Acquired Factor XI Deficiency during SARS-CoV-2 Infection: Not Only Thrombosis.

An 80-year-old woman with a positive RT-PCR (reverse transcription polymerase chain reaction) for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) on nasopharyngeal swab was admitted to our COVID-19 department during the pandemic with fever, dyspnea, and need for oxygen therapy; two large hematomas were present bilaterally in the axillary zones. Her past medical history included Crohn's disease diagnosed 30 years ago after perforation of the ascending colon, hypertension, bilateral hip replacement, appendectomy, and tonsillectomy; no history of bleeding complications was reported during or after surgical interventions or in the postpartum, and no history of menorrhagia or easy-bruising. One month and a half before admission, she underwent right knee replacement surgery with no complications, no blood units, or plasma transfusions were required. At that time complete blood count and screening clotting test were within normal range.

Investigations on admission were as follows: white blood cell 11,380/mm 3 (Neu 10,790; Ly 240); hemoglobin 8 g/dL; mean corpuscular volume 89 fL; platelets 251,000/mm 3 ; PTr 1.16 (n.v. 0.80–1.21); PTTr 1.49 (n.v. 0.80–1.18); D-dimer 2,161 μg/L (n.v. < 790); fibrinogen 455 mg/dL (n.v. 180–400); C-reactive protein 78.4 mg/L (n.v. < 5.0). A chest computed tomography scan outlined a picture of bilateral interstitial pneumonia with ground-glass areas and thickening of interstitium, consistent with a viral infection; moreover, the presence of hyperdense hematomas in the axillary zones was recorded (thickness 20–30 mm), no active bleeding was detected. Antiretroviral therapy with darunavir and ritonavir was started; hydroxychloroquine was not administered for prolonged QTc on electrocardiogram. Enoxaparin 4000 IU/day was introduced for prophylaxis and continued until hospital discharge.

The isolated prolonged PTTr (silica) was confirmed in day +5 and +7 from admission (1.30 and 1.42, respectively), PTTr (ellagic acid, reagent sensible to factors VIII, IX, XI deficiency) was prolonged too, so a PTT MIX test was required: the PTTr MIX at time 0 was normal but PTTr MIX at 2 hours at 37°C was above the normal range, consistent with the presence of an inhibitor ( Table 1 ); the fXI activity was 37% (n.v. 55–150) while fVIII, fIX, and fXII were in the normal range. Detection of antiphospholipid antibodies (lupus anticoagulant, anticardiolipin, and anti-β2 glycoprotein I) resulted negative. The PTT MIX test and clotting factors activity was retested 3 days later and the presence of an acquired deficiency of fXI was confirmed. Patient was discharged in good clinical conditions more than 1 month after admission, no thrombotic or bleeding complications occurred during hospitalization.

Table 1

Coagulatory and inflammatory parameters

	Unit	Ref.	March 17	March 21	March 23	March 26	April 7	April 12	April 18	April 21
Thrombin time	Ratio	[< 1.20]			1.08
aPTTr (silica)	Ratio	[0.80–1.18]	1.49	1.30	1.42	1.55	1.40	1.30	1.30	1.24
aPTT MIX time 0	Ratio	[0.80–1.18]			1.17	1.24
aPTT MIX 2 h 37°C	Ratio	[0.80–1.18]			1.20
aPTTr (ellagic acid)	Ratio	[0.70–1.18]			1.36
Factor VIII	%	[50–200]			120
Factor IX	%	[65–140]			101
Factor XI	%	[55–150]			37	38.5
Factor XII	%	[55–180]			63
D-dimer	μg/L	[< 790]	2,161		3,501		2,471	2,004	2,183
Fibrinogen	mg/dL	[180–400]	455	541	505			541	568
CRP	mg/L	[< 5.0]	78.4			55.6		12.7	34.7

Abbreviations: aPTTr, activated partial thromboplastin time ratio; CRP, C-reactive protein.

Inhibitors against clotting factors generally occur in patients with severe congenital deficiencies who underwent periods of replacement therapy. Acquired factor inhibitors in the absence of congenital deficiency are rare events, among them the production of autoantibodies inhibiting fVIII is the most frequent acquired bleeding disorder. On the contrary, inhibitors against fXI have been reported only anecdotally in literature mostly associated to autoimmune disorders, that is, systemic lupus erythematosus, 1 rheumatoid arthritis, 2 Crohn's disease, 3 membranoproliferative glomerulonephritis, 4 or in association with malignancies, that is, chronic lymphocytic leukemia, 5 gastrointestinal adenocarcinoma, 3 and thymoma 6 ; a case of transient acquired fXI deficiency after gynecological surgery has been reported. 7 The risk of bleeding in conditions of fXI deficiency is relatively low and correlation between factor levels and symptoms is very poor; in these persons hemorrhage is usually provoked, exacerbated by trauma or surgical procedures. 8 9

In the classical chart of the intrinsic pathway of coagulation waterfall the fXIIa activates the zymogen fXI which in turn activates fIX and so, through a series of enzymatic reactions, it conducts to thrombin formation. It is nowadays well documented that fXII has only a marginal role in the hemostasis in vivo, contributing more to inflammation and pathologic coagulation; in fact, fXI seems to be activated in a stronger way directly by thrombin, thus supporting and amplifying thrombin formation in loop. 10 Moreover, it has also been shown how fXIa activates fXII after tissue factor-initiated coagulation (via extrinsic pathway ). 11 Maybe coagulation is not a waterfall but a whirlpool!

But fXI is not only a component of the coagulation pathway, it also participates to the process of contact activation by mean of the bradykinin-generating kallikrein–kinin system (KKS) (including fXII, prekallikrein PK, high molecular weight-kininogen HK, and C1 esterase inhibitor C1INH) contributing in this way to several host-defenses and to the initial response to infection: in fact, the inflammatory response to viral infections is in part mediated by contact activation and it has already been demonstrated how hantavirus and herpes simplex virus 1 can lead to the activation of KKS pathway. 12 The fXI, together with fXII, seems to have a pivotal role in this context. 13 In a sepsis setting of fXI deficient mice, Bane et al have shown how a reduced early cytokine response and KKS activation conduct to a survival benefit compared with wild-type mice. 11 14

Reconnecting to the case reported above, the reason for the occurrence of inhibitors production against fXI during SARS-CoV-2 infection is still not clear to us, but maybe we can speculate that the inhibition of fXI could be a means of the immune system to try to turn off a sustained inflammatory overresponse against an infective agent which can otherwise conduct to deleterious effects for the host. Why not to consider this patient, with an autoimmune milieu (Crohn's disease), as evolutionary advantaged in this condition having the possibility to counterbalance the state of hypercoagulability and inflammation (the so-called, thromboinflammation 15 )? Why not to act directly against fXI, inhibiting it and on one side blocking the pathological coagulation loop with a mild risk of hemorrhage, as already demonstrated 16 17 and on the other side disrupting the mechanism driving to the uncontrolled inflammatory overresponse by blunting KKS and complement activation?

By our, at this point, everyday experience and from the still scanty literature in this regard it seems to us that patients with severe pneumonias by SARS-CoV-2 can develop a hypercoagulable condition driving to a higher risk for (micro)vascular thrombosis, still to demonstrate 18 19 20 21 ; if true, considering the interface role of fXI in vascular thrombosis and inflammation, the development of therapies inhibiting this factor could be an intriguing leverage on which to act for trying to improve microvascular perfusion, reduce inflammation, and protect organ function during severe infections.