Vitamin D supplementation to persistent carriers of MRSA—a randomized and placebo-controlled clinical trial

Methicillin-resistant Staphylococcus aureus (MRSA) are resistant to beta-lactam antibiotics and cephalosporins and can cause severe infections, especially in elderly and weak patients. Historically, MRSA infections were more prevalent among patients within the health care sector (i.e., hospital-acquired or hospital-associated MRSA). Recent surveys suggest that community-acquired or community-associated onsets are becoming more common. Findings of MRSA in screening or clinical cultures are notifiable to the Public Health Agency by Swedish Communicable Disease Act [1]. Official statistics show that the number of new cases increases each year; in 2016, 4397 cases of MRSA were reported in Sweden, an increase of 12% compared to 2015 and the total amount of new notifiable cases (colonization and infection) has doubled every fifth year since 2007 [2]. Precautions have been made using ordinary infection control methods to limit the spread. In addition, all MRSA-positive cases are tracked and bacterial clones are regularly subjected to genetic analysis at clinical bacteriology laboratories in order to identify spreading patterns. Notably, clinical disease with MRSA is only the tip of the iceberg with regard to prevalence. The majority of cases never present clinical symptoms but rather constitute a reservoir for the bacterium in society. Therefore, in Stockholm County, all patients with risk factors such as skin lesions, chronic wounds, and urinary catheters are cultured for the presence of MRSA on admission to hospital. In cases of a positive culture, the patients are placed separately from other patients, in order to limit the risk of contamination and spread to other patients and hospital staff.

The population can be divided in three groups with regard to carriage of S. aureus, including MRSA: “never carriers,” “intermittent carriers,” and “persistent carriers” [3]. It is possible to eradicate MRSA from mucosal surfaces using a combination of topical and systemic antibiotic treatment. However, results are varying and especially persistent carriers often experience that MRSA colonization reoccurs after some time (3–6 months after eradication). Long-term antibiotic treatment has been considered to be unfeasible due to effects on the normal flora, risk of developing allergy, and also for developing resistance against the used antibiotic treatment. Notably, the reason(s) why certain individuals are persistent carriers of MRSA remain elusive.

Thus, alternative treatment strategies or preventive measures are highly warranted. Here, we hypothesized that vitamin D could eliminate MRSA carrier status among a group of persistent carriers in Stockholm, Sweden. The rationale behind our hypothesis is based on several pieces of evidence: (i) carriers of S. aureus have lower vitamin D levels than non-carriers [4, 5]; (ii) vitamin D induces the expression of several antimicrobial peptides (AMPs) that can kill S. aureus, including MRSA [6]; (iii) atopic skin, which is highly susceptible to S. aureus colonization has impaired expression of AMPs [7]; and (iv) results from an interventional study previously performed by us showed that vitamin D-treated patients had significantly fewer bacterial cultures that were positive for S. aureus [8].

Here, we report the results from a randomized, placebo-controlled trial where we have tested the hypothesis that vitamin D supplementation could reduce MRSA carriage in a group of persistent carriers. The results showed no effect of 12-month intervention with 4000 IU vitamin D/day on MRSA carriage in this group of individuals. A subgroup analysis for baseline vitamin D levels was performed but did not reveal any significant interaction between baseline 25OHD levels and the primary endpoint. Other subgroup analyses were performed on complicating factors for MRSA carriage, including presence of eczema, pet in the home, concomitant skin disease, smoking, as well as five different genetic polymorphisms, without any significant interaction terms. Importantly, reported adverse events were few and judged as not being related to the study drug. The compliance in the study was excellent, as shown by increased 25OHD levels in the treatment group versus virtually non-changed 25OHD levels in the placebo group.

Results from previous observational and experimental studies provided evidence that formed the basis for the current trial. There are several reports in favor of a beneficial effect of vitamin D on MRSA carriage: (1) vitamin D is known to produce AMPs with anti-staphylococcal effects [6]; (2) MRSA carriers have lower 25OHD levels in serum than non-carriers [4, 5]; (3) results from our own RCT from 2012 showed that vitamin D-treated patients produced significantly fewer respiratory cultures with S. aureus than the placebo group did [8]; (4) a recent small study from Thailand (n = 20) reported that 2000 IU vitamin D/day reduced skin colonization in atopic dermatitis [10]. On the other hand, there are also data refuting a role for vitamin D in S. aureus carriage: (1) it is known that the LL-37, which is the major AMP being produced by vitamin D, requires quite high concentrations and a specific ionic environment to be active against S. aureus [11]; (2) an observational study in diabetic patients found no connection between 25OHD levels in plasma an S. aureus carriage [12]; (3) and finally, a large RCT from New Zealand including n = 322 healthy adults who were given monthly bolus doses of 100,000 IU failed to find an effect against S. aureus carriage [13].

This study has several strengths. First, we had access to a unique cohort of persistent MRSA carriers in the Stockholm County, Sweden, an area where all patients who are MRSA-positive are registered and followed up closely. The discovery of MRSA positivity most often occurs on admission to a hospital, since all risk factors, such as wounds, eczema, and catheters, are screened for MRSA. This arrangement in Stockholm enabled us to perform the current study, where we could strictly include persistent carriers, here defined as two MRSA-positive bacterial cultures from at least one location, at least 3 months apart and during 3 years prior to inclusion. Second, since baseline levels of 25OHD have been shown to be important for the effects of vitamin D supplementation, we only included patients with < 75 nmol/L. Finally, the study was a randomized, double-blind, and placebo-controlled study and the compliance was excellent with a clear increase of serum 25OHD levels in the vitamin D arm but not in the placebo group.

Along with its strengths, the study has several weaknesses, which need to be taken into account. First, and most important, the study comprised only n = 65 patients due to early termination of inclusion and was therefore underpowered. There were several reasons for the early termination of the study, including a major reorganization in the Stockholm health care system for MRSA carriers, slow inclusion rate, lack of staff, lack of eligible patients, and due to short duration date of the study drug. The new power calculation revealed that the power decreased from 80.2% down to 40.4% with an expected difference between intervention and placebo of 12%. This low power could lead to the fact that a true beneficial effect of vitamin D would be missed (type 2 error). Nevertheless, we performed stringent statistical analyses of the whole dataset and failed to produce any significant effects or any trends of beneficial effects. It should be noted that the study protocol included sampling at baseline and after every 3 months during the first year. In total, the vitamin D arm produced 318 cultures from nose, throat, and perineum over the study period, whereas the placebo group produced 393 cultures, which when combined provided a large dataset. The fraction of positive cultures was almost identical in the two groups (vitamin D, 32.4%; placebo, 34%; Fisher’s exact test, p = 0.94), which further lend credibility to the overall null result. Combined, it is not likely that a larger study would produce a different result.

Another problem was that there were few individuals with 25OHD levels < 25 nmol/L. It appears that the best effect of vitamin D supplementation occurs in individuals below this level [14]. Apparently, persistent MRSA carriers in the Stockholm area have reasonably good 25OHD status with a median level of 50 nmol/L, which is in line with another cohort of non-MRSA carriers in Stockholm with 49.5 nmol/L [8]. A strict inclusion criteria of < 25 nmol/l could have been used but the study should then have been carried out in another setting where vitamin D deficiency is more common.

The only parameter that was significantly associated with a negative carrier status after 12 months was the use of antibiotics with activity against MRSA but not for other antibiotics. Although this finding was expected, the eradication strategy was not always effective, since only 5/6 individuals cleared MRSA after antibiotic treatment. Thus, there are strong indications that also other factors are involved in MRSA carriage. For example, there could be intracellular reservoirs for the bacteria, protecting against antibiotics with poor intracellular penetrance [15]. In addition, host genetic factors could be involved. In fact, polymorphisms in a vitamin D-related gene have been shown to be associated with MRSA carriage [16]. Notably, no such effect could be found in the current work, but the sample size hampers drawing definite conclusions. Genome-wide association studies (GWASs) for carriage in humans identified SNPs in IL4, DEFB1, CRP, and VDR for persistent nasal carriage [17]. More recently, the composition of the microbiota was shown to play a major role in S. aureus carriage. For example, commensal Corynebacteria in the nose were shown to produce compounds with anti-S. aureus effects [18]. The skin commensal Staphylococcus lugdunensis was found to produce a novel antibacterial peptide (Lugdunin) with potent effects against S. aureus [19]. In addition, coagulase negative staphylococci (CoNS) were shown to produce other antimicrobial peptides that could kill S. aureus in human skin. These CoNS were absent in atopic skin, which contain high amount of S. aureus [20]. Combined, the presence of commensal staphylococci appears to protect against S. aureus and also—theoretically—against MRSA carriage. Despite the negative findings in this work, it is still possible that vitamin D could lead to beneficial changes on the microbiota composition (promoting commensal staphylococcal colonization) in MRSA carriers, but this needs to be investigated in a larger study, including a longer follow-up time.

The results presented here do not support the use of vitamin D supplementation to persistent MRSA carriers. Even though the study was small, it had a stringent RCT design with many sample points during a 12-month period, which lend credibility to the findings. Current eradication strategies with conventional antibiotics are still not 100% effective and MRSA carriage often reoccurs in a subset of the population. Thus, there is still an unmet medical need to find novel strategies to eradicate MRSA from mucosal surfaces. Future studies could involve modulation of the microbiota by introducing specific strains with anti-staphylococcal effects, staphylococcal vaccination, or other host-directed therapies in combination with conventional antibiotics.