Development of a human herpesvirus 8-negative effusion-based lymphoma during treatment with dasatinib for chronic myeloid leukemia

Abstract

We present the case of an 85-year-old male patient diagnosed with human herpesvirus 8 (HHV8)-negative effusion-based lymphoma (EBL) that developed from long-lasting pleural effusion (PE) induced by dasatinib treatment for chronic myeloid leukemia (CML). After the onset of this disorder, dasatinib treatment was discontinued and drainage was performed to regress the effusion. The major molecular response (MMR) was thus lost. The patient did not tolerate nilotinib treatment, but bosutinib was successful in restoring MMR. During these clinical courses, the patient suffered from a recurrence of EBL, which was treated with rituximab-based chemotherapy. The PE sample just before the 3^rd cycle of chemotherapy revealed the proliferation of CD57-positive T cells, along with the disappearance of lymphoma cells. Anti-tumor immunity may have been activated following the immunochemotherapy in the undisturbed immunological environment when both EBL and CML almost regressed. After four cycles of R-CVP therapy, the patient has been in remission for 16 months and no longer requires drainage.

INTRODUCTION

Primary effusion lymphoma (PEL) is defined as human herpesvirus 8 (HHV8)-positive effusion-based lymphoma (EBL) according to the fourth version of the World Health Organization (WHO) classification. However, PEL-like disease has also been reported and is referred to as “HHV8-unrelated PEL-like lymphoma” or “HHV8-negative EBL.” Most of the EBL cases reported in East Asia are not related to HHV8; in fact, 17 of 25 Taiwanese patients and 64 of 69 Japanese patients were negative for HHV8.^1,2

Approximately half of HHV8-negative EBL patients had underlying medical conditions leading to fluid overload, including cirrhosis and cardiac complications.^3,4 Immunosenescence may also play a role in the development of HHV8-negative EBL in the elderly.^1,2 Dasatinib, a tyrosine-kinase inhibitor (TKI) used for the treatment of chronic myeloid leukemia (CML), frequently induces pleural effusion (PE) (28-33%),⁵ which is caused by lymphocyte-predominant exudates.⁶ To date, six cases of HHV8-negative EBL developed during dasatinib treatment, have been reported.^7-11

Here, we report a rare case of HHV8-negative EBL that developed during long-lasting dasatinib-induced PE.

CASE REPORT

An 85-year-old man was diagnosed with CML in May 2016. He was treated with a TKI imatinib, which was discontinued due to facial and lower limb edema and appetite loss. In September 2016, imatinib was replaced by bosutinib, but treatment was discontinued because of the onset of diarrhea. As a third-line TKI, treatment with dasatinib was initiated in January 2017 with an initial dose of 50 mg/day. As bilateral PE developed one month later, the patient continued to receive a reduced dose of dasatinib (20 mg/day) with furosemide. A profound molecular response (International Scale (IS), 0.0055%) was achieved in August 2017.

He complained of increasing dyspnea on exertion in October 2020. Chest radiography and computed tomography (CT) revealed bilateral PE and pericardial effusion (Figure 1, A). Furosemide and tolvaptan were administered under the diagnosis of congestive heart failure. Dasatinib therapy was discontinued and drainage of the pericardial effusion, which contained no atypical cells, was then performed. Left PE was drained, and cytology showed large, pleomorphic, atypical cells with fine chromatin, irregular nuclear contours, lobulated nuclei with one to multiple nucleoli, and basophilic cytoplasm with a few vacuoles (Figure 2, A). On immunohistochemistry (IHC) and in situ hybridization (ISH), the cell block specimen tested positive for CD10, CD20, CD79a, PAX5, BCL2, BCL-6 and κ; negative for CD3, CD5, CD7, MUM-1, CD138, c-myc, CD30, λ, HHV8 and EBER-ISH (Figure 2, B-G). The Ki-67 proliferation index was high (Figure 2, H). Immunoglobulin heavy chain gene rearrangement analysis using PCR revealed a dominant monoclonal peak (Figure 2, I). Karyotype analysis revealed the following complex chromosomal abnormalities: 47, XY, +21 [2] /49, idem, +3, +18 [4] /49, idem, +3, +18, -21, +mar [2] /47, idem, der (1) t (1;2) (p36.1; q21), and add (2) (q11.2) [12]. Fluorescence in situ hybridization analysis for the breakpoint cluster region-Abelson (BCR-ABL) 1 gene yielded negative results. Positron emission tomography CT (PET-CT) showed the absence of lymphadenopathy, hepatosplenomegaly, tumor mass or 18F-fluorodeoxyglucose uptake. Therefore, the diagnosis of HHV8-negative EBL was confirmed.

Fig. 1.

Chest X-ray A: Upon admission, B: Before the 1^st R-CVP therapy

Fig. 2.

Cytology, immunohistochemistry (IHC) and in situ hybridization (ISH); (A) Cytology revealed large, pleomorphic, atypical cells, displaying fine chromatin, irregular nuclear contours, lobulated nuclei with one to multiple nucleoli, and basophilic cytoplasm with a few vacuoles (Giemsa stain ×400). (B-H) On IHC or ISH, the cell block specimen exhibited tumor cells positive for CD10 (B. IHC ×400), CD20 (C. IHC ×400), kappa (D. ISH ×400), and negative for lambda (E. ISH ×400), HHV8 (F. IHC ×400) and EBER-ISH (G. ISH ×400). The Ki67 proliferation index was high (H. IHC ×400). (I) Immunoglobulin heavy chain gene rearrangement analysis by PCR method revealed a dominant monoclonal peak. (J, K) In the PE sample just before the 3^rd R-CVP therapy (Fig. 3. C1), the atypical large cells were invisible, and many small lymphocytes were seen (J. Giemsa stain ×400). The small lymphocytes were negative for granzyme B (K. IHC ×400)

Although the pericardial effusion did not increase thereafter and the right PE regressed without drainage, the volume of the left PE continued to increase. However, PE drainage alone, without chemotherapy, was performed because of the patient’s advanced age and comorbidities. Although the atypical large cells gradually became invisible in the drained PE sample for cytology, PE constantly increased and required repeated drainage.

Loss of MMR (IS, 0.8207%) was reported in December 2020, and the patient was treated with nilotinib, which was eventually discontinued because of hyperglycemia and appetite loss, and IS increased to 27.8%. Bosutinib was initiated in January 2021 at an initial dose of 100 mg, which was gradually increased to 300 mg, resulting in MMR restoration (IS, 0.0980%) in April 2021 (Figure 3).

Fig. 3.

Clinical course: IS: International Scale. IS is used as a tumor marker for chronic myeloid leukemia (CML); IL-2R: Interleukin-2 receptor. IL-2R is used as a tumor marker for HHV8 negative effusion-based lymphoma (EBL); R-CVP: rituximab, cyclophosphamide, vincristine, prednisolone; PE: pleural effusion; A: first drainage after admission; B: drainage just before the 1^st cycle of R-CVP chemotherapy; C1: thoracentesis just before the 3^rd cycle of R-CVP chemotherapy; C2: thoracentesis just before the 4^th cycle of R-CVP chemotherapy

The reappearance of atypical large cells was recognized in March 2021. Left PE increased more and more (Figure 1, B). R-CVP (rituximab, cyclophosphamide, vincristine, and prednisolone) therapy was initiated. Subsequently, PE did not increase and did not require drainage, except for thoracentesis. PE samples taken just before the 3^rd cycle of R-CVP therapy revealed the proliferation of CD57-positive T cells (Figure 4), along with the disappearance of atypical large cells (Figure 2, J). The number of CD57+ T cells decreased dramatically in the next PE sample (Figure 4). After completing four cycles of R-CVP therapy, the patient has been in remission for 16 months, as of November 2022.

Fig. 4.

Flow cytometry. B: PE sample just before the 1^st R-CVP therapy; C1: PE sample just before the 3^rd R-CVP chemotherapy; C2: PE sample just before the 4^th R-CVP chemotherapy; Blue circle: Many κ-positive large cells presented in the B PE sample; Red circle: The percentage of CD57-positive cells in the small lymphocyte gating of the C1 PE sample increased significantly to 97(0+97) %.

DISCUSSION

Prolonged fluid overload and inflammatory stimulation were thought to have contributed to the clonal development of exudative B cells. A single case report of disease was published, which occurred in the peritoneal cavity after drainage of a ventriculoperitoneal shunt was inserted 30 years ago.¹² Most Hepatitis C virus (HCV)-associated HHV8-negative EBL cases involve the peritoneum, which could be related to liver cirrhosis caused by HCV infection.¹³ In our case, HHV8-negative EBL developed 45 months after dasatinib-induced PE.

The underlying mechanisms of dasatinib-induced PE are not clearly understood. This adverse event could be immune-mediated because lymphocyte counts are relatively high in PE of patients with this disease.¹⁴ The other proposed hypothesis involves the inhibition of platelet-derived growth factor receptor β, resulting in a reduction in interstitial fluid pressure,¹⁵ and inhibition of the Src-family kinase, leading to changes in vascular permeability.¹⁴ Phan et al. demonstrated that dasatinib increases both in vitro and in vivo the pulmonary endothelial permeability in response to reactive oxygen species, leading to PE.¹⁶

Several cases of HHV8-negative EBL were successfully treated with effusion drainage.^17-24 Of the six patients who developed HHV8-negative EBL during dasatinib treatment, three were treated with drainage. The disease resolved in one of the three patients treated with drainage, although he had three recurrences in 24 months.⁸ One of the other two patients treated with drainage achieved a complete metabolic response, although he relapsed with COVID-19 pneumonia at 9 months.⁹ The last of the three patients treated with drainage, maintained remission for 18 months.¹⁰ The state of molecular response (MR) after stopping dasatinib was not described in relation to one of the three patients. The other two patients maintained deep molecular response (DMR) with imatinib (Table 1). DMR is associated with maximal restoration of immune recovery in CML patients on TKI.²⁵ In our patient, HHV8-negative EBL recurred relatively soon after the disease was controlled by drainage, which may have been influenced by the compromised immune system of the patient owing to the exacerbation of CML.

Table 1. Previous and present cases of dasatinib-associated HHV8-negaive EBL.

Case / Reference	Age / Sex	Immunophenotype	EBER-ISH	Site	Time of dasatinib treatment to appearance of effusion / EBL diagnosis	Treatment	Outcome of treatment / prognosis from diagnosis	State of CML after stopping dasatinib
1 / [7]	69 / M	CD20, CD79a, BCL2, BCL6, MUM-1	-	Pleural effusion	1 y / 2 y	R-CHOP ×2, R-CVP ×3	CR / alive and in remission at the time of reporting	Lost MMR, achieved MMR with nilotinib
2 / [8]	56 / F	CD19, CD20, CD79a, CD138, BCL-2, PAX5	-	Pleural effusion	44 mon / 59 mon	Drainage, steroids	Resolved and recurred three times / recurred at 24 mon	No description of the MR #
3 / [9]	69 / M	CD20, CD79a, PAX5, IRF4	+	Pleural effusion	1 y / 8 y	Rituximab ×6	CR / alive at 17 mon, maintaining remission	Lost MMR, achieved DMR with ponatinib
4 / [9]	69 / M	CD20, CD79a, PAX5	-	Pleural effusion	4 y / 4 y	Drainage	CMR / relapsed and died from COVID-19 at 9 mon	Maintained DMR with imatinib
5 / [10]	55 / F	CD20, κ	+	Pericardial effusion	15 mon / 16 mon	Continuous drainage for 4 d	CMR / alive at 18 mon, maintaining remission	Maintained DMR with imatinib
6 / [11]	40 / M	CD20, CD79a, BCL2, MYC	+	Pericardial effusion	23 mon / 23 mon	R-CHOP ×4	CR / alive at 8 mon, maintaining remission	Maintained DMR with nilotinib
7 / Present case	85 / M	CD10, CD20, CD79a, PAX5, BCL2, BCL6, κ	-	Pleural effusion	1 mon / 46 mon	Drainage, R-CVP ×4	CR / alive at 24 mon, maintaining remission	Lost MMR, achieved DMR with bosutinib

EBER-ISH: Epstein-Barr virus-encoded RNA in situ hybridization; R-CHOP: rituximab, cyclophosphamide, doxorubicin, vincristine, and prednisolone; R-CVP: rituximab, cyclophosphamide, vincristine, and prednisolone; CR: complete response; CMR: complete metabolic response; COVID-19: coronavirus disease 2019; CML: chronic myeloid leukemia; MR: molecular response; MMR: major molecular response (IS ≤ 0.1%); DMR: deep molecular response (IS ≤ 0.01%); #: only described “CML without blastic transformation.”

CD57-positive cells were rarely observed in the PE sample just before the 1^st R-CVP therapy. However, they increased significantly in the PE sample just before the 3^rd R-CVP therapy. Most CD57-expressing cells were positive for CD3 and either CD4 or CD8 with the quivalent frequencies (Figure 4). CD57 expression in lymphocytes indicates terminally differentiated “senescent” cells and an inability to proliferate. These cells also exhibit high cytotoxic potential when intracellular expression of cytotoxic molecules, such as granzymes and perforin, is confirmed.²⁶ However, on the glass slide made of the PE sample just before the 3^rd R-CVP therapy for cytology, the cells were not stained with granzyme B (Figure 2, K). Anti-tumor immunity might have been activated following R-CVP therapy in the undisturbed immunological environment when both EBL and CML had almost regressed, leaving these senescent cells in the PE just before the 3^rd R-CVP therapy. Anti-tumor immunity seems to play an important role in the relatively favorable clinical course of HHV8-negative EBL in contrast to PEL.