Purpose Diffuse alveolar hemorrhage (DAH) is a rare and potentially lethal complication of systemic lupus erythematosus (SLE) with a high mortality rate. It occurs more frequently in patients with lupus nephritis (LN). The aim of our study is to explore the characteristics of patients that develop DAH with lupus nephritis, risk factors that predispose DAH, treatment response and outcomes.

Methods Multicenter retrospective cohort study was undertaken including 6 centers in Saudi Arabia from 2002 to 2018. SLE patients meeting the SLICC criteria with lupus nephritis (biopsy proven or proteinuria or renal impairment due to lupus) presenting with diffuse alveolar hemorrhage (fulfilling a predefined criteria) were included in the study. An identical number of control group with lupus nephritis was also studied. Data was obtained from medical records by using a data sheet: demographics including age, gender, diagnosis, date of diagnosis of lupus, date of presentation of alveolar hemorrhage, clinical presentation, detection of alveolar hemorrhage proved by radiology, lavage or biopsy and laboratory parameters: including level of hemoglobin before and during DAH, sign of activity, treatment and outcome of DAH. Identification of risk factors predisposing to DAH in lupus nephritis patients was analyzed.

Results We identified 23 cases of DAH with lupus nephritis, all fulfilling the criteria. Mean age at presentation of DAH was 31.09 ± 12.6 years ranging from 14–57 years, of which 87% were females. 13 patients 56.5% had Class 4 LN and 21.7% had Class 4 and 5 LN on renal pathology. DAH occurred at a mean of 6.5 years ±3.8 in 13/23 patients with LN. Shortness of breath 95%, new chest x ray finding 95.7% and mean drop of haemoglobin of 2.72 gm/dl ±0.97 were more frequent at presentation of DAH with LN patients. High SLE disease activity - SELENA SLEDAI 2K was 38.56 ±19.3 was present at the onset of DAH. All were treated with methyprednisone,15/23 (65.2%) underwent mechanical ventilation and plasmapheresis was done in 21/23 patients (91.3%). Cyclophosphamide was given in 14/21 patients (60.9%), Intravenous immunoglobulins were given in 14/23 patients (65.2%) and dialysis was done in 12/23 patients (52.2%). Mortality occurred in 8 patients 34.8%. In comparison with the LN group, a mean haemoglobin of 7.56 ± 1.3, CNS involvement, vasculitis and fever>38 °C were of statistically significance P value: <0.001,0.02,0.03 and 0.03 respectively.

Conclusion In this multicenter cohort series with DAH in LN patients CNS involvement, vasculitis and fever>38 °C were associated in the occurrence of DAH. Mortality was low in our cohort in comparison to previous series which may be explained by early diagnosis and use of aggressive management.