Extreme hypomagnesemia: characteristics of 119 consecutive inpatients.

Extreme hypomagnesemia (hypoMg) can be encountered in many situations, but little data currently exist. Our aim is to describe the epidemiological, clinical, etiological characteristics, and the biological abnormalities of consecutive inpatients with extreme hypomagnesemia. In our observational monocentric study, between 1st July 2000 and April 2015, all inpatients with extreme hypomagnesemia, defined by at least one plasma magnesium concentration (PMg) below 0.3 mmol/L, were included. Demographic, clinical, biological characteristics and the drugs prescribed before the qualifying PMg measurement were retrospectively collected. 41,069 patients had at least one PMg assessment. The prevalence of extreme hypomagnesemia is 0.3% (119 inpatients). The median age is 70 years, 52% are women. The patients were mainly hospitalized in intensive care (n = 37, 31.1%), oncology (n = 21, 17.6%), gastroenterology (n = 18, 15.1%) and internal medicine (n = 16, 13.4%) departments. One hundred patients (84%) had a medical history of gastrointestinal disease (39% with bowel resections, 24% with stoma), and 50 (42%) had a cancer history. The drugs most commonly prescribed (known to induce hypoMg) are proton pump inhibitors (PPI) (n = 77, 70%), immunosuppressive regimens (n = 25, 22.5%), platinum salt-based chemotherapies (n = 19, 17.1%), and diuretics (n = 22, 19.8%). The suspected causes of hypomagnesemia are often multiple, but drugs (46%, including PPI in 19%) and chronic gastrointestinal disorders (37%) are prominent. Associated electrolyte disturbances include hypocalcemia (77%) and mild hypokalemia (51%). The 1-month mortality from all causes is 16%. Extreme hypomagnesemia is rare in inpatients, and is frequently associated with severe hypocalcemia. Digestive disorders and drugs are the main contributory causes.

Plasma exchanges for the treatment of severe systemic necrotizing vasculitides in clinical daily practice: Data from the French Vasculitis Study Group.

The use of plasma exchanges (PLEX) in systemic necrotizing vasculitides (SNV) still need to be codified. To describe indications, efficacy and safety of PLEX for the treatment of SNV, we conducted a multicenter retrospective study on patients with ANCA-associated vasculitis (AAV) or non-viral polyarteritis nodosa (PAN) treated with PLEX. One hundred and fifty-two patients were included: GPA (n = 87), MPA (n = 56), EGPA (n = 4) and PAN (n = 5). PLEX were used for rapidly progressive glomerulonephritis (RPGN) in 126 cases (86%), alveolar hemorrhage in 64 cases (42%), and severe mononeuritis multiplex in 23 cases (15%). In patients with RPGN, there was a significant improvement in renal function compared to baseline value (P < 0.0001), the plateau being reached at month 3 after PLEX initiation, and estimated glomerular filtration rate improved especially as the number of PLEX increased. In patients with alveolar hemorrhage, mechanical ventilation was discontinued in all patients after a median time of 15 days. Patients treated for mononeuritis multiplex showed improvement of severe motor weakness. After a median follow of 22 months, 18 deaths (12%) were recorded, mainly in patients with RPGN and within the first 6 months. Incidence of end-stage renal disease and/or death was similar between groups of different baseline renal function, but was increased in MPO-ANCA compared to PR3-ANCA. Adverse events attributable to PLEX were recorded in 63%. No death occurred during PLEX. This large series describes indications, efficacy and safety of PLEX in daily practice. Randomized controlled studies are ongoing to define optimal indications, PLEX regimen and concomitant medications.

Central nervous system involvement of granulomatosis with polyangiitis: clinical-radiological presentation distinguishes different outcomes.

OBJECTIVE: The aim of this study was to describe the presentation and outcomes of patients with granulomatosis with polyangiitis (GPA) presenting with CNS involvement. METHODS: Patients were included in this nationwide retrospective study if they had GPA according to ACR criteria and/or the European Medicines Agency algorithm and CNS involvement. RESULTS: Thirty-five patients were included in the study. CNS involvement was observed in 51% of patients at GPA diagnosis. Headache (66%) was the main symptom, followed by sensory (43%) and motor impairment (31%). CNS involvement was characterized by pachymeningitis in 20, cerebral ischaemic lesions in 15 and haemorrhagic lesions in 2, with hypophyseal involvement in 2 patients. According to the clinical-radiological presentation, we distinguished granulomatous (G-CNS) and vasculitic (V-CNS) phenotypes. G-CNS patients more frequently had headaches, while V-CNS patients more frequently had motor impairment and renal involvement. Induction therapy produced clinical responses in 86% of patients. Baseline modified Rankin scale was higher for V-CNS than G-CNS patients (3 vs 2, P = 0.002). Initial spinal cord pachymeningitis was significantly associated with the need for a new induction regimen for relapsing/refractory disease (P = 0.01). Long-term neurological sequelae were noted in 51% of patients, including 35% with G-CNS and 69% with V-CNS (P = 0.08). Neurological sequelae were mainly noted in cases of spinal cord pachymeningitis (100%) and ischaemic or haemorrhagic lesions (73%). CONCLUSION: The clinical-radiological phenotype distinguished different long-term outcomes in patients with GPA and CNS involvement. Long-term neurological sequelae persisted in half of patients, mainly those with spinal cord pachymeningitis and vasculitic lesions.