Government Documents

Corneal Sensitivity and Slit Scanning In Vivo Confocal Microscopy of the Subbasal Nerve Plexus of the Normal Central and Peripheral Human Cornea

Corneal Sensitivity and Slit Scanning In Vivo Confocal Microscopy of the Subbasal Nerve Plexus of the Normal Central and Peripheral Human Cornea
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  C LINICAL S CIENCE Corneal Sensitivity and Slit Scanning In Vivo ConfocalMicroscopy of the Subbasal Nerve Plexus of theNormal Central and Peripheral Human Cornea  Dipika V. Patel, PhD, MRCOphth,* Mitra Tavakoli, MSc,† Jennifer P. Craig, PhD,* Nathan Efron, PhD, DSc,‡ and Charles N. J. McGhee, PhD, FRCOphth* Purpose: To determine the subbasal nerve density and tortuosity at 5 corneal locations and to investigate whether these microstructuralobservations correlate with corneal sensitivity. Method: Sixty eyes of 60 normal human subjects were recruited into 1 of 3 age groups, group 1: aged  , 35 years, group 2: aged 35–50years, and group 3: aged  . 50 years. All eyes were examined usingslit-lamp biomicroscopy, noncontact corneal esthesiometry, and slit scanning in vivo confocal microscopy. Results: The mean subbasal nerve density and the mean cornealsensitivity were greatest centrally (14,731 6 6056 m m/mm 2 and 0.38 6 0.21 millibars, respectively) and lowest in the nasal mid  periphery (7850 6 4947 m m/mm 2 and 0.49 6 0.25 millibars,respectively). The mean subbasal nerve tortuosity coefficient wasgreatest in the temporal mid periphery (27.3 6 6.4) and lowest in thesuperior mid periphery (19.3 6 14.1). There was no significant difference in mean total subbasal nerve density between age groups.However, corneal sensation (  P  = 0.001) and subbasal nerve tortuosity(  P  = 0.004) demonstrated significant differences between age groups.Subbasal nerve density only showed significant correlations withcorneal sensitivity threshold in the temporal cornea and with subbasalnerve tortuosity in the inferior and nasal cornea. However, thesecorrelations were weak. Conclusions: This study quantitatively analyzes living humancorneal nerve structure and an aspect of nerve function. There is nostrong correlation between subbasal nerve density and cornealsensation. This study provides useful baseline data for the normalliving human cornea at central and mid-peripheral locations. Key Words: cornea, confocal microscopy, nerves, sensitivity,tortuosity( Cornea 2009;28:735–740) T he cornea is one of the most densely innervated tissues inthe human body, its innervation srcinating from theophthalmic branch of the trigeminal nerve. As well as servinga protective function, corneal nerves also play important rolesin regulating corneal epithelial integrity, proliferation, and wound healing. 1 Corneal nerve architecture has been extensively studied using light and electron microscopy. 2–4 However, studies usingthese techniques may be unrepresentative and limited inaccuracy because human corneal nerves are known todegenerate within 13.5 hours of death. 3 In vivo confocalmicroscopy overcomes this problem by enabling noninvasiveexamination of the living human cornea.Unlike stromal corneal nerves, which vary in orienta-tion, the subbasal nerve plexus lies parallel to the cornealsurface between Bowman membrane and the basal epithelium.This enables reproducible imaging and quantification of nerve parameters in this layer. The nerve terminals that extend fromthe subbasal nerve plexus into the epithelium cannot beresolved by in vivo confocal microscopy. Therefore, this studysought to use the subbasal nerve plexus density as a marker for corneal epithelial innervation. Although a number of studieshave used in vivo confocal microscopy to quantitativelyanalyze the subbasal plexus in the human cornea, eachassessed only the central cornea. 1,5–9 The purpose of this study was to quantitatively analyzethe subbasal nerve plexus at 5 corneal locations using slit scanning in vivo confocal microscopy and to correlate thesemicrostructural observations with noncontact cornealsensitivity. MATERIALS AND METHODSSubjects After written informed consent, 60 human subjects wererecruited and 1 normal eye of each subject was included in thestudy. Adult subjects were recruited into 1 of 3 age bands,group 1: aged 20–34 years, group 2: aged 35–50 years, and group 3: aged 51–65 years. Received for publication Janurary 6, 2008; revision received October 2, 2008;accepted October 25, 2008.From the *Department of Ophthalmology, Faculty of Medical and Healthscience, University of Auckland, Auckland, New Zealand; † Division of Cardiovascular Medicine, University of Manchester, Manchester, United Kingdom; and  ‡ School of Optometry, Queensland University of Technology, Brisbane, Australia.Supported in part by an unrestricted award from The Maurice and PhyllisPaykel Trust.The authors have no proprietary interest in any of the devices mentioned inthis article.Reprints: Prof Charles N. J. McGhee, PhD, FRCOphth Department of Ophthalmology, Private Bag 92019, University of Auckland, Auckland, New Zealand 1142 (e-mail:  Ó 2009 by Lippincott Williams & Wilkins Cornea    Volume 28, Number 7, August 2009 | 735  Exclusion criteria were previous contact lens wear,history of ocular trauma or surgery, ocular disease, and systemic diseases that may affect the cornea. Informed writtenconsent was obtained from all subjects after full explanation of the nature and possible consequences of the study. The study protocol was approved by the Auckland Medical EthicsCommittee. Examinations All subjects were carefully examined by slit-lamp biomicroscopy to exclude corneal disease. Corneal sensitivitywas then measured by noncontact pneumatic cornealesthesiometry 10 using a standardized 10-mm working distanceand 0.9-second stimulus duration. Corneal sensitivity thresh-old measurements were made on the central cornea and in4 quadrants—the inferior, temporal, nasal, and superior mid  peripheries. The corneal sensitivity threshold was determined using the forced-choice double-staircase technique, that is, thesubject was initially presented with a suprathreshold stimulusthat was subsequently decreased until the subject could no longer detect it. The crossover point was noted, and asubthreshold stimulus presented and subsequently increased inintensity until a positive response was obtained. The mean of these crossover points was the corneal sensitivity threshold.In vivo confocal microscopy, using slit scanningtechnology (Confoscan 2; Fortune Technologies America,Greensboro, NC), was subsequently performed on all subjectsusing the previously described protocol. 11 For all subjects, thecentral cornea was examined using a standard setting of 4 passes, with a scanning range between 700 and 800 m m(throughout the z  axis) to image the full corneal thickness and a 200 m m scanning range to specifically image the subbasalnerve plexus centrally and in the inferior, temporal, nasal, and superior mid periphery (Fig. 1). To obtain tangential images at mid-peripheral locations, subjects were asked to gaze in eachdirection and the examiner aligned the microscope lensappropriately. Image Analysis For each corneal location in each subject, 2 imagescontaining the maximum number of nerves at the level of the FIGURE 1. In vivo confocal microscopy images of the corneal subbasal nerve plexus taken from the central cornea and in theinferior, temporal, nasal, and superior mid periphery of the same subject (Frame sizes 340 3 255 m m). 736 | q 2009 Lippincott Williams & Wilkins Patel et al  Cornea    Volume 28, Number 7, August 2009  subbasal plexus were selected. All frames were subsequentlyrandomized and encoded by an independent observer (J.P.C.).Measurements were performed using a caliper tool(analySIS 3.1; Soft Imaging System, Mu¨nster, Germany). For all images, a standard frame size of 283 3 212 m m (area 0.06mm 2 ) was selected. Nerve density was assessed by measuringthe total length of nerves per frame, and the mean of the 2nerve densities for each corneal location in each subject wasrecorded. Nerve tortuosity was computed using a novel mathe-matical program, which assessed the previously reported ‘‘tortuosity coefficient’’ (TC). 12 Images of the subbasal nerve plexus were processed using the Scion beta 4.02 image processing software. The red, green, and blue color imageswere converted to 8-bit indexed color. A grip pen was used tomanually trace, in black, 1 nerve at a time, along its axis, byselecting the pencil tool in the toolbox window. Each imagewas then threshold converted to a grayscale unit of 255 and thegrayscale image converted to binary, by setting pixels that had  been highlighted by thresholding to black (255) and all other  pixels to white (0), resulting in an image where nervesappeared black against a white background (Fig. 2A).The processed images were saved in tagged image file(TIF) format, and the TC was calculated using a MATLABfunction (MATLAB,version6.5;MathWorks) that was created specifically for this purpose. The average TC was calculated for all main nerves, within the frame. When nerves exhibited a branching pattern, only the thickest branch was considered asa continuation of the nerve. Using the MATLAB built-infunction ‘‘im2double,’’ the image was converted to an array(matrix) of numbers. The elements of the matrix were either 0 (background) or 1 (nerve). The coordinates of the nervewerethe indices of the ‘‘nonzero’’ entries in the matrix, which werereturned by the MATLAB built-in function ‘‘find.’’ A straight line that connected the end points of the nervewas plotted, and the image was translated and rotated to the srcin, to align thestraight line with the x axis (Figs. 2B, C). Statistical Analysis SPSS version 12 for windows (Chicago, IL) was used for statistical analysis. The Kruskal–Wallis test was used todetermine any significant differences between multiplegroups.The Mann–Whitney U  test was used to compare the means of variables between pairs of groups. P  values of  , 0.05 wereconsidered significant for all statistical tests. RESULTS The mean and range of ages in each group, and thesex ratios are provided in Table 1. A total of 600 images of thesubbasal nerve plexus were selected for analysis. Corneal Location The mean subbasal nerve density and the mean cornealsensitivity (inverse of sensitivity threshold) were greatest centrally and lowest in the nasal mid periphery. Subbasal nervetortuosity was greatest in the temporal mid periphery and lowest in the superior mid periphery. FIGURE 2. Measuring nerve tortuosity. A, A sample nerve. B,The straight line that connects the ends of the nerve is plotted,and the image is translated to the srcin. C, Image B after rotation and rescaling of the axis. q 2009 Lippincott Williams & Wilkins | 737 Cornea    Volume 28, Number 7, August 2009 Corneal Sensitivity and Subbasal Nerve Plexus   When analyzed according to corneal location using theKruskal–Wallis test, all parameters demonstrated significant differences between the locations (Table 2). Post hoc Mann– Whitney U  tests on pairs of location groups revealed that subbasal nerve density and corneal sensation were signifi-cantly greater in both the central cornea and temporal mid  periphery compared with all other peripheral locations (  P  # 0.01). Subbasal nerve tortuosity was significantly different for all pairs except central compared with inferior cornea (  P  =0.10) and nasal compared with superior cornea (  P  = 0.13). Age When analyzed according to age using the Kruskal– Wallis test, there was no significant difference in mean totalsubbasal nerve density with age. Corneal sensation and subbasal nerve tortuosity demonstrated significant differences between the age groups (Table 3). Post hoc Mann–Whitney U  tests on pairs of age groups revealed that corneal sensitivitythreshold was significantly lower in group 2 compared withgroups 1 (  P  = 0.01) and 3 (  P  , 0.01). However, there was nosignificant difference in corneal sensation between groups 1and 3 (  P  = 0.37). The subbasal nerve tortuosity wassignificantly greater in group 3 compared with group 1(  P  , 0.01) and group 2 (  P  = 0.04). There was no significant difference in nerve tortuosity between groups 1 and 2(  P  = 0.25). Correlations For each corneal location (N = 60), Spearman rhocorrelation was performed to determine any significant relationships between variables (Table 4). Subbasal nervedensity only showed significant correlations with cornealsensitivity threshold in the temporal cornea (Fig. 3) and withsubbasal nerve tortuosity in the inferior and nasal cornea.However, these correlations were weak. DISCUSSION The mean central subbasal nerve density in this study(14,731 m m/mm 2 ) was higher than that reported by other investigators using slit scanning confocal microscopy (11,110 m m/mm 2 ) 1 and tandem scanning confocal microscopy (8404 m m/mm 2 ). 6 The disparity is likely to be related to differencesin methodology. In the current study, images containing themaximum number of nerves at the level of the subbasal plexuswere selected, whereas other investigators have selected ‘‘the best-focused’’ and ‘‘most representative’’ images. 1 Another important difference is in the manner in which nerve density isdefined. The current study defined this parameter as the totallength of nerves visible within a defined frame. Someinvestigators have only included nerve branches longer than50 m m when measuring the total length of nerves and includethe whole area of the image, rather than a defined frame, in the TABLE 1. Summary of the 3 Subject Groups Categorized by Age Range No. Eyes Mean Age (Yrs) Age Range (Yrs) Males:Females Group 1 20 25.62 6 3.02 21.72–30.84 11:9Group 2 20 43.46 6 4.64 35.30–49.26 7:13Group 3 20 60.73 6 6.74 51.47–61.64 6:14 TABLE 2. Comparison of Subbasal Nerve Density, CornealSensitivity Threshold, and TC for Each of the 5 CornealLocations Assessed Corneal LocationMean SubbasalNerveDensity 6 SD( m m/mm 2 )Mean CornealSensitivityThreshold 6 SD(Millibars)MeanTC 6 SD Central 14,731 6 6056 0.38 6 0.21 23.8 6 14.5Inferior mid periphery 8477 6 6460 0.45 6 0.22 25.2 6 13.0 Nasal mid periphery 7850 6 4947 0.49 6 0.25 20.1 6 11.7Superior mid periphery 8566 6 6441 0.48 6 0.23 19.3 6 14.1Temporal mid periphery 12,556 6 6909 0.41 6 0.26 27.3 6 6.4Kruskal–Wallis test  P  , 0.001 P  , 0.001 P  , 0.001 Using the Kruskal–Wallis statistical test, all parameters demonstrated significant differences between the locations. TABLE 3. Summary of Data for All Parameters According to Age Group Age GroupMean TotalSubbasal NerveDensity 6 SD( m m/mm 2 )Mean CornealSensitivityThreshold 6 SD(Millibars)MeanTC 6 SD Group 1 10,641 6 6819 0.44 6 0.21 20.8 6 9.5Group 2 10,115 6 6766 0.39 6 0.18 23.2 6 13.7Group 3 10,552 6 6688 0.50 6 0.29 25.7 6 14.0Kruskal–Wallis test  P  = 0.70 P  = 0.001 P  = 0.004 Using the Kruskal–Wallis test, no significant difference was detected in mean totalsubbasal nerve density with age; however, corneal sensitivity and subbasal nervetortuosity demonstrated significant differences between the age groups. TABLE 4. Spearman Rho Correlation for Subbasal NerveDensity, Tortuosity, and Corneal Sensitivity for Each CornealLocation (N = 120) SensitivityThreshold Tortuosity Subbasalnerve densityCentral cornea Correlationcoefficient  2 0.005 2 0.115  P  0.959 0.217Inferior cornea Correlationcoefficient  2 0.163 0.298  P  0.077 0.003 Nasal cornea Correlationcoefficient 0.011 0.220  P  0.908 0.026Superior cornea Correlationcoefficient  2 0.026 0.057  P  0.779 0.573Temporal cornea Correlationcoefficient 0.180 2 0.025  P  0.050 0.796 738 | q 2009 Lippincott Williams & Wilkins Patel et al  Cornea    Volume 28, Number 7, August 2009  measurement. 6 This is likely to result in a lower measured nerve density than the current study because the regions alongthe vertical edges of the image, where there is a marked reduction in image contrast, are not excluded.The current study used in vivo confocal microscopy toquantitatively analyze the subbasal nerve plexus in the cornealmid periphery and to correlate these observations with cornealsensation. Although the Cochet-Bonnet esthesiometer is themost widely used technique for measuring corneal sensitivity,it suffers from several drawbacks. 10,13 In particular, it isinvasive and inevitably damages the superficial cornealepithelium. This may arguably produce an artificially in-creased sensitivity because of the presence of free nerveendings within the corneal epithelium. Another disadvantageis the limited range of stimulus intensities available. Thus, theminimum stimulus is often suprathreshold. The current studyused the noncontact pneumatic corneal esthesiometer toovercome these problems, with the added advantages that a large continuous range of stimulus intensities can be produced, and testing is not usually associated with patient apprehension. The forced-choice double-staircase techniqueemployed in the current study further refines the threshold stimulus.A limitation of the current study is the difficulty inensuring exact correspondence of locations tested byesthesiometry and in vivo confocal microscopy. Subjectivealignment was performed because the use of fixation targetswas not a viable option with the model of in vivo confocalmicroscope used. During image acquisition, the microscopehousing obscures the view from the subject’s contralateral eye.Additionally, placing targets on the microscope housing (infront of the contralateral eye) results in convergence of the eyeunder examination. Future studies may be improved by the useof the integrated fixation targets available in the latest Confoscan model. Despite this limitation, it would bereasonable to presume that images and measurements in thisstudy are representative samples of the local area, particularlyin light of recent data elucidating the 2-dimensionalarchitecture and dynamic nature of the human cornealsubbasal nerve plexus. 14,15 Both subbasal nerve density and corneal sensitivity werehighest in the central cornea, followed by the temporal mid  periphery. Studies using the Cochet-Bonnet esthesiometer or Larson–Millodot esthesiometer have revealed a similar var-iation in corneal sensitivity with corneal eccentricity. 13,16 Ascorneal sensation serves a protective function, it would seemappropriate that these locations (which are at greatest risk fromtrauma) should exhibit the greatest sensitivity and subbasalnerve density. It has been postulated that the low sensitivity of the superior mid periphery isbecause of neuronal adaptation tocontinuous pressure from the upper eyelid. 16 Increased nerve tortuosity may represent a morphologicmarker of nerve regeneration. A study in sciatic nerve crushexperiments has demonstrated increased tortuosity of regen-erating nerves, particularly in older animals. 17 The recent introduction of an objective method of grading subbasal nervetortuosity led to the observation that patients with diabeteswith greater severity of peripheral neuropathy exhibit greater tortuosity of subbasal nerves in the central cornea. 12 Increased tortuosity may therefore represent a morphologic marker of nerve regeneration. In the current study, subbasal nervetortuosity was, for the first time, analyzed in the corneal mid  periphery in addition to the central cornea. Subbasal nervetortuosity was observed to be greatest in the temporal mid  periphery and lowest in the superior mid periphery. These data provide a useful baseline when analyzing the effects of diseaseon the subbasal nerve plexus. The correlation between cornealsensitivity and subbasal nerve tortuosity was weak in theinferior and nasal cornea and insignificant at the other corneallocations. This may be attributed to the importance of factorsother than nerve morphology on corneal sensation.There was no significant change in total subbasal nervedensity with age. These results are in agreement with those of  previous studies investigating the central cornea. 6 Althougha significant difference in global corneal sensitivity betweenage groups was noted, there was no significant correlation between age and corneal sensation.Previous studies investigating the effect of age oncorneal sensation have concentrated on the central cornea.Draeger  18 observed that central corneal sensitivity is not significantly affected by age, whereas Millodot  19 demonstrated a relatively even level of sensitivity up to the fifth decade of life, after which corneal sensitivity rapidly decreases. Incontrast, a recent study using the noncontact pneumaticcorneal esthesiometer noted a gradual reduction in centralcorneal sensation with increasing age. 20 Interestingly, in thecurrent study, the subbasal nerve tortuosity increased significantly with age. The reason for this has yet to beelucidated, although it has been postulated that increased nervetortuosity may represent a morphologic marker of nerveregeneration. 12 Althoughnumerous studies have analyzedthe correlations between subbasal nerve density and tortuosity and corneal FIGURE 3. Relationship between corneal sensitivity thresholdand subbasal nerve density in the temporal cornea revealsa weak but statistically significant correlation (Spearmanrho = 2 0.18, P  = 0.05). q 2009 Lippincott Williams & Wilkins | 739 Cornea    Volume 28, Number 7, August 2009 Corneal Sensitivity and Subbasal Nerve Plexus 
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!