Acinetobacter pneumonia: Is the outcome different from the pneumonias caused by other agents.

BACKGROUND: The principal aim of the present study was to determine whether Acinetobacter spp. pneumonia differs from hospital-acquired pneumonias (HAPs) caused by other agents with respect to therapeutic success and survival rate.
METHODS: This study includes 140 adult patients diagnosed with HAPs caused by identified etiologic agents between March 2005 and February 2006. These patients were divided into two groups according to the agent responsible for their infection (Acinetobacter spp. [n = 63] or non-Acinetobacter spp. [n = 77]). The groups were compared in terms of risk factors, therapeutic success and six-week survival rates.
RESULTS: Previous antibiotic use and the risk of aspiration were independent factors responsible for the development of Acinetobacter spp. pneumonia. Hypoalbuminemia, steroid use and the use of a mechanical ventilator were determined to be mortality-associated independent risk factors for Acinetobacter spp. pneumonia. The clinical success rate at the end of therapy was 41.6% and, at the sixth week, the survival rate was 35% among patients in whom Acinetobacter spp. was the causative agent. Conversely, in the control group, these values were 43 and 32%, respectively (P > 0.05). We found that the use of the appropriate antibiotics for the treatment of Acinetobacter spp. pneumonia was an important factor in survival (P < 0.001).
CONCLUSION: The outcomes of Acinetobacter spp. pneumonia do not differ from HAPs associated with non-Acinetobacter spp. in terms of therapeutic success and survival rates.

Acinetobacter species are Gram-negative, nonfermentative, nonspore-forming, nonmotile, aerobic coccobacillary organisms. The prevalence of infections caused by Acinetobacter spp. has increased rapidly since the 1970s.[1]

Acinetobacter spp. are frequently encountered agents responsible for hospital-acquired pneumonia (HAP), especially the late-onset, ventilator-associated pneumonias (VAPs). Acinetobacter spp. rapidly acquire antibiotic-resistance mechanisms, which may contribute to its virulance. Hospital-acquired pneumonias caused by Acinetobacter spp. lead to significant mortality and morbidity because of both their intrinsic and acquired resistance.[2] However, there have been a limited number of studies comparing Acinetobacter spp. pneumonia and non-Acinetobacter spp. pneumonia.[34]

The primary aim of the present study was to determine whether Acinetobacter spp. pneumonia is different from HAPs caused by other agents in terms of therapeutic success and survival rate. The secondary aim was to assess the drug resistance characteristics of Acinetobacter spp. and the independent risk factors associated with the development and mortality of pneumonia caused by Acinetobacter spp. in our hospital.

Methods

This study included 140 adult patients with HAP of known etiologic agent, admitted to the Trakya University Medical Faculty Hospital between March 2005 and February 2006. When the etiologic agent of their pneumonia could be identified, the patients were divided into two groups: those with HAP caused by an Acinetobacter spp. [n = 63] and those with HAP caused by a non-Acinetobacter spp. [n = 77; control group].

Hospital-acquired pneumonia: Hospital-acquired pneumonia was defined according to the standard definitions of the American Thoracic Society guidelines for the management of adults with hospital-acquired pneumonia.[5]

Ventilator-associated pneumonia: Pneumonia that developed 48 h after being connected to a ventilator was accepted as VAP.[5]

HAP developed in patients undergoing immunosuppressive therapy: HAP developed in patients undergoing immunosuppressive therapy for solid organ tumors, hematologic malignancy, or rheumatoid disease. The use of >20 mg corticosteroid for at least 3 weeks was assumed the required criteria for being considered as immunosuppressive therapy.

Pneumonias that developed >4 days after hospitalization were considered late-onset pneumonias.[5]

Patients were considered to have severe HAP in the presence of one of the following criteria:[6]

arterial oxygen pressure (PaO2)/fractioned oxygen percentage (FiO2) < 250;

severe sepsis or septic shock findings; or

bilateral or multilobar involvement, cavitations, abscess, effusion and rapid progression.

Multidrug-resistance (MDR) was defined as resistance to more than one of the following five drug classes: antipseudomonal cephalosporins, antipseudomonal carbapenems, β-lactam/β-lactamase inhibitor combinations, antipseudomonal fluoroquinolones and aminoglycosides.[7]

Clinical success was marked by a decline or disappearance of symptoms (fever, cough, sputum, and dyspnea) in the patients receiving antibiotic therapy.[8]

Appropriate antibiotherapy: The patients were given at least three days of antibiotics appropriate to the antibiogram of the identified etiologic agent.

Study protocol

Data obtained from the patients included in the study were collected prospectively. The approval of the local Ethical Committee was obtained during the planning phase of the study and each patient (or his/her caregivers) gave informed consent prior to participation in the study. The patients' demographic data, risk factors and the severity and the day of onset of the pneumonia were recorded. Chest X-rays, blood count, biochemical parameters, arterial blood gases and CRP were obtained from each patient prior to the initiation of therapy. Blood and sputum/tracheal aspirate cultures (if possible) were also obtained. Sputum/tracheal aspirate specimens containing >25 leukocytes per field and <10 epithelial cells per field were deemed acceptable for culture. Pleural fluid samples were obtained from patients with previously detected pleural effusions. Computed thoracic tomography was performed as required.

The decisions pertaining to the diagnosis and treatment of HAP were managed multidisciplinarily by the clinician responsible for the care of the patient, the pulmonary disease specialist, and the infectious disease specialist. Clinically recovered patients were discharged after posttreatment evaluation. Discharged patients were called for follow-up at the sixth week.

Statistical methods

Descriptive statistics and frequency analyses were calculated for the different types of cases. Kaplan-Meier methods were used for the survival analysis. Univariate Cox regression analysis was used to assess factors that might independently affect mortality. After the univariate analysis, variables with P < 0.1 were analysed using a multivariate Cox regression model.

Univariate logistic regression analysis was used to examine independent risk factors on the outcome (Acinetobacter vs. non-Acinetobacter). Multivariate logistic regression analysis with a backward stepwise method was used to examine significant factors (P < 0.10) obtained from a univariate model.

A P value of <0.05 was considered to be statistically significant. Statistical analyses were conducted using SPSS 9.0 (SPSS Inc., Chicago, IL, USA) statistical software.

Results

Of the 63 patients in whom Acinetobacter spp. were isolated, 38 (60%) were male and 25 (40%) were female. The mean age was 64.43 ± 13.89 years, with a range of 35 to 95 years. Thirty-eight of these patients were hospitalized in the internal medicine service, whereas 25 patients were hospitalized in the surgery service. The clinical services from which Acinetobacter spp. were most frequently grown were neurology (n = 29, 46%) and brain surgery (n = 7, 11.1%). In patients in whom Acinetobacter spp. were isolated, 43 had HAP, 17 had VAP and 3 had pneumonia that developed during immunosuppressive therapy. In the group in which etiologic agents other than Acinetobacter spp. were isolated, 45 had HAP, 17 had VAP and 15 had pneumonia that developed during immunosuppressive therapy (P = 0.027).

Seventy Acinetobacter spp. infections were isolated from the 63 patients followed during the course of the study. Of these, 49 were isolated from tracheal aspirates, 5 from blood and tracheal aspirates, 1 from pleural fluid and 15 from blood cultures (in which other possible infection sources, such as central venous catheter, etc. had been excluded). Three different Acinetobacter spp. strains were isolated from 2 patients, whereas 2 different Acinetobacter spp. strains were isolated from 3 patients. When the demographic characteristics and risk factors of the patients were compared between the groups from which Acinetobacter spp. and other causative agents were isolated, the risk factors with a P- value of < 0.1 according to univariate analysis were examined by multivariate analysis [Table 1]. Previous antibiotic use (P = 0.02) and the risk for aspiration (P = 0.02) were determined to be significant risk factors in the group from which Acinetobacter spp. were isolated. Previous antibiotic use and aspiration risk increased the risk for Acinetobacter spp. infection nearly three-fold (95% CI, 1.15–7.32) and nearly 2.5-fold (95% CI, 1.10–5.51), respectively [Table 2]. (Aspiration risk: Patients with confusion due to any cause, especially neurological diseases and patients lying supinely.)

Table 1

Demographic variables and risk factors in HAP patients with Acinetobacter spp. and with non-Acinetobacter spp.

Variables factors	Acinetobacter spp. n (%)	Non-Acinetobacter spp. n (%)	P
Age (Mean ± SD)	64.43 ± 13.89	62.36 ± 16.9	0.43
Gender (M)	38 (60.3)	47 (61)	0.93
Late-onset pneumonia	55 (87.3)	66 (85.7)	0.78
Day of pneumonia	14.75 ± 16.07	17.39 ± 17.55	0.36
(Mean ± SD)
Severe pneumonia	48 (76.2)	51 (66.2)	0.20
Patient risk factors
COPD	5 (7.9)	5 (6.5)	0.74
Heart failure	16 (25.4)	20 (26)	0.93
Diabetes	14 (22.2)	9 (11.7)	0.09
CRF	12 (19)	13 (16.9)	0.74
CVD	38 (60.3)	33 (42.9)	0.04
Malignancy	9 (14.3)	22 (28.6)	0.04
Antibiotic use	55 (87.3)	49 (63.6)	<0.01
Hypoalbuminemia	57 (90.5)	66 (85.7)	0.39
Smoking	23 (36.5)	29 (37.7)	0.88
Alcohol intake	4 (6.3)	7 (9.1)	0.55
Risk for aspiration	50 (79.4)	42 (54.9)	<0.01
Risk factors due to medical interventions
Use of H2 blockers	33 (52.4)	38 (49.4)	0.72
Steroid use	29 (46)	36 (46.8)	0.93
Use of cytostatics	2 (3.2)	13 (16.9)	0.01
Use of sedatives	6 (9.5)	4 (5.2)	0.32
Risk factors due to invasive interventions
Previous operation	14 (22.2)	13 (16.9)	0.42
Emergent intubation	19 (30.2)	16 (20.8)	0.20
CPR	4 (6.3)	5 (6.5)	0.97
Being connected to MV	17 (27)	17 (22.1)	0.50
Urinary catheter	51 (81)	52 (67.5)	0.07
TPN	15 (23.8)	11 (14.3)	0.15
Central catheter	23 (36.5)	21 (27.3)	0.24
Nasogastric tube	45 (71.4)	39 (50.6)	0.01
Tracheostomy	15 (23.8)	7 (9.1)	0.02

COPD = Chronic obstructive pulmonary disease, CRF = Chronic renal failure, CVD = Cerebrovascular disease, CPR = Cardiopulmonary resuscitation, MV = Mechanical ventilator, TPN= Total parenteral nutrition, SD = Standard deviation

Table 2

Evaluation of risk factors via multivariate analysis found to be significant by univariate analysis

Variables	Univariate analysis			Multivariate analysis
	P	HR	%95 CI	P	HR	95% CI
Diabetes	0.090	2.15	0.86–5.38	0.070	0.40	0.15–1.08
CVD	0.040	2.02	1.03–3.98	0.950	1.02	0.39–2.68
Malignancy	0.040	2.40	1.01–5.68	0.780	0.85	0.27–2.68
Antibiotic use	<0.01	2.92	1.63–9.42	0.020	2.91	1.15–7.32
Aspiration risk	<0.01	3.20	1.50–6.83	0.020	2.47	1.10–5.51
Use of cytostatics	0.010	6.19	1.34–28.59	0.160	3.18	0.62–16.23
Urinary catheter	0.070	2.04	0.92–4.49	0.260	0.51	0.15–1.65
Nasogastric tube	0.010	2.43	1.20–4.93	0.520	1.35	0.52–3.45
Tracheostomy	0.021	3.12	1.18–8.23	0.090	2.40	0.86–6.70

HR = Hazard ratio, CI = Confidence interval, CVD = Cerebrovascular disease

Determination of Acinetobacter strains is based on drug susceptibility patterns. When antibiotic susceptibility was examined, the highest susceptibility was to netilmicin [Table 3]. Sixty-four of the 70 isolated Acinetobacter spp. strains were multidrug resistant. Since tigecycline and colistin were not available in our country during this study, sensitivity for these two drugs was not investigated.

Table 3

Antibiotic susceptibilities of Acinetobacter spp. strains

Antibiotics	%
Netilmicin	93
Cefepime	69
Piperacillin–tazobactam	65
Ceftazidime	50
Ampicillin–sulbactam	48
Imipenem	39
Meropenem	36
Amikacin	34
Cefoperazone	23

Clinical success after treatment was achieved in 26 patients (41.3%) from whom Acinetobacter spp. were isolated, but this rate was reduced to 22 patients (34.9%) after the follow-up (sixth week). In the group from which an agent other than Acinetobacter spp. was isolated, clinical success rates were 43 and 32%, respectively, and no significant difference was determined between these rates (P = 0.8).

Forty-one of 63 patients (65%) died during the six-week period. According to the Kaplan-Meier survival analysis, the survival rates at the 3rd, 7th, 14th and 42nd days were 87, 76, 65 and 35%, respectively. No significant difference was determined in terms of survival rates between the group in which Acinetobacter spp. were isolated and the groups in which agents other than Acinetobacter spp. were isolated [Figure 1]. When Acinetobacter spp. strains were evaluated for sensitivity and resistance to imipenem, no difference was demonstrated in terms of survival in patients with pneumonia in whom non-Acinetobacter spp. were grown (P = 0.77).

Figure 1

Survival analysis of Acinetobacter spp. and non-Acinetobacter spp. groups

In the Acinetobacter spp. group, only Acinetobacter spp. were isolated from 34 patients, and in 29 patients, the isolated agents were polymicrobial. In the control group, a single agent was isolated in 62 patients, while polymicrobial agents were isolated from 15 patients. In neither group were the effects on mortality dependent upon the causative agent(s) being single or polymicrobial.

In both the Acinetobacter and non-Acinetobacter spp. groups, there were no significant differences among the patients who were exitus before receiving the appropriate antibiotics (P = 0.57). It was found that the appropriate antibiotic treatment for Acinetobacter spp. was significant in the survival rate in the Acinetobacter spp. group (P < 0.001), while it was nearly significant in the non-Acinetobacter spp. group (P = 0.054). However, among the patients who were appropriately treated with antibiotics in both groups, no significant differences in terms of survival were found (P = 0.20). For the Acinetobacter spp. group, mortality increased 5.01 (95% CI, 2.47–10.18) times when the patient did not receive the appropriate antibiotics [Figure 2].

Figure 2

The effect on survival of taking an appropriate antibiotherapy in patients with Acinetobacter spp. pneumonia

The impact of the variables on the six-week survival in patients with Acinetobacter spp. pneumonia was analysed by univariate Cox regression analysis [Table 1]. Variables with a P-value of < 0.1, according to the univariate analysis, were evaluated with multivariate analysis. Hypoalbuminemia (P = 0.04), steroid use (P = 0.002), and the use of a mechanical ventilator (P = 0.036) were determined as factors that independently affected survival. The risk of mortality was increased 3.24-fold by hypoalbuminemia (95% CI, 1.05–9.96), 3.07-fold by steroid use (95% CI, 1.49–6.34) and 2.19-fold by the use of a mechanical ventilator (95% CI, 1.05–4.58; Table 4). The effect of Acinetobacter spp. bacteriemia on mortality was not found.

Table 4

Evaluation of risk factors via multivariate analysis found to be significant by univariate analysis which impacted mortality

Variables	Univariate analysis			Multivariate analysis
	P	HR	95% CI	P	HR	95% CI
Severe pneumonia	0.05	2.19	0.97–4.97	0.43	1.48	0.55–3.99
CRF	0.07	1.95	0.94–4.04	0.08	1.99	0.91–4.35
Hypoalbuminemia (<3.5 mg/dl)	0.08	2.26	0.88–5.81	0.04	3.24	1.05–9.96
Risk for aspiration	0.04	2.65	1.03–6.81	0.87	1.09	0.36–3.24
Use of H2 blockers	0.09	1.71	0.91–3.19	0.80	1.09	0.52–2.31
Use of steroids	<0.01	3.11	1.64–5.90	<0.01	3.07	1.49–6.34
Use of sedatives	0.02	2.86	1.11–7.36	0.25	1.92	0.62–5.95
MV	0.01	2.34	1.22–4.50	0.03	2.19	1.05–4.58

HR = Hazard ratio, CI = Confidence interval, CRF = Chronic renal failure, MV = Mechanical ventilator

Discussion

This prospective observational study investigated 140 patients with a diagnosis of hospital-acquired pneumonia (HAP). We intended to determine whether Acinetobacter spp. pneumonia is different from HAP caused by other agents with regard to therapeutic success and survival rate. We found that previous antibiotic use and the risk of aspiration were independent predictors of the development of Acinetobacter pneumonia, but we did not find differences in the clinical success or in the six-week survival rates.

We found only two studies comparing Acinetobacter spp. pneumonia and non-Acinetobacter spp. pneumonia in English literature.[34] However, all studies were performed in intensive care units (ICUs) on intubated patients. The difference between the present study and the other studies is that the present study not only included all HAPs developed within the entire hospital, but also investigated the patients with HAP, VAP and pneumonia occurring during immunosuppressive treatment.

Risk factors for Acinetobacter spp. pneumonia were defined as neurologic problems and aspiration, previous antibiotic use and being connected to a ventilator.[29-12] In the present study, aspiration increased the risk for Acinetobacter spp. pneumonia nearly 3-fold, and previous antibiotic use increased the risk for Acinetobacter spp. pneumonia nearly 2.5-fold. Similar to our study, another study which evaluated VAPs that grew Acinetobacter spp. (n = 41) versus non-Acinetobacter spp. (n = 40) found that previous antibiotic use was determined to be a risk factor for ventilator-associated Acinetobacter spp. pneumonia, according to multivariate analysis.[3] In another study, which analysed 46 VAPs associated with Acinetobacter spp. and 79 VAPs associated with other pathogens, previous ceftriaxone and ciprofloxacin use were determined to be significant risk factors.[4]

Acinetobacter spp. pneumonia was recently identified as an important cause of mortality, particularly among patients who acquire pneumonia in ICUs. We found the high proportion of Acinetobacter spp. pneumonia among HAP in our institution to be very significant. In 63 of the 140 patients with HAP included in the study, Acinetobacter spp. were the responsible agents. This situation shows that, in our hospital, there are problems with regard to infection control measures and antibiotic use.

Infections caused by MDR Acinetobacter spp. are difficult to treat and are associated with high mortality. Carbapenems are frequently used in such patients; however, resistance develops rapidly.[13] The Acinetobacter spp. strains grown in the present study were susceptible to netilmicin (93%) and cefepime (69%). The susceptibility rate to imipenem was 39%. MDR Acinetobacter spp. strains were grown in 64 (91%) patients. In a study previously performed in our hospital, the susceptibility of imipenem in Acinetobacter spp. strains between 1994 and 1995 was 100%, which was then reduced to 35% between 2003 and 2004.[14] In Turkey, in a study performed on VAPs caused by Acinetobacter spp. strains, resistance to ceftazidime, imipenem and ciprofloxacin was determined to be 60, 64 and 80%, respectively, and the most susceptible antibiotic was cefoperazone-sulbactam.[15] Since tigecycline and colistin were not available in our country at the time of this study, sensitivities for these two drugs were not investigated. When it is considered that most Acinetobacter spp. cases are resistant to most drugs, the use of these antibiotics could affect the results of Acinetobacter spp. pneumonia treatment.

In patients in whom Acinetobacter spp. was isolated, the clinical success rate after treatment was 41%, which was reduced to 35% after follow-up. No significant difference was shown between the groups in which Acinetobacter spp. and non-Acinetobacter spp. were isolated in terms of clinical success, after both treatment and follow-up. Although the number of patients in whom pneumonia developed while under immunosuppressive therapy was significantly higher in the non-Acinetobacter spp. agent group, no significant difference was found between the two groups in terms of survival rates. There was also no significant difference in terms of survival rates between the patients with pneumonia caused by imipenem-susceptible or -resistant Acinetobacter spp. strains and the patients with non-Acinetobacter spp. In a study performed in patients with VAP, survival was evaluated between groups in which Acinetobacter spp. (imipenem-susceptible - imipenem-resistant) and non-Acinetobacter spp. were isolated. Similar to the present study, no difference was found between the two groups in terms of survival.[31617] In the present study, we also found that receiving the appropriate antibiotics is a factor affecting survival in the Acinetobacter spp. group.

When the mortality-associated risk factors were evaluated, it was shown that hypoalbuminemia, steroid use and the use of a mechanical ventilator increased mortality. Although there are studies investigating the risk factors affecting in-hospital mortality in ventilator-associated pneumonias caused by Acinetobacter spp., no study evaluating mortality-associated risk factors in patients with only HAP and with Acinetobacter spp. growth exists in the literature.[318]

In conclusion, Acinetobacter spp. pneumonia does not differ from HAPs caused by non-Acinetobacter spp. agents in terms of therapeutic success and survival rate. Patients with HAPs caused by Acinetobacter spp. have a high risk of aspiration, the incidence of which is gradually increasing in patients who have previously received antibiotic therapy.