POEMS syndrome, also known as osteosclerotic myeloma, is a constellation of polyneuropathy, monoclonal plasma cell proliferative disorder and several other clinical features.1, 2 The three major diagnostic criteria include Castleman disease, sclerotic bone lesions and elevated serum levels of vascular endothelial growth factor. The rest of the acronym, that is, organomegaly, endocrinopathy and skin changes, form the minor criteria along with extravascular volume overload, papilledema and thrombocytosis/polycythemia. Diagnosis of POEMS syndrome depends on the presence of polyneuropathy, monoclonal gammopathy and one other major and one minor criterion.3

Treatment is aimed at eradicating the underlying plasma cell clone and the control of major symptoms. In cases of 1–3 plasmacytomas without BM infiltration, localized radiation therapy may be sufficient.4 Conversely, widespread disease requires systemic therapy, which is adapted largely from therapy for multiple myeloma.4 Owing to its rarity, current literature for POEMS syndrome comprises case reports and series, and no randomized trials have been conducted.5 In 2001, the first report of a successful autologous hematopoietic SCT (auto-HCT) in a patient with POEMS syndrome was published,6 followed by similar case series from different centers.7, 8, 9 Results suggest that high-dose chemotherapy and auto-HCT are associated with durable clinical response, but with significant post-transplant morbidity.7, 8, 9 Here we present our experience in seven patients with POEMS syndrome who underwent auto-HCT at the MD Anderson Cancer Center.

Diagnosis of POEMS was confirmed according to the criteria of Dispenzieri.3 Peripheral blood stem cells were collected with G-CSF in six patients, and with cyclophosphamide and G-CSF in one patient. The preparative regimen was melphalan 200 mg/m2 in six patients, whereas one patient received melphalan 180 mg/m2 due to renal insufficiency. Patients were evaluated for responses (hematologic, radiologic and clinical), toxicity, PFS and OS. Hematologic response was defined by the International Myeloma Working Group (IMWG) criteria.10 Radiologic data for bone lesions were assessed with bone surveys and positron emission tomography/computed tomography scans. Improvement of clinical symptoms was defined by amelioration in performance status, laboratory values and physical examination findings. Toxicities were defined as per the Common Terminology Criteria for Adverse Events version 4.0.

As described in Table 1, all seven patients had osteosclerotic bone lesions, monoclonal gammopathy and polyneuropathy. Three patients had biopsy-proven Castleman disease. Other features were: skin involvement in three patients (one with Raynaud’s, one with suspected plethora, and one with hyperpigmentation and scleroderma-like thickening), endocrinopathy with hypogonadism in three, ascites in two, anasarca in two, pulmonary hypertension in one, papilledema in one, pseudotumor cerebri in two and pericardial effusion in one. Three patients had thrombocytosis, which normalized after treatment. Adequate numbers of PBSC were collected from all patients (median 4.93 × 106/kg; range 2.4–7.37 × 106/kg). All patients received systemic therapy prior to auto-HCT; three patients underwent localized radiation for bone disease. Median follow-up after auto-HCT was 30 months (10–83 months). Median time for both neutrophil and platelet engraftment was 11 days. Only 1 (14%) patient had engraftment syndrome, consisting of a maculopapular skin rash, involving 85% of the body surface area on day 10 after auto-HCT, which resolved within 48 h of application of triamcinolone cream. Significant post auto-HCT complications were fungal pneumonia in two patients and pulmonary embolism in one patient. Patients fully recovered from these complications. One-year transplant-related mortality was 0%. Hematologic responses were measured at 3, 6 and 12 months after auto-HCT and then every 4–6 months. Best responses were achieved at 3–6 months: 3 patients (43%) had a complete response, 3 (43%) had a very good partial response and 1 (14%) had a partial response by IMWG criteria. All seven patients had complete or significant resolution of their clinical symptoms after auto-HCT: four patients had significant improvement in neuropathy, one had significant improvement in fever and night sweats and six had stable sclerotic bone lesions while one had mild improvement. Organomegaly normalized in both patients who had it, while four patients with anasarca, ascites, skin disease or papilledema had complete resolution of their problems. Median Karnofsky performance status improved from 80% (70–100) at the time of transplant to 90% (80–100) by 1 year post auto-HCT. Estimated 1- and 5-year PFS//OS were 100/100% and 86/100%; one patient relapsed 3 years after auto-HCT. Six patients are alive with controlled disease, whereas one patient died 6 years after auto-HCT due to gastrointestinal bleeding unrelated to his POEMS.

Table 1 Patient and transplantation characteristics (n=7)
Full size table

The natural history of POEMS syndrome is defined by progressive polyneuropathy and sclerotic bone disease, which leads to significant morbidity, along with mortality if respiratory compromise ensues.11 Both conventional and novel therapies, such as melphalan, dexamethasone, CHOP, lenalidomide and bortezomib, as single agents and in combination, lead to objective responses.12, 13, 14 Previous reports have described prolonged remission rates and significant improvement of clinical symptoms for POEMS after high-dose melphalan and auto-HCT.7, 8, 9 Our retrospective case series indicates a 100% response rate with 3 complete responses, 3 very good partial responses and 1 partial response, without significant auto-HCT-related morbidity or mortality. Importantly, a complete hematologic response was not required for significant clinical response.

We were able to mobilize>2 × 106 CD34+ PBSC from all seven patients, with only one patient requiring a combination of chemotherapy and growth factors. This is consistent with data from the Mayo clinic,8 but contrary to some reports that documented poor mobilization in these patients.15, 16 With small numbers and a heterogeneous population, it is difficult to comment on the causes of this discrepancy.

The response rates and durable PFS and OS were similar to previous reports of high-dose melphalan and auto-HCT. This favorable outcome may be attributed to a relatively younger patient population (median age 45) and a lower disease burden at auto-HCT as all patients had received induction therapy.

This case series of seven patients enhances the prior limited number of POEMS patients described in the literature who have successfully undergone auto-HCT, which has led to improved objective hematologic response, decrease in significant symptoms and recovery of functional status. Further studies are needed to investigate the role of proteasome inhibitors and immunomodulatory agents with or without auto-HCT.