Ceftazidime-Avibactam plus Aztreonam for the Treatment of Infections by VIM-Type-Producing Gram-Negative Bacteria

ABSTRACT

This is a retrospective single-center study of 24 patients who received ceftazidime-avibactam plus aztreonam (CZA/ATM) for the treatment of VIM-type-producing Gram-negative bacillus (GNB) infections. The bacteria isolated were Enterobacterales in 22 patients and Pseudomonas aeruginosa in 2. Sixteen out of 19 isolates showed synergistic activity. Two patients presented clinical failure at day 14, and the 30-day mortality was 17% (4/24). CZA/ATM could be considered an alternative therapy for VIM-type-producing GNB infections.

INTRODUCTION

Infections caused by metallo-β-lactamase (MBL)-producing Gram-negative bacilli (GNB) pose a therapeutic challenge due to the lack of therapeutic options and the high associated mortality (>30%) (1, 2). MBLs can hydrolyze all β-lactams except aztreonam (ATM) but usually coexist with other β-lactamases that inhibit its effect (3, 4).

The optimal treatment of MBL-producing GNB infections is not well defined (5, 6), and recent Infectious Diseases Society of America (IDSA) guidelines recommend ceftazidime-avibactam/aztreonam (CZA/ATM) as the preferred treatment option (2) based mostly on series of NDM-type MBLs (7).

Therefore, in our center, where VIM is the predominant type, we started to use CZA/ATM to treat VIM-type-producing GNB infections, especially severe infections, even when a synergy test was not available. The aim of this study is to describe the clinical cure rate of CZA/ATM in the treatment of infections caused by VIM-type-producing GNB.

This is an observational and retrospective study performed at Vall d’Hebron University Hospital, a 1,000-bed tertiary hospital in Barcelona, Spain. We included all adult patients diagnosed with a VIM-type-producing GNB infection treated with CZA/ATM for ≥48 h between November 2018 and November 2021. Patients were identified through the pharmacy database. Data were retrospectively collected from electronic medical records, and the study protocol was approved by the hospital ethics committee.

(This work was presented in part as an oral communication at the 32nd European Congress of Clinical Microbiology and infectious Diseases (ECCMID), 23 to 26 April 2022, Lisbon, Portugal [8], and in part as a poster at the 25th SEIMC Meeting, Granada, Spain, June 2022 [9]).

Definitions.

Infections were defined according to the Centers for Disease Control and Prevention criteria (https://www.cdc.gov/nhsn/PDFs/pscManual/17pscNosInfDef_current.pdf). CZA was administered at a dose of 2.5 g every 8 h (over 2 h), and ATM was administered at a dose of 2 g every 8 h (over 30 min). All dosages were adjusted for creatinine clearance if necessary. Additional therapy was defined as the administration of at least one other active drug for ≥72 h. Infection source control was determined by the research team and defined as removal of any preexisting contaminated intravascular device or drainage of intra-abdominal abscesses or other fluid collections thought to be the source of infection (7).

Primary outcomes were clinical failure at 14 days and 30-day all-cause mortality from the index culture (7). Secondary outcomes were infection relapse, microbiological recurrence, and adverse reactions. Clinical failure was defined as death, lack of clinical improvement, or switch to another drug because of lack of improvement (7). Patients were followed up for at least 90 days after the onset of infection. Infection relapse was defined as the onset of a second microbiologically documented MBL-producing GNB infection with the reappearance of clinical signs and symptoms of infection in the same site (10). Microbiological recurrence was defined as a second positive culture obtained after completing CZA/ATM therapy (when repeated cultures were available) irrespective of the presence of signs or symptoms of infection.

Microbiological methods.

Bacterial isolates were identified by mass spectrometry (Vitek-MS; bioMérieux). Antimicrobial susceptibility testing was performed by disk diffusion, and the MIC of specific antibiotics was determined by a gradient strip test (Etest; bioMérieux) or microdilution (Vitek-2; bioMérieux) according to the manufacturer's recommendations. Interpretation of the results was performed by applying the EUCAST clinical breakpoints available from 2018 to 2021, and carbapenemase production (KPC type, OXA-48 like, IMP type, VIM type, and NDM type) was confirmed by the immunochromatographic NG-Test Carba 5 assay (NG-Biotech) according to the manufacturer's recommendations. Synergy of CZA and ATM was determined by using both gradient strip tests at a right angle and calculating the fractional inhibitory concentration index (FICI). Values of ≤0.5 were considered to indicate synergy, values of >0.5 to 4 were considered to indicate indifference, and values of >4 were considered to indicate antagonism (11).

During the study period, there were 135 patients with a clinical isolation of VIM-type-producing GNB and 65 received an active antimicrobial treatment. Thirty patients received CZA/ATM, and 24 met the inclusion criteria (6 received CZA/ATM for <48 h). Table 1 describes each included patient in detail.

TABLE 1.

Demographic, clinical and outcome characteristics of the 24 patients with infections caused by VIM-type-producing Gram-negative bacilli treated with ceftazidime-avibactam plus aztreonam^a

Patient	Age, yr (sex)	Underlying disease	CCI	Type of infection	Bacterial species	Bacteremia	Drainable source/drainage performed	Additional therapy/antibiotic	Adverse event	Clinical failure at day 14	Mortality at day 30
1	64 (M)	Vasculopathy with foot amputation	12	Postoperative wound infection	C. braakii	No	Yes/yes	No	No	No	No
2	58 (F)	Ulcerative colitis treated with colectomy	2	Hepatic abscess	C. braakii	No	Yes/yes	No	No	No	No
3	65 (M)	Myocardial infarction	3	Tracheobronchitis	C. braakii	No	No	Nebulized tobramycin	Nonsevere diarrhea	No	No
4	60 (M)	Cardiogenic shock needing ECMO, combined liver-kidney transplant recipient	7	Wound infection	C. freundii	No	Yes/yes	No	No	No	No
5	65 (M)	Aortic valve infective endocarditis due to Aerococcus urinae	2	Ventilator-associated pneumonia	E. cloacae	No	No	No	No	Death due to respiratory failure	Yes
6	65 (M)	Lung transplant recipient	3	Postoperative wound infection	E. cloacae	No	Yes/no	No	No	No	No
7	65 (M)	SARS-CoV-2 pneumonia	4	Catheter-related bacteremia	E. cloacae	Yes	Yes/yes	No	No	No	No
8	75 (M)	Kidney transplant recipient	9	Postoperative wound infection	E. cloacae	Yes	No	No	Encephalopathy	Death due to aspiration pneumonia	Yes
9	30 (F)	Ovarian cancer with pulmonary embolism	2	Catheter-related bacteremia	E. cloacae	Yes	Yes/yes	Systemic amikacin	No	No	No
10	64 (M)	SARS-CoV-2 pneumonia and pulmonary embolism	6	Urinary tract infection	E. cloacae	Yes	No	No	No	No	No
11	54 (M)	Cardiogenic shock needing ECMO	4	Ventilator-associated pneumonia	E. cloacae and K. oxytoca	No	No	Nebulized colistin	No	No	No
12	40 (M)	Disseminated esophageal cancer	3	Primary bacteremia	E. cloacae	Yes	No	No	Nonsevere diarrhea	No	Death due to unknown causeb
13	38 (F)	Lung transplant recipient	1	Catheter-related bacteremia	E. cloacae	Yes	Yes/yes	Systemic amikacin	No	No	No
14	57 (F)	Follicular lymphoma	3	Primary bacteremia	E. cloacae	Yes	No	Systemic amikacin	No	No	No
15	78 (M)	Kidney transplant recipient	5	Catheter-associated urinary infection	E. cloacae	Yes	No	No	No	No	No
16	82 (M)	Inferior extremity vasculopathy treated with arterial angioplasty	10	Postoperative wound infection	E. coli	No	Yes/yes	No	No	No	No
17	70 (M)	Cholangiocarcinoma	4	Biliary tract infection	E. coli	Yes	No	No	No	No	No
18	66 (M)	Myocardial infarction	12	Tracheobronchitis	K. oxytoca	No	No	Nebulized colistin	No	No	No
19	60 (M)	Burkitt’s lymphoma	4	Soft tissue infection	K. oxytoca	Yes	Yes/yes	No	No	No	No
20	64 (F)	Myelodysplastic syndrome treated with an allogeneic HSCT	4	Primary bacteremia	K. oxytoca	Yes	No	No	No	No	No
21	71 (M)	Aortic valve stenosis treated with TAVI	5	Catheter-associated urinary infection	K. oxytoca	No	No	No	No	No	Death due to nosocomial pneumoniac
22	80 (M)	Aortic valve replacement	9	Tracheobronchitis	K. pneumoniae	No	No	Nebulized colistin	No	No	No
23	58 (M)	Lung transplant recipient	7	Tracheobronchitis	P. aeruginosa	No	No	Nebulized colistin	No	No	No
24	84 (F)	Hip fracture	6	Hip abscess	P. aeruginosa	No	Yes/yes	No	No	No	No

^aCCI, Charlson comorbidity index; ECMO, extracorporeal membrane oxygenation; F, female; HSCT, hematopoietic stem cell transplantation; M, male; SARS-CoV-2, severe acute respiratory coronavirus 2; TAVI, transcatheter aortic valve implantation.

^bThis patient presented sudden death due to an unknown cause 5 days after finishing antibiotic treatment.

^cThis patient presented respiratory insufficiency 6 days after finishing antibiotic treatment.

The most frequent infections were lower respiratory tract infections (6 [25%]), soft tissue infections (6 [25%]), catheter-related bacteremias (3 [13%]), primary bacteremias (3 [13%]), and urinary tract infections (3 [13%]). Eleven (46%) patients had positive blood cultures, and 7 (29%) patients presented septic shock. CZA/ATM was administered at a standard dose in 22 (92%) patients, and the median duration was 12 days (interquartile range [IQR], 9 to 28 days). CZA/ATM was combined with other in vitro-active antibiotics in eight (33%) patients.

Among the 24 patients, 23 VIM-type-producing Enterobacterales isolates and 2 VIM-type-producing Pseudomonas aeruginosa isolates were identified. (One patient showed a coinfection.) Table 2 shows the in vitro susceptibility patterns of each isolate. A synergy test was performed in 19 isolates: 16 isolates showed a synergistic effect (84%), and 3 showed indifference (16%).

TABLE 2.

In vitro susceptibility patterns of the 25 isolates from the 24 patients with infections caused by VIM-type-producing Gram-negative bacilli treated with ceftazidime-avibactam plus aztreonam

Patient	Bacterial species	Encoded carbapenemase(s)	MIC (mg/L) ofa:													Synergy test result
			CRO	CAZ	FEP	ATM	IPM	MEM	ETP	CZA	AMK	GEN	CIP	TGC	SXT
1	C. braakii	VIM type + KPC type	≥64	≥64	≥64	≥256	3	≥32	4	≥256	4	4	≥4	≤0.5	≤1	Synergy
2	C. braakii	VIM type	≥64	≥64	2	0.19	1.5	2	0.25	≥256	≤2	≤1	≥4	2	≥16	ND
3	C. braakii	VIM type	≥64	≥64	4	12	≥32	3	8	≥256	≤2	≤1	≥4	≤0.5	≥16	Synergy
4	C. freundii	OXA-48 type + VIM type	≥64	≥64	2	0.125	12	≥32	6	≥256	≤2	≥16	≥4	≤0.5	≤1	Indifference
5	E. cloacae	VIM type	≥64	≥64	≥64	0.75	1	6	0.75	≥256	≤2	8	≤0.25	1	≥16	Synergy
6	E. cloacae	VIM type	≥64	≥64	≥64	0.38	≥32	≥32	≥32	≥256	≤2	4	2	2	≥16	Synergy
7	E. cloacae	VIM type	≥64	≥64	2	0.25	4	0.75	0.75	≥256	≤2	≤1	0.5	1	≥16	Synergy
8	E. cloacae	VIM type	≥64	≥64	≥64	≥256	3	1	1.5	≥256	≤2	8	≥4	2	≥16	Synergy
9	E. cloacae	VIM type	≥64	≥64	≥64	≥256	≥16	≥32	≥8	≥256	4	8	≥4	2	≥16	Synergy
10	E. cloacae	VIM type	≥64	≥64	2	0.5	8	3	2	≥256	≤2	≤1	1	1	≥16	Synergy
11.1	E. cloacae	VIM type	≥64	≥64	4	0.75	8	3	≥32	≥256	≤2	≤1	≥4	2	≥16	ND
11.2	K. oxytoca	VIM type	≥64	≥64	2	0.38	2	0.75	0.047	≥256	≤2	≤1	2	1	≥16	Synergy
12	E. cloacae	VIM type	≥64	≥64	≥64	0.19	16	≥32	16	≥256	≤2	8	1	2	≥16	Synergy
13	E. cloacae	VIM type	≥64	≥64	2	0.38	≥32	≥32	8	≥256	≤2	≤1	1	1	≥16	Synergy
14	E. cloacae	VIM type	≥64	≥64	2	1	16	1.5	0.25	≥256	≤2	≤1	1	2	≤1	Synergy
15	E. cloacae	VIM type	≥64	≥64	16	1	2	1.5	1	≥256	≤2	2	≥4	2	≥16	Synergy
16	E. coli	VIM type	≥64	≥64	2	0.064	1.5	0.38	0.38	≥256	≤2	≥16	≥4	≤0.5	≥16	Indifference
17	E. coli	VIM type	≥64	64	2	≥256	2	0.75	0.047	64	≤2	2	2	≤0.5	≥16	ND
18	K. oxytoca	VIM type	≥64	≥64	2	0.125	1	0.094	0.5	≥256	≤2	≤1	≤0.25	≤0.5	≤1	ND
19	K. oxytoca	VIM type	≥64	≥64	2	0.016	16	1	0.19	≥256	≤2	8	≤0.25	2	≥16	ND
20	K. oxytoca	VIM type	≥64	≥64	2	0.016	16	2	1	≥256	≤2	8	≤0.25	1	≥16	Indifference
21	K. oxytoca	VIM type	≥64	32	2	0.19	0.75	0.064	0.047	32	4	8	≥4	1	≥16	Synergy
22	K. pneumoniae	VIM type	≥64	≥64	24	0.75	3	16	4	32	8	≤1	2	4	≤1	Synergy
23	P. aeruginosa	VIM type	NA	≥64	≥32	4	≥16	≥16	NA	≥256	4	≥16	≥4	NA	NA	ND
24	P. aeruginosa	VIM type	NA	≥64	≥32	12	≥16	≥16	NA	≥256	≥64	≥16	≥4	NA	NA	Synergy
Total susceptibility (%)			0	0	0	75	33	49	36	0	96	50	17	27	23

^aNumbers in boldface represent the resistant isolates according to the current EUCAST clinical breakpoint values. AMK, amikacin; ATM, aztreonam; CAZ, ceftazidime; CIP, ciprofloxacin; CRO, cefotaxime; CZA, ceftazidime-avibactam; ETP, ertapenem; FEP, cefepime; GEN, gentamicin; IPM, imipenem; MEM, meropenem; NA, not applicable; ND, not done; TGC, tigecycline; SXT, trimethoprim-sulfamethoxazole.

Two cases (8%) presented clinical failure at day 14, and the 30-day mortality was 17% (4/24). Both patients with VIM-type-producing P. aeruginosa presented good clinical outcomes, as did the three patients with isolates that did not show synergy. In patients with bacteremia, the 14-day clinical failure rate was 1/11 (9%), and 30-day mortality was 2/11 (18%).

Clinical relapse after completing treatment did not occur in any case, but microbiological recurrence occurred in a patient who had a new respiratory positive culture that was considered colonization.

CZA/ATM-associated adverse effects were reported in 3 (13%) patients, but CZA/ATM was not withdrawn: two with non-Clostridioides difficile diarrhea and one with encephalopathy.

In the largest reported cohort of patients treated with CZA/ATM, the 30-day mortality (19%) is very similar to our results (17%) (7). However, the 14-day clinical failure rate (25%) is much higher, probably because in the study by Falcone et al. (7), only bloodstream infections were included. In addition, they found a decrease in the risk of mortality and clinical failure in patients treated with CZA/ATM compared to those patients who received other treatments. In a recent systematic review of the literature describing 94 patients treated with CZA/ATM from 18 studies (12), clinical resolution within 30 days was achieved in 80% of patients (75/94). Notably, there were no additional cases of VIM-type-producing GNB infections treated with CZA/ATM apart from the 5 patients described in the series of Falcone et al.

The question regarding the optimal treatment for MBL-producing P. aeruginosa remains unanswered. In the literature, there is a case report of a patient with osteomyelitis due to a VIM-type-producing P. aeruginosa isolate successfully treated with surgical debridement plus CZA/ATM and amikacin (13). We report 2 cases of VIM-type-producing P. aeruginosa infections, both with clinical success.

On the other hand, there is no standardized antimicrobial susceptibility testing method recommended for CZA/ATM (14). In a systematic review of 2,209 MBL-producing GNB isolates, high antimicrobial activity of ATM when combined with AVI or CAZ-AVI was described in 79.6% of MBL-producing Enterobacterales isolates and 6.2% of MBL-producing P. aeruginosa isolates (12). In our study, 84% of the tested isolates showed synergistic effects.

This study has several limitations given its retrospective nature and single-center design. Furthermore, it has a small sample size, and there is no control group. Another limitation is the lack of a synergy test of all strains. Additionally, the identification of acquired resistance mechanisms to β-lactams other than carbapenemases was not performed. It is therefore possible that other β-lactamases were present, but not identified. However, to the best of our knowledge, this is the largest series of VIM-type-producing GNB infections treated with CZA/ATM.

In conclusion, CZA/ATM could be a useful alternative therapy for the treatment of VIM-type-producing GNB infections.