correlation between surface roughness data and the patients’ hematological data

correlation between surface roughness data and the patients’ hematological data

Abstract

The surface roughness due to the pore size distribution of dialysis membranes, surface texture, surface toughness and electrical potential distribution are some of important parameters of the membrane surface that directly or indirectly affects hemocompatibility due to its interaction with blood cells, proteins, and platelets of blood being filtered. We obtained hollow fiber dialysis membrane from four different company brands of dialysis cartridge generally used at dialysis centers across the kingdom of Saudi Arabia. The membrane fiber samples were prepared for nano-characterization of its outer and inner surfaces using atomic force microscope (AFM).  Apart from outer and inner surface of the membranes, we also imaged the cross section of the membrane fibers accessing it from the embedded domain at the blood inlet. The nano-images of the membrane surfaces were recorded using peak force quantitative nanomechanical (PF-QNM) mode. Using the offline software for AFM image analysis, we analyzed the membranes under study for the surface roughness and toughness that are supposedly important parameters affecting hemocompatibility. A comparative analysis of the nanomechanical properties obtained for different makes of the membrane fibers are presented here, with further aim to find a correlation with the dialysis patients’ relevant hematological data that may get affected due to a difference in nanomechanical properties.

Key Words: Dialysis Membrane, AFM, Surface Roughness, Nano-Characterization

Introduction:

Kidney failure patients require to undergo a periodical filtration of their bloods to get rid of excess water and toxic materials artificially which otherwise are naturally performed by two kidneys in normal conditions [1-3]. Dialysis membranes developed by healthcare researchers have been used commercially for decades to design and manufacture branded blood filtration cartridges or devices used to dialyze kidney failure patients [1].  Theoretically, the performance of these filtration devices is dependent on several macro and micro or nano parameters such as inlet flow patterns, hemodynamics of overall flow and microcirculation through the hollow membranes within the devices as well as the surface features specially on the inner surface of the hollow membranes which comes in direct contact with blood components. Several studies conducted during the last decade of the last century reported life threatening situations associated with hemolysis due to the then dialysis equipment in use [4], due to the kinked hemodialysis line for mechanical reasons [5], due to the influence of different hemolysis membranes [6, 7]. Recent studies have reviewed this hemolysis problems due to hemodialysis prevalent during first two decades of the current century that shows that the situation warrants improvement [8, 9]. A comparison has been made to account for the hemolysis due to varied hemodialysis treatment [10] however hemodynamics including high flow rate manipulations apparently did not affect the hemolysis [11]. Therefore, the nano level surfaces of hollow micro-fibers, the pore structures and surface roughness of the porous membranes may be of great importance in terms of overall hemocompatibility and specially the red cell damage. Atomic Force Microscopy (AFM) offers a host of techniques to measure such surface nano-characteristics[12]. Topological images of surfaces are processed, and quantitative parameters of topological features are calculated. Mathematically the roughness average is the arithmetic average (Ra) of the absolute values of the roughness profile ordinates. Ra is very common measures of surface roughness in engineering practice. It represents a very good approximation of the height variations in the surface. The units of Ra in nanotechnological metrology is generally in nanometers. The Root Mean Square (RMS) roughness is also called (Rq) and is defined as the root mean square average of the roughness profile ordinates. A model for roughness statistics of such materials have been put forward in a recent study that enlightens the subject in depth [13].

Our primary aim in this report, therefore, is to present our preliminary data that were obtained using atomic force microscope (AFM) to ascertain if any significant differences in surface roughness parameters exist among different makes of the hollow porous dialysis fibers used to develop branded devices. Additionally, our aim continues to find a correlation between surface roughness data and the patients’ hematological data if any (study in progress).

Materials and Methods:

Dialysis membrane cartridges of four different makes were obtained from a dialysis center in Jeddah city of Saudi Arabia. In order to keep it anonymous we do not write the brand names rather we denote them as dialysis membrane of make-A, B, C and D, namely, DM-Make-A, DM-Make-B, DM-Make-C and DM-Make-D.

The unused dialysis devices were cut open at two places. One cut was made at the blood inlet side to expose the inlet compartment in order to facilitate the access by AFM probes to the embedded hollow fibers. An optical image of the embedded hollow membrane fibers is shown in Figure-1a. Another cut was made in the body of the cartridge device to obtain individual pieces of hollow fibers to be used as samples for nanoimaging by AFM. The abluminal surfaces of the fibers (Figure-1b) were used for nanoimaging as such after gluing it to a glass slide while the luminal surfaces (Figure-1c) were exposed by a slant cut made to the glued fibers by a sharp blade and thus accessed by AFM tip probe for luminal surface nanoimaging.

We used Dimension ICON with ScanAsyst, and its proprietary PF-QNM mode, equipped with Nanoscope software (Bruker Corporation, Billerica, Massachusetts, United States) to investigate nanofeatures on the hollow and porous dialysis membrane fibers of different makes. Both the surfaces luminal as well as abluminal of the hollow dialysis fibers were imaged for comparison purpose even though the luminal surface of the fibers is of direct relevance as it comes in direct contact with the

patient’s blood during the dialysis period of almost 4 hours. The tips model used were the TESPA made up of 0.01-0.025 Ohm-cm Antimony (n) doped Si. The company specifications of the TESPA model tips were 3.5 – 4.5 µm for thickness, 110-140 µm for length, 25-35 µm for width with frequency range of 228-271 kHz and spring constant of 20-80 N/m. All the image data obtained were analyzed using offline Nanoscope Analysis software tools version 1.5. A student t-test was employed to determine the level of significance.

Results and Discussions:

The luminal and abluminal surfaces of all the four types of hollow dialysis fibers were investigated by analyzing the surface images of 25 to 100 µm² areas sampled at different locations of the fibers. The micro-sized images grabbed by AFM with visual nano-information are shown in the gallery of images as presented in Figure-2a&b. The figure-2a compares the abluminal surface roughness and toughness of the hollow fiber membranes of all four types. All images are representative of 100 µm² area on the membrane abluminal surface. The figure-2b compares the luminal surface area of same dimension for roughness and toughness of all four membranes under study. These visual impressions are quantitatively represented in the form of bar diagram in figure-3 and figure-4.

An average and standard deviation of the nano data obtained were plotted for all categories to compare the roughness and toughness represented by Rq which is indeed the Root Mean Square (RMS) average of the roughness profile ordinates. The plots for luminal roughness and toughness in terms of DMT modulus are shown in Figures-3 and 4. The luminal roughness of two of the fibers (DM-Make-B and DM-Make-C) are low and almost of similar magnitude (4.5±0.2 nm and 4.2±0.1 nm) as compared

to significantly higher values (9 ± 1 nm and 15.9 ± 0.6 nm)  for the fibers of other two makes DM-Make-A and DM-Make-D as shown in Figure-3a.

The luminal toughness or the DMT modulus obtained for all fiber types are shown in Figure-3b. The two fiber types (DM-Make-B and DM-Make-C) were found to exhibit lower and comparatively similar values of DMT modulus (0.8±0.2 GPa and 0.7±0.03 GPa) as against the other two membrane models (DM-Make-A and DM-Make-D) corresponding to a value of 196.0 ± 8.6 GPa and 1.7 ± 0.25 GPa. The average DMT modulus value of 196 GPa for DM-Make-A was exceptionally high as compared to others. Tough surfaces may cause a high stress to the flowing blood particles during the dialysis process and may cause damage especially to the vulnerable red blood cell membrane. This may in turn reduce the hematocrit value of the patients’ blood by the end of the dialysis session. We hope to find answer to this query when we compare patients’ data as part of our ongoing study.

The abluminal surface roughness of all four types (Figure-4a) do represent a trend of being opposite in roughness values. DM-Make-B and DM-Make-C are smoother luminally but rougher at abluminal surface while other two fibers show the opposite trend. However, the surface rigidity or DMT modulus values (Figure-4b) are a kind of diverse. The fiber DM-Make-A is tougher from inside while from outside it shows quite low DMT modulus values. On the other hand, DM-Make-C is quite smoother and softer from inside as compared to outside. The abluminal values of roughness is quite high for this fiber type while it is quite softer with a low DMT modulus values as compared to other abluminal fiber surfaces. We chose to ignore discussing much on the abluminal nanomechanical properties as it does not come in direct blood contact during dialysis hence of least relevance.

Conclusions:

We conclude that the fibers used by various manufacturers to make blood filtration devices for dialysis patients differ significantly in terms of nanomechanical properties and features which may represent important parameters that in turn may affect filtered blood quality of dialysis patients. A study to ascertain any statistically significant correlation of these nanomechanical parameters with the state of patients and patients blood data may divulge a guideline to improve upon the design and nanomechanical properties of hollow and porous dialysis membrane for a better result and dialysis patient satisfaction.