Negative dysphotopsia, a subjective dark shadow seen in the temporal field after cataract surgery, is an uncommon but well-described complication of cataract surgery.1–4 The prevalence of negative dysphotopsia has been reported to be as high as 19% immediately after surgery, decreasing to about 3% by 1 year.5,6 The negative dysphotopsia shadow is most noticeable with bright illumination in the temporal field, often improves with pupil dilation, and is reported with nearly all intraocular lens (IOL) designs.2,3 Mild negative dysphotopsia usually abates with time; however, severe cases might benefit from surgery. Surgical options in treating persistent negative dysphotopsia vary and include IOL exchange with ciliary sulcus placement, sulcus-fixated piggyback IOL implantation, reverse optic capture, and neodymium:YAG laser anterior capsulotomy.3,7–12
Several causative mechanisms of negative dysphotopsia have been proposed, although no consensus has been reached. Various studies3,4,8,13 found no clear correlation between negative dysphotopsia and IOL or biometric data, and the current opinion is that the negative dysphotopsia phenomenon is multifactorial in nature. Ray-tracing simulations have been used to try to determine the etiology of negative dysphotopsia.4,13–15 Analysis of light rays at large angles to the visual axis in schematic eyes show a nonilluminated area of peripheral retina that is located in a gap between rays that are refracted by the IOL optic and rays that miss the IOL entirely.13–15 This nonilluminated area of nasal retina has been implicated as the source of the temporal shadow seen by negative dysphotopsia patients.13,15
Although ray tracing is useful in depicting where light rays go, the effect of ray density on the relative intensity of the light is difficult to quantify from ray drawings. Although light reaching the far peripheral retina might not form a focused image, it does provide illumination. Previous studies have rarely included simulated brightness in the peripheral retina. Displaying ray-tracing results as a retina illumination profile is difficult for this visual region of the eye, in part because rays are focused on the curved surface of peripheral retina,16 but also because the image has to be scaled in a meaningful manner. This has recently been addressed by converting retinal locations to input visual angles and creating representative images on a polar plot.A
The purpose of our ray-tracing study was to use patient-specific biometric and IOL data to create simulated peripheral retina illumination profiles in patients with negative dysphotopsia to understand peripheral retina illumination before and after placement of a sulcus-fixated piggyback IOL.
Patients and methods
Patients with persistent negative dysphotopsia of more than 13 months were referred between January 2017 and January 2018. The use of retrospective clinical data was approved by the Institutional Review Board, Mayo Clinic, Rochester, Minnesota, USA.
All patients had uneventful phacoemulsification with capsular bag–fixated IOL implantation. Each patient had a different primary IOL type. Secondary piggyback IOLs were implanted by one of the authors (J.C.E.) at the Mayo Clinic. All piggyback IOLs were placed in the ciliary sulcus anterior to the primary IOL, and the power of the piggyback IOL was chosen to improve the final refractive error. A peripheral iridectomy was placed superiorly during surgery.
The patients’ biometric measurements preoperatively and 6 weeks after piggyback IOL placement were used for the construction of theoretic eye models. High-frequency ultrasound biomicroscopy (Quantel Medical) was used to measure the cornea–iris distance, iris thickness, iris–IOL distance, cornea–primary IOL distance, and cornea–secondary IOL distance (Figure 1). Partial coherence interferometry (Lenstar LS 900, Haag-Streit AG) was used to measure corneal power and axial length. Corneal astigmatism was not considered; however, the spheroequivalent anterior curvature used was calculated from the corneal power. The posterior corneal radius of curvature was estimated from the anterior surface using a typical radius ratio of 0.84 times the anterior radius.17 Measured IOL thicknesses were adjusted for differences in ultrasound velocities.
Patient-specific optical models were created using Zemax OpticStudio optical design software (Radiant Zemax). An extended light source was simulated using an array of object points at a distance of 6 m, with each point emitting light rays evenly spaced in the visual angle. This created uniform illumination for a region covering 60 to 110 degrees relative to the horizontal optical axis, and +10 degrees vertically to correctly illuminate a region covering 60 to 100 degrees by +3 degrees because a peripheral point-spread function was very large. A 2.5 mm diameter pupil (3.0 mm apparent pupil)4 that was decentered nasally by 0.25 mm was used for all cases, and additional calculations were done in 1 case with a set of different pupil diameters. Model eyes were centered on the optical axis, and 5 degrees were added to the horizontal input ray angles to account for the fovea location.
Because there is not a 1:1 relationship between object and image points at larger visual angles, locations on the retina were scaled directly to object space using the angle subtended at the second nodal point. This scaling concept has been used before15; however, the retinal intensities were used in a new manner to create a simulated image on a polar plot centered on the visual axis. The angular relationship is highly linear for the chief ray that passes through the center of the pupil to at least 70 degrees, and it is assumed that the relative values are approximately correct for even higher angles, although the limitations of this model are not known. Aberrations alter the effective ray locations for large angles for the light passing through the IOL. Light rays missing the IOL enter the eye at different angles to the IOL-focused light; thus, the 2 regions overlap for larger pupil diameters. It is assumed that the retina is symmetrical, and the same relationship is used for all azimuthal angles. The retinal region that was used for evaluations corresponded to 60 to 110 degrees in the temporal direction and +3 degrees vertically.
Zemax ray-tracing software does not directly generate the types of images that are of interest here, so the following methods were used: (1) Special optical surfaces were created for the IOL using the “user-defined surface” capability so that rays passing through the IOL and rays missing the IOL were all recorded for a sequential ray trace. (2) The model eye was set up so that it could be rotated about a vertical axis, and a macro was used to record the coordinates of ray intersections on the retina for a large number of object points at a 6 m distance. The macro takes advantage of the “ray aiming” capability of the ray-tracing software to find the location of the iris boundary for rays from each object point and then launches a set of rays that are equally spaced in angle. A separate “polarization” routine was used to calculate the fraction of each refracted light ray that reached the retina, and after Fresnel reflections were removed, this was also saved. (3) The retinal ray intersections and intensity values were then imported into MATLAB (MathWorks, Inc.) for analysis and display. The angles subtended at the second nodal point were calculated, using the nodal point value calculated by the ray-tracing software, and the weighted rays were added into an array that represented the relative polar locations in object space of the retinal locations.
Additional weighting of the rays was also necessary to equalize the energy as the angles increased. This is because the pupil becomes increasingly elliptical with increasing visual angle, resulting in the horizontal pupil opening becoming progressively smaller. This has an effect on the transmitting area, which falls off as an approximate cosine function.18 The mechanism by which the eye compensates for the image becoming dimmer in the periphery is unknown19; to counteract it, each ray was weighted by 1/(cos[0.8 × retinal angle]). This increases the relative intensity of the most peripheral rays and largely compensates for pupil obliquity on light intensity to provide a standardized image so that the relative image properties can be evaluated. The simulated images also use a gamma correction value of 0.3 to enhance the visibility of lower intensity levels, and each image was scaled to fill the full intensity range.
The exact IOL optic geometry and design for the 3 primary IOLs and for the sulcus-fixated piggyback IOL are unpublished. Approximations of the designs were estimated using information given on manufacturer websites and other published sources. The primary IOL optics have a 6.0 mm diameter; Figure 2 shows the estimated central cross-sectional profiles. The haptics and the IOL edges were not included in ray analysis. Primary and secondary IOLs were considered to be free-standing in aqueous humor.
The study evaluated 3 patients with persistent negative dysphotopsia of more than 13 months. Additional calculations for pupil decentration were performed in Case 1.
Table 1 shows the patient and IOL characteristics and the outcomes after implantation of a sulcus-fixated piggyback IOL. All were left eyes that had a corrected distance visual acuity of 20/20. Preoperatively, all patients reported a persistent temporal shadow that developed immediately after cataract surgery, increased in photopic conditions, and lessened or disappeared after pupil dilation. After secondary sulcus-fixated IOL placement, the negative dysphotopsia shadow resolved in 2 eyes and was unchanged in 1 eye.
Preoperatively, the mean space between the posterior iris surface and the primary IOL was 0.45 mm (range 0.35 to 0.63 mm). Placement of a piggyback IOL shifted the iris anteriorly (mean 0.41 mm; range 0.37 to 0.47 mm), shifted the primary IOL posteriorly (mean 0.11 mm; range 0.05 to 0.19 mm), and increased the mean distance between the iris and primary IOL to 0.75 mm.
Figure 3 shows a preoperative ray-tracing diagram of the pseudophakic eye in Case 1. It also shows the simulated retinal illumination profile and a polar plot of that profile.
Figure 4 shows simulated retina illumination profiles before and after piggyback IOL placement in all 3 eyes. In Cases 1 and 3, the temporal shadow was subjectively improved after piggyback IOL placement. The postoperative retinal illumination profiles for these cases (Figure 4, A and C) show increased illumination of the peripheral retina that included the preoperative dark area. In Case 2, in which negative dysphotopsia symptoms persisted after piggyback IOL placement, the postoperative retina illumination profile showed the dark retinal region to be largely unchanged and the far peripheral retina to be less illuminated compared with preoperatively (Figure 4, B).
The rays traced in Figure 5 for Case 1 show how a piggyback IOL alters peripheral retina illumination. The main effect comes from the piggyback IOL acting as a spacer and increasing the distance between the iris and the primary IOL by shifting the iris anteriorly compared with the preoperative iris. The refractive index of the piggyback IOL also has a small effect because it laterally moves the light rays that are incident on the primary IOL without changing their angles. (This would be slightly different in Case 2, in which where the piggyback IOL has some power.) The ray shifting slightly reduces the extent of the focused image for the piggyback situation (Figure 5). The effect this has on rays that miss the IOL is already included in Figure 5, but if the iris had been moved forward without the use of a piggyback IOL, input light from even lower visual angles would bypass the primary IOL.
Figure 6 shows a separate analysis of Case 1 before IOL placement. As the pupil diameter increases from 2.5 mm to 5.0 mm, the dark region of peripheral retina disappears. A simulated peripheral image is given with a radial intensity plot across each image.
In our series of 3 patients with persistent negative dysphotopsia after cataract surgery, retina illumination modeling showed a dim or dark region in the peripheral nasal retina that is consistent with previous ray-tracing diagrams and similar in location to the temporal shadow described by patients with negative dysphotopsia.4,13–15,20 Placement of a sulcus-fixated piggyback IOL shifts the iris anteriorly, and to a lesser extent the primary IOL posteriorly. This anatomic change nearly doubles the space between the iris and primary IOL compared with that preoperatively, although it also changed the refractive index of the region from aqueous to silicone. This additional physical space allows input light at lower visual angles to miss the primary IOL and illuminate peripheral retina that was not illuminated preoperatively. In 2 cases, increased illumination of the peripheral retina, including the preoperative dark area, was associated with improvement in the negative dysphotopsia shadow. In 1 case, the dark region in illumination changed little and the negative dysphotopsia shadow was subjectively unchanged.
At present, the mechanism underlying negative dysphotopsia has been considered to be multifactorial.3,4,8,13 The pseudophakic eye, however, is different from the phakic eye, and the differences were not identified in earlier discussions on the topic. The crystalline lens is much thicker and larger in diameter than an IOL optic, with the iris always in contact with the natural lens; thus, all light entering the pupil also passes through the lens. In contrast, after removal of a cataract, the iris shifts posteriorly by a mean of 0.8 mm21 and a new space between the iris and the capsular bag–fixated IOL is created.9 As a consequence, input light at large visual angles can completely miss the IOL and access this space to illuminate nasal retina. Ray-tracing diagrams consistently show a dim or nonilluminated region of peripheral nasal retina that is formed by the gap between the rays that miss the IOL and the rays focused by the IOL optic, and this likely corresponds to the negative dysphotopsia shadow.4,13,15,16,19
Although ray-tracing diagrams show where peripheral light rays go inside the eye, ray diagrams do not readily quantify brightness or light intensity changes. This is important, because variation in illumination across the peripheral nasal retina might be interpreted clinically as bright or dark variation in the temporal visual field. We used our new retina illumination profile in negative dysphotopsia patients to quantify light-intensity differences across the peripheral retina and confirmed previous ray-tracing diagrams, showing a dim or nonilluminated area of nasal retina. Furthermore, superimposing the illumination profile on a polar plot showed that the illumination gap was similar in location to the temporal shadow described by patients with negative dysphotopsia and suggested by other authors.4,13,20
Placement of a sulcus-fixated piggyback IOL is an accepted surgical treatment for negative dysphotopsia, although its success rate in improving the temporal shadow is approximately 73%.3,11 The mechanism by which a piggyback IOL alters illumination of the peripheral retina has not been previously explained. Our findings show that a piggyback IOL acts as a spacer to further expand the space between the iris and the primary IOL, primarily by shifting the iris anteriorly (mean 0.4 mm) and to a lesser degree by shifting the primary IOL posteriorly (mean 0.1 mm). This added space allows more input light at lower visual angles to miss the primary IOL and increase illumination of the peripheral retina. In our small series, increased illumination of the peripheral retina, including the preoperative dark region, was associated with subjective improvement of the temporal shadow. Shifting the iris anteriorly by a mean of 0.4 mm regains only 50% of the initial 0.8 mm mean posterior shift after removal of the cataract.21
Peripheral retina illumination in patients with negative dysphotopsia was recently studied by Makhotkina et al.8 Their optical modeling following a secondary sulcus-fixated IOL showed increased light irradiance in the peripheral retina in patients with negative dysphotopsia whose temporal shadow improved subjectively and on Goldmann perimetry.8,20 Our optical modeling differed because their irradiance measurements extended to approximately 80 degrees temporally, whereas our illumination measurements extended up to 100 degrees.
Our analysis included 3 primary IOLs with different optic materials, refractive indices, and cross-sectional geometries. All primary IOLs had different peripheral retina illumination profiles before and after piggyback IOL placement. The IOL cross-sectional profile in Case 2 is not widely discussed in the literature in regard to negative dysphotopsia; however, one method for achieving a small incision with an IOL that has a lower refractive index is to reduce the optic diameter on one or more surfaces. Our results suggest that different IOL styles might affect peripheral retina illumination differently and influence the success rate of a secondary piggyback IOL. Larger studies comprising more patients with negative dysphotopsia and separately analyzing of effect of IOL geometry are needed.
If an illumination gap in the peripheral nasal retina is responsible for the temporal shadow described by patients with negative dysphotopsia, successful treatments for negative dysphotopsia should also improve peripheral retina illumination and diminish an illumination gap. First, pupil dilation partially or completely improves the temporal shadow in some patients,2,3 and this was the situation in our Case 1. Retinal illumination calculations in this case documents that the illumination gap disappears after pupil dilation (Figure 6). Previous ray tracing shows that as the pupil diameter increases, input light at low visual angles is not blocked by the iris and accesses the space between the primary IOL and iris.15,A This results in increased illumination of the peripheral retina, often including illumination of a dim or dark retina region. Therefore, pharmacologic pupil dilation and an iris that shifted anteriorly by a piggyback IOL have a similar mechanism of action in that both allow more light rays to miss the IOL, illuminating more peripheral retina.
Second, negative dysphotopsia usually improves with time. It is thought that this is the result of time-related opacification of the capsule peripheral to the optic. The translucent capsule scatters light rays that miss the IOL onto a more posterior peripheral retina region, illuminating a larger area of peripheral retina, including the dark region.
Third, exchanging a capsular bag–fixated IOL with a sulcus-fixated IOL moves the IOL anteriorly and puts it in contact with the iris, and nearly always improves negative dysphotopsia symptoms.3,9 Similar to the crystalline lens, all light that enters the pupil now passes through and is refracted by the sulcus IOL. No light misses the IOL, although the extent of the visual image might still be limited by the diameter of the IOL. Vámosi et al.9 found no difference in the iris–IOL distance between patients with negative dysphotopsia and an asymptomatic control group. However, they also stated that the iris–IOL distance was an important parameter because narrowing the iris–IOL distance was so effective in reducing negative dysphotopsia. Our work shows that increasing the iris–primary IOL distance by using a piggyback IOL can reduce negative dysphotopsia symptoms in some patients. We believe that a space between the iris and IOL is necessary to create an illumination gap on the nasal retina. However, whether the patient is symptomatic from variable illumination of the peripheral retina depends on other factors, such as the extent of functional peripheral nasal retina.
Accurately estimating the intensity of peripheral light reaching the pseudophakic retina can be difficult. First, eccentric light entering the pseudophakic eye is subject to polarization sensitive transmission losses described by Fresnel reflection equations. Reflections increase with higher incident angles and higher refractive index material, which are complicated effects because the lens shape is adjusted to achieve correct IOL powers based on the particular optic material. Second, the pupil becomes increasingly elliptical with increasingly eccentric light, and this reduces the pupil area and limits the entry of peripheral rays into the eye. To compensate for this, we approximated rays by a cosine function, scaled the image to fill the available intensity range, and adjusted the visual image by a gamma value of 0.3.15
Our study has limitations. First, our case series is small. Second, the visual angle of the shadow before the second surgery was not measured. Characteristics of the temporal shadow have been recorded by using Goldmann perimetry for light input at different angles.20 However, measurement of the shadow location is more difficult than it might seem, and questions have been raised about what is actually being measured at large visual angles. Goldmann perimetry extends to 90 degrees of the visual angle, with limited resolution at larger angles. A recent studiesB showed that the temporal visual field can extend to a mean of 100 degrees (range 88 to 108 degrees). Third, at some angle, the light no longer passes through the IOL and instead passes through the peripheral capsule. Our modeling did not account for variable opacification of the anterior or peripheral capsule. Third, in the phakic eye, input visual angles typically correspond directly with a single retinal region. In the pseudophakic eye, it is possible for light to miss the IOL at large angles and the illumination angle does not correspond with the retinal location that is expected. Fourth, approximations were used in the modeling. The actual asphericity of the cornea was unknown, and an average asphericity was used for the central anterior cornea, with a radius of curvature estimated from the keratometry value. The asphericity of the peripheral cornea was not known or included in calculations. The posterior corneal surface radius was assumed to be a typical value of 0.84 of the anterior radius, with the asphericity assumed in earlier modeling.17 A spherical retina was assumed using a typical radius of 12.0 mm. However, because the main comparisons were before and after surgery, the relative changes should not be affected.
Finally, in a conventional IOL, light at large visual angles might be spread across the retina because of imaging aberrations. This is different from other optical systems, which in general are limited to a region where the image quality is high and each image point is matched to a distinct object location. In the periphery of the eye, light can be highly defocused. In addition to the aberration effect, light can miss the IOL completely and reach the retina at a location and image size that is consistent with light focused by the cornea only and not by the IOL. These effects are addressed by evaluating retinal locations rather than input angles for light rays. This is something that can be done using a model for the eye but that is not possible in standard clinical testing.
In summary, our ray-tracing and retina illumination models provide additional support for a dim or nonilluminated region in the nasal peripheral retina as the reason for the temporal shadow seen by patients with negative dysphotopsia. In addition, placement of a secondary sulcus-fixated piggyback IOL appears to improve negative dysphotopsia by increasing illumination of the peripheral retina, including the dim or dark preoperative illumination gap. Although persistent negative dysphotopsia is uncommon after cataract surgery, there is no method to currently predict which patients might be bothered by a temporal shadow. The work here suggests that the distance between the iris and the IOL is an important parameter in the illumination of peripheral retina and in negative dysphotopsia. Further work is needed to design IOLs that more uniformly illuminate the peripheral retina, reduce the illumination gap, and potentially prevent negative dysphotopsia.
What Was Known
- In negative dysphotopsia after cataract surgery, there is no consensus on the etiology of the temporal shadow or the mechanism by which a secondary sulcus-fixated intraocular lens (IOL) is sometimes beneficial in eliminating symptoms.
What This Paper Adds
- Our study provides evidence to support that a dim or dark region in the nasal peripheral retina is the reason for the temporal shadow in negative dysphotopsia.
- A sulcus-fixated piggyback IOL improved negative dysphotopsia by shifting the iris anteriorly, allowing additional light at lower visual angles to miss the primary IOL and increase illumination of peripheral retina, including an area poorly or nonilluminated preoperatively.
1. Davison JA. Positive and negative dysphotopsia in patients with acrylic intraocular lenses. J Cataract Refract Surg
2. Masket S, Fram NR. Pseudophakic negative dysphotopsia: surgical management and new theory of etiology. J Cataract Refract Surg
3. Masket S, Fram NR, Cho A, Park I, Pham D. Surgical management of negative dysphotopsia. J Cataract Refract Surg
4. Holladay JT, Simpson MJ. Negative dysphotopsia: causes and rationale for prevention and treatment. J Cataract Refract Surg
5. Osher RH. Negative dysphotopsia: long-term study and possible explanation for transient symptoms. J Cataract Refract Surg
6. Makhotkina NY, Nijkamp MD, Berendschot TTJM, van den Borne B, Nuijts RMMA. Effect of active evaluation on the detection of negative dysphotopsia after sequential cataract surgery: discrepancy between incidences of unsolicited and solicited complaints. Acta Ophthalmol
7. Makhotkina NY, Berendschot TTJM, Beckers HJM, Nuijits RMMA. Treatment of negative dysphotopsia with supplementary implantation of a sulcus-fixated intraocular lens. Graefes Arch Clin Exp Ophthalmol
8. Makhotkina NY, Dugrain V, Purchase D, Berendschot TTJM, Nuijts RMMA. Effect of supplementary implantation of a sulcus-fixated intraocular lens in patients with negative dysphotopsia. J Cataract Refract Surg
9. Vámosi P, Csákány B, Németh J. Intraocular lens exchange in patients with negative dysphotopsia symptoms. J Cataract Refract Surg
10. Cooke DL, Kasko S, Platt LO. Resolution of negative dysphotopsia after laser anterior capsulotomy. J Cataract Refract Surg
11. Burke TR, Benjamin L. Sulcus-fixated intraocular lens implantation for management of negative dysphotopsia. J Cataract Refract Surg
12. Folden DV. Neodynium:YAG laser anterior capsulotomy: surgical option in the management of negative dysphotopsia. J Cataract Refract Surg
13. Holladay JT, Zhao H, Reisin CR. Negative dysphotopsia: the enigmatic penumbra. J Cataract Refract Surg
14. Hong X, Liu Y, Karakelle M, Masket S, Fram NR. Ray-tracing optical modeling of negative dysphotopsia. J Biomed Opt. 2011;16(12):125001.
15. Simpson MJ. Double image in far peripheral vision of pseudophakic eye as source of negative dysphotopsia. J Opt Soc Am A Opt Image Sci Vis
16. Simpson MJ. Vignetting and negative dysphotopsia with intraocular lenses in “far peripheral vision”. J Opt Soc Am A Opt Image Sci Vis
17. Goncharov AV, Dainty C. Wide-field schematic eye models with gradient-index lens. J Opt Soc Am A Opt Image Sci Vis
18. Mathur A, Gehrmann J, Atchison DA. Pupil shape as viewed along the horizontal visual field. J Vis. 2013;13(6):1-8.
19. Simpson MJ. Mini-review: far peripheral vision. Vision Res
20. Makhotkina NY, Berendschot TTJM, Nuijts RMMA. Objective evaluation of negative dysphotopsia with Goldmann kinetic perimetry. J Cataract Refract Surg
21. Simpson MJ, Muzyka-Woźniak M. Iris characteristics affecting far peripheral vision and negative dysphotopsia. J Cataract Refract Surg
None of the authors has a financial or proprietary interest in any material or method mentioned.
Other Cited Material
A. Simpson MJ, “Modeling Intraocular Lens Negative Dysphotopsia and Visual Phenomenon,” presented at the 9th European Meeting on Visual and Physiological Optics. Athens, Greece, August 2018.
B. Bain C, Marín-Franch I, McNaught AI, Artes PH, “The Limits of the Far Peripheral Visual Field,” presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Honolulu, Hawaii, USA, May 2018. Abstract in Invest Ophthalmol Vis Sci 2018; 59(9):1272. Available at: https://iovs.arvojournals.org/article.aspx?articleid=2693356&resultClick=1
. Accessed November 16, 2018.