We found that neither MP nor ZL could lead to an early clinical remission as compared to TCC alone in active CN. There was a significant decrease of pro-inflammatory cytokines with MP, but ongoing osteolysis could not be abated, indicating the role of cytokine-independent pathways in the progression of bone destruction, as elaborated in Fig. 5. Similarly, there was significant suppression of osteoclast activity with ZN, but could not translate into earlier remission.
The criteria used to define remission of active CN are based on reduction of signs of inflammation including redness, swelling and normalisation of temperature difference (< 2°C) between both feet, which is the most consistent and objective parameter. These clinical criteria usually correspond to resolution of marrow edema on T2W MRI images and normalisation of radiotracer uptake on TcMDP bone scan at ROI.[24, 25] Therefore, most studies available in the literature have utilised and recommended only clinical criteria to predict the remission. This apparently seems in consonance with the pathogenesis of disease as clinical signs of inflammation are predominantly contributed by preceding local ‘cytokine storm’ and consequent release of prostaglandins coupled with exaggerated vasoreactivity in response to cumulative minor trauma to an insensate foot. Further, ongoing osteolysis as a result of cytokine-mediated activation of osteoclastogenesis, also contributes to ongoing signs of inflammation. Thus, the diminution of clinical signs of inflammation and stabilisation/cessation of ongoing osteolysis may not always be concordant and usually takes a prolonged period of time varying from 6 to 12 months.
Off-loading the inflamed foot in active CN with a non-walking TCC is considered as the “gold standard” of treatment.[12] A non-walking TCC not only helps by preventing further trauma to the insensate foot thus abating inflammation and subsequent osteoclast activation but is also instrumental in reducing and favourably redistributing the abnormal plantar pressure distribution.[26] However, it usually takes more than 6–12 months for clinical remission.[13] During this long period of non-ambulation, BMD of foot bones may be adversely affected. Therefore, efforts were made to identify pharmacotherapeutic agents that could target the pathophysiology of CN of foot in patients with diabetes.
The recent understanding of pathophysiology of active CN focuses on the role of inflammatory cytokines that are postulated to incite RANKL-NF-кB activation and consequently, local osteoclastogenesis in the affected foot bones. [4, 5, 21] There is some pre-clinical evidence demonstrating reduction in resorption with anti-TNF-α antibodies [27] but no clinical evidence for the use of anti-inflammatory agents in active CN, till date. The current study used methylprednisolone in high dose pulse therapy to abruptly curtail the local ‘cytokine storm’. We observed a delay in resolution of active CN with MP, despite cogent suppression of cytokines. Although circulating cytokine levels were assessed, they were regarded as indicative of local cytokine concentration, owing to the fact they were assessed in the same individuals at baseline and end of follow-up. The delay in resolution of clinical activity of CN despite suppression of cytokines can be explained by uninhibited activation of cytokine-independent RANKL pathway, suppression of favourable anti-inflammatory cytokines involved in bone healing (IL-4, IL-10) and worsening of hyperglycemia which, in turn, directly exacerbates the RANKL-NF-κB activity and subsequent osteoclastogenesis.[28] Also, the increase in bone resorption and decrease in bone formation, resulting in ongoing osteolysis of foot bones could have contributed to a delay in clinical remission.
Lack of substantial evidence regarding efficacy of ZL in achieving clinical remission in active CN may be attributed to the doses used in previous studies (4mg versus 5mg), frequency of admission (single use versus multiple monthly doses), longer lag period and a very modest effect on ‘cytokine storm’.[13, 16] Initial studies with alendronate and pamidronate showed gain in terms of improvement in symptom score and suppression of bone turnover but their effect on clinical remission was either not found to be significant or not investigated. A systematic review[29] and few recent studies [13, 16] suggested that bisphosphonates may not reduce time to remission in patients with active CN or even increase the time in cast. The current study shows a remarkable decrease in bone resorption markers and consequent increase in BMC with ZL suggesting effective suppression of ongoing osteoclastogenesis. This could be pertinent in the prevention of deformities and fractures over long-term, due to long retention of the drug in bone matrix But, monotherapy does not seem to be sufficient for clinical resolution as ‘cytokine storm’ may continue unabated due to only a modest immunomodulatory efficacy of ZL. Similarly, denosumab, a monoclonal antibody against RANK-L, decreased time of fracture healing and clinical resolution of active CN in an open-label trial using historical controls.[18] Although denosumab is a potent anti-resorptive, the drug necessitates long-term usage as there is a high rate of rebound fractures on drug withdrawal. In fact, a recent meta-analysis of randomized controlled trials of all the studied medical therapies for active CN found no difference in outcomes in comparison to struct offloading with TCC.[30]
Periodic analysis of bone turnover markers (P1NP and CTX) suggested maximum osteolysis at baseline in all groups, in accordance with clinical activity of active CN as demonstrated previously.[21] Both P1NP and CTX declined in the zoledronate and placebo groups on follow-up without any significant difference between the groups, suggesting an overall decrease in bone turnover because of off-loading. On follow-up, maximum bone mass accrual at ROI was noted with zoledronate despite a similar alteration in cytokines as compared to placebo, suggesting its direct beneficial effect on bone parameters. A similar improvement in foot BMD has been demonstrated earlier with oral bisphosphonate using DEXA.[15] However, whether the initial gain in BMD persists after the quiescence of clinical activity needs to be assessed in long term follow-up studies. Though teriparatide has been shown to increase foot bone BMD in patients of chronic CN of foot,[22] its efficacy needs to be evaluated in patients of active CN.
Adverse events were noted in all three groups. Patients in the zoledronate group developed a transient flu-like reaction, which recovered with the use of analgesics and supportive care. The incidence of flu-like reaction in the current study was higher than previously reported (31.6%) where any one component of acute phase response (pyrexia, myalgia, headache, arthralgia, influenza-like symptoms) was considered. Two patients (n = 2, 16%) developed acute kidney injury after the second dose of zoledronate which is comparable to the incidence following standard dosing regimen of bisphosphonates (8 to 15%).[31] Cast-related tissue injury was observed in two patients from either group with literature evidence showing an overall incidence of 5.7%. A significant proportion of the patients in the methylprednisolone group developed worsening of glycemic profile that was managed by intensification of subcutaneous insulin therapy.
The strengths of the current study include proof-of-concept for the use of anti-inflammatory agent in active CN, an RCT design, homogenised use of standard-of-care TCC in all and precisely defined criteria for remission of active CN. The observations from the present study enable proposition of newer aspects of etiopathogenesis of active CN. We perceived certain limitations of the current study including a small cohort and a shorter duration of follow-up. Longer follow-up could provide more patient centric information like fractures, deformities and recurrences, keeping in mind the anticipated benefit of increased BMC due to tendency for long retention of bisphosphonates in the skeleton. RANKL and anti-inflammatory cytokines (IL-4, IL-10), a dorsal venous arch sample (rather than systemic sample) and in-vitro assessment using a bone biopsy sample could provide newer perspectives in the pathogenesis of active CN.