International Journal of Hematology

, Volume 95, Issue 3, pp 311–314

Generalized purpuric drug exanthem with hemorrhagic plaques following bendamustine chemotherapy in a patient with B-prolymphocytic leukemia


  • Aruna Gavini
    • Department of Hematology/OncologyRoger Williams Medical Center
  • Gladys H. Telang
    • Department of Dermatology, Rhode Island HospitalAlpert Medical School of Brown University
    • Division of Hematology/Oncology, Memorial Hospital of Rhode IslandAlpert Medical School of Brown University
Case Report

DOI: 10.1007/s12185-012-1012-2

Cite this article as:
Gavini, A., Telang, G.H. & Olszewski, A.J. Int J Hematol (2012) 95: 311. doi:10.1007/s12185-012-1012-2


A 72-year-old woman presented with dyspnea and lower extremity edema. Extreme lymphocytosis, cytopenia, and splenomegaly were found, and she was diagnosed with B-prolymphocytic leukemia. Following the first dose of therapy with bendamustine, the patient developed severe generalized maculopapular rash, which subsequently progressed to exuberant, non-blanching palpable purpura with hemorrhagic plaques suspicious for leukocytoclastic vasculitis. These events coincided with severe chemotherapy-induced neutropenia and thrombocytopenia, but there were no clinical symptoms of infection. Skin punch biopsy revealed perivascular and diffuse upper dermal lymphocytic infiltrate with eosinophils and marked erythrocyte extravasation consistent with a purpuric drug exanthem. The patient was treated with steroids, with complete resolution of the findings. This new form of cutaneous toxicity of bendamustine is presented along with a review of previous experience with the drug.


BendamustineDrug eruptionProlymphocytic leukemiaVasculitisChronic lymphoid leukemia


Management of B-prolymphocytic leukemia (B-PLL) remains challenging, with most patients achieving short remissions after standard chemotherapy. Treatment is associated with immunosuppression and cytopenia exacerbating baseline disease-related abnormalities. Cutaneous toxicities may further complicate the course and need to be carefully evaluated in the context of clinical circumstances. In this report, we describe a patient with a severe purpuric drug exanthema caused by bendamustine, complicated by large hemorrhagic plaques mimicking leukocytoclastic vasculitis.

Case presentation

A previously healthy 72-year-old woman, with no significant medical history or medication use, presented with escalating lower extremity edema, dyspnea on exertion, and weight loss. She denied fevers, night sweats, abdominal discomfort, or bowel movement abnormalities. Her examination was notable for massive splenomegaly and no lymphadenopathy or skin lesions. The initial workup demonstrated anemia and thrombocytopenia with extreme lymphocytosis (absolute lymphocyte count 1116 × 103/μL, neutrophil count 11.3 × 103/μL, hemoglobin 7.4 g/dL, platelet count 95 × 103/μL, LDH 378 IU/L). CT scan additionally showed mediastinal, retroperitoneal and iliac adenopathy with spleen spanning 26 cm in craniocaudal dimension. Blood smear contained abundant prolymphocytes and no smudge cells (Fig. 2a). Flow cytometry confirmed B-cell prolymphocytic leukemia with immunophenotype positive for CD19 and CD20, but negative for CD5, CD10, CD23, CD103, CD11c and CD25. Additional cytogenetic and fluorescence in situ hybridization (FISH) tests revealed deletion of 17p chromosome in all malignant lymphocytes, while translocation t(11;14) was ruled out.

Due to the heavy burden of disease, the patient was treated initially with a cytoreductive course of cyclophosphamide and dexamethasone, which yielded no significant improvement in the blood cell counts. We subsequently proceeded to therapy with bendamustine 120 mg/m2 on two consecutive days, with a plan to add rituximab during subsequent cycles. On day 4 of the first cycle, the patient developed maculopapular rash over her trunk and extremities. It was initially ascribed to prophylactic allopurinol, which was discontinued. By day 16, the eruption rapidly worsened with extension onto the entire body including palmoplantar regions. Large plaques of palpable purpura developed as well as bullae and ulcerations on her edematous lower extremities (Fig. 1). The patient denied any pain, fever, pruritus or other symptoms. Her vital signs were normal except for sinus tachycardia (blood pressure of 110/68 mmHg, heart rate 110/min, temperature 36.7°C), and there were no signs of infection. There were no ocular, oral, or genital lesions. However, laboratory workup demonstrated severe chemotherapy-induced neutropenia with absolute neutrophil count of 0 per μL, eosinophilia with absolute number of 14.6 × 103/μL, platelet count of 19 × 103/μL, normal renal function, mild elevation of hepatic transaminases (aspartate aminotransferase 45 IU/L, alanine aminotransferase 39 IU/L) and no evidence of disseminated intravascular coagulation. Tests for antinuclear and anti-neutrophilic cytoplasmic antibodies (ANCA), cryoglobulin, and rheumatoid factor were negative. The leukemic count remained very high (873 × 103/μL), indicating poor response to the alkylating agent. The clinical diagnosis was one of leukocytoclastic hypersensitivity vasculitis, since the scenario met four out of five 1990 American College of Rheumatology Criteria for the condition [1].
Fig. 1

Skin examination revealing generalized and confluent purpura of the lower extremities, palpable purpuric lesions on upper back and hemorrhagic plaques with rare bullae on the forearms

A punch biopsy of the skin from the left shoulder was performed (Fig. 2). It revealed no typical features of leukocytoclastic vasculitis (such as granulocytic infiltrate, fibrinoid necrosis of the vessel walls etc.), but rather a dense upper dermal lymphocytic infiltrate, both scattered and perivascular, comprised of small CD20-positive, CD5 and CD23-negative lymphocytes with occasional eosinophils. This immunophenotype was identical to the leukemic infiltrate present in the peripheral blood. There were no neutrophils identified and the periodic acid-Schiff (PAS) stain was negative for fungal organisms. Gram stain and bacterial culture showed scant skin flora including methicillin-sensitive Staphylococcus aureus. The final pathologic diagnosis was one of severe purpuric drug exanthem with prominent dermal hemorrhage in a setting of thrombocytopenia and infiltration with malignant lymphocytes.
Fig. 2

a Peripheral blood smear with numerous prolymphocytes (×100). b, c Skin biopsy reveals dermal malignant lymphocytic infiltrate with eosinophils (arrows) and marked erythrocyte extravasation (b ×10, c ×40). d Immunohistochemistry highlights CD20-positive prolymphocytic dermal infiltrate

The patient was treated with oral pulse dexamethasone 40 mg daily for 4 days. The rash resolved gradually over 2 weeks with no residual hyperpigmentation. Following recovery of the neutropenia, the oncologic treatment was resumed with monoclonal antibody ofatumumab. During the therapy, the patient was rechallenged with allopurinol with no recurrent rash.


B-cell prolymphocytic leukemia (B-PLL) was initially described as a rare variant of chronic lymphoid leukemia (CLL). Currently it is recognized as a distinct disorder of mature B lymphocytes affecting the elderly [2], characterized by high peripheral blood leukemic counts, typically over 100 × 103/μL, as well as splenomegaly without significant lymphadenopathy. It is traditionally defined by the presence of 55% or more prolymphocytes on the blood smear. In contrast to CLL cells, B-prolymphocytes strongly express surface IgM immunoglobulin as well as pan-B-cell antigens such as CD19 and CD20/FMC-7, while CD5 and CD23 are typically absent, indicating later stages of maturation. Molecular studies are useful for differentiation from other mature B-cell leukemias. Previous studies reported high incidence of abnormalities in the chromosome 14, although many of those cases are now classified as variants of mantle cell lymphoma [3]. TP53 mutations in the chromosome 17p locus have been demonstrated in 50-75% of patients, a proportion higher than in any other lymphoproliferative disorder [4]. This finding underscores difficulties in treatment, with relatively low response rates and short duration of remissions after standard therapies using purine analogues or alkylating agents, akin to high-risk CLL with 17p deletion [5].

Bendamustine is a novel agent, synthesized with the intent of combining the alkylating properties of mechlorethamine and the purine antimetabolite properties of benzimidazole [6]. In the pivotal phase III trial, bendamustine demonstrated improved efficacy in front-line therapy of advanced CLL over chlorambucil with manageable, primarily hematologic toxicity [7]. Preclinical data suggest that bendamustine may induce cytotoxic effects independent of the presence TP53 mutation, although this finding awaits clinical confirmation [8]. Based on current data, for patients with ultra-high risk lymphoid leukemia harboring the 17p deletion, treatment with experimental agents or monoclonal antibodies is favored over traditional chemotherapy [9].

Cutaneous eruptions have been reported in 9–16% of patients with CLL and non-Hodgkin lymphoma treated with bendamustine in clinical trials. However, the incidence of grade 3 rash (by the National Cancer Institute Common Terminology Criteria for Adverse Events) was below 3% and there was no grade 4 hypersensitivity noted. In the post-marketing experience, cases of toxic epidermal necrolysis and Stevens–Johnson syndrome occurred. Their relationship to the drug was unclear due to concomitant use of allopurinol or rituximab, agents previously implicated with those idiosyncratic reactions [10]. In our case, the etiologic link between the severe eruption and bendamustine was strengthened by no concurrent administration of rituximab and subsequent retreatment with allopurinol. Cutaneous drug eruptions may have divergent histopathology. Exanthematous drug eruptions which account for the vast majority of drug eruptions, such as in this case, are characterized by predominantly perivascular inflammation with or without eosinophils and interface dermatitis Other classifications include nodular and diffuse dermatitis, vesiculobullous drug eruptions, vasculitis, pustular, follicular, sclerodermoid and panniculitis forms [11, 12]. The only previous report of severe skin reaction to bendamustine involved a pruritic, desquamating rash with vacuolar-interface dermatitis, dyskeratotic keratinocytes, and lymphocyte exocytosis [13]. In contrast to that reference, our case exhibited no features of desquamating interface dermatitis, but rather a dense perivascular lymphocytic infiltrate with extensive intradermal hemorrhage and systemic eosinophilia. The infiltrate consisted of extravasated leukemic cells rather than inflammatory polyclonal population. This histology can be best characterized as a drug exanthem of the superficial perivascular and interstitial type without interface changes, a form of Gell-Coombs type 4 hypersensitivity reaction [14]. Conversely, hypersensitivity vasculitis (cutaneous necrotizing vasculitis) is a small-vessel process characterized by leukocytoclastic histology and its pathophysiology is related to immune complexes (Gell-Coombs type 3) [15]. “Lymphocytic vasculitis” is a controversial subcategory, occasionally seen in drug hypersensitivity [16]. Severe forms may present as drug-induced hypersensitivity syndrome or drug rash with eosinophilia and systemic symptoms (DRESS) [17]. In case of chemotherapeutic agents such as bendamustine, the picture may be obscured by the presence of thrombocytopenia and neutropenia predisposing to hemorrhage and infection, while generally exerting immunosuppressive effect.

In summary, severe drug eruption may rarely be provoked by treatment with bendamustine. The clinical differential diagnosis of purpura fulminans should be considered in severely neutropenic patients. Concomitant thrombocytopenia may lead to accentuated dermal hemorrhage with plaques clinically mimicking leukocytoclastic vasculitis. This case underscores the need for a comprehensive assessment and communication between the hematologist, pathologist, and dermatologist in order to establish the correct diagnosis, taking into consideration the underlying disease process, typical hematologic effects of chemotherapy and histology.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© The Japanese Society of Hematology 2012