Management of deep-seated Gram-positive infections poses a challenge for clinicians. Amongst such entities are osteomyelitis, endocarditis, deep abscesses, pyomyositis, bacteraemia and septic arthritis [1, 2]. Several factors conspire to make these conditions difficult to treat. Presence of devitalised tissue such as bony sequestrum, necrotic debris and pus provides an immune-privileged nidus of infection [3]. The avascularity, lowered pH and redox potential of these infective loci oftentimes impedes access and bactericidal activity of extraneous antibiotics in addition to components of the immune system [3, 4]. Foreign bodies, including indwelling prostheses are also frequently involved in deep-seated infections, for the same reasons [5, 6]. These infections typically involve a large inoculum of organisms many of which may be metabolically dormant and/or incorporated within protective biofilm, rendering them less vulnerable to destruction by antibiotics and the host immune response [3, 4]. Where surgery or other mechanical means of source control are not feasible, antibiotic therapy is not typically curative and the goal of management is merely suppressive, aiming to improve quality of life for the patient [7].
Although there is arguably a wider choice of agents available to treat Gram-positive infections than their Gram-negative counterparts, the available options are still not ideal [8, 9]. Few oral agents offer sufficient bioavailability to be utilised in these infections and those that do present substantial limitations. Linezolid, fusidic acid and rifampicin have each been associated with haematological dyscrasias and are all potentially hepatotoxic, particularly with prolonged use, the latter also being unsuitable for monotherapy owing to the potential for rapid selection of mutational resistance during therapy [10, 11, 12, 13, 14]. Rifampicin also acts as a potent inducer of hepatic cytochrome P450 enzymes, reducing the effects of partnered drugs which are substrates of these, such as fusidic acid, resulting in de facto monotherapy and potentially selecting for rifampicin resistance when such combinations are utilised [15]. Pristinamycin has been associated with occasional reports of fatal toxic epidermal necrolysis [16]. Use of clindamycin and fluoroquinolones is jeopardised by a relatively high background rate of resistance amongst staphylococci and the attendant risk of Clostridioides difficile infection [17, 18]. Fluoroquinolones are approved only for use in adults and their use also entails the risk of rare yet serious adverse events including QT interval prolongation, ruptured aortic aneurysm, tendonitis and retinal detachment [19]. Co-trimoxazole, doxycycline and minocycline are not ideal from an antimicrobial stewardship perspective given that they each exert unnecessary selective pressure for resistance in Gram-negative organisms, as do fluoroquinolones [20, 21, 22]. Many of these oral options, though generally active against staphylococci, are not consistently active against streptococci and other organisms that may be involved in complex infections [23]. In any case, there is always a risk of non-concordance with oral therapy especially amongst patients with complicating psychosocial factors [24]. As a result, intravenous agents are typically advised in deep seated infections, at least in the initial intensive phase of therapy; commonly employed agents include β-lactams, vancomycin, teicoplanin and daptomycin [25]. Such intravenous therapy poses several logistical challenges including inpatient admission or the need for regular attendance at OPAT clinics, central line insertion and oftentimes, therapeutic drug monitoring [8, 25]. This is costly and must be undertaken by skilled personnel. Moreover, the invasive nature of central line placement can leave patients vulnerable to further iatrogenic harms including thrombosis and Gram-negative line-related infections [8, 25].
Dalbavancin, a lipoglycopeptide antibiotic for intravenous administration, has the potential to obviate most of these problems. It is the dimethylaminopropyl amide derivative of the naturally occurring compound A40926, itself a secondary metabolite of an actinomycete, Nonomuraea gerenzanensis ATCC 39727, isolated from an Indian soil sample in the 1980s [26]. Dalbavancin has an in vitro breath of spectrum qualitatively like that of the conventional glycopeptide antibiotics vancomycin and teicoplanin though with quantitatively greater potency [27]. Although all share a common mechanism of action, disrupting cell wall biosynthesis by binding to the terminal D-alanyl-D-alanine residues on murein precursor chains, modal MICs for Staphylococcus and Streptococcus spp. are 0.03 – 0.06 µg/ml 8-32-fold lower than those of vancomycin and teicoplanin (0.125 – 0.5 µg/ml) [28,29]. The increased potency of dalbavancin owes to additional physicochemical mechanisms, such as the ability to dimerize and hydrophobically anchor into bacterial cytoplasmic membranes via its lipid side chains [27,28,29]. Animal studies have indicated that AUC/MIC is the parameter which most closely dictates the bactericidal action of dalbavancin and that it exhibits, to a degree, concentration dependent kill kinetics [29]. This contrasts with vancomycin which displays time dependent kill kinetics, despite sharing a common target [29]. In the case of S. aureus optimal AUC24hr/MIC ratios for dalbavancin are predicted to be 100-300, much lower than the 400-600 target suggested for vancomycin [30,31]. Some in vitro data suggest that dalbavancin might possess other advantages over classical glycopeptides including greater potency against organisms in the biofilm state and the ability to suppress bacterial exotoxin production, though this has not been corroborated in clinical studies [32,33]. Dalbavancin is highly active against most clinically relevant Gram-positive organisms with the notable exception of vancomycin resistant enterococci, or more rarely vancomycin resistant Staphylococcus aureus (VRSA) carrying the vanA gene, as well as some species seldom encountered as pathogens such as Enterocloster clostridioformis [26, 27, 28]. Vancomycin intermediate S. aureus (VISA) strains typically have elevated minimum inhibitory concentrations (MICs) for dalbavancin though these still tend to be well below the susceptibility breakpoint [36]. Similarly to the lipopeptide antibiotic daptomycin, dalbavancin has been found to be strongly synergistic with β-lactams in vitro through a phenomenon of collateral sensitivity known as the ‘see-saw’ effect whereby increasing MIC for one agent brings about a corresponding drop in the MIC for the other, though it remains unclear whether this can be capitalised upon for enhanced clinical effect [37].
Dalbavancin displays linear, dose-dependent pharmacokinetics with a prolonged serum half-life of around 346 hours thus maintaining therapeutic unbound concentrations in blood, bone and synovial fluid for at least 6 weeks following a single intravenous administration [38,39]. Previously reported concentrations measured 14 days after a single 1g infusion were 4.1, 15.9, 13.8 µg/g in cortical bone, synovium and skin, respectively, as compared to a plasma concentration of 15.3 µg/ml [35,38]. Concentrations in blister fluid approximated 30 µg/ml, 7 days after an infusion of 1g in another study. Dalbavancin is highly (~93 %) protein bound in vivo, principally to serum albumin [35, 38, 39]. It does not interfere with hepatic cytochrome-P450 enzymes and therefore has a low potential for deleterious drug interactions [40, 41]. Unlike vancomycin, dalbavancin has not been associated with ototoxicity or nephrotoxicity and ‘red man’ syndrome appears to be rarely, if ever, associated with its use [40, 41]. Two infusions of 1.5g administered one week apart should provide 6-8 weeks of antibiotic activity that is theoretically comparable to that afforded by daily dosing with vancomycin or teicoplanin [40]. Dalbavancin was approved in 2015 for ABSSSI but its use in deep seated infections remains off-label though clinical experience backing this usage is now mounting [42-50]. The present study retrospectively evaluates clinical outcomes for patients with various infection types (ABSSSI, osteomyelitis, vascular graft infection, bacteraemia, septic arthritis and prosthetic joint infections) receiving dalbavancin either as primary or consolidation therapy.