Melanin in the trabecular meshwork is associated with age, POAG but not Latanoprost treatment. A masked morphometric study

Melanin in the trabecular meshwork is associated with age, POAG but not Latanoprost treatment. A masked morphometric study

Experimental Eye Research 82 (2006) 986–993 www.elsevier.com/locate/yexer Melanin in the trabecular meshwork is associated with age, POAG but not Lat...

471KB Sizes 0 Downloads 0 Views

Experimental Eye Research 82 (2006) 986–993 www.elsevier.com/locate/yexer

Melanin in the trabecular meshwork is associated with age, POAG but not Latanoprost treatment. A masked morphometric study Kathryn P.B. Cracknell, Ian Grierson *, Penny Hogg, Ajesola A. Majekodunmi, Peter Watson, Vincent Marmion St Paul’s Unit of Ophthalmology, Department of Medicine, Royal Liverpool University Hospital, Liverpool L69 3GA, UK Received 12 August 2005; in revised form 26 September 2005; accepted in revised form 4 October 2005 Available online 17 November 2005

Abstract We wished to conduct a light and electron microscopic investigation of pigmentation within the trabecular meshwork of normals and primary open angle glaucoma (POAG) patients. In particular we wished to get a precise determination of whether there was a relationship between pigmentation and age. In addition we wanted to know if there was a difference between normals and POAGs and whether trabecular meshwork hyperpigmentation was associated with topical latanoprost medication. A total of 25 sham trabeculectomies conducted on post mortem donor eyes provided the age-matched normals and there were 62 trabeculectomy specimens from POAG patients. These were masked and the meshwork subjected to qualitative and quantitative morphological investigation. Light and electron microscopy confirmed that most of the trabecular meshwork melanin was phagocytosed and within meshwork cells. The granules were measured and found to be of the large iris epithelial type. Light microscopic morphometric analysis showed that the number of meshwork cell profiles that contained melanin increased both in normals and POAGs with age. However there was nearly three times more pigmented meshwork cells in the POAGs than the normals. The POAGs were divided into three groups of (1) minimal or no medication prior to surgery, (2) maximal medical therapy and (3) maximum medical therapy including latanoprost (12 specimens). All groups were significantly greater that the normals but of the three it was the maximal medical therapy group (without latanoprost) that had the highest pigmentation. We concluded that pigmentation of the meshwork is age-related and it is elevated in POAG by mechanisms unknown. The melanin accumulation seems to be partly due to the disease process, partly as a consequence of chronic antiglaucoma medication but interestingly not due to latanoprost even in patients where there is iris darkening (four specimens). q 2006 Elsevier Ltd. All rights reserved. Keywords: Trabeculectomy; Glaucoma; Ageing; Trabecular meshwork; Outflow system; Melanin; Phagocytosis; Latanoprost; Prostaglandins; Morphometry

1. Introduction The normal and glaucomatous trabecular meshwork is variably pigmented between individuals and within the 3608 meshwork circumference of a given individual. Although this pigmentation is poorly understood there is not thought to be local melanogenesis. It has been suggested that melanin granules, released from the iris, are present in the circulating aqueous humour of the human anterior chamber in low but quantifiable numbers (Kuchle et al., 1998). Exercise and pupil * Correponding author. Address: Unit of Ophthalmology, Department of Medicine, University Clinical Departments, University of Liverpool, The Duncan Building, Daulby Street, Liverpool, L69 3GA, UK. Tel.: C44 151 706 4912; fax: C44 151 706 5934. E-mail address: [email protected] (I. Grierson).

0014-4835/$ - see front matter q 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2005.10.009

dilation contribute to the process. On entering the outflow system, some get trapped in the drainage pathways but many of the retained granules are phagocytosed by trabecular meshwork (TM) cells (Richardson et al., 1977; Buller et al., 1990) and even by histiocytic macrophages in pathological situations of massive debris overload (Lee, 1995). Meshwork cells are facultative macrophages with a poor selectivity for the types of material they incorporate into their cytoplasm (Richardson et al., 1977; Buller et al., 1990; Grierson and Hogg, 1995; Lee, 1995) they also have an equally inadequate digestive process (Grierson and Hogg, 1995). There is little evidence to suggest that melanin is digested in situ, hence it remains within the cells and gradually leads to pigmentation in the ageing meshwork as can be seen by gonioscopy. If, as in pigment dispersion syndrome, there is marked pigmentation of the outflow tissues in early life, then at least some patients seem to show some depigmentation with age (Lichter and Shaffer, 1970) presumably because with

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

sufficient time, excess melanin can escape via Schlemm’s canal. There has been only one limited quantitative morphological study by our own group (Grierson, 1987) of pigmentation in the normal but ageing Caucasian trabecular meshwork. In addition to our earlier observations, the morphological research into pigmentation in the trabecular meshwork of primary open angle glaucoma (POAG) patients is limited to the electron microscopy work of Rodriques et al. (1976). So there is some morphological evidence of pigment build up with age but no quantitative evidence about what happens in POAG. Much more morphology work has gone on with respect to the effects of the large pigment overload associated with pigmentary glaucoma and its effect on the architecture of the outflow system. There is a deleterious morphological and cellular effect of melanin on the trabecular meshwork tissue (Richardson et al., 1977; Robinson et al., 1981; Farrar and Shields, 1993) but the JCT region seems to be spared (Alvarado and Murphy, 1992; Murphy et al., 1992). The widespread use of prostaglandins as IOP lowering drugs and their iris-darkening side effect (Watson et al., 1996; Wistrand et al., 1997; Pappas et al., 1998; Camras et al., 2000; Hara, 2001; Sjernschantz, 2001; Grierson et al., 2002; Teus et al., 2002; Cracknell et al., 2003; Chou et al., 2004; ArranzMarquez et al., 2004) highlights the case for possible increased trabecular pigmentation through drug action. Although there has been anecdotal evidence of latanoprost induced darkening of the trabecular meshwork as observed by goinoscopy the only published gonioscopy study to date has shown no evidence of trabecular meshwork darkening (Nakamura et al., 2004). Investigations into the iris-darkening side effect of latanoprost have shown that the increase in iris colouration appears to be due to subtle changes in the size of the iris stromal melanin granules themselves rather than any increase in granule numbers or proliferation of the melanocytes (Grierson et al., 2002; Cracknell et al., 2003) although adverse morphological effects in the iris has been reported by at least one group (Arranz-Marquez et al., 2004). Prostanoid-induced up-regulation of melanogenesis if extreme could be a cause for concern. If it continues unchecked then it is possible that the cytoplasm of the iris melanocytes could become packed with mature melanin granules. A time may come when the melanocytes rupture so releasing their granules into the stroma and the circulating aqueous humour. In such a circumstance, large amounts of melanin in the aqueous humour might be expected to give rise to more

987

marked pigmentation of the outflow tissues and even a possible secondary glaucoma. What little evidence there is on the morphology of the outflow system in the latanoprost-treated eye is that substantive melanin accumulation is not evident but the evidence is limited and subjective (Grierson et al., 2002). In this present study, a quantitative light and electron microscopy investigation was conducted on the melanin within the trabecular meshwork of the outflow system, to determine more precisely whether this varied with age and whether the melanin content of the meshwork was significantly different in POAG. We have also carried out a detailed investigation to determine if there was any morphological evidence for increased trabecular meshwork pigmentation in a latanoprost treated POAG group. 2. Material and methods 2.1. Subjects Human tissue was handled according to the Declaration of Helsinki. The specimens were collected following trabeculectomy procedures (these totalled 87). The criteria used to select trabeculectomy blocks for inclusion in this study were that they contained a substantive portion of trabecular meshwork (Fig. 1) and Schlemm’s canal. The trabeculectomies for morphological investigation were restricted to Caucasians and had either homogenous (blue or brown) or heterogeneous (hazel) eye colour. The samples that made up the normal (non-glaucoma) group were collected by performing a trabeculectomy-like procedure on enucleated eyes that were made available for research and came from our local eye bank. All the individuals in this series were over 40 years of age and we selected a total of 25 specimens age-matched to the POAG patients. The POAG specimens were obtained following conventional trabeculectomy procedures and this provided 62 trabeculectomy blocks (Table 1). Around 45% of the POAG trabeculectomy specimens had either no antiglaucoma medication or had timoptol for the 2–4 week period prior to surgery. These trabeculectomies were performed during the late 1970s and the 1980s in the Cambridge area of the UK where early trabeculectomy at the time was a treatment of choice. The surgery was done mainly by one of us (P.W.) and some of these trabeculectomy specimens have been the subjects of earlier reports on meshwork cellularity in POAG (Grierson and Hogg, 1995). The remaining 55% of the trabeculectomy specimens were from patients who had been on the maximal medical therapy

Fig. 1. Trabeculectomy specimen (A) a macro photograph showing a pigmented trabecular meshwork (arrow). (B) An LM of a semi-thin meridional section through a trabeculectomy specimen showing the trabecular meshwork demarcated by a yellow line.

988

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

Table 1 Trabeculectomy groups used for analysis Sample size

Condition

Tissue block

Glaucoma treatment

25

Normal

None

28 22 12

POAG POAG POAG

Trabeculectomy dissection Trabeculectomy Trabeculectomy Trabeculectomy

a

Medication less than 4 weeks Maximal medical therapy (pilocarpineCb blockers) Maximal medical therapyClatanoprosta

Treatment on period latanoprost 3 months–2 years 6 months and of these six underwent color change and six did not.

prior to glaucoma surgery. Of these, 19% had been exposed to latanoprost for periods that ranged between 3 months and 2 years 6 months. Four had been positively identified as having undergone the latanoprost induced iris darkening that we refer to as LIID (Table 1). The bulk of the trabeculectomy specimens were collected prior to 1993, with the exceptions being those treated with latanoprost, which were collected from 1998 to 2003. We chose to use a retrospective collection of POAG trabeculectomies rather than recruit while we were collecting our latanoprost series to be absolutely sure that latanoprost was not part of the maximal medical therapy prior to surgery. The POAG trabeculectomy collection was chosen with a cut off date that preceded the use of latanoprost in the UK (Watson et al., 1996). 2.2. Tissue preparation All the trabeculectomy specimens were processed for routine light microscopy (LM) and transmission electron microscopy (TEM). This involved a primary fixation in 2.5% glutaraldehyde in 0.1 M Sorensons phosphate buffer, followed by washing in phosphate buffer, then post-fixing for 1 h in 1% buffered osmium tetroxide. They were then buffer-washed once more, dehydrated through a graded series of alcohols, cleared in propylene oxide and finally embedded in an Araldite/Epon mixture. The mixture was cured for 48 h at 80 8C. Semi thin sections for LM of 0$5–1$0 mm thickness and

ultra thin sections for TEM of 90 nm were cut using an ultra microtome (Reichert Ultracut E). 2.3. Tissue analysis At this stage the samples were masked to ensure that the subsequent analysis was un-biased. 2.3.1. LM Semi thin sections of 1 mm in thickness were cut from the sample blocks. The sections were stained with 1% toluidine blue and viewed on a Nikon Optiphot-2 microscope. The whole of each section was examined systematically under a 40! objective lens and 100! oil immersion objective lens. The images were recorded digitally using a JVC TK-1280E colour video camera and stored on Aequitas IDA software archiving system (DDL Ltd, Cambridge, UK). The images were imported into an image analysis software package (Aequitas IA) for evaluation. The system was then calibrated using a micrograph of a 1 mm graticule taken at the appropriate magnifications. The sections were analyzed to determine the incidence of TM cells based on a count of nuclei and the percentage of the TM cells within the sections that contained melanin granules (a cut off was set so that only cells that contained three or more melanin granules were counted as being pigmented). In addition semi thin sections from a subset of 10 POAG

Fig. 2. Measurement of TM melanin granules using Aequitas image analysis. Granules being identified and measured are seen as blue ringed with red.

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

trabeculectomies were subjected to detailed analysis of melanin granule area and projected diameter.

989

3. Results 3.1. LM

2.3.2. TEM For each sample 90 nm (ultra thin) sections were cut and mounted on copper grids and then counter-stained with uranyl acetate and lead citrate. The sections were viewed on a Phillips CM10 transmission electron microscope. A random series of micrographs (magnification!5,000) were taken for each sample. The microscope was calibrated using a diffraction, cross grating replica (Agar S106). Contact prints were made and the images were digitized by scanning the contact prints on a flat bed scanner at 150 dpi. The images were then imported in to an image analysis package (Image pro 4.5). Each micrograph was examined and the image analysis software was used to investigate whether or not the melanin granules were intra or extra cellular and also the characteristics of the TM cell melanin content. To enable the image analysis to identify the individual granules it was necessary for an experienced observer to note melanin granules that were touching or in very close proximity to each other and manually separate them (Fig. 2). Measurements were taken of the melanin granules within the TM cell cytoplasm. For each a determination was made of—the maximum diameter of each granule, the area of the granule and the total number of granules within each micrograph. 2.4. Statistical analysis The data was checked to see if it was normally distributed using the Shapiro-Wilk test. The Mann–Whitney U-test was then used to examine the data sets to determine whether there were any significant differences between them. With the scatter graphs that were plotted the square of product moment correlation was calculated and these values were tested to determine the significance (if any) of these values.

A masked evaluation of the pigmentation within the meshwork samples was carried out. The distribution of granules between trabeculectomy specimens was very variable. Mature melanin granules were identified in every region of the meshwork but there did seem to be less in the JCT than either the corneoscleral or uveal portions of the trabecular part of the meshwork. We found that over 90% of the melanin granules within the TM tissue were incorporated within the cytoplasm of the TM cells (Fig. 3). The figure of 90% held good whether we examined specimens from the normal group, the POAG patients or the latanoprost-treated sub group of patients. The only other melanin-containing cells found in the meshwork where occasional tissue histiocytes and the much larger irisderived clump cells. Both were rare findings and no more prominent in POAGs than normals. The size of the melanin granules was variable and did not show, whether it was based on masked estimate or measurement, any difference in granule size between our various groups including the latanoprost-treated subgroup. A selection was made of 10 POAG trabeculectomies that had the largest area of meshwork to study and these were subjected to extensive image analysis on captured images of the tissue sections. We were able to determine the area of melanin within each image, and a count of the number of granules present was conducted. From these we could determine the average area of the melanin granule profiles; this area measurement was then converted into an equivalent diameter. The mean area of the granules was 0.412G0.129 mm2 and the projected diameter of the melanin granules was found to be 0.79G0.13 mm. The LM data on the percentage of meshwork cell profiles that contained melanin was examined to determine if there was a relationship between TM cell pigmentation and age. In the normal control group (Fig. 4) there was definite positive correlation between age and pigmentation with an R2Z0.36

Fig. 3. A light micrograph of the trabecular meshwork from a latanoprost-treated POAG patient showing melanin granules in TM cells without (A) and with (B) colour enhancement.

990

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

and compared it was the maximum medical therapy group without latanoprost that stood out from the others where on average over 40% of the cells were found to be pigmented (P! 0.03 or better) (Fig. 6). The latanoprost group and the minimal medication group both had an average pigmentation just below 30% there being no significant difference between the two. This is especially note worthy because four of the latanoprost group were known to have undergone the LIID colour change. Surprisingly the four LIIDs had a lower than average pigmentation for POAGs of only 21.3%! 3.2. TEM

Fig. 4. Relationship between age and percentage of melanin containing TM cells profiles observed in the normal population.

(significant at PZ0.001). Within the POAG specimens (Fig. 5) a positive correlation (R2Z0.079; significant at PZ0.05) was found. In this case the relationship was much weaker there being a much more pronounced spread within the data with a range that was as low as 10% but was around 70% for a small number of specimens. Using the regression equations (Figs. 4 and 5) it was calculated that in the normals the % pigmented TM cells increased from 7.2% at 50 years of age to 18.7% at 80 years of age whereas in the POAGs at 50 years of age the % of pigmented TM cells was 23.3% going up to 42.9% at 80 years of age. In addition we would like to determine whether there was a relationship between duration of disease and % pigmented TM cells. Unfortunately we did not have suitable information on duration of glaucoma but we did have a subset of patients (From the non-latanoprost treated groups) where the full duration of antiglaucoma medication was evident. In this unfortunately small subset no relationship was evident despite the treatment range being from 1 month to 19 years. A simple mean percentage comparison between the normals and the various POAG groups was particularly revealing. The mean percentage of TM cell profiles that contained melanin in the normal series was 12.9%G8.1 whereas for the entire POAG grouping the percentage was significantly higher at 33.5%G17.9 (P!0.001). When the POAGs were subdivided

Fig. 5. Relationship between age and percentage of melanin containing TM cells in POAG (excludes treatment with prostaglandins).

Ultra thin sections were examined by TEM to determine whether or not there were any differences at the fine structural level between the melanin granules found in the meshwork of normals and POAGs and also POAGs with and without latanoprost. It turned out that the appearance, size, location and pattern of distribution of melanin granules were similar in all groups. Qualitative masked systematic examination of the specimens revealed, as had been seen by LM, a patchy distribution of melanin granules with the vast majority, about 90% in all cases, incorporated within meshwork cell cytoplasm (Fig. 7(A)). Of the meshwork cells that had melanin granules in their cytoplasm, some cells had a particular abundance however the vast majority had only a very sparse pigmentation. It appeared that some TM cells were more avidly phagocytic or alternatively had more opportunity for phagocytosis than others. In these more heavily pigmented cells, the melanin was either distributed as isolated granules or the melanin formed aggregates of granules (Fig. 7(B)). A surrounding single membrane was more obvious around the aggregates than the single isolated granules indicating the aggregated granules at least were within phagosomes. The melanin granules frequently were found in association with lipid inclusions. The granules irrespective of whether they were isolated or aggregated often had a ragged outline. They showed evidence of degeneration by variation in electron density and a minority

Fig. 6. The difference in incidence of pigmented TM cells between normals and POAGs that have various medical end points prior to trabeculectomy. All three POAG graphs have a significantly greater % incidence of pigmented TM cells than normals (P%0.05 or better) but the latanoprost-treated group was not significantly different than the other two POAG groups.

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

991

Fig. 7. TEM micrographs showing the distribution of melanin granules in the meshwork of a POAG patient (treated with latanoprost). The size of the granules is variable but some are quite large. Low (A) and high power (B).

of granules were vacuolated. The JCT region had a relatively low incidence of cell profiles containing melanin granules. Careful examination of TM cell cytoplasm in all regions of the trabecular meshwork from our specimen series provided no evidence of there being premelanosomes in TM cells. This held true whether the TM cells had or had not cytoplasmic melanin granules. Random measurements of melanin granule size from all the groups of specimens produced a range of particle diameters between 0.2 and 1.2 mm with less than 5% of granules falling into the lower range. The vast majority were large granules and there was no significant differences between any of our specimen groups. When a sample of in excess of 600 granules was measured from a randomly selected series of three specimens, the mean granule diameter was found to be 0.61G0.14 mm. 4. Discussion The present study shows that there is a morphologically recordable pigmentation of the normal and POAG meshwork that clearly increases with age. The pigmentation of the POAG meshwork is significantly greater than normal and this holds true whether the POAG patients had minimal or maximal medical therapy prior to surgery. Pigmentation of the trabecular meshwork has been mentioned in early morphological investigations (Hogan et al., 1971; Rohen and LutjenDrecoll, 1971) and the relationship between increased pigmentation and age at least in the normal eye has been noted (Bron et al., 1997). That pigmentation in the POAG meshwork is of some interest is evident from the work of Rodriques et al. (1976). The present study brings a quantitative basis to a body of knowledge that was essentially subjective and confirmed the clear difference in pigmentation levels between normals and POAGs that was indicated by Grierson (1987). The trabecular meshwork is not known to have a population of melanin synthesising cells; the nearest melanocytes are located in the iris root and pectinate ligaments (Hogan et al., 1971). The concept that debris passing along with the aqueous flow gets phagocytosed by the TM cells is long established (Rohen and Van der Zypen, 1968; Grierson and Hogg, 1995).

The process has been considered as a self cleaning mechanism to keep the outflow pathways patent by many researchers including Polansky et al. (1984); Rohen and Lutjen-Drecoll (1991). As a trickle of melanin granules is present in the normal and glaucomatous aqueous humour (Kuchle et al., 1998), it is reasonable to assume that at least some are cleaned, by TM cell phagocytosis, from the aqueous as the granules pass through the drainage pathways. Incorporation of melanin into TM cells has been observed both in experimental studies in situ (Sherwood and Richardson, 1988) and in tissue culture (Polansky et al., 1984; Grierson and Hogg, 1995). In support of the likelihood that the meshwork melanin is iris derived is our own size measurement data. We know that the mature melanin granules in melanocytes have quite a small diameter (mean size 0.3 mm2) whereas our measurements from the meshwork indicated that the granules are for the most part far larger. Our TEM value gave a mean of 0.61G0.14 mm that went up to 0.79G0.13 mm for our LM calculations. The latter value is virtually identical to the known average diameter for mature iris pigment epithelial melanin granules as measured by us in an earlier study that gave a mean size of 0.8 mm2 (Cracknell et al., 2003). That our meshwork values are slightly smaller than the optimal iris epithelial values in part might be explained by degenerative alterations following phagocytosis that were apparent by TEM. The pigmentation of TM cells in both the normals and POAGs is substantial for a Caucasian population. Probably no more than otherwise had been suspected from earlier work (Rodriques et al., 1976; Grierson, 1987) but far less than occurs in a West African POAG population (unpublished observations by one of us A.A.M). The question arises whether such a pigment load is of functional significance to TM cell behaviour. That would seem to be unlikely, on the basis of cellularity and morphology at least (Johnson, 1989). TM cells have been shown both in vivo (Johnson, 1989) and vitro (Grierson and Hogg, 1995) to be amazingly resilient to the effects of melanin. Further in pathological conditions such as pigment dispersion syndrome where the iris epithelium releases massive amounts of mature melanin granules into the outflow system, trabecular meshwork dysfunction and pigmentary glaucoma is far from being inevitable. Pigment dispersion is a risk factor for

992

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

pigmentary glaucoma but only a minority convert (Farrar and Shields, 1993). In pigmentary glaucoma the influx of melanin into the meshwork does not seem to cause a simple blockage of the JCT region (Murphy et al., 1992). Most of the meshwork cells have a heavy pigment burden (Grierson, 1987), which some authors think is lethal to the TM cells when the burden becomes too high (Richardson et al., 1977). We still do not know whether the relatively low melanin burdens seen in the normal ageing and POAG meshwork have a TM cell behaviour modifying effect. Phagocytosis studies in vitro, initiated and stimulated by the tissue culture work of Jon Polansky (Polansky et al., 1984), perhaps indicate that any behavioural changes might be subtle (Shirato et al., 1989; Grierson and Hogg, 1995; Matsumoto and Johnson, 1997 among others). It seems unlikely that the significantly increased melanin granule content of the POAG meshwork contributes substantially to cell and tissue decline but what might be the reason for the increase in the first place? There is no clear-cut and self evident explanation. First of all we are not dealing with simply a medication effect. Certainly there is a significant increase in % pigmented TM cells from the group where surgery is the first treatment of choice to the maximal medical therapy group (P!0.001). However there is also a clear increase in % pigmentation from the normals to the early surgery group (P!0.003) that has to be related to the disease process. Of course care needs to be taken when comparing small specimens like trabeculectomies especially as it is known that pigment distribution round the meshwork in health and disease is segmental (Johnson, 1989; Farrar and Shields, 1993). On the other hand the superior quadrant, where all the trabeculectomies were taken, is considered the least heavily pigmented region and so the difference between normal and the POAGs might be even greater than has come out in this investigation. The more pronounced meshwork pigmentation associated with POAG might be explained in a number of ways that include there being an increased TM cell phagocytosis rate in the glaucomatous outflow system, a greater release of iris melanin or there is a more sluggish passage of melanin particles through the system creating more opportunity for phagocytosis to take place. We do know from the work of Matsumoto and Johnson (1997) that phagocytosis upregulation is less likely because the authors have shown that the phagocytosis of latex microspheres by the normal and the POAG meshwork is virtually identical. However, there is at least one study that has identified transillumination defects in patients with POAG (Christen et al., 2003) so more pigment availability coupled with a sluggish passage through the meshwork is at least a credible explanation for the higher melanin content of the POAG meshwork. We were surprised that the maximal medical therapy group stood out from all the rest of the POAGs for their level of pigmentation. Particularly as the maximal medical therapy group had a higher % pigmented TM cell population than even the maximal medical therapy group with latanoprost (subgroup analysis P!0.03). What surprised us was that the maximal medical therapy trabeculectomies were collected at a time

when pilocarpine was a common second line antiglaucoma treatment. Virtually all of the patients in this group had pilocarpine for a substantive period and most were on pilocarpine prior to surgery. Pilocarpine can be used in pigment dispersion and pigmentary glaucoma patients to narrow the pupil and reduce the amount of anterior chamber pigment (Haynes et al., 1992). Perhaps there is a difference between selective use of pilocarpine and chronic pilocarpine administration. Again that we are not just dealing with a simple treatment effect is evident from our modest attempt at relating % pigmented TM cells to duration of treatment. The absence of any relationship is baffling. Although it needs to be pointed out that the subset with clear data was only 20 and probably to small for confident analysis. Topical latanoprost treatment is associated with the LIID side effect (Watson et al., 1996; Wistrand et al., 1997; Pappas et al., 1998; Lindquist et al., 1999; Camras et al., 2000; Hara, 2001; Sjernschantz, 2001; Grierson et al., 2002; Teus et al., 2002), darkening of periocular skin (Herndon et al., 2003; Wand et al., 2001) and eyelash changes (Johnstone, 1997; Chiba et al., 2004) so it is not surprising that questions have been raised concerning a possible effect on pigmentation in the trabecular meshwork (Grierson et al., 2002). There have been some anecdotal reports of increased meshwork pigmentation, but a recent investigation of chamber angle pigmentation following long-term exposure to latanoprost in Japanese patients (Nakamura et al., 2004) showed no gonioscopic evidence of increased pigmentation in the meshwork region. The mechanism that causes latanoprost to bring about an increase in iris pigmentation is still not fully understood. Indeed there is still controversy over the exact detail of the morphological changes associated with LIID. We have been unable to find much in the way of iris pathology (Grierson et al., 2002); only an enlargement of the existing melanin granules within the melanocyte cells (Cracknell et al., 2003). On the other hand Arranz-Marquez et al. (2004) did identify degenerative changes including release of melanin by melanocytes into the iris stroma. In this latter situation there exists at least the theoretical potential for an extra burden of free iris melanin granules to contribute to meshwork pigmentation. The present study provides no evidence to support the scenario of there being extra pigmentation of the POAG meshwork as a result of topical latanoprost treatment. Although the % of TM cell profiles containing pigment in the latanoprost treated group of specimens is significantly higher than the controls, it is significantly lower than the other maximal medical therapy group. Why there is such a difference between these two POAG group remains unclear. None the less the relatively low pigmentation in the meshwork of the whole latanoprost group of specimens including the four LIIDs, coupled with the fact that the vast majority of the meshwork melanin is iris epithelial rather than iris stromal derived (as is the case for all the specimens in our series), makes it improbable that enhanced meshwork pigmentation is a common, or even plausible, latanoprost side effect.

K.P.B. Cracknell et al. / Experimental Eye Research 82 (2006) 986–993

Acknowledgements We would like to thank Miss Gillian Hogarth and Ms Juliet Trowell for their help with this work. Support: This investigation was funded by grants provided by the van Neste Foundation, the Sandra Charitable Trust and Pfizer Global Pharmaceuticals. References Alvarado, J.A., Murphy, C.G., 1992. Outflow obstruction in pigmentary and primary open angle glaucoma. Arch. Ophthalmol. 110, 1185–1779. Arranz-Marquez, E., Teus, M.A., Saornil, M.A., Mendez, M.C., Gil, R., 2004. Analysis of irises with a latanoprost-induced change in iris color. Am. J. Ophthalmol. 138, 625–630. Bron, A.J., Tripathi, R.C., Tripathi, B.J., 1997. Woolfe’s Anatomy of the Eye and Orbit, eighth Ed. Chapman &Hall, London. Buller, C., Johnson, D.H., Tschumper, R.C., 1990. Human trabecular meshwork phagocytosis. Observations in an organ culture system. Invest. Ophthalmol. Vis. Sci. 31, 2156–2163. Camras, C.B., Neely, D.G., Weiss, E.L., 2000. Latanoprost-induced iris color darkening: a case report with long-term follow-up. J. Glaucoma 9, 95–98. Chiba, T., Kashiwagi, K., Ishijima, K., et al., 2004. A prospective study of iridial pigmentation and eyelash changes due to ophthalmic treatment with latanoprost. Jpn. J. Ophthalmol. 48, 141–147. Chou, S.Y., Chou, C.K., Kuang, T.M., Hsu, W.M., 2004. Incidence and severity of iris pigmentation on latanoprost-treated glaucoma eyes. Eye 19, 784– 787. Christen, R., Pache, M., Tenchner, B., Meyer, P., Prunte, C., Flammer, J., 2003. Iris transillumination defects in patients with primary open angle glaucoma. Eur. J. Ophthalmol. 13, 365–369. Cracknell, K.P., Grierson, I., Hogg, P., Appleton, P., Pfeiffer, N., 2003. Latanoprost-induced iris darkening: a morphometric study of human peripheral iridectomies. Exp. Eye Res. 77, 721–730. Farrar, S.M., Shields, M.B., 1993. Current concepts in pigmentary glaucoma. Surv. Ophthalmol. 37, 233–252. Grierson, I., 1987. What is open angle glaucoma? Eye 1, 15–28. Grierson, I., Hogg, P., 1995. The proliferative and migratory activities of trabecular meshwork cells. Prog. Retin. Eye Res. 15, 33–67. Grierson, I., Pfeiffer, N., Cracknell, K.P.B., Appleton, P., 2002. Histology and fine structure of the iris and outflow system following latanoprost therapy. Surv. Ophthalmol. 47, S176–S184. Hara, T., 2001. Increased iris pigmentation after use of latanoprost in Japanese brown eyes. Nippon Ganka Gakkai Zasshi 105, 314–321. Haynes, W.L., Johnson, A.T., Alward, W.L.M., 1992. Effects of jogging exercise on patients with the pigmentary dispersion syndrome and pigmentary glaucoma. Ophthalmology 99, 1096–1103. Herndon, L.W., Robert, D.W., Wand, M., Asrani, S., 2003. Increased periocular pigmentation with ocular hypotensive lipid use in African Americans. Am. J. Ophthalmol. 135, 713–715. Hogan, M.J., Alvarado, J.A., Weddell, J.E., 1971. Histology of the Human Eye. Saunders, Philadelphia. Johnson, D.H., 1989. Does pigmentation affect the trabecular meshwork? Arch. Ophthalmol. 107, 250–254. Johnstone, M.A., 1997. Hypertrichosis and increased pigmentation of eyelashes and adjacent hair in the region of the ipsilateral eyelids of patients treated with unilateral topical latanoprost. Am. J. Ophthalmol. 124, 544–547. Kuchle, M., Mardin, C.Y., Nguyen, N.X., Wartus, P., Naumann, G.O., 1998. Quantification of aqueous melanin granules in primary pigment dispersion syndrome. Am. J. Ophthalmol. 126, 442–445.

993

Lee, W.R., 1995. Doyne lecture. The pathology of the outflow system in primary and secondary glaucoma. Eye 9, 1–23. Lichter, P.R., Shaffer, R.N., 1970. Diagnostic and prognostic signs in pigmentary glaucoma. Trans. Am. Acad. Ophthalmol. Otolaryngol. 74, 984–998. Lindquist, N.G., Larsson, B.S., Stjernschantz, J., 1999. Increased pigmentation of iridial melanocytes in primates induced by a prostaglandin analogue. Exp. Eye Res. 69, 431–436. Matsumoto, Y., Johnson, D.H., 1997. Trabecular meshwork phagocytosis in glaucomatous eyes. Ophthalmologica 211, 147–152. Murphy, C.G., Johnson, M., Alvarado, J.A., 1992. Juxtacanalicular tissue in pigmentary and primary open angle glaucoma. Arch. Ophthalmol. 110, 1779–1785. Nakamura, Y., Nakamura, Y., Morine-Shinjo, S., Sakai, H., Sawaguchi, S., 2004. Assessment of chamber angle pigmentation during longterm latanoprost treatment for open-angle glaucoma. Acta Ophthalmol. Scand. 82, 158–160. Pappas, R.M., Pusin, S., Higginbotham, E.J., 1998. Evidence of early change in iris color with latanoprost use. Arch. Ophthalmol. 116, 1115–1116. Polansky, J.R., Wood, I.S., Maglio, M.T., Alvarado, J.A., 1984. Trabecular meshwork cell culture in glaucoma research: evaluation of biological activity and structural properties of human trabecular cells in vitro. Ophthalmology 91, 580–594. Richardson, T.M., Hutchinson, B.T., Grant, W.M., 1977. The outflow tract in pigmentary glaucoma: a light and electron microscopic study. Arch. Ophthalmol. 95, 1015–1025. Robinson, C.H., Nopanitaya, W., McPherson, S.D., 1981. Pigmentary glaucoma: an ultrastructural study. Ann. Ophthalmol. 13, 49–51. Rodriques, M.M., Spaeth, G.L., Weinreb, S., Sivalingam, E., 1976. Spectrum of trabecular pigmentation in open-angle glaucoma: a clinicopathologic study. Trans. Am. Acad. Ophthalmol. Otolaryngol. 81, 258–276. Rohen, J.W., Lutjen-Drecoll, E., 1971. Age changes in the trabecular meshwork in human and monkey eyes. In: Aging and Development. Schattaur Verlag Stuttgart, pp. 1–36. Rohen, J.W., Lutjen-Drecoll, E., 1991. Ageing processes in the anterior segment of the eye. In: Lutjen-Drecoll, E., Rohen, J.W. (Eds.), Basic Aspects of Glaucoma Research III. Schattauer, Stuttgart, pp. 191–224. Rohen, J.W., Van der Zypen, E., 1968. The phagocytic activity of the trabecular meshwork endothelium. Graefe’s Arch. Clin. Exp. Ophthalmol. 175, 143– 160. Sherwood, M.E., Richardson, T.M., 1988. Phagocytosis by trabecular meshwork cells: sequence of events in cats and monkeys. Exp. Eye Res. 46, 881–895. Shirato, S., Murphy, C., Bloom, E., Franse-Carman, L., Maglio, M., Polansky, J.R., Alvarado, J., 1989. Kinetics of phagocytosis in trabecular meshwork cells. Invest. Ophthalmol. Vis. Sci. 30, 2499–2511. Sjernschantz, J.W., 2001. From PGF2alpha-isopropyl ester to latanoprost: a review of the development of xalatan. The proctor lecture. Invest. Ophthalmol. Vis. Sci. 42, 1134–1145. Teus, M.A., Arranz-Marquez, E., Lucea-Suescun, P., 2002. Incidence of iris colour change in latanoprost treated eyes. Br. J. Ophthalmol. 86, 1085– 1088. Wand, M., Ritch, R., Isbey, E.K., Zimmerman Jr.., T.J., 2001. Latanoprost and periocular skin color changes. Arch. Ophthalmol. 119, 614–615. Watson, P., Sjernschantz, J.W., the latanoprost study group, 1996. A six-month, randomized, double-masked study comparing latanoprost with timolol in open-angle glaucoma and ocular hypertension. Ophthalmology 103, 126– 137. Wistrand, P.J., Stjernschantz, J., Olsson, K., 1997. The incidence and timecourse of latanoprost-induced iridial pigmentation as a function of eye color. Surv. Ophthalmol. 41, S129–S138.