We read the case report from Badros and coworkers with great interest.1 A patient with POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, monoclonal gammopathy, and skin changes) was successfully treated with bevacizumab. Increased level of vascular-endothelial growth factor (VEGF) is thought to play a role in the pathophysiology of POEMS syndrome.2,3 Here we describe a POEMS syndrome patient with high VEGF levels, treated with bevacizumab, without clinical improvement.
A 41-year-old male presented with slowly progressing, distal and symmetric, sensory, and motor neuropathy. Bone marrow aspiration showed a normal plasma-cell count. Magnetic resonance imaging (MRI) of the pelvis showed a solitary mixed lytic and sclerotic pelvic lesion, and a biopsy showed infiltrating plasma cells. He also met the remaining diagnostic criteria for POEMS syndrome.3 After 2 months of therapy with 500 mg cyclophosphamide given every 14th day, the patient presented with a World Health Organization (WHO) performance status III and the progression of the polyneuropathy left him bedridden. His plasma VEGF (p-VEGF) was 765 pg/mL (normal, < 50 pg/mL). Based on the rapid response of bevacizumab in the patient reported by Badros et al,1 we decided to give 2 courses of 5 mg/kg bevacizumab q 14 days. This approach was chosen to try to improve on the neurologic symptoms by reducing the level of p-VEGF before starting radiation therapy toward the pelvic plasmocytoma. After the first bevacizumab infusion p-VEGF decreased from 765 to 62 pg/mL, and remained low (median, 52 pg/mL; range, 34–89 pg/mL). Still, the condition worsened with increased paresis. Bevacizumab was stopped and radiation therapy was started. Five weeks after the last bevacizumab infusion, the patient experienced multiorgan failure and severe capillary leak syndrome with edema, ascites, and pleural effusion. After 2 weeks of intensive care, he died of pneumonia.
VEGF may affect the blood-nerve barrier by increasing micro-vascular hyperpermeability, thereby increasing the endoneural pressure subsequent to edema and nerve damage.2 This makes VEGF a logical target in POEMS syndrome. Still, the pathogenesis of the polyneuropathy has not been fully clarified, and several other cytokines like interleukin-1β (IL-1β), IL-6, and tumor necrosis factor-α (TNF-α) might contribute.3 Furthermore, the neuroprotective role of VEGF in neurodegenerative disorders4 might imply that a rapid removal of a chronic high level of VEGF might result in increased apoptotic activity in motor neurons under stress and hypoxia. Similarly, little is known about how endothelial cells that have been stimulated by a pathologically high level of VEGF over a long period of time react on the rapid removal of VEGF in plasma. It is known, though, that VEGF is an important survival factor for newly formed blood vessels, and that the removal of VEGF is followed by apoptosis of endothelial cells.5,6 The sudden collapse of a large volume of newly formed blood vessels following bevacizumab treatment might even lead to an increase of capillary leakiness and temporary impairment of the clinical condition, as seems to be the case in our patient. Following this, we suggest that, even if encouraging results have been reported, careful consideration should be taken before introducing bevacizumab in POEMS syndrome patients.
Bevacizumab therapy for POEMS syndrome
In response to our letter describing clinical improvement in a POEMS patient after bevacizumab, Straume et al describe a patient treated with bevacizumab and radiation therapy who, 5 weeks later, developed capillary leak and died. Although capillary leak cannot be directly attributed to bevacizumab, the report emphasizes the need for caution when using bevacizumab in POEMS patients.
Here we report on another POEMS patient who has been treated successfully with bevacizumab in Limoges, France. A 60-year-old woman was diagnosed with POEMS after 2 years of progressive sensory/motor neuropathy that left her bedridden. She had monoclonal immunoglobulin A lambda (0.1 mg/dL), normal marrow examination, and no bone lesions. POEMS manifestations included hyperpigmentation, hemangiomas, edema, diabetes mellitus, and hypothyroidism. She was hospitalized with pulmonary hypertension and bilateral pleural effusion (pulmonary artery [PA] pressure of 70 mm Hg). Considering her poor condition and the absence of bone lesion to irradiate she received an infusion of 300 mg bevacizumab without any improvement for 2 weeks, and mechanical ventilation was initiated. She then received 1 g cyclophosphamide and 40 mg dexamethasone 4 times daily with a second infusion of bevacizumab. After the second infusion, her edema and pulmonary hypertension resolved (she lost 31 kg in 15 days). Most interestingly, progressive disappearance of hyperpigmentation and whitening of hemangiomas was seen. Currently, she has a good clinical status without edema or clinical sign of pulmonary hypertension, 2 months after a third infusion of bevacizumab. Such rapid clinical improvement in POEMS syndrome is unusual and is probably attributed to bevacizumab.
There is growing evidence that VEGF is critical in the pathogenesis of POEMS manifestations. VEGF levels correlate with POEMS clinical activity, thus providing a biochemical and therapeutic target to diagnose and monitor response. Several studies have shown that responses to therapy, including high-dose chemotherapy, were associated with inhibition of VEGF production and/or activity.1,2 Bevacizumab induces a rapid decrease in VEGF levels, making it an appropriate candidate in therapy of POEMS patients.
Straume et al raise a concern about the deleterious effects of rapid reduction of chronically high VEGF levels on the neuronal cells. Although this is an interesting hypothesis, there is currently no proof that low VEGF levels contribute to neurodegeneration. It is possible that low levels of VEGF are neuroprotective, while high levels cause neuronal sheath edema and ischemic polyneuropathy.3 VEGF's effects on the endothelial cells are more complicated. VEGF increases vascular permeability and promotes angiogenesis by stimulating endothelial-cell growth.4 Extensive apoptosis of endothelial cells as suggested by Straume et al could have catastrophic consequences such as bleeding, a known side effect of bevacizumab. Although the data are limited, we hypothesize that endothelial damage stimulates platelet aggregation and release of VEGF that is stored at high levels in the platelets, thus maintaining adequate VEGF levels locally to preserve the newly formed blood vessels. It is worth noting that inhibition of platelet function has resulted in decreased VEGF levels and clinical improvement in POEMS.5
Understating the cause and source of VEGF overproduction in POEMS will help define the role of bevacizumab, if any, in the treatment of this disorder. In the absence of clinical trials, bevacizumab should be carefully considered in selected POEMS patients as adjuvant to conventional chemotherapy, and when used, patients must be monitored carefully for potential toxicities.
Correspondence: Ashraf Badros, Greenebaum Cancer Center, University of Maryland, 22 South Greene Street, Baltimore, MD 21201; e-mail: abadros@umm.edu.
