Abstract Castleman disease (CD) describes a group of heterogeneous hematologic disorders with characteristic histopathological features. CD can present with unicentric or multicentric (MCD) regions of lymph node enlargement. Some cases of MCD are caused by human herpesvirus-8 (HHV-8), whereas others are HHV-8–negative/idiopathic (iMCD). Treatment of iMCD is challenging, and outcomes can be poor because no uniform treatment guidelines exist, few systematic studies have been conducted, and no agreed upon response criteria have been described. The purpose of this paper is to establish consensus, evidence-based treatment guidelines based on the severity of iMCD to improve outcomes. An international Working Group of 42 experts from 10 countries was convened by the Castleman Disease Collaborative Network to establish consensus guidelines for the management of iMCD based on published literature, review of treatment effectiveness for 344 cases, and expert opinion. The anti–interleukin-6 monoclonal antibody siltuximab (or tocilizumab, if siltuximab is not available) with or without corticosteroids is the preferred first-line therapy for iMCD. In the most severe cases, adjuvant combination chemotherapy is recommended. Additional agents are recommended, tailored by disease severity, as second- and third-line therapies for treatment failures. Response criteria were formulated to facilitate the evaluation of treatment failure or success. These guidelines should help treating physicians to stratify patients based on disease severity in order to select the best available therapeutic option. An international registry for patients with CD (ACCELERATE, #NCT02817997) was established in October 2016 to collect patient outcomes to increase the evidence base for selection of therapies in the future.

Castleman disease (CD) includes a group of rare and heterogeneous disorders with characteristic lymph node histopathological abnormalities. CD can occur in a single lymph node station, which is referred to as unicentric CD (UCD). CD can also involve multicentric lymphadenopathy and inflammatory symptoms (multicentric CD [MCD]). MCD includes human herpesvirus-8 (HHV-8)–associated MCD, POEMS-associated MCD, and HHV-8−/idiopathic MCD (iMCD). The first-ever diagnostic and treatment guidelines were recently developed for iMCD by an international expert consortium convened by the Castleman Disease Collaborative Network (CDCN). The focus of this report is to establish similar guidelines for the management of UCD. To this purpose, an international working group of 42 experts from 10 countries was convened to establish consensus recommendations based on review of treatment in published cases of UCD, the CDCN ACCELERATE registry, and expert opinion. Complete surgical resection is often curative and is therefore the preferred first-line therapy, if possible. The management of unresectable UCD is more challenging. Existing evidence supports that asymptomatic unresectable UCD may be observed. The anti–interleukin-6 monoclonal antibody siltuximab should be considered for unresectable UCD patients with an inflammatory syndrome. Unresectable UCD that is symptomatic as a result of compression of vital neighboring structures may be rendered amenable to resection by medical therapy (eg, rituximab, steroids), radiotherapy, or embolization. Further research is needed in UCD patients with persisting constitutional symptoms despite complete excision and normal laboratory markers. We hope that these guidelines will improve outcomes in UCD and help treating physicians decide the best therapeutic approach for their patients.

Nine of 56 (20% actuarial) patients receiving a T celldepleted, HLA-identical sibling BMT for hematological malignancy developed hemorrhagic cystitis (HC) 15-368 days post BMT. Hematuria was severe and prolonged (median duration 18 days). In eight patients (89%), a viral etiology was confirmed (four adenovirus, four polyomavirus). HC was associated with significant morbidity, with all patients requiring continuous bladder irrigation and transfusion support for blood loss and thrombocytopenia. HC occurring before day 100 was significantly associated with a reduction in longterm survival: 1/7 (14.3%) patients developing HC before day 100 became long-term survivors vs 21/49 (42.8%) without HC by day 100 (P = 0.034). In univariate analysis, HC was associated with a diagnosis of multiple myeloma (P = 0.02). There was a trend towards a higher incidence of HC in patients reactivating cytomegalovirus (CMV) compared with those remaining CMV negative (18.4 vs 5.5% respectively, P = 0.17). HC was not associated with graft-versus-host disease, or with the transplant dose of CD34 + progenitors or CD3 + cells, patient age or sex. Life-threatening, viral-induced HC and the unusually high incidence of adenovirus-induced HC may have been caused by immune deficiency associated with T cell depletion in this series.

Peripheral blood stem cell (PBSC) transplantation is successful in improving engraftment without increasing acute graft-versus-host disease (GVHD), despite much larger numbers of T cells in unmanipulated PBSCs than in bone marrow grafts. In mouse models and retrospective human studies, granulocyte colony-stimulating factor (G-CSF) therapy has been associated with less acute GVHD. We studied the effect of G-CSF on interferon (IFN)-γ and IL-4 expression in CD4+lymphocytes. CD4+ cells co-cultivated with G-CSF and stimulated with PHA or CD3 monoclonal antibodies showed significant decreases in IFN-γ and increases in IL-4 expression (n = 13;P < .01). G-CSF appeared to have a direct effect on CD4+ cells independent of monocytes present in the culture because purified CD4+ cells exposed to G-CSF, washed, and cocultivated with untreated monocytes demonstrated similar changes in IFN-γ and IL-4 expression, whereas untreated CD4+ cells cocultured with G-CSF–stimulated monocytes behaved as controls. We then studied peripheral blood mononuclear cells (PBMCs) from G-CSF–mobilized PBSC donors. When their PBMCs were cultured with PHA or CD3 monoclonal antibody, the percent of IFN-γ–expressing cells decreased by a mean of 55% and 42%, respectively, whereas the percent of IL-4–containing cells increased by a mean of 39% and 58%, respectively, following G-CSF stimulation. Increased apoptosis of IFN-γ–producing CD4+ cells was not responsible for the shift in TH1/TH2 subsets. G-CSF-R mRNA was present in both CD4+ and CD8+ cells. These results suggest that G-CSF decreases IFN-γ and increases IL-4 production in vitro and in vivo and likely modulates a balance between TH1 and TH2 cells, an effect that may be important in PBSC transplantation.