Our patients had similar demographics to those in previously published literature on aortic vascular endograft infections, insofar as they were mostly male, older and with multiple co-morbidities, and these latter factors correlated with mortality [2, 5, 9, 10]. The median time to presentation with infection in our cohort was similar to that reported previously [2]. Fever, pain and anorexia were common presenting features in our data and have been previously reported [2, 9]. Notably, most of our patients had a normal white blood cell count at presentation with infection, in contrast to prior cohorts [2]. Where infecting organisms are identified for AGIs, they are predominantly endogenous, and include a broad range of species [10, 12,13,14]. The incidence of polymicrobial infection was notable here, in line with other series [2, 15]. The potential for more than one culprit organism should be considered when treating empirically. A relatively high proportion (4/11, 36%) had suspected aorto-enteric fistulae [2, 10]. Fistulae should be suspected in the case of polymicrobial infections, especially with Gram negative or fungal organisms. A significant proportion of patients had negative cultures, in agreement with previously described patients, possibly related to prior antimicrobial use [2, 8,9,10, 12]. This complicates the selection of antibiotics, leads to the use of more broad-spectrum antibiotics and/or multiple antibiotics, increasing the chances of antibiotic side effects or adverse events. This needs to be taken into consideration when assessing patients’ suitability for treatment on OPAT. It does not, however, preclude the use of OPAT; all four patients with no guiding microbiology were successfully treated.
Expert consensus favours surgical removal of an infected graft where possible [2, 8]. One of the most notable findings here was the longer-than-expected survival, especially given the low rate of surgical intervention. In our series 8/11 (72%) of all patients (including suspected cases) and 5/8 (62%) of definite cases were alive at a minimum of 25 months (median 36 months) of follow up post-diagnosis of infection, despite only one of these patients having undergone graft explantation. In the published literature, despite a high (~ 30%) peri-operative mortality rate, patients who have undergone radical surgical intervention have survived longer than those who have not [2, 5, 6, 10]. Indeed, in one case series, all patients who did not undergo explantation had died within 2 years of presentation with infection [2]. A review of published literature looking specifically at non-explanted AGI showed slightly increased survival in those who had some sort of temporising procedure (drainage, surgical debridement, sac irrigation, and/or omentoplasty) as well as antimicrobial therapy, as compared with antimicrobial therapy alone (59% vs 50%) [14]. Our overall (27%) and medically managed (20%) all-cause mortality compares favourably, and there was no infection-attributable mortality despite less use of temporizing or definitive procedures.
However, this small selective cohort cannot be used to compare surgical versus medical therapy for AGI, but only to describe outcomes in those who survive and are considered candidates for outpatient management. There is an inherent bias in the selection of patients for analysis from an OPAT registry, as those who presented in extremis or died during the first presentation of infection are not included. By definition, the patients in our cohort were well enough for hospital discharge at least once with OPAT, skewing overall survival as compared with cohorts including all patients who present with AGI. In a series including all presenting AGI patients, 17/22 (77%) presented with an urgent surgical indication (rupture or bleeding), as compared with only two of our patients presenting with bleeding and none with rupture [2]. Perhaps these emergent events represented a late presentation of infection, and our cohort may have been at an earlier point in AGI, more amenable to medical therapy. Co-morbidity indices have not consistently been reported for comparison and could be useful for stratifying risks. Overall, however, we think it is unlikely that our patients were a much fitter cohort than those previously described; they were elderly with multiple co-morbidities, and potentially less fit for surgery. A substantial proportion also had presumed aorto-enteric fistulae, a recognised risk factor for poor outcome. In addition, the experience of individual surgical centres with, and skill in, performing primary explantation and in-situ allo/autograft reconstruction may also impact decisions around intervention in those who would elsewhere be considered for radical surgery. Finally, mortality is a crude outcome measure. Patients who survive explant and reconstruction may return to normal function and activity, whereas some survivors in the medically managed group may continue to decondition in the setting of chronic infection. Such qualitative outcomes should be measured to add more meaningful outcome data to decisions about the best course of action.
To the best of our knowledge there is no other published data on OPAT outcomes for AGI. As with other complex infections, OPAT can deliver significant savings in inpatient bed-days and healthcare costs [10]. This study did not involve a formal cost analysis, but with more than half of the total intravenous antibiotics for these infections being delivered in the outpatient setting, it is likely that the use of OPAT incurred cost savings to the healthcare system. Delivering antibiotics in the outpatient setting also has psychological and emotional benefits for patients and their families, especially those for whom this is a life-limiting condition [11]. The nature of AGIs necessitates regular ongoing assessment of response to antimicrobial therapy, whether intravenous or oral, and the goals of therapy should be made clear to patients. Proposed changes to British Society for Antimicrobial Chemotherapy (BSAC) OPAT outcome measures suggest starting with defining a treatment goal (cure, improvement, palliation) and then assigning an OPAT outcome (aim attained, aim not attained, indeterminate, death) [11]. These proposed changes would improve multi-disciplinary planning and patient involvement in shared decision-making in complex infections such as AGI. Adverse events and need for readmission in this cohort were picked up with standard weekly face-to-face outpatient review by an ID physician, demonstrating appropriate safety-netting. While no deaths attributable to vascular graft infection occurred during the follow-up period in our series, AGI carries a high mortality rate as described above and this should be addressed early in the treatment course. OPAT has been used indefinitely in certain palliative situations, and some patients with AGI may be appropriate candidates for this approach [16, 17]. Discussions about re-admission and preparedness for potential vascular catastrophes in the community may also be appropriate for certain patients, and palliative care referral should be offered where appropriate.