Carbapenemase Of Intestinal Rods – The Beginning Of Post-Antibiotic Era?

One of the most clinically and epidemiologically important mechanism of resistance to carbapenems is β-class metallo-beta-lactamases (MBL) produced mainly by notfermentative rods e.g. Pseudomonas aeruginosa and less frequently by bacteria of the Enterobacteriaceae family, e.g. K. pneumoniae or E. coli. Furthermore, MBL-producing strains are classified as resistant to penicillins (including β-lactamase inhibitors), cephalosporins and carbapenems with the exception of aztreonam, although the use of monobactams in therapy should be based on the result of microbiological quantification. Importantly, metallo-β-lactamases require zinc ions as cofactors for the hydrolysis reaction of β-lactam antibiotics [27, 53, 57, 73]. Three subclasses are currently distinguished on the basis of differences in the amino acid sequence and the structure of the active centre. The B1 and B3 subclasses have two bivalent zinc ions in the active centre, while the B2 subclass only has one zinc ion [48]. Genes responsible for the production of metallo-β-lactamases are found on the chromosome, plasmids or integrons. In environmental bacteria, e.g. Stenotrophomonas maltophilia, Bacillus cereus, Caulobacter crescentus or Aeromonas hydrophila, many species-specific MBL chromosomally encoded enzymes have been discovered [53, 57]. At present, the following types of MBL with clinical significance are distinguished, i.e. IMP, VIM, NDM, SPM, IND, GIM, SIM, KHM, AIM or DIM [12, 16, 53, 73].

The first case of bacteria with the acquired MBL mechanism, i.e. IMP-1 was detected in P. aeruginosa in 1991 in Japan. Subsequently, in 1997 the VIM-1 type was identified in Italy, and SPM-1 in Brazil. It should be stressed that since then, and especially since the second half of the 1990s, the number of MBL (+) strains causing epidemic outbreaks (Greece, Italy, Canada, Korea, Kenya) has been growing steadily [16, 29, 53, 80, 85, 89]. In addition, over the period of 2002–2006 in Tunisia, 35 cases of P. aeruginosa strains producing MBL mechanisms of VIM-2 type were recorded. Importantly, the gene cassette bla_VIM-2 was located within the integron of class I [29].

In turn, Papa Ezdra R. et al. characterized the bacterium P aeruginosa VIM-2 (+) isolated in Uruguay (2011–2013). Based on the MLST method, most of the tested bacteria were classified to the following sequential type, i.e. ST155 and, to a lesser extent, ST1565 or ST1195. It should be mentioned that P. aeruginosa ST155 bacterium producing VIM-2 was also detected in Spain [65].

The mechanism of MBL was identified for the first time in Poland among P. aeruginosa bacilli in 1998–2000 [53]. In turn, during the period of 1998–2006, a total of 20 strains of the genus Pseudomonas spp. (P. aeruginosa n = 18, Pseudomnas putida n = 2) producing MB-type carbapenemases were isolated at the Centrum Zdrowia Dziecka in Warszawa [Children’s Health Institute in Warsaw]. Using the PCR method among the majority of bacteria the bla_VIM-4 genes were found while in two isolates bla_VIM-2. In addition, the genes encoding MBL were located on the chromosome, and in all analysed bacteria, class 1 integrons were detected [66]. In north-eastern Poland (September 2012 – December 2013), 45 strains of P. aeruginosa MBL (+) were detected in a succession from patients hospitalized at the University Clinical Hospital in Białystok. Eventually, resistance genes i.e. bla_VIM-2, were identified in three isolates, whereas bla_VIM-4 was identified in only one strain. Importantly, in the bacterium P. aeruginosa, the bla_VIM genes were located in the class 1 integron. The analysis of genetic relatedness using the PFGE method allowed for classifying the tested VIM (+) strains into four unrelated pulsotypes (AD). In turn, the MLST technique enabled the isolation of the following sequential types, i.e. ST111, ST27, ST17 and interestingly ST2342. It is worth mentioning that in the isolate of Ps21 (ST2342) VIM-2 (+) a unique sequence of gene cassettes was found, previously not described in P. aeruginosa [52]. In turn, the first strain of K. pneumoniae producing the VIM-4 type was identified in 2008 in Bydgoszcz in a 61-year-old patient undergoing a surgical operation [53, 79].

In summary, eight strains of MBL (+) were isolated in Poland from 2006 to 2008, and in 2009, 2010 and 2011 respectively: 22, 23, 31 of 39 hospitals located in 24 cities. The dominant producers of the MBL mechanism were: E. cloacae – 50.0%, S. marcescens – 17.9%, K. oxytoca – 16.7% or K. pneumoniae – 11.9%. The vast majority of isolates produced the VIM type MBL mechanism (93.8%), whereas in the three S. marcescens strains, the IMP type was identified [12].

Another mechanism, i.e. KHM-1 was first reported in Citrobacter freundii in Japan (1997) while GIM-1 in P. aeruginosa in Germany (2002). Recently, the percentage of the isolation of GIM-1 strains has also been increasing, this mechanism has been identified, among others, in the following bacteria, i.e. Serratia marcescens, E. cloacae, P. putida, Acinetobacter pittii, and, what is more dangerous, in E. coli, Klebsiella oxytoca or C. freundii. All cases of GIM-1 (+) strains were registered in Germany, mostly in Düsseldorf or in locations lying 40 km away from that city [38, 40, 60]. Other variants of metallo-β-lactamases, i.e. SPM-1, SIM-1, DIM-1, TMB-1 and AIM-1 are produced by notfermentative rods, i.e. Pseudomonas spp., Acinetobacter spp. and until now these mechanisms have not been detected in the bacilli of the order Enterobacterales [48, 60].

In 2008, the first case of bacterium from the family Enterobacteriaceae was recorded with a new type of MBL resistance – NDM-1 enzyme (New Delhi Metallo-β-lactamase), encoded by the gene bla_NDM-1, hydrolysing all β-lactam antibiotics except aztreonam [10, 57, 68]. It is also worth emphasizing that currently as many as 16 variants of NDM are distinguished, with the presence of NDM-1 in microorganisms being established in the overwhelming majority of cases [45]. NDM-1 is a monomer with a molecular weight of 28 kDa which, compared to the VIM-1/VIM-2 type, exhibits only 32.4% similarity. In addition, NDM-1 binds cephalosporins more tightly than VIM-2, and in particular cefuroxime, cefotaxime and cephalothin, as well as penicillin and, which is interesting, does not bind to carbapenems as strongly as IMP-1 or VIM-2, but still hydrolyses them in a similar amount [48, 91]. Moreover, in the bacilli of K. pneumoniae bla_NDM genes are located mainly on the following plasmids, i.e. IncA/C, IncF, IncR, IncH, IncN, IncL/M or IncX [45].

It should be noted that among the NDM (+) strains, the presence of other determinants of antibiotic resistance has been established a number of times, among others resistance to aminoglycosides (16 S RNA methylase), quinolones (Qnr), macrolides (esterases), rifampicin, chloramphenicol or sulfamethoxazole [60]. For example, carbapenemase production in K. pneumoniae strains is presented in Tab. I.

The new NDM-1 enzyme was first detected in K. pneumoniae isolated from the urine of a 59-year-old male (Sweden), hospitalized in India in January 2008 [10, 57, 68]. In addition, the patient was found to be an E. coli NDM-1 (+) carrier, with the bacterium present in faeces [10]. After the first report there were further reports testifying to the spread of NDM-1 (+) strains within various species of Enterobacterales in countries such as India, Pakistan and Bangladesh, at the same time indicating the above-mentioned countries as the primary source from which NDM strains −1 (+) reached European countries, USA, Australia or African countries [44, 57, 59]. Most likely, the first bacterium to acquire the bla_NDM-1 gene was Acinetobacter baumannii [57, 68]. In Poland, the first strain of New Delhi Metallo-β-lactamase appeared in August 2011 in E. coli isolated from a 53-year-old man of Polish citizenship transported from Congo to an Intensive care Unit (ICU) in Warsaw. The isolated E. coli 5428/11 strain produced simultaneously NDM-1, CTX-M-15, TEM-1 and OXA-1 and was characterised by susceptibility to only to tigecycline and colistin. Despite 12 days of colistin treatment, the patient died as a result of multiple organ failure. What is significant, transmission of E. coli MBL (+) to other patients at the ICU ward did not take place at the hospital [21]. In Poland, successive strains of NDM (+) (K. pneumoniae ST11) were isolated from 4 patients hospitalized in 2012 (November-December) in Poznań. What is more, the above-mentioned patients did not travel in 2012 [4, 33]. Since 2013, the number of NDM (+) strains isolated in Poland has increased at an alarming rate. In 2013–103 cases were recorded, while in 2014–247. It is worth noting that from 2012 to 2014, two main outbreaks were observed in Poland with the epicentre in Poznań (n = 176) and Warsaw (n = 191), where in most cases the pandemic clone of K. pneumoniae ST11 was the alarm factor [4]. Then, in 2015, 470 NDM (+) strains were recorded successively, and in 2016 as many as 1771. In summary, in total over the period of 2011–2016, 2,596 cases of isolating bacteria with the NDM-1 resistance mechanism were established. Moreover, 93.3% of these patients were hospitalized, 4.35% were outpatients and only 2.3% were people staying at care and curative institutions, nursing homes or hospices. In addition, carriers were found in 57.3% of cases, and in 42.7% of patients additionally signs of infection occurred. Taking into account the age of patients, up to 65% of patients were people > 65 years of age [94]. In 2016, the highest percentage of the isolation of NDM (+) strains in the country was recorded in Masovian (n = 1394) and Podlaskie (n = 260) provinces [94].

It should be pointed out that in the first quarter of 2017, the number of 785 patients in Poland in whom Enterobacterales occurrence was found, mainly K. pneumoniae strains producing New Delhi type carbapenemase, raised great concern. In the provinces: Masovian n = 545 cases, followed by Podlaskie n = 186, Warmian-Masurian n = 18, and Świętokrzyskie n = 10 cases were recorded. In addition, the number of centres in which NDM bacteria were identified in and around Warsaw was n = 74, while in Podlaskie Province n = 17. It should be noted that in the first quarter of 2017, compared to the first quarter of 2016, throughout the country an alarming increase was recorded in isolating bacteria of the order Enterobacterales, mainly NDM-producing K. pneumoniae from 310 to 785 (150%). It is worth emphasizing that the number of provinces in which NDM (+) bacteria were found to rose from 8 to 13. The most significant changes were recorded in Masovian Province, where the incidence of NDM-producing strains increased from 273 to 545 cases, while in Podlaskie Province from 26 to 186. In addition, it should be concluded that the epidemic spread of NDM-producing K. pneumoniae occurs in these provinces. The number of patients with infections in the first quarter of 2017 in Masovian Province amounted to 384 (n = 193 – the first quarter of 2016), while in Podlaskie 121 (n = 13 – the first quarter of 2016). What is more, there was also a sharp increase in the cases of isolating NDM-producing Enterobacterales in the following provinces: Warmian-Masurian and Świętokrzyskie [95]. In turn, in the second quarter of 2017, there was a disturbing increase in the frequency of isolating NDM (+) bacteria in the following provinces: Pomeranian, Lower Silesian, West Pomeranian and Wielkopolskie. While in Małopolskie Province it took place in the third quarter of 2017.

In summary, in the first three quarters of 2017, a record number of NDM (+) isolates in Poland, i.e. n=2512, was confirmed by KORLD (National Reference Centre for Microbial Susceptibility to Medications). In total, in the first quarter n = 991 strains producing the NDM mechanism were recorded, and in the second and third quarters n = 1016 and n = 505 cases, respectively. The highest incidence of the rods of Enterobacterales NDM (+) was recorded in Warsaw, i.e. I quarter n = 365, II quarter n=422, III quarter n = 137. In turn, in Masovian Province (excluding Warsaw), in the first quarter the number of NDM carbepenemase-producing bacteria was n = 353, in the second quarter n = 304 and in the third quarter n = 137. Whereas in Podlaskie Province, n = 207 NDM (+) bacteria were identified in the first quarter, and in the second and third quarters, n = 168 and n = 196 strains respectively [97].

The report by KORLD points to the conclusion whereby in Poland in 2017 the majority of the bacilli of Enterobacterales NDM (+) were K. pneumoniae bacteria (n = 2291, 91.2%). The remaining species of Gramnegative bacteria were isolated with a much smaller percentage share, i.e. E. coli (n = 18, 0.7%), E. cloacae (n = 3, 0.1%) or C. freundii (n = 1, 0.03%) [97].

The incidence of strains producing the NDM mechanism in Poland over the period of 2011–2017 is shown in Fig. 1.

Fig. 1.

Undoubtedly, a huge increase in the isolation of strains producing the New Delhi-type metallo-β-lactamase in Poland, and more frequent occurrence of epidemic centres in various regions of the country, or finally new clones of bacteria (e.g. K. pneumoniae – ST147, E. coli – ST410, ST448, ST405, ST131) producing the above mechanism of resistance, caused it to become one of the most significant current problems in the field of drug resistance in our country [94]. The main producers of NDM (+) are bacteria from the Enterobacteriaceae family, i.e. K. pneumoniae and E. coli [8, 48]. Nevertheless, carbapenemases were also detected in the following bacteria: K. oxytoca, E. cloacae, C. freundii, Proteus mirabilis, Salmonella spp., Providencia spp. and with a slightly smaller percentage share in Acinetobacter spp. or P. aeruginosa [18].

European reports show that in 2013, only two countries recorded single hospital outbreaks caused by microorganisms producing the NDM mechanism. In turn, in 2015, five countries reported the occurrence of single nosocomial epidemic centres caused by these strains, while regional and transregional dissemination was reported in seven countries [2]. What is more, the data published by the Chief Sanitary Inspectorate indicate that in 2016 there were 35 outbreaks found to be caused by the NDM-producing K. pneumoniae bacterium. In addition, compared to 2015, there was a significant increase in the number of epidemic centres caused by these microorganisms from 1.9% to 6.6% [25]. In turn, in 2017, 63 epidemic centres caused by NDM-producing K. pneumoniae were reported (a total of 557 people were infected). In addition, taking into account the percentage of individual etiological factors of the epidemic centres of K. pneumoniae MBL (+) constituted 4.97% of isolates, including strains producing NDM-type carbapenemases of 2.57% [26]. Cases of the occurrence of K. pneumoniae NDM (+) strains in the world are presented in Tab. I. In turn E. coli NMD (+) strains have been isolated, e.g. in the following countries, i.e. India, Canada, Cameroon and Europe (Great Britain, Belgium, Sweden, France, Austria, Norway, Germany, Poland) [21, 48].

An intriguing case of isolating carbapenemase-producing Enterobacteriaceae family strains was reported in the case of two Polish tourists with gunshot wounds after terrorist attacks at the Bardo Museum in Tunis in March 2015. The above-mentioned patients were transported to Warsaw after hospitalization in Tunisia. From the first of these patients, K. pneumoniae resistant to carbapenems was isolated. The microbiological analysis of the studied strain demonstrated a positive result in the Carba NP and MBL phenotypic test with EDTA as well as resistance to temocillin, which suggested that OXA-48 was produced. Nevertheless, the PCR analysis only confirmed the presence of the bla_NDM gene within the Tn125 transposon. In addition, the mLSt technique classified the strain under study as ST147. Moreover, in the hospitalized patient, other carbapenemase-producing strains were not isolated at the time of admission and hospitalisation [36]. In the case of the second patient, as a result of a gunshot, there was a serious damage to the subcutaneous tissue within the femoral trochanter. At the time of admitting the patient to the hospital, a carbapenem-resistant strain of K. pneumoniae was found in the rectal swab. It should be noted that the Carba NP test result was positive, furthermore resistance to temocillin was demonstrated, whereas the MBL phenotype tests were negative. The PCR method and sequencing confirmed the presence of the gene bla_OXA-48 located within the transposon Tn1999.2. Nevertheless, 10 days after admitting the patient to the hospital, NDM-1 (+) strains were cultured from the wound swab, i.e. K. pneumoniae and E. coli. Further MLST analysis demonstrated that these E. coli strains belonged to ST410, while K. pneumoniae to ST147. It should be emphasized that K. pneumoniae isolates originating from the two Polish tourists produced the NDM-1 mechanism and also belonged to the same sequential type, i.e. ST147, indicating that the colonization of the patients during the hospitalization in Tunisia was most likely in both cases [36]. The susceptibility tests of the above-mentioned strains showed multi-drug resistance, and the only effective antibiotic was colistin. In addition, K. pneumoniae isolates were found to be susceptible to chloramphenicol. In the case of amikacin, considering the MIC values, the strains under study were classified as susceptible or intermediate [36].

An equally unusual case of simultaneous isolation of several carbapenemase-producing strains from a patient was recorded in China. A 46-year-old man was admitted to hospital in June 2012 because of headaches, nausea, vomiting and eventually suspicion of meningitis. It should be added here that in the following days, the patient was transported to Shanghai, where during a two-month stay in hospital, a total of 34 strains of Gram-negative bacteria were isolated. Interestingly, the ones identified were, among others, KPC-2 producing K. pneumoniae species, followed by E. coli NDM-1 (+), IMP producing Enterobacter aerogenes and A. baumannii OXA-23 (+) [17].

Another noteworthy case was reported in 2015. A 74-year-old woman of Danish nationality was admitted to a hospital in India, where she was hospitalized with myocardial infarction and hypercholesterolemia. Then 10 days later the patient was transferred to a hospital in Denmark. Importantly, using the MALDI-TOF, PCR and MLST techniques among the isolates tested the following were identified: K. pneumoniae ST147 bacterium producing NDM-7 and OXA-181, E. coli ST1284 NDM-5 (+) and A. baumannii ST2 strains producing OXA-23 [30].

Class A carbapenemases represented by 62 β-lactamases were divided into 6 subtypes i.e. IMI/NMC-A enzymes, SME enzymes, GES enzymes, KPC enzymes, SFC-1 and SHV-38. The KPC (Klebsiella pneumoniae carbapenemases) enzymes currently play the greatest role in all of the class A carbapenemases. Strains carrying the bla_KPC gene are characterized by possession of resistance to β-lactam antibiotics, as well as additionally other mechanisms, which ultimately leads to therapeutic failures [38]. The main producer of KPC-type carbapenemases is K. pneumoniae, although these enzymes were also found in other bacterial species belonging to the order Enterobacterales, e.g. K. oxytoca, E. coli, Enterobacter spp., C. freundii, S. marcescens, Salmonella enterica, and even in non-fermenting bacilli such as P. aeruginosa or P. putida [27, 57, 72, 78]. As in the case of bacteria that produce other mechanisms, a diversified level of resistance is observed here as well. However, the KPC enzymes of which 22 are currently known have the widest substrate spectrum among the described β-lactamases [57, 58, 78]. The global occurrence of K.pneumoniae strain KPC (+) is shown in Tab. I.

Literature analysis indicates that K. pneumoniae strains producing the KPC mechanism are the most common etiologic factor of vascular bed (52%) infections, followed by respiratory tract (30%) or urinary tract (10%) infections [83]. In turn, data obtained by Campos A.C. et al. demonstrated that K. pneumoniae KPC (+) strains are mainly isolated from rectal swabs (32%), followed by circulatory and urinary systems (24%), lower respiratory tracts (21%) or postoperative wounds (10%) [11]. The KPC mechanism was first identified in K. pneumoniae in 1996 in North Carolina (USA). The KPC (+) strains were rarely isolated in the United States until 2005, when several outbreaks were discovered in hospitals in New York and New Jersey. Since then, we have witnessed a rapid spread of these bacteria in the population, as evidenced by the isolation of > 1,200 KPC-producing strains from blood samples at a hospital in New York in 2012 [61].

It is worth mentioning that in Israel, as early as in 2006, an increasing number of epidemic outbreaks in hospitals, caused mainly by K. pneumoniae KPC strains were found. The average number of new cases was as follows: 2005 – 6 cases (1.9 cases per 100,000 patients); first half of 2006 – 39.5 cases (11.8 cases per 100,000 patients); second half of 2006 – 89 cases (27 cases per 100,000 patients); first quarter of 2007 – 143 cases (41.9 cases per 100,000 patients). Importantly, the number of new clinical isolations of CRE increased sharply in the second half of 2006 and in the first quarter of 2007, the peak value was recorded in March, i.e. 55.5 cases per 100,000 patients. In connection with the above, the Ministry of Health (MH) issued a regulation in March 2007 aimed at limiting the scale of this phenomenon. A specially appointed infection control team assumed supervision over 27 Medical Care Hospitals. It is worth noting that by 31 March 2007, a total of 1,275 patients from the aforementioned 27 hospitals (175 cases per 1,000,000 population) had been found to carry carbapenemase-producing strains. Prior to MH intervention, the monthly frequency of CRE isolation was 55.5 per 100,000 patients, respectively. Thanks to the implementation of the Ministry of Health guidelines, a sharp increase in the number of outbreaks was halted, and by May 2008 the monthly number of new cases had decreased, eventually reaching the level of 11.7 per 100,000 patients (p < 0.001). The above effect was achieved primarily by following the recommendations of the MH, i.e. by isolating patients infected or colonized with CRE strains, and the separation of individual personnel and medical equipment, which prevented the further transmission of the bacteria. Of particular importance was the fact that infection control teams also played an enormous role, conducting a series of visits to supervise the use of the above-mentioned procedures or compliance with mandatory laboratory tests [77].

In Europe, the first KPC (+) strains appeared in Greece (Crete) in 2007, where an epidemic outbreak occurred, involving a total of 22 patients [50]. In turn, over the period from January 2007 to December 2008, another epidemic outbreak was found in a hospital in Greece (Athens) involving a total of 50 patients colonized or infected with K. pneumoniae KPC-2. Finally, the mortality rate among the patients of the ICU was 58.8% (n = 34), while in patients hospitalised in other wards it was 37.5% (n = 16) [81]. Moreover, the literature data shows that in 2007–2008 an epidemic outbreak caused by bacteria producing KPC-2 was observed in two more hospitals (Crete, Thessaloniki). The analysis carried out from February to December 2008 showed that in 18 hospitals in Greece (Athens n = 14, Crete n = 3, Thessaloniki n = 1) the occurrence of CRE bacteria was determined in 173 patients, i.e. K. pneumoniae KPC-2 (+) [23].

Analysing the data obtained through the MOSAR project (Mastering Hospital Antimicrobial Resistance in Europe) it was found that over the years 2008–2011 in patients hospitalized in Intensive Care or Rehabilitation Wards in 18 hospitals in Europe, the strains of K. pneumoniae KPC (+) constituted the largest group among the carbapenemase-producing bacteria. Analysing the data in detail, it was shown that in Greece the ST258 clone (KPC-2) prevailed, while in Italy ST512 (KPC-3) was dominant. In turn, in Israel, a huge variation of K. pneumoniae bacterium was found, i.e. ST512, ST36, ST258, ST383, ST833, ST17 or ST34 [5].

In addition, the data obtained by Agodi A. et al. showed the isolation of 24 strains of K. pneumoniae from 16 patients hospitalized in an ICU in Italy (Catania) in 2009. It should be noted that all the strains under study produced KPC-3 carbapenemase and were classified as ST258. At the ICU ward, thanks to the implemented preventive measures and strict monitoring of the compliance of the medical staff with the recommended procedures, the occurrence of CRE strains was curbed even despite the constant admission of new patients. The above fact clearly emphasizes that optimized measures are effective in reducing the spread of antibiotic-resistant bacteria [1].

In our country (Warsaw) in 2008, the first strains of K. pneumoniae KPC (+) were isolated from a 56-year-old patient (who had not been travelling recently before that time), admitted to the Cardiology Ward from another Warsaw hospital with pneumonia of undetermined aetiology. The isolated bacteria belonged to the ST258 clone, producing both KPC and SHV-12 [6, 12]. Then in Poland 33 strains with the KPC mechanism (K. pneumoniae n = 30, K. oxytoca n = 3) from 32 patients in five hospitals in Warsaw had been isolated by the end of 2008. From that moment on, a rapid spread of the bacterium throughout the country took place. Moreover, 86 KPC (+) strains were identified in 82 patients in 2009 (K. pneumoniae n = 84, E. coli n = 2, in the case of E. coli, the bacteria were isolated from patients simultaneously colonized by K. pneumoniae). Molecular analysis showed that 97.4% of K. pneumoniae strains (2008–2009) belonged to the ST258 clone and the bacteria were classified as ST11 or ST23 only sporadically. In turn, two strains of E. coli belonged to ST93 and ST224 and most likely acquired bla_KPC genes from K. pneumoniae. As many as 153 bacteria (mostly K. pneumoniae, with single E. coli) producing the KPC mechanism were identified in 2010, of which the majority of isolates came from Mazovia (n = 126 bacteria from 32 centres). The KPC (+) strains also appeared at the time in Świętokrzyskie Province (April 2010 – February 2011) in both medical centres and nursing homes. Thanks to the implementation of the national guidelines on infection control in 2011, the total number of isolations of KPC (+) strains decreased, ultimately amounting to 104. In March 2011, a new threat was observed in Podlaskie Province, where 29 strains from 10 hospitals had been isolated by the end of the year, mainly in Białystok [12, 96]. In summary, KPC strains were isolated from 371 patients over the years 2008–2011, from at least 58 hospitals in 34 locations, in addition, three regional outbreaks from the epicentre in Warsaw and Białystok were observed. In our country, the most cases of strains with the KPC mechanism appeared in centres in Warsaw and other cities of Masovian Province. However, it should be noted that the recorded cases also concerned the patients of clinics, dialysis centres or people staying in nursing homes. The reports from 2012 show that outbreaks of K. pneumoniae KPC (+) infections were recorded in four regions of the country; for comparison 5 outbreaks were recorded in 2013 [33]. In summary, KPC (+) strains were isolated in a total of 608 patients over the years 2010–2014. The vast majority of the above-mentioned patients were hospitalized in Masovian Province, with a slightly smaller percentage in Lublin, Silesian, Podlaskie and Świętokrzyskie Provinces. Moreover, in 21 cases patients came from other regions of Poland, among others Olsztyn, Łódź or Kraków. In addition, the status of KPC (+) strain carrier was observed in 209 patients, while in 399 patients there occurred symptoms of infection, i.e. urinary tract (51.6%), respiratory system (21.6%), skin and soft tissue (15.0%), blood and vascular bed (10.1%) and the so-called intra-abdominal infections (1.7%) [7]. Genetic analysis has shown that over the period of 2008–2009 epidemic outbreaks in Warsaw and its vicinity were caused by the strain K. pneumoniae ST258 producing the KPC-2 mechanism. Importantly, in other regions of Poland, infections were caused by K. pneumoniae ST258 or ST512, KPC-3 (+). Moreover, in hospitals in Warszawa, the KPC-2 mechanism was also identified in C. freundii ST17 and E. cloacae ST254 [7].

It is worth mentioning that in 2010 a case of isolating 4 strains of K. pneumoniae producing simultaneously KPC carbapenemase and 16S rRNA ArmA methylase was described. It was the first report of the occurrence of these strains in Poland as well as in Europe. The isolated bacteria carried bla_KPC-2 and armA genes on 50-kb and 90-kb plasmids. The researchers suggest that the simultaneous production of KPC-2 and 16S rRNA ArmA methylase is a new strategy enabling K. pneumoniae to survive in hospital conditions even despite the application of aminoglycosides and carbapenems in treatment [92]. It should be noticed that in February 2014, the E. coli ST479 strain producing the KPC-3 mechanism was identified for the first time in Poland. The tested isolate was characterized by resistance to trimethoprim/sulfamethoxazole, ciprofloxacin, with susceptibility to colistin and tigecycline. In the case of aminoglycosides (gentamicin), E. coli was classified as intermediate [62]. The occurrence of the strains of the order Enterobacterales producing KPC in Poland is shown in Fig. 1.

Currently in our country, the microorganisms producing KPC-type carbapenemases raise great concerns, mainly because there are no antibiotics with proven effectiveness in the treatment of infections caused by these strains, in addition, apart from that mechanism these bacteria often produce other β-lactamases, like among others ESBL and are MDR. As a rule, apart from β-lactam antibiotics, KPC (+) bacteria are insensitive to most aminoglycosides, fluoroquinolones, tetracyclines or cotrimoxazole. Moreover, the genes encoding the KPC carbapenemases are located mainly on plasmids, thanks to which they easily transfer from one bacterium to another belonging to the same or different species [31, 34]. The best known KPC (+) clone is K. pneumoniae ST258 possessing the so-called increased epidemic potential. The strain first appeared in the USA, then in Israel, Greece, and currently it has been identified in most countries of the world, including Poland, where it is the dominant clone [31, 56]. Infections caused by KPC (+) are burdened with high mortality, in the case of K. pneumoniae it is even 50% [31, 34].

Class D carbapenemases, referred to as oxacillinases (OXA), based on the amino acid sequence, have been divided into 12 subgroups, i.e. OXA-23, OXA-24/40, OXA-48, OXA-51, OXA-58, OXA-134a, OXA-143, OXA-211, OXA-213, OXA-214, OXA-229 and OXA-235. However, only a few of them have been identified in K. pneumoniae, i.e. OXA-23, OXA-48, OXA-51 and OXA-58. It should be noted that OXA-48 is the most widespread D-class carbapenemase. In 92.5% of strains isolated in Europe, as well as in North Africa, the gene bla_OXA-48 is located on IncL/M, but there have also been reports of the presence of this gene within IncA/C, IncH or Tn1999. What is more, currently 10 variants of this gene are distinguished [45].

The main producer of OXA-48 is the species K.pneumoniae, nevertheless the majority of the bacteria of the order Enterobacterales have been found to produce this type of carbapenemase. It is important to know that the molecular and epidemiological analysis carried out in Germany, demonstrated a horizontal transfer of the gene bla_OXA-48 between K. pneumoniae and E. coli. In addition, apart from E. coli, the OXA-48 enzymes have also been identified in the following bacteria, i.e. K. oxytoca, Enterobacter spp., Providencia rettgeri, C. freundii or S. marcescens [45]. The OXA-48 mechanism was first detected in a strain of K. pneumoniae in 2003 in Turkey. Since that year, the endemic occurrence of the rods producing OXA-48 has been found, among others, in countries such as Turkey, Morocco, Libya, Egypt, Tunisia or India [45]. It should be noted that in some European countries the percentage of OXA-48 producing strains is equally high. For example, the data from Spain (2013) and France (2011–2012) indicates a 74.7% and 78% incidence of this mechanism in carbapenemase-producing strains [45, 64, 74]. In addition, the results obtained in Spain (2013) from 83 hospitals point out that K. pneumoniae produced the OXA-48 mechanism in as much as 63%. Furthermore, in 2013-2015 at Hospital Universitario de Canarias, 267 carriers of class D carbapenemase-producing bacteria were reported. In 100 patients (116 episodes), OXA-48 (+) strains were isolated from the following clinical materials, i.e. in 43.42% of samples from urinary tract infections, subsequently from the operated site (17.17%) or blood (17.0%) [47]. It should be emphasized that in EU countries such as Spain, France, the United Kingdom or Germany, hospital outbreaks caused by OXA-48 (+) bacteria have been registered [7]. The list of countries in which the strains of OXA-48 (+) K. pneumoniae have been isolated are presented in Tab. I.

In conclusion, the European ECDC report shows that in 2013, only one European country, i.e. Malta, reported an endemic situation of strains producing OXA-48, while in 2015 two countries (Malta, Turkey) registered an endemic situation, and four countries reported transregional spreading (Spain, France, Belgium, Romania) [2].

In recent years, the first occurrences of the rods of Enterobacterales producing OXA-48 have been observed in Poland. In 2012, in Białystok, an E. cloacae strain ST89 producing carbapenemase OXA-48 was first identified in a 76-year-old patient after cardiac surgery, i.e. coronary artery bypass and mitral valve and tricuspid valve repair. After the surgery, the patient was transferred to the ICU, where after 38 days he died due to heart decompensation, clotting disorders, multiple organ failure and finally circulatory arrest. It should be highlighted that the first E. cloacae strains resistant to carbapenems were isolated from the patient on the 24^th day after the surgery from such materials like blood or bronchial secretions. Moreover, the bacterium E. cloacae also exhibited resistance to other antibiotics, i.e. tigecycline and colistin. Microbiological analysis including the Carba NP Test confirmed that the tested isolates produced carbapenemase, the OXA-48 mechanism was determined using the PCR method, and the MLST analysis demonstrated that E. cloacae MDR strains belonged to ST89 [49].

In turn, on the basis of the data obtained by Izdebski R. et al., between 2013 and January 2017, the isolation of 54 class D carbapenemase-producing strains was established in 52 patients from various Polish cities (Biała Podlaska, Bochnia, Bydgoszcz, Ełk, Grodzisk Mazowiecki, Kielce, Kraków, Poznań, Siemianowice Śląskie, Słupsk, Sosnowiec, Szczecin, Warsaw). In addition, it was reported that 14 patients had been previously abroad, while in two cases the family members had travelled. What is more, 34 patients eventually developed an infection, while in 18 only colonization was demonstrated. The isolated strains were extremely genetically diverse and represented the following species, i.e. K. pneumoniae (n = 37), where ST395 was dominant, ST11, ST15 and ST101; E. coli (n = 14) ST38, ST410, ST648; E. cloacae (n = 1) ST78; C. freundii (n = 1) ST124 and E. aerogenes (n = 1). The bacteria mentioned above produced OXA-48 (n=49), OXA-181 (n = 4) and OXA-232 (n = 1). The bla_OXA-48 genes were located mainly within Tn1999.1 (n = 29) and Tn1999.2 (n = 15). It should also be added that in 43 isolates, bla_OXA-48/181 genes were located within plasmids, whereas in 11 strains of E. coli ST38 and ST648 and one K.pneumoniae ST336 bla_OXA-48 on the chromosome [35].

It should be underlined that the bla_OXA-48 gene is most often identified in the K.pneumoniae STU clone, isolated in many countries around the world, including Spain, Greece, Taiwan, Libya, Turkey or Argentina. In addition, in 2013 in Spain, an outbreak occurred, including 44 patients caused by K. pneumoniae ST11 OXA-48 (+). It should be added that in the following clones, i.e. ST14, ST15, ST101, ST147 and ST405 the bla_OXA-48 gene has also been detected in such countries like the USA, Spain, the Czech Republic, Germany, Finland, France, India, Libya and Japan [45].

Substitution or deletion of single amino acids has led to the formation of a group of OXA-48 derivatives, which includes OXA-181, OXA-204, OXA-232, OXA-163, OXA-244 or OXA-245. For example, OXA-181 is widespread in many countries of the world, i.e. the United Kingdom, Romania, Canada, Singapore, South Korea, Japan, Australia and New Zealand. However, most of the infections have been reported in India. In turn, the OXA-204 and OXA-163 mechanisms were identified, among others, in Tunisia, France and Argentina, while OXA-244 and OXA-245 have been identified in Spain. Moreover, K. pneumoniae strains producing OXA-232 have been reported in the USA, Singapore, India or South Korea [45].

It needs be noted that currently other class D carbapenemases have also been detected, i.e. OXA-23, OXA-24/40, OXA-51, OXA-58, OXA-134a, OXA-143, OXA-211, OXA-213, OXA-214, OXA-229 and OXA-235 mainly in Acinetobacter bacteria and in particular A. baumannii, while they have not been detected in the species K. pneumoniae [45].

Colistin was discovered over 60 years ago. At first, however, it was not widely used due to its side effects, i.e. nephro- and neurotoxicity. Nevertheless, the increasing resistance to carbapenems has caused it to be currently recommended in the treatment of severe infections caused by CRE strains [45, 71]. The mechanism of action of colistin consists in its being attached to the negatively charged phosphate groups of lipid A, which is a component of lipopolysaccharide (LPS), and this results in the loss of cell membrane integrity and ultimately leads to cell death. Colistin resistance is mainly caused by the modification of the LPS molecule through the attachment of L-Ara4-N and PEtN, which results in decreasing the negative charge of the outer membrane and in the inhibition of colistin binding. These changes are mainly caused by mutations within the genes of regulatory system components, i.e. mgrB, phoP/phoQ,pmrA,pmrB,pmrC or crrABC [45, 71]. In particular, the inactivation of the mgrB gene, encoding the negative feedback of the PhoQ/PhoP regulatory system, is the most common cause of resistance to colistin among K. pneumoniae strains. Additionally, recent studies have shown that the conversion of a single amino acid in the PmrB protein leads to the overexpression of pmrCAB and pmrHFIJKLM operons which are involved in LPS modifications, eventually leading to polymyxin resistance. Moreover, recently a new plasmid has been identified, the so-called Plasmid Mediated Colistin Resistance carrying gene – mcr-1, which encodes phosphoethanolamine transferase, catalysing the binding of phosphoethanolamine to lipid A, which turned out to be one of the causes of resistance to colistin in K.pneumoniae and E. coli [45]. In July 2016 in Belgium a new mcr-2, gene was discovered, exhibiting 76.7% nucleotide similarity to mcr-1. The newly discovered mcr-2 gene was located on the IncX4 plasmid of the E. coli ST10 and ST167 strains, in which mcr-1 was not simultaneously detected [70]. The literature data indicates that 28- and 30-day mortality of patients with CRE infection (n = 221) was statistically significantly lower in the case of applying the combined polymyxin therapy, compared to the monotherapy (OR – Odds ratio = 0.36; 95% CI – Confidence interval = 0.19–0.68, p < 0.01) [55].

In Italy, which is the location of the endemic occurrence of K.pneumoniae producing KPC (+), a dramatic increase in resistance to colistin was established, from 12% (2011) to 65% (2012) in recent years. In addition, the data from 21 hospital laboratories shows that over the period of 2013–2014, the percentage share of colistin-resistant strains constituted as much as 43% [45, 88]. Moreover, other multicentre studies which have been conducted in this country have demonstrated that resistance to colistin in K. pneumoniae KPC (+), being the etiological factor of blood infections, has increased more than 3 times, and the so-called 30-day mortality among those patients has risen to 51% [22].

Additionally, the data from the Netherlands indicates that the bacteria K. pneumoniae ST258 KPC (+) which were isolated from the epidemic outbreak in 2013, were characterized by 100% resistance to colistin [45, 90].

Due to the increasing resistance of bacteria to antibiotics, fosfomycin which is mainly administered orally in the treatment of uncomplicated urinary system infections, has now gained immense importance and is recommended in the treatment of infections caused by CRE in the form of i.v. [20].

Fosfomycin is an antibiotic with a broad spectrum discovered in 1969, whose mechanism of action consists in inhibiting the first stage of cell wall synthesis by inactivating the transferase of UDP-N-acetylglucosamin-3-enolpyruval, also known as MurA [45]. Fosfomycin resistance is related to the gene fosA encoding glutathione transferase, which modifies this antibiotic. The gene fosA3 was first described in Japan, in a strain of E. coli producing CTX-M. It should be emphasized that at present there are new subtypes of the mentioned gene, i.e. fosA2, fosA3, fosA4 or fosA5 [20, 45]. In turn, in Gram-positive bacteria, e.g. Staphylococcus spp., Enterococcus spp. or Bacillus subtilis, the enzyme FosB with 48% similarity in the amino acid sequence to FosA, catalysing the reaction between cysteine and fosfomycin has been described [20].

In China, the incidence of the gene fosA3 in the bacilli of KPC (+) K. pneumoniae resistant to fosfomycin was 55.6%, respectively. It should be highlighted that especially in this country a high percentage of resistance to this antibiotic is reported. The studies conducted in one of the hospitals demonstrated that only 43.4% of K. pneumoniae strains producing the KPC-type carbapenemase were susceptible to fosfomycin. Moreover, a similar percentage of susceptibility to this antibiotic, i.e. 39.2% has been recorded in 12 other hospitals in this country [45]. In turn, the data from Europe (Italy) indicates as much as 90.2% resistance to fosfomycin among the strains of K. pneumoniae producing carbapenemases [15].

In the treatment of infections caused by the bacterium K. pneumoniae resistant to carbapenems, a synergistic effect of fosfomycin with carbapenems (70%), colistin (36%) or tigecycline (30%) has been reported. However, in the treatment of infections, where the etiological factor is OXA-48 producing strains, the literature data indicates antagonism between fosfomycin and colistin [20].

Tigecycline is an antibiotic classified as a glycylcycline, whose mechanism of action consists in inhibiting protein synthesis by affecting the interaction of aminoacyl-tRNA with the ribosome A locus [45, 76].

Overproduction of efflux pumps, i.e. AcrAB, as well as the overexpression of RamA, which are positive regulators of the AcrAB efflux system, is one of the main causes of decreased susceptibility of the bacterium K. pneumoniae to tigecycline. It should be added here, that the recent studies conducted in China indicate that OqxAB efflux pumps also play an important role in increasing the resistance to this antibiotic [45].

When analysing susceptibility to tigecycline, it should be noted that in Greece, over the period of 2004–2010, as many as 11.3% (n = 34/301) of KPC carbapenemase-producing strains were resistant to tigecycline. In turn, the data from years of 2013–2014 from the hospital laboratories in Italy showed that carbapenemase-producing K. pneumoniae strains were characterized by 6.0% resistance to tigecycline [45]. Even higher effectiveness was recorded in 2015 in India, where only 2.2% of E. coli strains and 4.7% of K. pneumoniae bacteria were resistant to this antibiotic [42]. In addition, there was a statistically significant lower 30-day mortality rate among the patients infected with CR strains in the case of administering a combined tigecycline therapy compared to the monotherapy (OR = 1.83, 95% CI = 1.07–3.12, p = 0.03) [56].

Due to the fact that carbapenem-resistant strains sometimes exhibit susceptibility to aminoglycosides, it is a group of antibiotics recommended in monotherapy or combined therapy for the treatment of infections caused by the bacteria producing KPC, NDM or OXA-48. Recently, it has also been demonstrated that the gentamicin therapy or a combined tigecycline therapy, reduces the mortality rate (20.7% vs. 61.9%) among patients with sepsis caused by K. pneumoniae ST512 clone which produces the following resistance mechanisms, i.e. KPC-3, SHV-11 and TEM-1 [28, 45]. The resistance to aminoglycosides may be associated with the gene rmtB, being a 16s rRNA methylase, located on the following plasmids, i.e. IncF, IncA/C, IncK or IncN. The data obtained over the years 2012–2014 in China demonstrated that 95% (n = 72/74) of the bacteria K. pneumoniae ST11, resistant to carbapenems, produced the mechanism KPC-2, and 2 strains produced NDM-1. Moreover, 34% (n = 25/74) of K. pneumoniae isolates were found to possess rmtB, which determined resistance to aminoglycosides. Recently a new rmtF gene has been identified i.e. 16S rRNA methyltransferase, present in the strains of Enterobacterales isolated among others in India or the United Kingdom. Moreover, as many as 20 out of 34 described aminoglycoside-resistant microorganisms, simultaneously produced the NDM-1 mechanism. In turn, the rmtF gene was detected in 6 producers of NDM-1 isolated in the United Kingdom [70].

Carbapenems, in accordance with recommendations, may be a therapeutic option in the treatment of infections caused by CR bacteria, provided that the investigates strain is sensitive to an antibiotic from this group. In addition, it is recommended that the therapy with carbapenem combined with another antibiotic which the strain is also susceptible to should be applied. Nevertheless, researchers have proved that with the carbapenem MIC values > 8 mg/l, a combined therapy has a high risk of failure, with mortality above 35% [43]. The studies conducted by De Pascale G. et al. on a group of patients of the ICU who were treated because of infections with K. pneumoniae CR strains according to the scheme, i.e. standard non-carbapenem treatment (ST-Standard Treatment) vs. application of carbapenems (DC-Double Carbapenem), mainly ertapenem, demonstrated that the occurrence of septic shock as well as higher levels of procalcitonin were statistically significantly more frequent in patients receiving the DC therapy. In turn, the 28-day mortality rate was statistically higher among patients treated with ST compared to DC (47.9% vs. 29.2%). Additionally, in colistin-resistant CR strains statistically significantly higher eradication was observed when using the DC therapy [14]. It should also be noted that the studies by Poirel L. et al. demonstrated the synergistic effect of imipenem in conjunction with ertapenem/doripenem or doripenem with meropenem/ ertapenem in the treatment of infections caused by KPC strains. In contrast, when analysing OXA-48 bacteria, a synergistic effect was observed between imipenemertapenem and imipenem-doripenem. What is important, in the case of NDM-1 strains and simultaneous NDM-1 and OXA-48 producers, there was no synergistic effect between carbapenems [69].

In the case of K. pneumoniae strains producing NDM-type carbapenemases, the synergistic effect of the combined therapy of colistin and fosfomycin was observed in vitro. Moreover, the combination of polymyxin B and chloramphenicol was also characterized by a higher effectiveness in the treatment of infections caused by these microorganisms, and prevented the development of resistance to the polymyxin group [45]. In turn, other scientists have noted a synergistic effect between aztrenam-meropenem-colistin, successively meropenem-colistin or fosfomycin-meropenem-colistin in the treatment of infections with NDM-1 strains [84]. In addition, recent studies indicate that the combined therapy with aztreonam and avibactam (a new inhibitor of β-lactamases) is effective in the treatment of infections, where the etiological factor is the strains producing metallo-β-lactamases [45].

The literature data indicates that high doses, as well as prolonged intravenous administering of ertapenem or doripenem, decreased the bacterial density in the case of infections caused by K. pneumoniae NDM-1 (+). It is worth mentioning that similar effects were also observed when using standard ertapenem doses, i.e. 1 g, every 24 h or doripenem, i.e. 500 mg every 8 hours. Importantly, the above conclusions were drawn only in the case of carbapenem-resistant strains producing the NDM-1 mechanism [45].

Recently, special attention has been paid to the high percentage of hospitalized patients colonized by K. pneumoniae strains producing KPC-type carbapenemases. This results in multiple hospital outbreaks, which increases the risk of mortality among patients, especially those hospitalised at the ICU. Mortality rates referring to infections with KPC (+) strains range from 22% to even 72%, respectively. Moreover, due to the multidrug resistance of microorganisms which produce the KPC mechanism, it is extremely difficult to treat patients effectively. In order to achieve the maximum bactericidal effect and minimize the spread of resistance, combined therapy is frequently recommended [87]. The literature data indicates that the combination of carbapenem with tigecycline, colistin or meropenem proved effective in the treatment of infections caused by the strains of K. pneumoniae KPC (+) [45, 87]. Multicentre studies conducted in Italy demonstrated that the combined therapy involving at least two antibiotics exhibiting in vitro activity against the bacterium K. pneumoniae KPC (+) was associated with lower mortality, especially among patients with bacteraemia, pneumonia, or septic shock. The percentage of mortality based on the so-called 14-day rate among patients with bacteraemia caused by the strains of K. pneumoniae KPC (+) when using a combined therapy was 32.0%, and in the case of monotherapy it was as high as 51.3%. The similar data was obtained among patients with pneumonia, where the combined therapy was associated with a 25.0% mortality rate, whereas the use of a monotherapy resulted in as much as 49.1% mortality [87]. Nevertheless, it should be highlighted that the data from other centres do not confirm the higher effectiveness of combined therapy over monotherapy. In connection with the above facts, continuing further research on the effectiveness of antibiotic therapy in the case of infections caused by the strains of K. pneumoniae producing the KPC mechanism is recommended [45].

When discussing the OXA-48-type carbapenemase-producing strains of K. pneumoniae, it should be noted that the carbapenem monotherapy is not recommended for the treatment of infections caused by these microorganisms. The literature data indicates that a combined therapy with sulbactam, meropenem and colistin is more effective in the treatment of infections, where the etiological factor is NDM-1 (+) microorganisms, in contrast to the bacteria producing OXA-48. In addition, the combination of fosfomycin with imipenem, meropenem and tigecycline demonstrated a synergistic effect in the case of the strains of K. pneumoniae OXA-48 [45].

It should be emphasized that the value of the so-called 30-day survival rate among patients with blood infections caused by the strains of Enterobacterales OXA-48 (+) is as high as 50.0%. Similar results were reported in Spain, where despite the high susceptibility of K. pneumoniae OXA-48-producing strains to antibiotics, i.e. amikacin (97.2%), colistin (90.1%), tigecycline (73%) or fosfomycin (66.2%), mortality among these patients was 43.5%, respectively [45].