Search
Search
Close this search box.

Treatment outcome and survival status among adult patients treated for lupus nephritis in selected tertiary hospitals of Ethiopia – Scientific Reports

This study is one of the few that assesses the treatment outcome, survival status, and associated factors of LN patients in the sub-Saharan Africa (SSA) contexts, particularly in Ethiopia. It is also the pioneer study to profoundly examine multiple factors associated with LN treatment outcome, and survival status by using a dual-center setting containing a relatively larger sample.

In this study, remission rate to treatment (complete plus partial response) is around 92.5% at the end of induction therapy with immunosuppressant drugs. This finding is in line with a study done in China in which nearly 90% of patients achieved remission after induction therapy consisting of prednisolone with either MMF or CYC21. However, it was relatively higher than the studies done in Ibn Sina University Hospital of Morocco, and Aristide Le Dante University Hospital of Senegal, in which 66% and 57.14% of patients achieved remission, respectively22,23. The discrepancy could be attributed to the differences in the selection of treatment regimen (MMF with prednisolone vs CYC alone or AZA alone), male-to-female sex ratio (0.09 vs 0.23 and 0.09 vs 0.128), and sample size (200 vs 114 and 200 vs 93). For instance, CYC alone was utilized in almost all cases of LN from Morocco due to low cost, whereas MMF was mainly utilized in our study by combining with prednisolone. Moreover, the current finding was relatively lower than a study done in India24, which reported 94.1% of remission (complete plus partial). The variation could also be attributed to differences in the inclusion criteria (age; 15–60 years vs < 18 years), minimum follow-up time (1 year vs 5 years), genetic difference, and definition of response.

At the last visit of follow-up, about 84.5% LN patients achieved remission (complete plus partial response). Out of these, 66.5% of them had a complete response [CR] while 18% had a partial response [PR]. A similar observation was noted in a study done in Italy in which 82.8% of LN patients achieved remission at the last follow-up25. However, this finding was relatively higher than the finding of a study done in Libya which reported a complete, partial, and non-response in 64.5%, 13.3%, and 22.3%, respectively26. This incongruity might be ascribed to the variation in the definition of response, magnitude of comorbidity like hypertension (38.5% vs 89.5%), choice of treatment protocol followed (ACR vs EULAR), sample size (200 vs 76), and use of adjuvant chloroquine (100% vs 0%)27. The current finding was relatively lower compared to a previous study done in Russia28, where overall remission (CR plus PR) was reported to be around 95.7%. Likewise, this is mostly due to differences in length of follow-up (5 years vs 23 years), genetics, socioeconomic status, and type of regimen selected for maintenance therapy.

Among the baseline laboratory tests, lupus serology (positive ANA and anti-dsDNA), serum creatinine, and 24-h urine protein were not significantly associated with either CR or PR in this study. These findings were consistent with a study done in Turkey8. In this study, the male gender had no statistically significant contribution to non-response, even though several studies29,30,31 identified it as a risk factor. This could be due to the small number of male participants (N = 17) in our study, and genetic or hormonal differences. Similar to a study done in Senegal, the presence of leucopenia at baseline was significantly associated with non-response23. However, this finding is not consistent with many previous studies22,26,32,33. This could be partly due to variations in laboratory test control, sample size, and the presence of infection.

In the present study, LN patients who had a severe SLE disease activity index were less likely to achieve a CR than NR as compared with patients with a moderate disease activity index. This real-world finding conforms with studies conducted in Taiwan tertiary referral center, West China Hospital of Sichuan University, and India33,34,35 but is incongruent with a study conducted in Egypt36. The incongruity might be ascribed to the higher proportion of patients receiving three days high dose pulse steroid therapy (24.5% vs 100%) and the smaller sample size (85 vs 200).

Our study differs from other studies23,26 concerning the impact of pulse steroid therapy on treatment outcomes. In our study, the use of pulse steroid therapy was significantly associated with CR though only a few patients (24.5%) received it. This finding is in agreement with studies conducted in Japan and Senegal37,38. In contrast, a study conducted in Pakistan found that pulse steroids had no statistically significant effect on LN treatment outcomes. This dispute is likely due to nearly 90% of patients receiving only oral corticosteroids after pulse steroid therapy in Pakistan while either CYC or MMF was initiated in our case39.

Our findings revealed that the presence of comorbidity is significantly associated with LN non-response. This finding was supported by a prospective study done in Egypt36, in which 81.1% of non-responders had comorbidity at the initial. Conversely, patients with comorbidity were not significantly associated with a reduction in PR in two previous studies23,33. This discrepancy could be attributed to the merging of complete and partial responses.

In this study, partial response is more likely to occur in patients taking MMF with prednisolone than in patients taking CYC with prednisolone during induction therapy. This finding was similar to a study done in the United States of America in which most patients were black African40. This is because MMF is more effective than CYC in black African patients. Prolonged duration of induction therapy (more than 6 months) favors NR when compared with 6-month therapy, possibly because of non-adherence, drug toxicity, and medication error. Using either MMF or AZA as maintenance therapy is likely associated with a statistically significant CR compared with prednisolone alone. This finding is consistent with various international guidelines4,13.

LN patients who did not respond completely or partially to induction therapy after 6 to 12 months were significantly more likely to have non-response after long-term follow-up than their counterparts. Related studies have found that patients who achieve a partial or complete response after induction therapy had a greater CR than non-responders after long-term follow-up28,41. In this study, LN patients taking prophylactic cotrimoxazole achieved a statistically significant CR better than NR as compared to patients not taking prophylaxis. This is due to the benefit of cotrimoxazole in preventing infections caused by the immunosuppressant drugs used in LN42. In various studies, the intention of using cotrimoxazole was as a prophylaxis for opportunistic infections, but in the current study, we come up with a new hypothesis of using cotrimoxazole to increase the complete response in patients due to the reduction of antibodies related to B-cell productions secondary to bone marrow suppressions.

Survival of patients attaining complete response at the last follow-up visit was more than 90%. This finding was in keeping with two previous studies16,43. In this study, 14.4% of patients were non-responders at the end of the follow-up visit. This finding is lower than previous studies done in Italy25 (17.2%) and Chicago16 (32.0%) but higher than a study done in Egypt36 (12.9%). The incongruity could be ascribed to the differences in sample size, class of LN, the definition of remission, diagnosis by biopsy, age, and genetics.

Our survival analysis revealed that patients attaining complete response after induction therapy were shown to have an excellent prognosis. The current study finding unveiled that those patients achieving complete and partial responses to induction therapy decreased the risk of non-responders by 93.1% and 75.1%, respectively. So, non-response to induction therapy negatively influenced patient survival. This finding was consistent with three previous studies16,25,28.

Our study had the strength of being a dual center with a relatively higher number of patients compared with most other studies. Moreover, most of our patients have stayed on follow-up for more than 24 months and treatment outcomes were evaluated individually. Nevertheless, our study may have been limited by its retrospective study design and variation of treatment regimens used to treat LN which resulted in variations in treatment outcomes. Besides this, around 28% of patients were diagnosed with LN without renal biopsy in our study even though renal biopsy was the gold standard diagnostic method for LN. Furthermore, the reasons for those patients who have been lost to follow-up have not been captured. Lastly, due to the high cost and interrupted supply of immunosuppressive medications, some LN patients were shifted from one regimen to another which may have impacted treatment outcomes.