, Volume 26, Issue 6, pp 935–939

Gallium nitrate induces fibrinogen flocculation: an explanation for its hemostatic effect?


  • A. Bauters
    • Department of Hematology and TransfusionUniversity of Lille Nord de France
  • D. J. Holt
    • MedInsight Research Institute
  • P. Zerbib
    • Department of Gastrointestinal Surgery and TransplantationUniversity of Lille Nord de France
    • MedInsight Research Institute

DOI: 10.1007/s10534-013-9669-4

Cite this article as:
Bauters, A., Holt, D.J., Zerbib, P. et al. Biometals (2013) 26: 935. doi:10.1007/s10534-013-9669-4


A novel hemostatic effect of gallium nitrate has recently been discovered. Our aim was to perform a preliminary investigation into its mode of action. Thromboelastography® showed no effect on coagulation but pointed instead to changes in fibrinogen concentration. We measured functional fibrinogen in whole blood after addition of gallium nitrate and nitric acid. We found that gallium nitrate induces fibrinogen precipitation in whole blood to a significantly higher degree than solutions of nitric acid alone. This precipitate is not primarily pH driven, and appears to occur via flocculation. This behavior is in line with the generally observed ability of metals to induce fibrinogen precipitation. Further investigation is required into this novel phenomenon.


FibrinogenFlocculationAggregationGalliumGallium nitrateHemostasis


Rapid cessation of bleeding is critical in cases of injury with severe blood loss, particularly where proper, rapid medical treatment is inaccessible, as in military combat situations. This requires the presence of an agent that can induce hemostasis in order to promote healing.

Many current wound-hemostasis products are based on the properties of kaolinite (Pahari et al. 2010), a clay that has been shown to induce the coagulation cascade (Margolis 1958).

A recent report (Goodley and Rogosnitzky 2011) has demonstrated that gallium nitrate possesses hitherto undiscovered properties that significantly reduce the bleeding time of wounds. The mechanism by which it operates is unknown.

This report demonstrates that the previously observed effect of gallium nitrate on hemostasis: (i) occurs via inducement to flocculation, and (ii) this process is dependent, at least in part, upon the presence of the gallium ion. This behavior is in line with the generally observed ability of other metals, such as zinc, to induce fibrinogen precipitation. A mechanistic explanation for this phenomenon, which would be desirable, cannot be advanced without further research.

Materials and methods

Gallium nitrate solution 14 % w/w (1.6 M) (purchased from Eby Pharma LLC, Dripping Springs, TX, USA) was used as received. Nitric acid solution 65 % (Merck KgaA, Germany) was used at a concentration of 10 % w/w (1.6 M).

Blood from healthy volunteers was taken into sodium citrate 0.129 M (3.8 %).

The experimental solutions of gallium nitrate and nitric acid described above were added in increasing volume: 25, 50 and 100 μl/ml in citrated whole blood samples (for final concentrations of 39, 75 and 145 mM, respectively). The results were compared to the addition of 25, 50 or 100 μl of Owren’s buffer.

Plasma was obtained by centrifugation at 2,000×g for 20 min at 20 °C before analysis.

Thromboelastography® (TEG®) was first employed to determine the effect of gallium nitrate on citrated whole blood. TEG® is a global test that provides a profile of the entire coagulation process: initiation, propagation and final clot strength measurements. TEG® measures the visco elastic properties of clotting whole blood activated with 10 μl of tissue factor and kaolin. The results of the TEG® process are reported in terms of R (reaction time), the time until the first evidence of clotting is detected, K (speed of clot formation), angle (curve made as K is reached) and MA (maximum amplitude of the clot). K and angle represent fibrin activity and MA represents the combined activity of fibrin and platelets in the clot formation.

The effects of experimental solutions on whole blood were measured via functional fibrinogen determination in plasma (Dade Thrombin Reagent, Siemens, Marburg Germany) using a Sysmex CA-7000 (Siemens, Germany). Results are given as percentage of the original reading without addition of aliquoted solution.

The assay uses optical absorbance to measure the time required to convert the soluble plasma protein fibrinogen into its insoluble polymer, fibrin. The time lag is dependent upon the initial fibrinogen concentration in plasma. The clotting time obtained in this manner is then compared with that of a standardized fibrinogen preparation.

In order to address the possibility that addition of gallium to citrated whole blood results in complexation of gallium with citrate (causing calcium unbinding from the citrate and inducing the coagulation cascade), we compared the TEG® results of 340 μl citrated whole blood both with and without addition of 20 μl/ml gallium nitrate. No signal was observed in either case, thus no induction of the coagulation cascade occurred; conversely, addition of 10 μl calcium chloride 0.2 M was necessary to set off the coagulation cascade. This suggests that citrate remains bound to calcium in the whole blood, and does not dissociate from the calcium upon addition of gallium nitrate.

Results and discussion


The TEG® results of the addition of gallium nitrate to citrated solutions of whole blood, and then addition of tissue factor and kaolin, are described in Fig. 1.
Fig. 1

Thromboelastograph plot of clotting time (R, min) and maximum amplitude (MA, mm) for increasing concentrations of gallium nitrate in citrated whole blood spiked with tissue factor and kaolin

We found that as gallium nitrate concentration in citrated whole blood is increased, clotting time (R, min) is relatively invariant and remains within normal values (2–8 min), while the maximum clotting strength (maximum amplitude, MA, mm) decreases by a third, taking it well below the expected range of values (51–69 mm). This indicates that the presence of gallium nitrate does not activate the coagulation cascade (which would be expected to lower R significantly) and renders fibrinogen unavailable for clot formation (evidenced by decreasing MA). At 20 mM gallium nitrate concentration (not shown), no fibrinogen was available to form a clot. TEG® therefore showed no effect on coagulation but pointed instead to changes in fibrinogen availability. We therefore pursued investigations into the fibrinogen concentration.

Fibrinogen determination

Percentage change of plasma fibrinogen concentrations after addition of Owren’s buffer, gallium nitrate, and nitric acid solutions to whole blood are shown in Fig. 2.
Fig. 2

Plasma fibrinogen concentration (FIB) relative to original sample reading for increasing concentrations of buffer, nitric acid and gallium nitrate in whole blood samples

The results demonstrate that the nitric acid solution had a significant effect on fibrinogen concentration when added to blood at greater than 75 mM concentration. A small decrease in fibrinogen (4 % on average) is seen at 39 mM, a concentration of 76 mM causes an average 30 % decrease, and 145 mM causes a concentration below the limit of detection (LOD).Visual inspection of the nitric acid samples showed that flocculation was induced.

By contrast, the effect of gallium nitrate on the fibrinogen concentration was significantly greater than that of nitric acid; a concentration of 15 mM resulted in an approximate 30 % reduction in fibrinogen concentration. Increasing the concentration to 39 mM brought the fibrinogen below the LOD. Precipitation was observed on addition of the gallium nitrate to the sample (see Fig. 3).
Fig. 3

Representative images of plasma samples with and without fibrinogen precipitate: (left) plasma sample with buffer added, no flocculation visible; (right) plasma sample with 25 μl gallium nitrate solution added. Flocculation effects are clearly present in the gallium nitrate sample. These flocculation effects disappear from the sample upon agitation, indicating that they are not due to activation of the coagulation cascade, whose products would remain compact. The scale bar is 20 mm

Visual examination

Visual inspection of the samples after analysis (not shown here) confirmed increasing precipitation effects with increasing concentration of reagent for both nitric acid and gallium nitrate solutions. Importantly, the precipitation effects were reversed upon agitation of the samples, indicating that precipitation was not a result of activation of the coagulation cascade, but instead a flocculation mechanism.

The effects of pH

The pH of the nitric acid solution was measured to be 0 (vs. a theoretical pH of −0.2). The pH of the gallium nitrate solution was measured to be 1.4, which accords with the theoretical dissociation constant (−1.4).

If the flocculation effect was due to pH, we would have expected the amount of flocculation occurring to be inversely proportional to pH. In light of these results, then, the greater flocculation occurring due to gallium nitrate cannot be due to the pH (since it is higher than the nitric acid solution), nor is it due to the presence of the counterion (since the nitrate concentrations are almost identical in our nitric acid and gallium nitrate solutions, at 1.6 M, and both solutions are shown by their pH readings to be almost completely dissociated in the test solutions).

Possible mechanisms of action

We conclude therefore that the presence of the gallium ion appears to be driving the flocculation. The mechanism and physical properties by which this occurs are unknown and require further investigation. It seems unlikely that ionic strength plays a role (the ionic strength of the gallium nitrate solution is 3.28 M, relative to an ionic strength of 1.59 M for the nitric acid solution), since solutions of different ions of greater ionic strength are known to not produce fibrinogen aggregation [for example, Al2(SO4)3, whose ionic strength is more than double that of gallium nitrate (Steven et al. 1982)].

Much of the literature on aggregation of fibrinogen due to metal ions concentrates on divalent cations (Steven et al. 1982) that have a variety of outcomes, from no influence whatsoever (in the cases of calcium and magnesium), to significant precipitation at micromolar concentrations (in the cases of nickel and zinc). However, the mechanisms by which flocculation, precipitation or aggregation are induced are poorly understood (Sangani et al. 2010).

We have found no description for this behavior of gallium [even in comprehensive reviews of its biological properties such as Bernstein (1998)].

The focus of our study was to uncover the mechanism through which gallium nitrate induces hemostasis. With the uncovering of its fibrinogen flocculation property [similar to that observed with zinc (Steven et al. 1982)], it remains to be investigated whether other metals, such as zinc, could be used for inducing hemostasis. Zinc is known to induce fibrinogen aggregation at lower concentrations than gallium nitrate. We have not found any reports of low concentration zinc being useful for hemostasis. However, Mohs’ paste, a very high concentration zinc chloride preparation has been shown to stop bleeding from malignant wounds (Kakimoto et al. 2010; Yanazume et al. 2013), although an escharotic affect at these concentrations has been observed. It may be useful to investigate whether non-escharotic low-concentration zinc salts can be used to induce hemostasis.

By contrast, gallium nitrate has no reported escharotic behavior even at molar concentration. Case reports exist about its topical transdermal use, at molar concentrations, in arthritis (Eby 2005). It has been studied in porcine models in millimolar concentrations for its positive effects on wound repair (Goncalves et al. 2002). A recent report suggested that due to its potential systemic absorption after topical exposure, its dermal toxicity should be re-evaluated (Staff et al. 2011). In the absence of dermal toxicity, however, its lack of coagulative properties may render it an ideal option for hemostasis where avoidance of the thromboembolic risk of some of the currently used hemostatic agents (Kheirabadi et al. 2010) is essential.


Gallium nitrate solution induces fibrinogen precipitation in whole blood samples via a flocculation pathway, and such behavior appears to be dependent upon the presence of the gallium ion in solution. This may explain the previously reported finding regarding the hemostatic effect of gallium nitrate.

Further research should explore whether fibrinogen flocculation also occurs in the presence of other water soluble gallium salts such as gallium chloride. Addition of excess EDTA may be useful to confirm that the gallium ion is causing the flocculation. Extensive further investigation is required in order to fully understand the mechanism behind the flocculation of fibrinogen and its significance.


The authors gratefully acknowledge the assistance of Dr. Lawrence Bernstein, George Eby, and Todd B. Hall to enable this research. Thank you to Carol A. Bienstock for assistance with manuscript preparation.

Conflict of interest

M. Rogosnitzky owns a patent on the use of gallium for hemostasis.

Copyright information

© Springer Science+Business Media New York 2013