10 — Blood

Blood is a class of its own.

+ The questionWhy does blood contact need its own questions?

A tissue that flows and clots

Blood is a tissue that flows, and it carries with it everything needed to turn solid. Its plasma holds the clotting proteins in an inactive form, and its platelets, white cells and red cells travel with them past every surface the blood meets. The tissue around an implant in Chapter 9 responds over weeks, but blood brings its proteins and cells to a surface at once and renews them with every pass of the flow. Blood also carries what forms at the surface on to the rest of the body, so blood contact raises questions of its own.

The running device meets no blood: its water chamber, hose and foam sit in the air path, and its mask cushion rests on skin. This chapter therefore takes its contrast from a device that does, a line that carries blood out of the body and back, as in hemodialysis. The inner wall of such a line touches the blood itself, and whatever the wall releases enters the blood at once. A line that only delivers saline or a drug touches blood through the fluid it carries, and the last section shows how testing treats the two kinds of line.

The first event at the line's wall is the exchange of proteins that Chapter 8 describes: on a surface that sets off clotting by contact, high-molecular-weight kininogen and factor XII replace fibrinogen, the protein that clotting turns into fibrin. Which protein holds the wall matters, because platelets adhere preferentially to a coat of fibrinogen, as Vroman and colleagues found in 1980. A coat of kininogen and factor XII instead starts a chain of reactions in the plasma, which the next section follows.

A surface can start the cascade

Coagulation is the turning of liquid blood into a solid clot. It runs as a cascade, in which each protein, once activated, activates the next, and the chain ends in a burst of thrombin, the enzyme that converts fibrinogen into fibrin. Kudzin and colleagues trace the steps at a surface in a 2026 review of how fibrous polymer materials promote or prevent clotting. Factor XII binds to the surface through high-molecular-weight kininogen and activates itself. Activated factor XII converts prekallikrein into kallikrein and activates factor XI, and activated factor XI recruits factor IX, so the chain runs on to the thrombin burst. This route is contact activation: it needs no injury, only a surface.

The wall's own properties enter at this first step. Kallikrein and activated factor XII bind efficiently to negatively charged or nanostructured surfaces, the review states. The electrical charge and the fine texture of a wall therefore affect how strongly the cascade starts. The body's other route begins at an injury, where tissue factor and activated factor VII set the chain going, and both routes lead to the thrombin burst. The wall of a blood line has no injury, so clotting there starts at the surface, and the product is a solid in a tube built to carry a liquid.

Platelets and complement call the white cells

Platelets respond to the coat directly. In Kudzin and colleagues' account, a platelet engages adsorbed fibrinogen through GPIIb/IIIa, a receptor on its surface, and the binding triggers an influx of calcium and the secretion of the platelet's α-granules, its internal stores. The activated platelet also exposes P-selectin, a protein that makes leukocytes, the white cells, adhere to the platelet, and the exposure amplifies inflammation. Fast flow adds a further step. Shear stress is the drag that faster layers of blood exert on slower layers and on the wall, and at high shear the platelets' activation and aggregation depend on von Willebrand factor, a large plasma protein.

A foreign surface can also set off complement, a set of plasma proteins of the immune defense. Wang and colleagues at Qingdao Central Hospital describe complement on the membranes of hemodialysis in a 2026 review of how blood and membranes interact. Once a surface sets complement off, the reaction can feed itself: fragments of the protein C3, together with another plasma protein, properdin, can support the amplification of complement's alternative pathway. Two small fragments, C3a and C5a, act as signals that recruit white cells and prime them, readying them to respond. The chain ends in the terminal complement complex, sC5b-9, which forms in the plasma or on the surface and which laboratories measure (last section). The review adds that complement and contact activation interact with platelets and white cells in one network of clotting and inflammation.

The network reaches beyond the wall. With strongly complement-activating membranes, the patient's white-cell count falls early in a treatment and then recovers, a transient leukopenia in Wang and colleagues' term. The same review notes that the complement response today is usually smaller than with the older cellulose membranes. What forms at a membrane inside a machine thus changes a count taken from the patient's whole circulation.

26 — Three responses start at one blood-contacting surface
BLOOD FLOW PLATELETS CONTACT ACTIVATION COMPLEMENT one network of clotting and inflammation BLOOD LINE WALL — NO INJURY protein coat (Chapter 8) fibrinogen kininogen + factor XII (they replace fibrinogen on contact-activating surfaces) at high shear: von Willebrand factor GPIIb/IIIa calcium in; α-granule contents out P-selectin WHITE CELL DRAWN IN factor XII → XIIa prekallikrein → kallikrein factor XI → XIa factor IX thrombin burst fibrinogen → fibrin clot injury route (tissue factor + factor VIIa) also leads to thrombin — not present on this wall C3 fragments + properdin: amplification C3a, C5a recruit and prime white cells sC5b-9 terminal complex: in plasma or on the surface carried into the circulation: with strongly complement-activating membranes, an early, transient fall in the white-cell count clotting starts at the surface: no injury Schematic. Mechanisms: Kudzin and colleagues, 2026 review; complement on dialysis membranes: Wang and colleagues, 2026 review; platelets and fibrinogen: Vroman and colleagues, 1980.
Device materialTissue and cellsGas or fluid pathFocal detail
Plate 26 — A surface in blood starts three responses that interact. Contact activation runs from factor XII to a burst of thrombin with no injury, platelets bind the fibrinogen in the coat and draw in white cells, and complement fragments recruit white cells, whose count across the circulation can fall for a time.

Red cells break from the flow or from the material

Hemolysis is the breaking of red cells, which releases their hemoglobin into the plasma. ISO 10993-4's amended test table separates two kinds by their cause, material-induced and mechanically induced (last section). In Kudzin and colleagues' account, extremely high shear stresses damage red cells mechanically, and a damaged red cell releases ADP, a small molecule that further stimulates platelets and coagulation. Broken red cells therefore feed the cascade and the platelets alike. The shear at a wall rises with the flow and with the narrowness of the passage, so mechanical hemolysis depends on how a device moves blood, not only on its material. The review gives no shear stress or exposure time at which the damage begins, so a test at the device's own flow shows whether that flow breaks red cells (last section).

Material-induced hemolysis is measured by putting blood in contact with the material. Malinauskas and colleagues examined the standard method for it, ASTM F756-17, in a bench study shown on an undated poster. The authors are scientists in FDA's device laboratories. They placed solid test articles in direct contact with diluted rabbit blood at 37 °C for up to 3 hours, with three gentle inversions every half hour. The blood was anticoagulated with citrate and pooled from three donors. A color reaction read at 540 nm then measured the hemoglobin freed into the plasma. Hemolysis rose with time, which supports the method's contact time of 3 hours.

As in the implantation test of Chapter 9, each run carries a negative control, here a material expected to break almost no red cells. It also carries a positive control, a material expected to break many. FDA's poster cites ASTM F756-17's condition for a valid run: the negative control must release less than 2 % of the blood's hemoglobin, and the positive control more than 5 %. The poster gives no formula for that share, but a 2019 report from a contract laboratory gives one: the hemoglobin a sample releases into the plasma, divided by all the hemoglobin the diluted blood holds, corrected for a blank of buffered saline alone.

The laboratory reads a material against three bands counted above the negative control: 0 to 2 % is non-hemolytic, 2 to 5 % slightly hemolytic and above 5 % hemolytic. It tested a thermoplastic elastomer with pooled human blood from three donors, by direct contact and through an extract, and the elastomer gave 0.00 %. The elastomer's maker publishes the report in its product database.

Solved example: whether a glove brand can serve as the positive control
  • The criterion. Under ASTM F756-17, as FDA's poster cites it, a positive control must release more than 5 % of the blood's hemoglobin.
  • The two results. Two of the nitrile glove brands that FDA's scientists tried as positive controls gave means of 93.3 % and 6.3 %, with standard deviations of 8.6 and 4.2 percentage points over three samples.
  • The first brand. One standard deviation below its mean, 93.3 minus 8.6, is 84.7 %, about 17 times the criterion, so its runs clear 5 % by a wide margin.
  • The second brand. One standard deviation either side of its mean runs from 6.3 minus 4.2, which is 2.1 %, to 6.3 plus 4.2, which is 10.5 %. The criterion lies inside that range, so single runs below 5 % are to be expected, and those runs would be rejected.
  • The result. Two gloves of the same kind of material gave means about 15 times apart: 93.3 divided by 6.3 is 14.8. The poster concludes that only the solvent DMSO, at 15 % or more, and one brand of nitrile glove were suitable positive controls, so a control is chosen by its measured hemolysis, not by what it is made of.
  • Assumptions. The ± is a standard deviation of three samples, as the poster's table states; rabbit blood at 37 °C for up to 3 hours.

Six families of test, one mechanism each

ISO 10993-4 is the standard for choosing tests of blood interaction. Its 2017 edition lists devices under three groups of its own: devices that communicate with the outside of the body and touch blood indirectly, those that communicate with the outside and touch it directly, and implants. These are not the tissue groups of ISO 10993-1:2025 that Chapter 2 uses. ISO 10993-4's Amendment 1, published in January 2025, moves devices that collect, store or give blood from the indirect-contact list to the direct-contact list, as a standards seller's summary of the amendment reports. A bag that stores blood holds the blood itself for weeks (Chapter 4), so the move puts it with the devices whose walls meet blood. In the same summary's account, the amendment adds blood monitors and devices for saline or therapeutics to the indirect list, and an indirect-contact device generally needs an indirect hemolysis test.

The amended table lists six categories of test, again in the seller's summary. Read against this chapter, each category reads one mechanism. Coagulation reads the cascade, platelet activation the platelets, and complement activation the fragments of complement. Hematology counts the blood's cells, and so reads the fall in white cells that complement can cause. Hemolysis, material-induced or mechanically induced, reads the red cells. Thrombosis reads the end of the chain, a clot formed on the device, and it is tested in the body, in blood led outside the body, or in the laboratory. The group a device falls in sets which categories its plan must address.

A testing laboratory's own undated document groups five categories for testing in the laboratory, with what each measures. Complement, for instance, is read by sC5b-9, the terminal complex that ends the complement chain. Hemolysis is tested by direct contact or through an extract in saline or phosphate-buffered saline. Thrombosis is not among the five, although the seller's summary lists laboratory tests for it as well as tests in the body.

27 — Six test categories and the mechanisms they read
INDIRECT BLOOD CONTACT — EXTERNALLY COMMUNICATING DIRECT BLOOD CONTACT — EXTERNALLY COMMUNICATING IMPLANT added by Amendment 1: blood monitors; devices for saline or therapeutics generally an indirect hemolysis test (seller's summary) the device's wall meets the blood itself ISO 10993-4:2017's own groups, not the tissue groups of ISO 10993-1:2025 (Chapter 2) devices that collect, store or give blood: moved to direct contact by Amendment 1, published January 2025 CATEGORY WHAT IT READS WHAT A LABORATORY MEASURES WHERE TESTED Hemolysis (material-induced; mechanically induced) red cells broken hemoglobin released, as a share of the blood's hemoglobin; direct contact or an extract in saline or PBS laboratory Coagulation the cascade partial thromboplastin time; thrombin–antithrombin complex laboratory Platelet activation platelets β-thromboglobulin; platelet count laboratory Complement activation complement sC5b-9 laboratory Hematology the blood's cells complete blood count; PMN elastase laboratory Thrombosis a clot on the device not in the laboratory's in vitro list in the body, ex vivo or laboratory Device groups and six categories: ISO 10993-4 Table 1 as replaced by Amendment 1, from a standards seller's summary. Measurements: a testing laboratory's own undated document (five in vitro categories, no thrombosis).
Device materialTissue and cellsFocal detail
Plate 27 — ISO 10993-4 sorts devices into three groups of its own and lists six test categories, which this chapter reads against the mechanisms of blood contact. Its Amendment 1, published in January 2025, moves devices that collect, store or give blood from indirect to direct contact, as a standards seller's summary reports.

How the blood moves during a test decides which mechanisms the test can show. Weber and colleagues, at the University of Tübingen, state in a 2018 review that extensive laboratory tests under ISO 10993-4 are required before a blood-contacting material is used clinically. In those tests, the review adds, the blood is held still, agitated or put in shear flow. Von Willebrand factor's part in platelet activation and the mechanical breaking of red cells both come with high shear. A still sample has no such shear, so a line with fast flow calls for a shear-flow model in those tests.

In practice

List every part of a device that touches blood, or a liquid that then enters the bloodstream, and sort each into ISO 10993-4's groups: indirect contact, direct contact or implant. For the running device the list is empty (Chapter 2). For a blood line, plan the categories its group lists, and run the platelet and hemolysis tests under flow where the line's shear is high. Test the finished line as made, since a resin maker's hemolysis report covers one category of six, for the material as its maker supplied it (Chapter 6).

+ What this chapter established
  • Blood brings clotting proteins and cells to a surface at once and carries what forms there to the whole body.
  • A surface starts the clotting cascade through factor XII and kininogen and calls platelets through adsorbed fibrinogen, and fast flow adds a step that depends on von Willebrand factor.
  • Complement fragments from a surface recruit white cells and can lower the patient's white-cell count for a time, and red cells break from the material or from extreme shear.
  • ISO 10993-4 sorts devices into three groups of its own. Its Amendment 1 of January 2025 lists six test categories and moves devices that collect, store or give blood to direct contact, as a standards seller's summary reports.
Created with Claude (Anthropic).
Educational reference material, not engineering, regulatory or clinical advice. Product and company names referenced are the property of their respective owners.