Statistics from Altmetric.com
Latent human herpes virus 8 (HHV8) infection has been implicated in malignancies occurring predominantly in HIV+ patients, including Kaposi sarcoma (KS), primary effusion lymphoma (PEL), plasmablastic lymphoma and multicentric Castleman disease (MCD).1 However, the oncogenic mechanisms remain enigmatic. Whether different serotypes of HHV8 have the same oncogenic potential, and in patients with concurrent HHV8-related malignancies whether intrapatient genotype variability exists remains to be studied. Such knowledge will be important in understanding the pathogenesis of HHV8-associated neoplasms. We report a case of an HIV+ male with disseminated KS and PEL who subsequently developed recurrent extracavitary relapse of PEL. HHV8 subtyping revealed that both the KS and extracavitary PEL were caused by HHV8 serotype C.
A Chinese man aged 26 years who had sex with men presented with a non-healing left foot lesion, suspected to be a callus. The lesion did not respond to local therapies and progressed to multiple violaceous papules over both his shins. Skin biopsy confirmed HHV8-positive KS. Blood work revealed HIV positivity with a CD4 of 57/μL and viral load of 1.4 × 105 copies/mL. He was started on combination antiretroviral therapy comprising Kaletra, Saquinavir and Truvada. Three months later, an increase in CD4 to 318/μL and drop in viral load to 270 copies/mL heralded progression of KS, consistent with immune reconstitution inflammatory syndrome. While awaiting chemotherapy, the patient was admitted with progressive shortness of breath, increasing bilateral leg oedema, fatigue and night sweats. His performance status declined and he remained bedbound. On examination, extensive mucosal and skin lesions of KS were noted. Bilateral enlarged inguinal lymph nodes were palpable. Bilateral breath sounds were diminished in the middle and lower zones of the lungs. Whole body CT showed multiple enlarged lymph nodes in the neck, mediastinum, axillae and inguinal regions with a conglomerate nodal soft tissue mass along the small bowel mesentery. Multiple lung nodules were seen with an area of consolidation in the left lung apex measuring 8.2×5.7×5.7 cm. Bilateral large pleural effusions and trace ascites were noted. Inguinal lymph node biopsy confirmed KS. Bronchoscopy ruled out other concurrent opportunistic infections as a cause of the pulmonary findings. Left pleural tap was exudative, cytology showed medium to large atypical lymphoid cells with deeply basophilic cytoplasm and irregular nuclei with small nucleoli (figure 1).
Flow cytometry and immunohistochemistry showed these abnormal lymphoid cells to be positive for LCA/CD45, CD38, HHV8 and Epstein-Barr virus encoded RNA (EBER), while negative for other B and T cell markers. Bone marrow showed trilineage haematopoiesis without any involvement by KS or lymphoma. A diagnosis of PEL with systemic KS was made.
The patient received six cycles of cyclophosphamide, adriamycin, vincristine and prednisolone (CHOP) chemotherapy. After one cycle of treatment, his effusions improved significantly and he was weaned down to room air. After three cycles of chemotherapy, CT scan showed good partial response to treatment. After six cycles, complete resolution of cutaneous KS lesions was noted on physical examination with some residual hyperpigmentation. Two months after completion of chemotherapy, the patient developed new subcutaneous nodules in his thighs. Fludeoxyglucose-positron emission tomography CT scan showed the appearance of multiple new hypermetabolic soft tissue masses in the bilateral gluteal regions and upper thighs with the largest lesion measuring about 4 cm and standardised uptake value of the lesions ranging between 11 and 17. Core biopsy of the gluteal mass showed hematolymphoid malignant cells staining positively for LCA/CD45, HHV8 and EBER, but negative for other B and T cell markers, consistent with extracavitary relapse of PEL (figure 2).
Salvage chemotherapy with ifosfamide, carboplatin and etoposide was started. After two cycles of therapy, complete resolution of the ulcerated left thigh mass was noted with resolution of systemic symptoms. PET CT scan done after six cycles showed responding nodal, lung, liver and splenic disease. However, the subcutaneous lesions in the lower extremities showed variable response. The patient was initially asymptomatic and opted for observation. Three months later, he presented with an enlarging swelling of the right gluteal region with pain, erythema and fevers. CT scan of the abdomen and pelvis showed further increase in the size of the right gluteal hypoenhancing soft tissue mass to 8.8 cm × 5.3 cm × 6.7 cm, inseparable from the right obturator externus muscle with no bony erosion. Biopsy of the right gluteal mass confirmed relapsed extracavitary PEL. The patient failed dexamethasone, cytarabine and cisplatin (DHAP) salvage. Interferon-α was offered but he opted for supportive care and succumbed to his disease 3 months later.
To understand the possible pathogenic mechanism of HHV8 in our patient, DNA was extracted from formalin-fixed paraffin-embedded tissue of the KS lymph node biopsy and the extracavitary PEL biopsy specimens. Nested PCR was used to amplify the open reading frame 26 (ORF26) conserved region to detect the presence of HHV8 and ORF K1 variable region 1 (VR1) for subtyping. Genotyping was performed on the PCR products of ORF K1 VR1 and were sequenced by Applied Biosystems 3730xl DNA Analyzer (Applied Biosystems, USA). Both lesions were found to contain HHV8 genotype C (figure 3).
ORF26 is a highly conserved region of HHV8. The ORF K1 of HHV8 is at the left end of the genome and has two variable regions, VR1 and VR2. Sequencing of this region has identified five major genotypes of HHV8, A–E, minor subtypes F and Z, all with multiple internal clades.2 These genotypes have different geographical prevalence with genotype C being the most prevalent in the Chinese and South Asians. Genotypes A–E have demonstrated a relationship to KS lesions.3 Limited data are available on HHV8 genotypes causing PEL.
HHV8 can establish latency in endothelial cells and hematopoietic progenitors, however this itself is insufficient for malignant transformation of cells.4 Limited expression of early lytic genes in HHV8-infected cells, including growth promoter genes like vIL6 and angiogenic factors like vascular endothelial growth factor and platelet-derived growth factor receptor-α may be involved in paracrine oncogenesis.5 6 In patients with multiple concurrent HHV8-related malignancies, it is unclear whether the same genotype is causative of the different malignancies and if so whether the mechanism of oncogenesis is the same or different between the tumour types. This is of special interest in patients with concurrent KS and PEL since the cell of origin of both malignancies is speculative.
KS spindle cells have different markers including those of vascular and lymphatic endothelial cells, dendritic cells, macrophages and smooth muscle cells suggesting a multifocal origin of the tumour.7 8 Transcriptional reprogramming of HHV8-infected endothelial cells may also be involved in the pathogenesis.9 Studies have suggested that B cells or monocytes may also act as a reservoir for HHV8 and transmit them to the lymphatic endothelial cells resulting in KS development.10 PEL is a disease of monoclonal B cells, although the precise stage of differentiation of the malignant B cells is unclear. While some data suggest that the malignant B cells are a preterminal stage of differentiation, others suggest that the tumour may develop by transformation of B cells at different stages of maturation.11 12 In one patient with HIV- HHV8+ concurrent malignancies of KS, PEL and MCD, genotype analysis of HHV8 showed no variability between the three malignancies.13 In our case, similarly, no intrapatient genotype variability was noted between KS and PEL. It is possible that persistent HHV8 latency in B cells may drive oncogenesis through different mechanisms in both KS and PEL. Due to sample limitations, we were unable to perform episome clonality studies and therefore we were unable to demonstrate with certainty that the same virus strain caused both cancers.14 Further work in this regard would be of much interest to understand HHV8-mediated oncogenesis.
The authors thank the Li Ka Shing Institute of Health Sciences for technical support.
Contributors KHF prepared the clinical write up and discussion. RR envisioned the study, coordinated the work and edited the manuscript. DPC, CY Lam and MC performed the laboratory work. CY Lui and NL were responsible for clinical care for the patient and manuscript writing. AWHC reviewed the histopathology and helped with manuscript writing. SSL envisioned the study, oversaw the laboratory work and manuscript writing.
Competing interests None declared.
Patient consent Obtained.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.