Neurotoxicity of Bortezomib Therapy in Multiple Myeloma (MM): A Single Center Experience.

Bortezomib has demonstrated activity in heavily pretreated MM patients. The dose limiting toxicity is peripheral neuropathy (PN) that affects up to 35% of the patients (Richardson et al JCO 2006). We retrospectively reviewed the incidence and severity of PN in 78 patients who received bortezomib at our institution. The median age was 57 years (range: 33–80), 62% were men, and 37% were African Americans (AA). Risk factors for PN included prior use of thalidomide in 53 patients (68%) and vincristine in 31 patients (40%). Seventeen patients (22%) had diabetes mellitus. Before bortezomib treatment 29 patients (37%) reported subjective grade 1–2 PN. Patients received bortezomib alone (n=10) or in combination with dexamethasone (n=36) and thalidomide (n=20) or chemotherapy (n=12). Responses included complete, near complete and partial responses in 5%, 10% and 27%, respectively. Grade 2 and higher PN occurred in 52% of the patients including Grade grades 3 and 4 PN in 15% and 7%, respectively. Twelve patients stopped bortezomib due to side effects that included PN (n=8), diarrhea (n=2) and CMV pneumonia (n=1), and 11 patients had dose reductions because of PN and fatigue. Grade 4 PN affected 6 patients (sensory n=4, and motor/sensory, n=2), 5 of whom were AA and 4 of whom had DM. The onset of grade 4 PN was acute rather than cumulative as noted with lower grades and took a median of 8 months (range: 4–16+) to improve compared to 3–4 months for lower grade PN. Neuropathy grade was not associated with age, sex, race or renal function (10 patients had creatinine > 2 mg/dl, including 2 on dialysis). Factors predictive of PN were baseline neuropathy (p=0.002) in addition to prior thalidomide use (p=0.03) and presence of DM (p=0.03). Responses were independent of neuropathy grade and whether bortezomib was combined with chemotherapy or thalidomide. The duration of therapy in responding patients was longer in patients receiving bortezomib in combination with thalidomide with a median of 5 cycles (range: 2–36) versus those who received other bortezomib combinations 3 cycles (range 1–19). Two of four patients with grade 4 sensory PN had their best response ever after 3 cycles of bortezomib; both had failed transplant and thalidomide based therapies; one remains in continuous complete remission for 24 months. PN therapy was mostly supportive including combinations of analgesics duloxetine, gabapentin, and pregabalin. Interestingly, 6 of 9 patients with PN who received lenalidomide as salvage therapy after progression on bortezomib had significant improvement in their symptoms; 3 of them stopped analgesics. In conclusion, the highest risk and grade of bortezomib neurotoxicity was seen in patients with baseline PN secondary to prior thalidomide use and DM. On the other hand, responding patients who received bortezomib in combination with thalidomide remained longer on therapy with minimal dose reductions. Although this may represent a section bias it is possible that thalidomide’s known anti-inflammatory and anti-angiogenesis properties protect against bortezomib-induced PN. Similar effects may explain the unexpected symptomatic improvement of PN on lenalidomide.