A Protean Protein

Thrombocytopenia that develops between the fifth and tenth day following heparin exposure in a patient with new thromboses is consistent with HIT. However, the patient’s infrarenal aortic thrombus preceded the initiation of heparin, and negative functional testing undermines the diagnosis of HIT in this case. Therefore, the arterial thromboses may be related to an underlying unifying diagnosis.

A third SPEP showed a 0.1 g/dL M-spike in the gamma region, but standard immunofixation did not reveal a monoclonal protein (Figure 2). However, a specific request for immunofixation testing using IgD antisera detected an IgD heavy chain. A lambda chain comprising 3% of urine protein was detected on 24-hour urine immunofixation but was not detectable by serum immunofixation. Bone marrow biopsy demonstrated plasma cells comprising 5% of bone marrow cellularity (Figure 3); flow cytometry of the aspirate demonstrated an abnormal lambda-restricted plasma cell population.

When a monoclonal protein is identified but does not react with standard antisera to detect IgG, IgM, and IgA, immunofixation with IgD and IgE antisera are necessary to rule out a monoclonal IgD or IgE protein. The underlying IgD isotype coupled with its low abundance made detection of this monoclonal protein especially challenging. With the discovery of a monoclonal protein in the context of polyneuropathy, the mandatory criteria of POEMS syndrome are met. The elevated VEGF level and hypothyroidism meet major and minor criteria, respectively. Arterial thromboses and heart failure are other features that may be observed in cases of POEMS syndrome.

POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, M protein, and skin changes) was diagnosed. Prednisone was continued, and weekly cyclophosphamide was initiated. After six weeks, the VEGF level remained elevated, and a neurologic examination showed minimal improvement. Due to poor respiratory muscle strength and difficulty managing secretions, he underwent percutaneous tracheostomy and gastrostomy tube placement. Unfortunately, his condition further deteriorated and he subsequently died of sepsis from pneumonia.

An autopsy revealed acute bronchopneumonia and multiple acute and subacute cerebral infarctions. There was extensive peripheral mixed axonal/demyelinating neuropathy, hepatosplenomegaly, atrophy of the thyroid and adrenal glands, hyperpigmented patches and thickened integument, and severe aortic and coronary atherosclerotic disease with a healed myocardial infarction.

POEMS syndrome¹ is a rare constellation of clinical and laboratory findings resulting from an underlying plasma cell proliferative disorder. This paraneoplastic syndrome is characterized by the chronic overproduction of proinflammatory and proangiogenic cytokines, including VEGF, which are postulated to drive its manifestations,² though the exact pathogenesis is not understood. Some of the disease’s most common features are summarized by its name: polyneuropathy, organomegaly, endocrinopathy, monoclonal plasma cell disorder, and skin changes.³

The International Myeloma Working Group (IMWG) diagnostic criteria¹ (Table) require the presence of both mandatory criteria (polyneuropathy and monoclonal plasma cell proliferation), plus at least one major and one minor criterion. Delayed diagnosis or misdiagnosis of this protean disorder is often driven by its rarity and clinical overlap with other paraprotein-associated polyneuropathies. These include amyloidosis, cryoglobulinemia, and monoclonal gammopathy of undetermined significance (MGUS), which can all produce antibodies directed against neural antigens. In addition, polyneuropathy is often the first and most striking manifestation of POEMS syndrome, fostering confusion with CIDP as both disorders are subacute, symmetric, motor-dominant, mixed axonal/demyelinating polyneuropathies.⁴

IgD and IgE monoclonal gammopathies are extremely rare. IgD myeloma, for instance, accounts for 2% of multiple myeloma cases, and IgE myeloma has been reported fewer than 50 times.⁵ IgD is secreted only in very small amounts, ordinarily representing 0.25% of the immunoglobulins in serum, while the majority is found in the plasma membranes of mature B-cells.⁶ These monoclonal gammopathies often escape detection for two reasons: (1) the very low paraprotein concentration produces undetectable or small M-protein levels on electrophoresis,⁵ and (2) immunofixation is routinely performed without antisera against IgD and IgE heavy chains.⁷

While this case depicts a rare manifestation of a rare disease, the principles underlying its elusive diagnosis are routinely encountered. Recognition of the specific limitations of the SPEP, UPEP, sFLC, and immunofixation tests, outlined below, can assist the hospitalist when suspicion for paraproteinemia is high.

First, low levels of monoclonal proteins may be associated with a normal SPEP. Accordingly, suspicion of a plasma cell dyscrasia should prompt serum immunofixation, even when the electrophoretic pattern appears normal.⁸

Second, laboratories routinely perform immunofixation with antisera against IgG, IgA, and IgM heavy chains and kappa and lambda light chains, whereas testing with IgD or IgE antisera must be specifically requested. Thus, clinicians should screen for the presence of IgD and IgE in patients with an apparently free monoclonal immunoglobulin light chain in the serum or with a monoclonal serum protein and negative immunofixation. In this case, the paraprotein was not detected on the first two serum electrophoreses, likely due to a low serum concentration, then missed on immunofixation due to a lack of IgD antiserum. On admission to the hospital, this patient had a very low paraprotein concentration (0.1 g/dL) on SPEP, and the lab initially reported negative immunofixation. When asked to test specifically for IgD and IgE, the lab ran a more comprehensive immunofixation revealing IgD heavy chain paraprotein.

Third, this case illustrates the limitations of the sFLC assay. IMWG guidelines specify that sFLC assay in combination with SPEP and serum immunofixation is sufficient to screen for monoclonal plasma cell proliferative disorders other than light chain amyloidosis (which requires all the serum tests as well as 24-hour urine immunofixation).⁹ Though the sFLC assay has been demonstrated to be more sensitive than urine analysis for detecting monoclonal free light chains,¹⁰ it is still subject to false negatives. Polyclonal gammopathy or reduced renal clearance with accumulation of free light chains in the serum may mask the presence of low levels of monoclonal sFLC,¹¹ the latter of which likely explains why the sFLC ratio was repeatedly normal in this case. In these circumstances, monoclonal free light chains can be identified by urine studies.¹¹ In this case, 24-hour urine immunofixation detected the excess light chain that was not evident on the sFLC assay. Even with these pitfalls in mind, there is still no evident explanation as to why the 24-hour urine studies done prior to the patient’s hospital admission did not reveal a monoclonal light chain.

This case also highlights the thrombotic diathesis in POEMS syndrome. Although the patient was treated with argatroban for suspected HIT, it is likely that the HIT antibody result was a false positive, and his thrombi were better explained by POEMS syndrome in and of itself. Coronary, limb, and cerebral artery thromboses have been linked to POEMS syndrome,^12,13 all of which were present in this case. Laboratory testing for HIT involves an immunoassay to detect circulating HIT antibody and a functional assay to measure platelet activity in the presence of patient serum and heparin. The immunoassay binds anti-PF4/heparin complex irrespective of its ability to activate platelets. The presence of nonspecific antibodies may lead to cross-reactions with the immunoassay test components, which has been demonstrated in cases of MGUS.¹⁴ In this case, elevated production of monoclonal antibodies by plasma cells may have led to false-positive results. With moderate to high clinical suspicion of HIT, the combination of a positive immunoassay and negative functional assay (as in this case) make the diagnosis of HIT indeterminate.¹⁵