Venous limb gangrene and pulseless electrical activity (PEA) cardiac arrest during management of deep‐vein thrombosis and progressive limb ischemic necrosis following vascular surgery

A 55-year-old male former smoker was admitted urgently to hospital by his vascular surgeon for further investigation and treatment of an ischemic right lower limb. The patient had undergone 34 days earlier right femoral-posterior tibial bypass surgery using the greater saphenous vein as a graft. On postoperative day (POD) 1, revision surgery was required because of early graft occlusion. He was discharged to home on POD4 on oral aspirin (81 mg/day). At the first postoperative visit (POD20), the patient had a painful, discolored right foot; arterial pulses were intact, but venous ultrasound revealed right soleal/peroneal deep-vein thrombosis (DVT). Rivaroxaban, 15 mg twice-daily orally, was started, and aspirin continued. Despite this treatment, 2 weeks later (POD34) the patient had ongoing pain with progression to foot necrosis (Figure 1), prompting vascular surgeon reassessment and hospitalization. The patient stated he was compliant with taking rivaroxaban, including specific mention of taking the medication with food, as recommended. Admission hemoglobin was 12.3 g/dL, white blood count 14.0 × 109/L (neutrophils, 10.6 × 109/L), and the platelet count was 1042 × 109/L. This patient has ischemic limb necrosis in the setting of recently diagnosed DVT that occurred within the first few weeks following arterial bypass graft surgery performed for peripheral arterial occlusive disease. There are several unusual aspects to the case. For example, despite preceding arterial revascularization and clinically evident right foot necrosis, the arterial pulses are palpable. This raises the possibility of small artery disease, perhaps due to cholesterol embolism syndrome or localized recurrent graft-related arterial microcirculatory occlusion. However, as lower limb DVT was recently diagnosed in the same limb, another possible diagnosis is venous limb gangrene, that is, distal ischemic necrosis secondary to microvascular thrombosis in a limb with acute DVT.1 However, venous limb gangrene usually occurs in the setting of disseminated intravascular coagulation (DIC) secondary to heparin-induced thrombocytopenia (HIT)2 or metastatic adenocarcinoma.3 Moreover, the patient's platelet count was unexpectedly high (1042 × 109/L). Although the platelet count typically increases post-surgery, a phenomenon called postoperative thrombocytosis, peak platelet counts are usually observed by approximately day 14, with return to baseline (preoperative) platelet count values by about day 30.4 Thus, the patient's greatly elevated platelet count on POD34 is striking and requires explanation. The patient underwent CT angiography, which showed patency of the arterial bypass and the distal arteries. However, there was an incidental finding of a new left popliteal DVT that was asymptomatic. Repeat right leg venous duplex ultrasound showed unchanged soleal/peroneal DVT, but there was new right distal posterior tibial vein thrombosis. Although absence of large artery occlusion on imaging does not completely exclude the possibility of distal arterial (microcirculatory) occlusion, the demonstration of new DVT in the ischemic limb continues to suggest venous limb gangrene. Another puzzling issue is the additional finding of a new contralateral limb DVT, suggesting an underlying prothrombotic state refractory to treatment with rivaroxaban and aspirin. Moreover, the thrombocytosis remains unexplained. Review of previous medical records showed that his preoperative platelet count, obtained 49 days prior to his recent vascular surgery, was 716 × 109/L. Further, he had been diagnosed 6 months earlier as having a Jak2-positive myeloproliferative neoplasm (MPN). This was based on persistent mild elevation of hemoglobin and neutrophils, and more marked elevation of platelet counts, for at least 2 years, and with 35% of peripheral blood cells possessing the Jak2 mutation. The patient was lost to follow-up, and no treatment for his MPN had been initiated. Review of the postoperative platelet counts showed the expected early postoperative platelet count decline (to a nadir value of 567 × 109/L measured soon after the revision arterial surgery) with subsequent platelet count increase to 636 the following day. No further postoperative platelet count values were available prior to discharge on POD4 or at any time prior to POD34. The documentation of a Jak2-positive MPN in this patient provides a plausible explanation for the markedly elevated platelet count value on POD34. Furthermore, the presence of an untreated ET may also explain the early postoperative graft occlusion, requiring revision surgery on POD1, as well as the possible subsequent occurrence of microthrombosis refractory to rivaroxaban/aspirin treatment in that same limb. The thrombotic complication rate following surgery in the setting of MPNs is at least 5-fold increased, despite cytoreductive, antiplatelet or anticoagulant drugs.5 For example, ET predisposes mainly to postoperative arterial thrombosis, especially if baseline cardiovascular risk factors are present.5 The risk is highest over the first month post-surgery, with cumulative incidence of approximately 5% for death or thromboembolism.5 However, it seems less clear whether untreated MPN and thrombocytosis is a suitable explanation for the patient's venous thrombosis manifesting as bilateral DVT (right soleal/peroneal/posterior tibial veins; left popliteal vein). The patient received intravenous heparin during the initial vascular surgery (16 000 U) as well as for the revision (13 000 U). He also received approximately 72 hours of post-surgery therapeutic-dose heparin infusion. A clinical "pearl" is that one should always consider a diagnosis of immune HIT in a patient who develops venous or arterial thrombosis 5 to 30 days following exposure to heparin, particularly when the heparin is given intra-/perioperatively (eg, for cardiac or vascular surgery) or when given for postoperative thromboprophylaxis.6-8 This is because HIT is strongly associated with thrombosis (12 to 15-fold increase in thrombosis vs controls),9 which can involve arteries, veins, or even the microcirculation. Moreover, HIT can present several days or even weeks after discontinuation of heparin because of the presence of HIT antibodies that possess both heparin-dependent as well as heparin-independent platelet-activating properties.7, 10 This disorder has been called "delayed-onset HIT,"7, 8 and is now classified as one of the "autoimmune HIT" (aHIT) disorders.10 In this patient, a platelet count of 1042 × 109/L measured on POD34 might seem to argue against an episode of recent postoperative HIT, although the absence of available platelet counts during the preceding month means that HIT cannot be ruled out. A potential diagnosis of HIT and HIT-associated limb ischemia was not considered. Rather, the patient was considered to have right foot ischemic necrosis secondary to (non-HIT-related) small artery thrombosis/embolization with concomitant DVT refractory to rivaroxaban/aspirin as a consequence of untreated MPN and recent arterial revascularization surgery in the same limb. Rivaroxaban was discontinued, and intravenous unfractionated heparin was started, with an initial bolus of 80 units (U)/kg (9100 U for 114 kg patient weight), followed by heparin infusion at 2050 U/hour. Within 25 minutes of receiving the heparin bolus, the patient developed a pulseless electrical activity (PEA) cardiac arrest. After resuscitation with epinephrine (total, 4 mg) and alteplase thrombolysis (total, 100 mg), he was admitted to an intensive care unit. Angiography post-resuscitation showed pulmonary embolism, but the clot burden (right lower lobe segmental embolism; left upper-lobe and left lower-lobe subsegmental embolism) was not thought to be sufficient to explain the PEA cardiac arrest. Moreover, an echocardiogram revealed only a moderately hypokinetic right ventricle. The immediate post-cardiac arrest CBC showed a "normal" platelet count of 163 × 109/L, with a subsequent rebound in the platelet count 3 hours later to 593 x 10^9^/L. Heparin was continued post-PEA arrest, but over the next 56 hours, the platelet count fell once again, reaching a nadir value of 148. At this time, the diagnosis of HIT was first suspected. The occurrence of a cardiac or respiratory arrest between 5 and 30 minutes after receipt of a heparin bolus is one of the potential clinical manifestations of "rapid-onset HIT".11, 12 Known as HIT-associated "anaphylactoid reaction," patients can develop a wide range of symptoms and signs, ranging from relatively benign (fever, chills/rigors, flushing, headache, tachycardia, hypertension) to severe (dyspnea/chest constriction, cardiopulmonary arrest). As seen here, a large magnitude platelet count fall is typically observed in post-heparin bolus HIT-associated anaphylactoid reaction, reflecting abrupt heparin-induced platelet-activating effects of high HIT antibody levels resulting from recent exposure to heparin. Moreover, the clinical picture of PEA cardiac arrest as a manifestation of HIT-associated anaphylactoid reaction has been reported.13 The concept of post-heparin bolus cardiac arrest as a manifestation of HIT was proposed by Ansell and colleagues in 198614; these workers noted that the extent and magnitude of pulmonary embolism seemed insufficient to explain their patients' cardiopulmonary demise, similar to what was observed in our patient. Nevertheless, these reactions can closely mimic pulmonary embolism, prompting one group to refer to them as "pseudo-pulmonary embolism".15 A rapid test for HIT antibodies, the "latex immunoturbidimetric assay," or LIA, was ordered. The result was positive, at 2.6 U/mL (reference range, <1.0 U/mL). The patient was switched from heparin to argatroban anticoagulation. The LIA is a rapid automated immunoassay that detects HIT antibodies based upon the principle of competition inhibition.16 In essence, a positive test result indicates the patient's plasma contains antibodies (IgG, IgA, and/or IgM) that are able to inhibit agglutination of nanoparticles bearing a HIT-mimicking monoclonal antibody in the presence of added HIT antigen (complexes of platelet factor 4 [PF4]/polyvinyl sulfonate [PVS]). Evaluation of the LIA by the McMaster Platelet Immunology Laboratory showed that this rapid assay has approximately 95% sensitivity and 95% specificity for a diagnosis of HIT.16 However, a result of 2.6 U/mL is a "weak" positive result; in the McMaster study, such a result was associated with a confirmed diagnosis of HIT in approximately 30% of cases. Given the patient's atypical clinical features, further investigations for HIT, including assessing patient blood for presence of platelet-activating HIT antibodies by serotonin-release assay (SRA), a more specific test for HIT,17 are warranted. The patient's condition did not improve after 5 days of intensive care management. He remained comatose, acidotic, with a "septic" clinical picture, and was pronounced dead at this time. No autopsy was performed, per family wishes. Unfortunately, the plasma sample that tested positive by LIA had been discarded. Although it was not possible to obtain any additional blood work, previously acquired heparin-anticoagulated plasma (green-top chemistry tube) was still available, and the residual plasma was sent to the McMaster Platelet Immunology Laboratory. This plasma sample was "deheparinized" using Ecteola cellulose,18 and tested by three PF4-dependent immunoassays, as well as the SRA. All four tests were positive, including the McMaster in-house IgG-specific PF4/heparin-EIA (EIA-G) (1.83 OD units; reference range, >0.45), as well as the SRA (serotonin-release, 82% and 81% at 0.1 and 0.3 U/mL, respectively [reference range, <20%]; with 0% serotonin-release seen at 100 U/mL heparin). Moreover, the patient's plasma also yielded strong serotonin-release (86% release) at 0 U/mL heparin ("buffer control"). Figure 2 summarizes the clinical course and key laboratory test results. The strong positive PF4-dependent EIAs and SRA are diagnostic of HIT in this clinical context of post-heparin bolus HIT-associated anaphylactoid reaction with PEA cardiac arrest. The demonstration of strong plasma-induced serotonin-release at buffer control (ie, heparin not present in the reaction well) raises the possibility of a recent episode of aHIT. As mentioned earlier, aHIT indicates a subset of HIT in which patients exhibit atypical clinical features, such as delayed-onset HIT (HIT that begins or progresses after stopping heparin), persisting (or refractory) HIT (thrombocytopenia that continues more than 1 week after stopping heparin), heparin "flush" HIT (HIT triggered only by exposure to small quantities of heparin), and "spontaneous HIT syndrome" (a disorder clinically and serologically indistinguishable from HIT but which is not explained by preceding exposure to heparin) 10. Further potential manifestations of aHIT include severe thrombocytopenia (platelet count <20 × 109/L) with overt decompensated DIC, or HIT complicated by multiple venous and arterial thrombotic events. The laboratory marker of aHIT is the presence of HIT antibodies that have both heparin-dependent as well as heparin-independent platelet-activating antibodies.10 The latter group of antibodies is shown most clearly by strong patient serum- (or plasma-induced) platelet activation even in the absence of heparin. For example, our patient's plasma exhibited 86% serotonin-release at buffer control, which is the serological hallmark of aHIT antibodies. One way to demonstrate concomitant heparin-dependent platelet-activating properties is to repeat the SRA with serial plasma/serum dilutions; in general, diluted patient serum/plasma will exhibit heparin-dependent platelet-activating properties. This was shown in a repeat SRA performed using neat, 1:2, and 1:4 diluted plasma (see Figure 3). As there were no available platelet count measurements between POD5 and POD30, we cannot exclude the intriguing possibility that our patient may have had a recent episode of postoperative thrombocytopenia pointing to the aHIT variant known as delayed-onset HIT. For example, the patient could have developed a period of (relative) thrombocytopenia between POD5 to POD21 (the usual period in which thrombocytopenia manifests in delayed-onset HIT7, 8) which subsequently recovered to 1042 × 109/L, as documented on POD34 (see dashed lines corresponding to "Theory #1" in Figure 2). Occurrence of delayed-onset HIT would plausibly explain why the patient developed postoperative (HIT-associated) DVT with venous limb gangrene. Given the generally favorable clinical outcomes reported using rivaroxaban and other direct oral anticoagulants for management of HIT,19 perhaps concurrence of delayed-onset HIT with untreated MPN could offer an explanation for the unusual picture of progressive limb necrosis despite rivaroxaban/aspirin—potentially reflecting either small artery/arteriole thrombosis/embolization and/or DVT-associated venous limb gangrene. Alternatively, the patient might never have had "thrombocytopenia" during the period from POD5 to POD30 ("Theory #2). However, the patient could still have had strong HIT antibodies that became clinically significant only when he received the 9100 U bolus of UFH, thus triggering the ultimately fatal acute anaphylactoid reaction. Although either scenario is plausible, the more likely possibility is occurrence of delayed-onset HIT with platelet count recovery by POD34, as this would help explain the occurrence of bilateral DVT (symptomatic in the right lower limb), the progression to limb necrosis in association with presumed microvascular thrombosis, as well as the serological positivity of HIT antibodies with heparin-independent platelet-activating properties. Another interesting feature of this patient case is the occurrence of HIT in a patient with underlying MPN. In recent years, several reports have suggested MPNs to be a risk factor for developing HIT.20-24 Bovet and colleagues found a tenfold increased incidence of HIT in ET.22 In addition, presence of JAK2 mutation in ET further increases the risk for HIT, which occurred in 17% (10/58) of ET patients treated with heparin in one retrospective study.24 Perhaps the elevated platelet counts in ET, contributing to greater amounts of platelet-derived PF4, predispose patients to forming higher quantities of immunogenic PF4/heparin complexes during heparin treatment. We believe our patient developed HIT based upon both clinical and serological perspectives. The 4Ts score25 was eventually 8 points (maximum) based upon: Thrombocytopenia (2 points for 84% platelet count falling to a nadir of 163 × 109/L) following a bolus of UFH; appropriate Timing (rapid onset of thrombocytopenia within 1 day with the patient previously exposed to a 4-day course of heparin 30 to 34 days earlier); Thrombosis (2 points for right soleal/peroneal/tibial vein and left popliteal vein DVTs with component of right lower limb microvascular thrombosis; post-heparin bolus anaphylactoid reaction with PEA cardiac arrest); and no apparent o Ther diagnosis (2 points) that plausibly explains these aforementioned clinical and laboratory features. Moreover, the patient tested strongly positive in four different PF4-dependent immunoassays (LIA; PF4/PVS IgG-specific EIA; PF4/PVS polyspecific EIA; and in-house PF4/heparin IgG-specific EIA), as well as strongly positive in the SRA, a sensitive and specific test for HIT. Our case illustrates several intriguing features of HIT, including its putative association with MPN, as well as the importance of obtaining a platelet count whenever preceding exposure to heparin is complicated by the development of venous or arterial thrombosis 5 to 30 days later. The strong possibility of an unrecognized postoperative period of thrombocytopenia, as a consequence of delayed-onset form of aHIT, indicates a likely missed diagnosis of HIT, underscoring the need to obtain a platelet count when evaluating thrombotic complications in the postoperative period. T. E. Warkentin has received lecture honoraria from Alexion; has provided consulting services to Aspen Global, Bayer, CSL Behring, Instrumentation Laboratory, has received research funding from Instrumentation Laboratory, and has provided expert witness testimony relating to HIT and non-HIT thrombocytopenic and coagulopathic disorders. A. M. Hirsch has given lectures on behalf of Bayer. The other authors declare no relevant conflicts of interest.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.