2019 Volume 67 Issue 3 Pages 277-283
The purpose of this study is to evaluate the usefulness of the rheological properties and adhesive force of ophthalmic viscosurgical devices (OVDs) as parameters for understanding and identifying the surgical behavior of cohesive and dispersive OVDs. The apparent viscosity, and the storage and loss moduli (dynamic rheological parameters) of 50% chondroitin sulfate (CS), 3% sodium hyaluronate (HA), Shellgan (the combination of 3% HA and 4% CS), Opegan (1% HA with a low molecular mass) and Opegan-Hi (1% HA with a high molecular mass) were obtained with a rheometer. The adhesive force of each sample was measured by using a texture analyzer. Opegan-Hi showed a solid-like behavior, while 50% CS showed a fluid-like behavior from their apparent viscosity and dynamic rheological parameters. Shellgan, 3% HA, and Opegan exhibited similar rheological properties and intermediate characteristics between Opegan-Hi and 50% CS, although their respective values were slightly different. Among these OVD samples, the adhesive force was higher in the order of 50% CS > Shellgan, 3% HA > Opegan > Opegan-Hi. The adhesive force of dispersive OVDs tended to be higher than that of cohesive OVDs, which correlated well with the removal times of OVDs from the eye that have previously been reported. In conclusion, we demonstrated that cohesive OVDs and dispersive OVDs have particular rheological and adhesive properties that can be applied to identify both types. These parameters obtained in this study provide useful information for a greater understanding and prediction of the behavior of OVDs in the eye during surgery.
Ophthalmic viscosurgical devices (OVDs) are viscoelastic solutions used to create and maintain space in the eye during various ocular surgeries. Two different types of OVDs (dispersive and cohesive) have been developed to support a particular surgical technique.1,2) The OVDs containing sodium hyaluronate (HA) as a viscoelastic substance show various rheological properties depending on the concentration and molecular mass of HA. To address the requirements for each patient, therefore, it is very important to sufficiently understand the properties of OVDs and exploit them accordingly. Cohesive OVDs protect intraocular tissues from invasions of surgical instruments and intraocular lenses during surgery by maintaining a deep anterior chamber.3,4) Additionally, they are rapidly removed from the affected area. Dispersive OVDs tend to lie adjacent to the corneal endothelium and be retained in the eye during phacoemulsification and aspiration (PEA), which results in difficulty in removing them from the affected area.5–8) To take advantage of their positive attributes, the dispersive-cohesive viscoelastic soft shell technique has evolved for cataract surgery, which involves using dispersive and cohesive OVDs together in sequence.2)
In the classification based on the cohesion-dispersion index (CDI), the value for cohesive OVDs is 30 or more and that for dispersive OVDs is less than 30.1) Thus, OVDs have peculiar rheological properties for each product, and each property tends to be considered separately for the OVD classification. For example, Opegan has been treated as a cohesive OVD in several reports though it has been classified as a dispersive OVD based on the CDI value.9,10) Different standards seem to co-exist for the classification of OVDs. Therefore, the current method for OVD classification is insufficient for an adequate description of their usability and operability. Many studies have demonstrated that dispersive OVDs show high tissue adhesiveness in the eye during PEA compared with cohesive OVDs. However, a suitable discussion of the mechanism for this adhesiveness has not yet been achieved.
To understand the physical characteristics of cohesive OVDs and dispersive OVDs, we compare the apparent viscosity, dynamic viscoelasticity, and adhesive force with mainly the following OVD samples: 50% chondroitin sulfate (CS), 3% HA, Opegan, Opegan-Hi, and Shellgan. This is the first study to measure the adhesive force of OVDs by using a texture analyzer, although it has already been applied in the fields of pharmaceuticals and food.11,12) We succeed in determining the difference of the adhesive force, as well as rheological properties, between cohesive OVDs and dispersive OVDs. To understand the behavior of OVDs in the eye during surgery, not only the CDI classification but also the rheological properties and adhesive force will be useful indexes.
The investigated OVD samples are shown in Table 1. Opegan®, Opegan-Hi®, and Shellgan® were from Santen Pharmaceutical Co., Ltd. (Japan). Three percent HA or 50% CS were prepared by dissolving powdered HA or CS in phosphate-buffered saline, respectively. Viscoat®, Discovisc®, and Provisc® were purchased from Alcon Inc. (Hünenberg, Switzerland). Healon® was purchased from Abbott Medical Optics Inc. (CA, U.S.A.). The average molecular weight (MW) of HA was estimated by the intrinsic viscosity according to the methods reported previously.13,14) The MW of HA or CS of the OVD products containing HA and CS, was estimated by size exclusion chromatography coupled with multi-angle light scattering, according to the method of Wyatt.15)
| OVD | Manufacturer or Distributor | Polysaccharide (PS) content | CDI %asp/mmHg | ||||
|---|---|---|---|---|---|---|---|
| Hyaluronic acid | Chondroitin sulfate | Total PS | |||||
| % | MW (kDa) | % | MW (kDa) | % | |||
| Opegan | Santen (Japan) | 1 | 1100 | — | — | 1 | <30 |
| Opegan-Hi | Santen (Japan) | 1 | 2500 | — | — | 1 | >30 |
| Shellgan | Santen (Japan) | 3 | 700 | 4 | 50 | 7 | — |
| 3% HA | Homemade | 3 | 700 | — | — | 3 | — |
| 50% CS | Homemade | — | — | 50 | 50 | 50 | — |
| Viscoat | Alcon (Switzerland) | 3 | 700 | 4 | 20 | 7 | <30 |
| Discovisc | Alcon (Switzerland) | 1.65 | 1600 | 4 | 20 | 5.65 | <30 |
| Provisc | Alcon (Switzerland) | 1 | 2200 | — | — | 1 | >30 |
| Healon | AMO (U.S.A.) | 1 | 2300 | — | — | 1 | >30 |
Rheological measurements (apparent viscosity and dynamic rheological parameters) of each OVD sample were performed by using a rheometer (MCR302, Anton Paar Japan K.K., Tokyo, Japan) equipped with a parallel-plate (PP25, Anton Paar Japan K.K.) and a gap of 0.5 mm at 25°C. The apparent viscosity (Pa·s) was determined with a shear rate in the range of 0.01–100 (1/s).
To evaluate the dynamic rheological parameters of the OVD samples, the profiles of the storage modulus G′ (Pa) and the loss modulus G″ (Pa) were obtained from stress sweep tests and frequency sweep tests. To determine the strain condition for the frequency sweep tests, the upper limit of the linear viscoelastic regimen was evaluated by stress sweep procedures at a fixed frequency (1 Hz). The frequency responses of G′ and G″ were determined at a constant strain, in the frequency range of 0.1–100 Hz.
Measurements of Adhesive ForceAdhesive force (N/mm2) of each OVD sample was measured by using a texture analyzer (TA.XT plus, Stable Micro Systems, U.K.) equipped with a cylindrical probe (10 mm in diameter) made of polyoxymethylene. The probe was moved downward at 0.5 mm/sec and attached to a 0.1-mL OVD sample, and then withdrawn at 0.5 mm/s. The peak of the minimum detachment force required to separate the probe from the OVD samples was defined as the adhesive force (Fig. 1). To discuss the behavior of OVDs in the eye during surgery, we evaluated the adhesive forces of Viscoat, Discovisc, Provisc, and Healon because the residence times of these products in the eye have already been determined.5)
The graph shows the profile of Shellgan as an example. (Color figure can be accessed in the online version.)
In the measurements of the adhesive force, he general linear models with Tukey–Kramer multiple comparison were performed for statistical comparisons between two groups of OVD samples. p-values of less than 0.05 were considered to be statistically significant.
The apparent viscosity of the OVD samples showed different values owing to the difference in molecular mass, concentration, and composition of the polysaccharide contained (Fig. 2). The CS samples showed a constant viscosity over the range of investigated shear rates: the values of 50% CS and 4% CS were 50 Pa·s and <0.1 Pa·s, respectively. For a shear rate value below 1 s−1, the viscosity of Discovisc, which has dispersive and cohesive properties, exhibited the highest value among the OVD samples. Cohesive OVDs; Opegan-Hi, Healon and Provisc, also exhibited higher viscosity compared with dispersive OVDs; Shellgan, 50% CS, 3% HA and Viscoat. Except for CS samples, the viscosity of OVD samples reduced depending on the share rate, which was in the range of 1–100 s−1. The viscosity of Discovisc was the most dependent on the shear rate among the OVD samples tested. Interestingly, the viscosities of Shellgan and Viscoat, which are combination OVDs consisting of 3% HA and 4% CS, were much higher than the total of each viscosity of these single agent solutions in the range of the measured shear rates.
(a) ■ Opegan-Hi; ● Shellgan; ◆ 3% HA; ▲ Opegan; ＊ 50% CS; × 4% CS, (b) ■ Healon; ● Provisc; ◆ Discovisc; ▲ Viscoat, (c) ● Shellgan; ▲ Viscoat; ◆ 3% HA. (Color figure can be accessed in the online version.)
To evaluate the dynamic rheological parameters of OVD samples, first, we investigated the profiles of the storage modulus (G′) and loss modulus (G″) of each OVD sample using stress sweep tests (Fig. 3). The G′ and G″ values of all OVD samples were independent of the applied strain in the range of 0.1–5%, which indicated that the linear region was less than 5%. The G′ and G″ values started to decrease when the strain exceeded 5% (nonlinear region), except for 50% CS. For 50% CS, the transition from linear to nonlinear was much higher than 100%.
(a) Storage modulus G′ of each OVD. (b) Loss modulus G″ of each OVD. ■ Opegan-Hi; ● Shellgan; ◆ 3% HA; ▲ Opegan; ＊ 50% CS, (c) Storage modulus G′ of each OVD. (d) Loss modulus G″ of each OVD. ■ Healon; ● Provisc; ◆ Discovisc; ▲ Viscoat. (Color figure can be accessed in the online version.)
Next, we performed the frequency sweep tests at 1% strain. This strain was chosen as it was shown to be in the linear region from the above tests. The results are shown in Fig. 4. For Opegan-Hi, Healon and Provisc, the storage modulus G′ depended on the frequency, and the profile of the G′ values were higher than that of the G″ values in the measured frequency range (Figs. 4a–c). This indicated that these cohesive OVDs stably exhibited a solid-like behavior and almost no fluidity. In contrast, 50% CS exhibited a fluid-like behavior because the profile of the frequency sweeps showed G″ > G′ over the entire measurement range (Fig. 4i). For Discovisc, Shellgan, Viscoat, 3% HA, and Opegan, the frequency sweep of these samples showed a more fluid-like behavior at low frequencies (G″ > G′), and more solid-like at high frequencies (G′ > G″) (Figs. 4d–h). The frequency of the transition points for each OVD sample was as follows: Discovisc 0.3 Hz; Shellgan 3.2 Hz; Viscoat 3.2 Hz; 3% HA 5.0 Hz; and Opegan 7.9 Hz.
(a) Opegan-Hi, (b) Healon, (c) Provisc, (d) Discovisc, (e) Shellgan, (f) Viscoat, (g) 3% HA, (h) Opegan and (i) 50% CS. Frequency sweep profile of the storage modulus G′ (closed circles) and the loss modulus G″ (open circles). (Color figure can be accessed in the online version.)
For OVD samples of more than 0.1 mL, the adhesive force could not be measured accurately, because the sample would attach to the side of the probe. For OVD samples of less than 0.1 mL, there was a variation in the measured data (data not shown). Thus, we have determined that 0.1-mL OVD sample was the proper quantity to evaluate adhesive force in this study. Figure 5 shows the adhesive force of each OVD sample detected as the detachment force required to withdraw the probe from the samples. 50% CS exhibited the strongest adhesive force among the OVD samples investigated. Shellgan and 3% HA showed a higher adhesive force compared with Opegan-Hi, which is classified as a cohesive OVD based on the CDI. The adhesive force of Opegan was also slightly higher than that of Opegan-Hi (p < 0.05, Tukey–Kramer multiple comparison).
Except for the comparison of Shellgan and 3% HA, the significance was determined as less than 0.05 in all combinations of OVD samples using Tukey–Kramer multiple comparison. (Monochrome imge.)
The HA molecules stack upon each other by a hydrophobic interaction or a hydrogen bond between the carboxyl and acetamide groups on the HA chains. The hydrophobic patches partly assist in the self-aggregation of HA molecules, and they form a meshwork structure like a honeycomb.16,17) The interactions of HA molecules, which form the meshwork, are fairly weak, so they can result in breakage of the mesh depending on certain conditions. For example, long-term flow duration can partially disentangle the meshwork, and allows the HA solution to exhibit a dominant viscous property.16,17) In Figs. 2 and 3, the rheological properties of the OVD samples containing HA changed depending on the applied stress. These results suggested that the cause might be the dissociation of the meshwork structure in the HA molecules. HA molecules also interact with CS molecules through hydrogen bonds when co-existing in an aqueous solution.18,19) The apparent viscosity of Shellgan and Viscoat, a combination OVDs consisting of 3% HA and 4% CS, dramatically increased, compared with the total of each viscosity of these single agent solutions. This was likely to result from the meshwork structure of the HA molecules being strengthened by the interaction with the CS molecules. The viscosity of Shellgan was approximately 10–20% higher than that of Viscoat at the low shear rate in the range of 0.01–1 s−1. As shown in Table 1, the molecular mass of CS contained in both OVDs was slightly different. We supposed that this small difference may affect the viscosity of these OVDs, and it is likely to influence on the usability or operability of both OVDs.
Classification of OVDs has been based on the CDI and their zero-shear viscosity range.1) The CDI (% asp/mmHg) is defined as the volume of OVD aspirated by pipette tip: a CDI value of 30 means 30% of OVD is aspirated when the vacuum is 100 mmHg.20) OVDs for which the CDI exceeds 30 are defined as cohesive type, whereas those for which the CDI is less than 30 are defined as dispersive type. Healon, Provisc, and Opegan-Hi are classified as cohesive OVDs, and Discovisc, Viscoat, and Opegan are classified as dispersive OVDs. There are no data for the classification of Shellgan, 50% CS, and 3% HA. The subtype of OVDs based on zero-share viscosity range is classified as follows: viscous; 100–1000 Pa·s, medium viscosity; 10–100 Pa·s, very low viscosity; 1–10 Pa·s.1) Opegan-Hi, Healon, Provisc, and Discovisc are viscous types, Viscoat is of medium viscous type, and Opegan is of very low viscous type. As shown in Fig. 2, the maximum apparent viscosities of these OVDs at the low shear rate were in the same ranges as the subtypes mentioned above. Therefore, we classified Shellgan, 50% CS, and 3% HA as medium viscous type in accordance with the subtype classification. The zero-shear viscosity of OVDs is related to the ability to maintain a space anterior chamber, because “zero shear” is a term used to describe a condition in which there is no fluid movement within the eye.21) It is also described that cohesive OVDs tend to intertwine during conditions of zero shear; this property helps these OVDs to maintain intraocular space very effectively. Except for Discovisc, the apparent viscosities of cohesive OVDs exhibited higher values than those of dispersive OVDs: this suggests the possibility that the space-maintenance ability of OVDs can be predicted by measuring the maximum apparent viscosity at the low shear rate. Meanwhile, PEA provides fluid-flow in the eye, like a parallel-plate with higher shear rate applying stress to OVDs. The decline of the viscosity of dispersive OVDs was less compared with cohesive OVDs when the shear rate exceeded around 1 s−1. The results suggested that dispersive OVDs resist pressure from fluid-flow in the eye which occurred by PEA. The apparent viscosity of OVDs under different shear rates provides their rheological properties in detail, which cannot be represented by zero-shear viscosity alone.
The dynamic rheological parameters determined from Figs. 3 and 4 indicated that behavior of cohesive OVDs such as Opegan-Hi was solid-like, whereas that of 50% CS was fluid-like. In the frequency sweep tests, Shellgan, Viscoat, 3% HA, and Opegan each had a particular transition point and these four OVDs exhibited similar rheological properties, which were just intermediate between Opegan-Hi and 50% CS (Figs. 3, 4). Discovisc also had a particular transition point; however, its rheological property was also similar to that of Opegan-Hi (Fig. 4d). Discovisc is classified as dispersive type based on the CDI classification, whereas Tognetto et al. showed that mixed properties of dispersive and cohesive types were exhibited in a single OVD.22) Our results might reflect the dual nature of Dsicovisc. Viscoat and Opegan are classified as dispersive OVDs in the CDI classification.1) Arshinoff and Gulati have defined a cohesive OVD containing HA as follows: they are high molecular weight (greater than 1000 kDa) and possess high zero-shear viscosity.23) Therefore, two different standards seem to exist for the classification of OVDs. According to the latter classification, we may be able to say that Opegan belongs to the cohesive type. This means that the classification of Opegan is very complicated. Healon EndCoat, a 3% HA OVD product (Abbott Medical Optics Inc., CA, U.S.A.), is classified as a dispersive OVD in the CDI classification, and the molecular mass of HA in the product is 800 kDa.24) Subsequently, our home-made 3% HA can also be classified as a dispersive OVD. The physical behavior of Shellgan is similar to that of Viscoat and 3% HA. Consequently, we propose that Shellgan might belong to the dispersive OVD class though we have not evaluated its CDI value. Therefore, our findings for Viscoat, Shellgan, and 3% HA suggested that their rheological properties might be typical properties of dispersive OVDs.
To understand the behavior of OVDs in the eye during surgery, their rheological parameters associated with the CDI would provide useful information. Although these parameters provide useful information to select a suitable OVD for individual cases, they are insufficient to completely explain the behavior of the OVDs in the eye during surgery because each OVD exhibits peculiar properties. We presumed that some of these properties would affect the behavior of the OVDs. Actually, many surgeons agree that dispersive OVDs are difficult to remove from the eye compared with cohesive OVDs.5–8) We presumed that this resulted from the difference in the adhesive force between two types of OVDs. However, there are no reports to compare the adhesive forces of the two types of OVDs. Thus, we focused on the relationship between the removal times of OVDs from the eye and the adhesive force. We investigated the adhesive force of Healon, Provisc, Viscoat, and Discovisc because their retention times in the eye have already been evaluated.5,6,25) As shown in Table 2, the adhesive force of four OVDs correlated well with the removal times of each OVD after intraocular lens implantation (Healon 3.6 s; Provisc 3.66 s; Discovisc 10.68 s; Viscoat 53.23 s).5) According to the CDI classification, Healon and Provisc are cohesive type and Viscoat is a dispersive type OVD.1) As shown in Table 2, our results also suggested that Discovisc has an adhesive force intermediate between the cohesive and dispersive types similar to the rheological property described above. The overall comparison in Fig. 5 indicated that dispersive OVDs tend to have a higher adhesive force when compared with cohesive OVDs. By evaluating the adhesive force, we were able to numerically observe an interesting aspect of the behavior of OVDs in the eye, which indicated that it can be a useful index to identify both types of OVDs.
| Healon | Provisc | Discovisc | Viscoat | |
|---|---|---|---|---|
| CDI classification | Cohesive | Cohesive | Dispersive | Dispersive |
| Adhesive force (N/mm2) | −(1.79 ± 0.06) × 10−4 | −(1.71 ± 0.05) × 10−4 | −(2.15 ± 0.17) × 10−4 | −(3.36 ± 0.12) × 10−4 |
Adhesive force was detected as the detachment force when the probe was withdrawn from the OVD sample. The values represent the mean ± standard deviations (n = 3). Except for the comparison of Healon and Provisc, the significance was determined as less than 0.05 in all combinations of OVD samples using Tukey–Kramer multiple comparison.
Shellgan, Viscoat, and 3% HA showed almost the same adhesive force, but the apparent viscosities of these OVDs were slightly different (Fig. 2c). Healon EndCoat, an OVD product containing 3% HA, demonstrated non-inferiority against Viscoat in corneal endothelial cell protection,24) whereas the viscosity of home-made 3% HA was significantly lower, compared with those of Viscoat or Shellgan (Fig. 2c). Although the apparent viscosity of Shellgan, Viscoat, and 3% HA showed a particular profile, as described above, their adhesive forces were similar (Fig. 5 and Table 2). Our results indicated that the adhesive force might be helpful to understand and predict the behavior of OVDs in the eye.
As described above, the classification of Opegan is very complicated: some reports have assigned it as a cohesive OVD and others as a very low viscosity dispersive OVD. Miyauchi et al. demonstrated that 1% HA with a molecular mass of 1100 kDa exhibits longer retention in the porcine anterior chamber after PEA, compared with 1% HA with a molecular mass of 2420 kDa.26) Particularly, the retention facilitated HA adhesion onto the corneal endothelium with uniformity. As shown in Fig. 5, the adhesive force of Opegan was higher than that of Opegan-Hi. The results suggested that we might be able to predict the strength of the phenomenon reported by Miyauchi et al. by comparing the adhesive force of OVDs. Taking the physical properties observed in this study into consideration, very low viscosity dispersive seems to be an appropriate classification for Opegan, as was the assignment using CDI.
Various kinds of OVD products are commercially available and their rheological properties play an important role in surgery. Arshinoff suggested that the classification of OVDs gives an understanding of their properties and provides a logical foundation to the surgeons, which results in assisting in the selection of a suitable OVD for individual situations.1) More in-depth information about their properties is needed to select appropriate OVDs in various surgeries. As a result, we have demonstrated the effectiveness of determining the adhesive force of OVDs using a texture analyzer to understand the difference between cohesive and dispersive types.
The rheological properties and adhesive force obtained in this study might provide useful information to support the decision of surgeons to select a suitable OVD. These parameters can be useful indexes to understand and predict the behavior of OVDs in the eye during surgery.
The authors are grateful to Takayuki Seo, Ph.D. (Seikagaku Corp.) for advice concerning the statistical analyses. We also thank Jun Takeuchi, Ph.D. (Seikagaku Corp.) for carefully reviewing the manuscript and useful suggestions.
The authors declare no conflict of interest.
