Universal methylprednisolone pulses for the induction therapy of lupus nephritis: why not?

Lupus nephritis (LN) is the main organic manifestation of systemic lupus erythematosus (SLE). The earlier complete renal response (CRR) is achieved, the fewer nephrons are lost and the better long-term prognosis is expected. Methylprednisolone pulses (MP) have the potential to initiate rapid and mighty anti-inflammatory effects, facilitating the subsequent response to induction therapy. Whether MP should be generally used in patients with class III, IV and V LN would be the result of balancing their additional clinical efficacy and potential extra toxicity. In their thorough literature review, Gutowski and Belmont concluded that ‘the risks exceed the benefits for routine administration in all forms of LN’.¹ Since my position is exactly the opposite (always talking about classes III, IV and V), I will try to explain why, basing my disagreement on the same evidence they compiled.

The authors are right in pointing out the lack of randomised controlled trials specific to this topic; however, this does not necessarily mean either lack of or low-quality evidence—well exemplified by the recommendations for hydroxychloroquine (HCQ) use in SLE. As shown in table 2,¹ a substantial amount of data supporting the efficacy of MP comes from observational studies using the Lupus-Cruces-Nephritis (LCN) protocol. Combining repeated MP (median 10 pulses per patient) with each bi-weekly dose of iv 500 mg cyclophosphamide, initial prednisone ≤30 mg/d—with a fixed tapering to 5 mg/d in 12 weeks—and universal HCQ resulted in CRR rates >80% at 12 months and reduced progression to chronic kidney disease (10.6% after a mean follow-up of 6.5 years), both comparing favourably with standard of care (SOC).² Moreover, the number of MPs given during induction has been found to be an independent predictor for CRR.³ These results were adjusted by a large number of clinical and therapeutic variables, including propensity score analysis, and were constant across classes III–IV and V LN.^{2 3} Indeed, the systematic review by Figueroa-Parra et al⁴ confirmed that the addition of MP to regimes with prednisone doses <40 mg/d increased CRR rates without compromising safety. But, while all the above studies were among those selected in their review, Gutowski et al concluded that ‘.although several studies, such as those by Ruiz-Irastorza and the recent meta-analysis by Figueroa-Parra, demonstrate benefits of pulse therapy in terms of efficacy and reduced cumulative steroid exposure, these data are largely derived from observational studies and may not fully account for confounding factors such as histopathologic severity’.¹ This seemed enough to disregard MP in terms of efficacy.

About toxicity, eyes should focus on table 5.¹ While the potential benefits of MP were conditional (‘greater CRR in some cohorts’, ‘can be followed by lower dose oral prednisone’), the risks are presented as absolute: higher mortality, infections, osteoporosis, fractures and osteonecrosis. However, going through the evidence allegedly supporting such strong statements, increased mortality and infections in the meta-analysis by Figueroa-Parra⁴ were seen only if combining MP with prednisone ≥40 mg/d; in the study by Zonana et al, as shown in table 3,¹ ‘Total prednisone dose increased fracture risk (RR 2.5). Pulses showed a weaker, nonsignificant association (RR 1.3)’¹; the increased risk of osteonecrosis was supported by quoting a 2015 systematic review, summarised on table 3¹ as ‘Steroid doses≥2 g prednisone equivalent (eg, 2 days of 1 g MP) were strongly associated with AVN, with incidence rising with cumulative dose’,¹ but what the original paper actually refers to are cumulative glucocorticoid (GC) doses, without mentioning MP in this section.⁵ Therefore, all the evidence supporting MP toxicity actually points to genomic-mediated effects instead. The authors forgot to reference in table 5¹ several studies, many shown in table 3,¹ not finding any associations between MP and GC-induced toxicity, including the systematic review of Ugarte-Gil et al.⁶ The reduction of GC-induced complications, also infections and osteonecrosis, has been constant among patients treated with the LCN protocol, who, in addition to MP, received a reduced prednisone load compared with SOC.² Thus, implementing non-genomic mechanisms and reducing genomic effects spares GC-mediated toxicity.

As a conclusion, the authors state that ‘the benefits of pulse therapy must be weighed against the increased risks of infection, adverse metabolic bone effects and mortality with pulse dosing’.¹ After re-interpreting their data, we could rather say that the benefits of pulse therapy, if combined with lower prednisone doses, include higher CRR, better long-term preservation of renal function, reduction of infections and adverse metabolic bone effects and, therefore, improved general prognosis. So, patients with classes III, IV and V LN deserve regimes with MP and reduced-dose prednisone; why not?