Severe pancytopenia a clue to Waldenström’s macroglobulinemia
A clinical scenario that often presents to a hematologist, and perhaps equally often with a broad differential diagnosis, is that of pancytopenia. An instructive case was recently encountered that helps to illustrate the need for a detailed evaluation in such a patient.
The patient was a 61-year-old woman, a food caterer, who was in her usual state of health until 2 weeks before presentation.
For 2 weeks, she had a marked increase in fatigue, causing her to stop during daily routine tasks to rest. This profound fatigue was not associated with any fever or chills, but the patient occasionally had night sweats. There was no noticeable change in her weight. She reported no sick contacts or recent travel. She had been working out in the woods and in her garden but did not recall any tick bites or other exposures.
About 5 months before the presentation, she was also evaluated for pale white and painful fingers, attributed to a possible Raynaud’s phenomenon. Her workup at the time also uncovered a positive antinuclear antibody titer, as well as positive serologies associated with Sjögren’s syndrome. She did not have any rashes then or at the current presentation.
She reported no bruising or bleeding symptoms at home. She had not had any melena, hematochezia, hematuria or hemoptysis. She recalled being told when she was pregnant that she was temporarily anemic, but she never required any treatment or blood transfusion. Otherwise, there was no history in the patient or her family of any blood disorders.
Her past medical history was remarkable for lobular carcinoma in situ of the right breast in 1998. She was treated for this, and subsequent mammograms through the current year have been fine.
Her history also included hypertension and hyperlipidemia. This symptomatology led to her presenting to the ED. Evaluation there revealed a total white blood cell count of 3,300 per mcL, hemoglobin of 4.8 gm/dL, mean corpuscular volume of 85, red cell distribution width of 13.1% and platelet count of 7,000 per mcL. The differential of the white blood cell count showed 13% neutrophils, 1% bands, 67% lymphocytes, 8% monocytes, 1% eosinophils and 10% atypical lymphocytes. Inspection of the peripheral blood film demonstrated large lymphocytes with irregularly shaped nuclei, suggestive of a reactive lymphocytosis. The N:C ratio was not markedly increased, and no nucleoli were observed. The serum creatinine, liver function tests (including bilirubin) and electrolytes were within normal limits.
Upon admission to the hospital, the patient was transfused 1 U of single-donor platelets. Her platelet count increased from 7,000 to 58,000 per mcL. A hematology consultation was then requested. Exam of the patient did not reveal any lymphadenopathy or splenomegaly.
Given her recent outdoor work, as well as having encountered many individuals in the past few weeks as part of her work, it was suspected that she may have a viral-mediated pancytopenia. The atypical lymphocytes observed, besides the high lymphocyte fraction in the differential, seemed consistent with this possibility. Laboratory investigations at that point revealed negative studies for hepatitis and HIV. Testing was also negative for parvovirus B19, Epstein-Barr virus immunoglobulin M and cytomegalovirus IgM. Vitamin B12 level was 405 pg/mL.
Coagulation parameters showed a prothrombin time of 10.8 seconds and a partial thromboplastin time of 21.7 seconds, both of which are within normal limits at our laboratory. D-dimer testing was positive at 6.66 mg/L, with a fibrinogen of 232.7 mg/dL. Schistocytes were not observed on the peripheral blood film, arguing against significant disseminated intravascular coagulation. Along with a normal total bilirubin, the lactate dehydrogenase level was within normal limits at 197 units/L, ruling out hemolysis.
At this point, a viral etiology appeared less likely, although still possible, given the atypical lymphocytosis. However, as an atypical lymphocyte cannot always be conclusively distinguished morphologically from a clonal or malignant process, further investigations were pursued.
Given the prior positive autoimmune serologies from earlier in the year, a repeat antinuclear antibody was low-positive, at 1:160, with a speckled pattern. Other autoimmune analyses, including anti-double stranded DNA, anti-Smith, anti-Ro, and anti-La, were all negative. These results, coupled with the lack of other clinical symptoms, indicated a systemic autoimmune disease was less likely.
The patient was being supported with blood transfusions, as needed, and with a stable hemoglobin value after transfusion, but primary hematologic disease appeared much higher in the differential diagnosis. Serum protein electrophoresis and urine protein electrophoresis with immunofixation were ordered, along with flow cytometry from the peripheral blood.
The serum protein electrophoresis and immunofixation demonstrated the presence of an IgM-kappa monoclonal protein, at a quantity of 2.03 g/dL. Urine protein electrophoresis also confirmed the kappa light chain monoclonal protein. Beta-2 microglobulin was modestly increased at 2.7 mg/L. The presence of an IgM monoclonal protein narrowed the differential substantially. Besides Waldenström’s macroglobulinemia, diagnosis possibilities included IgM-myeloma, IgM monoclonal gammopathy of undetermined significance, chronic lymphocytic leukemia and mantle cell lymphoma.
The flow cytometric analysis confirmed a monoclonal population of B cells that were negative for CD5, CD10 and CD38 while being positive for CD20, CD22 and CD23. Therefore, in concert with the IgM monoclonal protein, this highly suggested the diagnosis of Waldenström’s macroglobulinemia. Retinal exam uncovered no abnormalities, such as hemorrhages or dilated veins. CT scanning of the chest, abdomen and pelvis demonstrated no lymphadenopathy. No lytic lesions were identified.
Next, a bone marrow aspiration and biopsy were performed. The flow cytometry was consistent with that of the peripheral blood. Cytogenetic studies found trisomy 3 in two of 20 cells. The biopsy specimen had a 90% cellularity, with reduced trilineage hematopoiesis, along with moderate reticulin fibrosis. The infiltrate was morphologically that of lymphoplasmacytic lymphoma.
Central to the diagnosis of Waldenström’s macroglobulinemia is the demonstration of a lymphoplasmacytic cell population, along with the identification of an IgM monoclonal protein in the blood. Often, patients can have hyperviscosity, and the physical examination should include an assessment of this (eg, a retinal exam).
In this case, severe pancytopenia combined with the IgM monoclonal protein were key clues to the diagnosis. With this information, the atypical lymphocytes were likely malignant lymphocytes. Additionally, recognition of a CD5-negative process by flow cytometry also helped narrow the differential diagnosis. The more common hematological entities that comprise this subset, other than lymphoplasmacytic lymphoma, are hairy cell leukemia, marginal zone lymphoma and prolymphocytic leukemia. Generally, clinical and laboratory findings can readily differentiate the specific CD5-negative process.
In terms of treatment, one approach is to consider whether a patient is either asymptomatic, symptomatic but not requiring immediate therapy, and those who are symptomatic and require immediate treatment. There are multiple indications for treatment, as summarized in Table 1.
Patients who present with symptomatic hyperviscosity, cryoglobulinemia and moderate or severe cytopenias are among those who typically require immediate disease control. In such a situation, the aim of treatment is to achieve rapid reduction of the monoclonal protein. For those with symptomatic hyperviscosity, one option is plasmapheresis because 80% of the IgM protein is intravascular. Two to three sessions can reduce the serum IgM by up to 60%.
In our patient’s case, she had severe cytopenias, with replacement of most of the bone marrow’s normal hematopoietic elements. There is no FDA-approved therapy for Waldenström’s, partly because of the rarity of the condition and the lack of head-to-head drug trials. A large number of chemotherapeutic agents and regimens have shown efficacy, as shown in Table 2.
One of the most recent regimens that appears promising is the combination of bortezomib (Velcade, Millennium Pharmaceuticals), dexamethasone and rituximab. In a trial of first-line therapy, 23 patients received treatment with bortezomib 1.3 mg/m2 on days 1, 4, 8 and 11, dexamethasone 40 mg orally on days 1, 4, 8 and 11 and rituximab 375 mg/m2 on day 11. Patients received four consecutive cycles (one cycle=2 weeks), followed by four more cycles at 3-month intervals.
In this study, all patients achieved a 90% or greater reduction in the IgM protein. The overall response rate was 96%, and at a median follow-up of 22.8 months, 18 of 23 patients remained free of disease progression. Prophylactic antiviral therapy for herpes zoster was strongly recommended based on rates of infection without prophylaxis. Other than this regimen, treatment options such as rituximab with CHOP; rituximab with cyclophosphamide, vincristine and prednisone; and fludarabine with rituximab have all shown favorable results.
An important clinical point is to be aware of the well-described phenomenon of the rituximab-induced IgM “flare.” With rituximab monotherapy, the flare rate can be as high as 50%, resulting in symptomatic hyperviscosity and worsening of other IgM-related symptoms. Therefore, the serum IgM should be monitored closely during the initial phase of therapy.
Amit Mehta, MD, is an attending physician at Regional Cancer Care in Durham, N.C., and is a member of the HemOnc Today Editorial Board.
For more information:
- Ghobrial IM. Cancer. 2004;101:2593-2598.
- Leleu X. Cancer Lett. 2008;270:95-107.
- Treon SP. Blood. 2009;114:2375-2385.
- Treon SP. J Clin Oncol. 2009;27:3830-3835.