methods) Reported power calculation? (Y/N), if yes, sample size and power Unusual study design/issues
Eyes or unit of randomisation/ unit of analysis	· One eye included in study, specifying how the eye was selected · Two eyes included in study, both eyes received the same treatment, specifying how analysed (best/worst/average/both, and adjusted for within‐person correlation/both, and not adjusted for within‐person correlation) and specifying if mixture of one eye and two eye · Two eyes included in study, eyes received different treatments,specifying if correct pair‐matched analysis was performed
Participants
Total number of participants	This information will be collected for the total study population recruited into the study. If these data are only reported for the participants who were followed up, this will be indicated.	Ethnic group Equivalence of baseline characteristics (Y/N)
Number (%) of men and women
Average age and age range
Inclusion criteria
Exclusion criteria
Interventions
Intervention (N = ) Comparator (N = )	· Number of people randomised to each group · Intervention name and details
Outcomes
Primary and secondary outcomes as defined in the study reports	Outcomes (as specified in the study report)  Adverse effects reported (Y/N)  Length of follow‐up and intervals at which outcomes were assessed	Planned/actual length of follow‐up
Identification
Country		Setting
Dates conducted	Specify dates of recruitment of participants mm/yr to mm/yr	Full study name: (if applicable) Reported subgroup analyses (Y/N)   Were trial investigators contacted?
Sources of funding
Declaration of interest
Included on trials registry?	Y/N (including registration number if available)

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Singh 2016.

Study characteristics
Methods	Study design: randomised controlled trial Exclusions after randomisation: not reported. How missing data were handled (e.g. available case analysis, imputation, etc.): not reported Losses to follow‐up: not reported Other comments (e.g. unusual study design, issues): none Reported power calculation: no Trial duration: participants were followed up 1 week, 4 weeks, 3 months, and 6 months after surgery Unit of randomisation/unit of analysis: two eyes were included in the study, and eyes received the same treatments. Participants were divided into two groups, which received different treatments.
Participants	Baseline Characteristics “Large retro‐equatorial recession of horizontal rectus muscles 9 mm on the medial rectus and 12 mm on the lateral rectus (Group 1).” Age (mean + SD): not reported Number of participants: 5 Gender: not reported Review period: participants were followed up 1 week, 4 weeks, 3 months, and 6 months after surgery Participants with INS and sensory disorders: not reported Participants with congenital idiopathic INS: not reported “Simple tenotomy and resuturing of the 4 horizontal rectus muscles (Group 2).” Age (mean + SD): not reported Number of participants: 5 Gender: not reported Review period: participants were followed up 1 week, 4 weeks, 3 months, and 6 months after surgery Participants with INS and sensory disorders: not reported Participants with congenital idiopathic INS: not reported Overall Age (mean + SD): not reported Number of participants: 10 Gender: not reported Review period: 6 months Participants with INS and sensory disorders: not reported Participants with congenital idiopathic INS: not reported Inclusion criteria: aged at least 5 years; diagnosis of nystagmus in infancy; null in primary position, and visual acuity of at least 6/60 Exclusion criteria: previous extraocular muscle surgery; associated strabismus; unco‐operative for nystagmus recording, convergence damping of nystagmus, a null position of from primary position horizontally and vertically, on systemic medication known to affect ocular oscillations, concurrent medical condition or known risk that would increase chances of an adverse event due to general anaesthesias (including a family history of malignant hyperthermia). Baseline between‐group differences: not reported Additional treatment(s) not relevant to this review: not reported Patient satisfaction assessment: "No patient had restricted extraocular movements. No significant complications were noted. Based on a questionnaire of subjective postoperative improvement, most patients noticed an improvement in their quality of life after the surgery in both the groups."
Interventions	Intervention Characteristics “Large retro‐equatorial recession of horizontal rectus muscles 9 mm on the medial rectus and 12 mm on the lateral rectus (Group 1).” “horizontal rectus muscles were either recessed or tenotomised.”: large retroe‐quatorial recession of horizontal rectus muscle of 9 mm on the medial rectus and 12 mm on the lateral rectus “Simple tenotomy and resuturing of the 4 horizontal rectus muscles (Group 2).” “Horizontal rectus muscles were either recessed or tenotomised.”: simple tenotomy and resuturing of the 4 horizontal rectus muscles
Outcomes	Change in binocular best‐corrected distance visual acuity Outcome type: continuous outcome Reporting: partially reported (no variability measures reported) Data value: change from baseline Notes: “Group 1 patients showed 1 line of improvement in their binocular visual acuity for distance on the logMAR chart and on the ETDRS chart. There was no statistical significance to this change in our study. There was no improvement in binocular visual acuity on the logMAR chart or ETDRS chart in patients in group 2. On comparison there was no statistically significant difference between the groups.” Amplitude of the nystagmus waveform Outcome type: continuous outcome Reporting: partially reported (no variability measures reported) Data value: change from baseline Intensity (amplitude multiplied by frequency) of the waveform Outcome type: continuous outcome Reporting: partially reported (no variability measures reported) Data value: endpoint Notes: “The intensity decreased in both the groups in all gazes.” Quality of life and self‐reported outcome measures Outcome type: adverse event Reporting: partially reported (no quantitative data provided) Data value: endpoint Notes: “No patient had restricted extraocular movements. No significant complications were noted. Based on a questionnaire of subjective postoperative improvement, most patients noticed an improvement in their quality of life after the surgery in both the groups.” Frequency of the waveform Outcome type: continuous outcome Reporting: partially reported (no variability measures reported) Data value: change from baseline Notes: “The frequency of nystagmus did not show any significant change in either group. Intergroup and intragroup P values showed no difference.” No permanent adverse effects after surgery Outcome type: dichotomous outcome Reporting: fully reported Data value: endpoint Notes: "No participant had restricted extraocular movements. No significant complications were noted."
Identification	Sponsorship source: funding sources: not reported Declarations of interest: not reported Country: India Setting: Strabismus clinic of the Rajendra Prasad Centre for Ophthalmic Services Comments: Dates study conducted: not reported Trial registration number: not reported Full study name (if applicable): not reported Authors name: Pradeep Sharma Institution: All India Institute of Medical Sciences Email: drpsharma57@yahoo.com Address: Rajendra Prasad Centre for Ophthalmic Sciences, All India Institute of Medical Sciences, New Delhi 110029, India
Notes	Contacting study investigators: one review author (Laura Downie) emailed the trial corresponding author on 25 August 2020 to ask for: (i) details regarding the variance parameters for both treatment groups for the following outcome measures: change in binocular best‐corrected visual acuity, change in amplitude of the nystagmus waveform, change in intensity of the nystagmus waveform, and change in frequency of the nystagmus waveform; and (ii) quantitative data relating to quality of life outcomes. As of 8 February 2021, we had not received a response, thus, we did not use any additional information for the review.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "The cases were assigned by generating a random number table." Judgement comment: random number table used to generate randomisation sequence
Allocation concealment (selection bias)	Unclear risk	Quote: "Ten patients selected from the strabismus clinic of the Rajendra Prasad Centre for Ophthalmic Services were randomly assigned to two groups," Judgement comment: did not report how allocation was administered; trial was described as randomised, but without further details
Blinding of participants and personnel (performance bias) All outcomes	High risk	Judgement comment: no mention of masking; in the absence of reporting, we assume that masking was not performed
Blinding of outcome assessment (detection bias) All outcomes	High risk	Judgement comment: no mention of masking; in the absence of reporting, we assume that masking was not performed
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Judgement comment: follow‐up not clearly reported
Selective reporting (reporting bias)	Unclear risk	Judgement comment: no access to clinical trial registry or protocol
Other bias	Low risk	Judgement comment: no other apparent sources of bias

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Akbari 2013	Non‐RCT study design
Alió 2003	Non‐RCT study design
Atilla 1999	Non‐RCT study design
Bagheri 2008	Non‐RCT study design
Bilska 1995	Non‐RCT study design
Cherednichenko 1989	Non‐RCT study design
D'Esposito 1989	Non‐RCT study design
Daktaravichene 1989	Non‐RCT study design
Dorn 1982	Non‐RCT study design
Dubner 2016	Non‐RCT study design
EUCTR2007‐002595‐34‐GB	Evaluation of an alternative intervention
EUCTR2007‐007663‐25‐GB	Evaluation of an alternative intervention
Felius 2009	Non‐RCT study design
Felius 2012	Non‐RCT study design
Flynn 1979	Non‐RCT study design
Graf 2016	Non‐RCT study design
Graf 2019	Non‐RCT study design
Gupta 2006	Non‐RCT study design
Helveston 1991	Non‐RCT study design
Hertle 2006	Non‐RCT study design
Hertle 2009	Non‐RCT study design
ISRCTN65414827	Evaluation of an alternative intervention
Jiang 2006	Non‐RCT study design
Kose 2003	Non‐RCT study design
Krzystkowa 1995	Non‐RCT study design
Kumar 2011	Non‐RCT study design
Li 2006	Non‐RCT study design
NCT00001866	Evaluation of an alternative intervention
NCT00661440	Evaluation of an alternative intervention
NCT00799942	Evaluation of an alternative intervention
NCT01312402	Evaluation of an alternative intervention
Pratt Johnson 1991	Non‐RCT study design
Sendler 1990	Non‐RCT study design
Su 2005	Non‐RCT study design
Taylor 1987	Non‐RCT study design
Tomonari 2017	Non‐RCT study design
Wagdy 2017	Non‐RCT study design
Wang 1996	Non‐RCT study design
Wang 2008	Non‐RCT study design
Wang 2011	Non‐RCT study design
Zhao 2009	Non‐RCT study design
Zubcov 1993	Non‐RCT study design

Differences between protocol and review

Methods not implemented

Due to only identifying a single eligible randomised controlled trial (RCT) in the review, several aspects of the protocol could not be undertaken. We were unable to quantitatively report on treatment effects due to a limited number of eligible trials (N = 1) and paucity of quantitative data. Assessments of study heterogeneity and reporting biases were not able to be performed. We were unable to conduct meta‐analyses for any of the prespecified outcome measures, or subgroup analyses for factors considered to potentially affect outcomes. We could not perform a sensitivity analysis for the primary outcome measure as only one RCT was included in the review.

We had specified that we would generate 'Summary of findings' tables for the following prespecified comparisons:

• any surgical intervention versus no intervention;

• any surgical intervention versus any non‐surgical intervention;

• four horizontal rectus muscle tenotomy with reattachment versus no intervention;

• myectomy of the extraocular muscles without reattachment versus no intervention;

• myectomy of the extraocular muscles without reattachment versus any other surgical intervention.

As we did not identify any RCTs evaluating the clinical effects of these comparisons, we developed a 'Summary of findings' table for the two surgical interventions compared in the only eligible RCT, since we had listed the procedures as possible interventions in the Methods section of our protocol (Cham 2019).

Contributions of authors

Kwang M Cham screened the abstracts and full‐text articles. He assessed the risk of bias and extracted the data. He wrote the first draft of the review. He also revised the draft after receiving input from the full review team, and complied the final draft for submission. 

Laura E Downie assessed risk of bias and extracted data. She also revised the first draft of the full review. 

Ljoudmila Busija had substantial input in editing the first draft of the full review. 

Lionel Kowal had substantial input in editing the first draft of the full review. 

Anat Bachar Zipori had substantial input in editing the first draft of the full review. 

Larry A Abel screened the abstracts and full‐text articles. He had substantial input in editing the first draft of the full review.  

All authors provided final approval of the submitted review.

Sources of support

Internal sources

• The University of Melbourne, Monash University, Tel Aviv University, Australia

  The authors acknowledge the in‐kind support of their institutions for the undertaking of this review.

External sources

• National Institute for Health Research (NIHR), UK

  This review was supported by the NIHR, via Cochrane Infrastructure funding to the CEV UK editorial base.

  The views and opinions expressed therein are those of the authors and do not necessarily reflect those of the Systematic Reviews Programme, NIHR, NHS or the Department of Health.

Declarations of interest

KMC: none known
LAA: no direct conflicts of interest; one reimbursement by a non‐profit‐organised research workshop and one hour's expert witness testimony on nystagmus and stress
LB: Monash University received payments from Jesuit Social Services Limited (Victoria, Australia) and Charité – Universitätsmedizin Berlin for statistical consultancy work performed by LB
LK: none known
ABZ: none known
LED: has previously received funding to undertake clinical trials in the fields of dry eye disease and contact lenses, which are unrelated to this work, from Allergan Pty Ltd, Alcon Pty Ltd, Azura Ophthalmics Pty Ltd and Coopervision Pty Ltd. 

New