Karin Massaro1 and Silvia Figueiredo Costa1,2
1 Department of Infectious Diseases, Faculty of Medicine, University of Sao Paulo, Sao Paulo, Brazil
2 Laboratory of Bacteriology-LIM54, Hospital das Clinicas, University of Sao Paulo, Sao Paulo, Brazil
| This is an Open Access article distributed
under the terms of the Creative Commons Attribution License
(http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
An ideal marker in the neutropenic population after HSCT is the one which positivetes at the onset of fever, or at most up to 24 hours after its onset, the patients at potential risk for infection due to bacterial and fungi and mortality. Several biomarkers have been used in HSCT patients in the last decade. However, it seems that C-RP and Il-6 are the most useful markers to early detected infection and risk for death.
The number of hematopoietic stem cell transplantations has increased
significantly over the past decade. The morbidity and mortality due to
infectious complications are the major clinical issue in HSCT
recipients.[1,2] However, the identification of these infectious
complications is still based on clinical criteria, though, mainly in
the occurrence of fever.
In patients after HSCT, it's hard to set infection apart from other causes of fever, such as acute Graft versus host diseases (GVHD), or veno-occlusive disease (VOD) and sinusoidal occlusion syndrome. The detection of these major transplantation-related complications is essential for the early and proper introduction of antibiotics, immunomodulators or hemostatic treatment. The diagnosis and the evolution of infectious complications may be optimized with the help of early sensitive and specific markers of bacterial and fungal infections. An ideal marker should preferably either precede significant microbiological findings or justify an additional intensive search for an infection focus even in afebrile. The early prediction of serious bacteremia helps to identify those patients who are more likely to benefit from a combination of drugs and reduce the unnecessary toxicity of additional treatments when there is no need. At the same time, it helps the identification of patients at low risk of complications, which may be treated at outpatient units.[5,6]
Studies of Biomarkers of Infection in Febrile Neutropenia after HSCT
PCT, IL-6 and CRP cannot be utilized to differentiate Systemic
Inflammatory Response Syndrome (SIRS) from infection, in patients who
receive ATG or similar within up to three days after the application of
the drug. PCT, CRP and IL-6 have limited value in the diagnosis of
infection during the administration of ATG or other anti T-cell
antibodies, such as OKT-3. On the other hand, Blijlevens et al.
(2000) found increased levels of PCT in isolated cases of GVHD.
Procalcitonin. There can be a slight induction of PCT after HSCT, as when it occurs in chemotherapy, with levels rarely above 0.5 to 1 ng/mL. That was observed, for instance, in children with ALL, AML, NHL with B and LH cells. In cases of serious infections or sepsis, these patients may also induce high levels of PCT. If neutropenia is severe, the induction of PCT is reduced; however, the synthesis is not totally suppressed.[26-27] Some considerations must be made to justify the different results in two of the studies mentioned. In the study by Blijlevens et al. (2000), the limited number of patients harmed the statistical analysis of the results. Zintl et al. (1996) showed, in a study with patients submitted to HSCT with prior myeloablative chemotherapy, that PCT increases in systemic bacterial infections and is strictly related to the septic shock and also that higher levels indicate bad prognosis. A study that evaluated PCT in neutropenic hematological patients with mucositis revealed that it represents an important tool to predict the occurrence of major infections in patients with febrile neutropenia, even in the presence of associated complicating conditions, such as mucositis and GVHD. A lower threshold value (e.g. PCT > 0.25 ng/mL) increases the diagnostic reliability in case of a serious infection, even in patients with neutropenia.[30-31] This reduced increase may happen since the adhering monocytes, among other factors, are necessary for the induction of PCT. Even if the immune response is overall weakened in neutropenic patients, and also by the use of corticosteroids or cyclosporine for the treatment of autoimmune diseases or after transplantation, the induction of PCT is not totally suppressed. In case of neutropenia, the severity of the systemic inflammatory response cannot often be evaluated, since the “fever” symptom is weakly associated with the gravity of the sepsis or prognosis of the disease. In an international multicentric study, 158 patients with neutropenia and fever were evaluated. A cut-off value of 5 ng/mL determined a sensitivity and specificity for the diagnosis of severe sepsis of 83% and 100% respectively. In concentrations below 0.5 ng/mL, the severe infection would be unlikely. In case of fungal infections, the levels of PCT were 1.17 ± 0.44 ng/mL on the first day of diagnosis, with a quick downward trend. Twelve patients died. On the first day of fever, they had levels of PCT of 20.45 ± 4.48 ng/mL (average, standard deviation). Similar data were reported in other studies. [24,26,33] A study that compared the value of PCT, neopterin, CRP, IL-6 and IL-8, as diagnostic markers for such purpose, showed an area under the ROC curve in the distinction of bacteremia by Gram-negative from all the other causes of fever of 0.863 (IC 95% 0.74-0.98) and results not significant for the other markers.
Ortega et al. (2004) showed PCT levels in patients after HSCT with FUO of 0.3 ± 0.2 ng/mL (average, standard deviation), and with infections microbiologically documented in the range of 0.5 ± 0.7 ng/mL. In cases of extended fever (over five days) and on suspicion of aspergillosis, the levels of PCT were even higher (10.1 ± 6.7 ng/mL).
In a recent study that evaluated the relevance of ultrasensitive CRP and PCT levels in patients after HSCT, it was verified that the ultrasensitive CRP was higher in the patients with aGVHD and also in the cases of sepsis. The increased levels of PCT were associated only with a bacterial infection. Only the levels of PCT could be used to differentiate infection from other complications related to transplantation. In the study mentioned, PCT, besides not increasing in GVHD, did not increase in fungal infections, although the bias of that study was the size of the sample (n = 35).
C-reactive protein. A study with adult patients submitted to allogeneic HSCT (n = 137) showed that the pre transplantation CRP and serum ferritin levels can be predictors of bacterial complications. This study has the limitations of being retrospective, without standardization of different hematological diseases and limited number of samples. Likewise, Pihusch et al. (2006) found high CRP levels both in infections after HSCT and in aGVHD cases. In this study, which analyzed the levels of PCT, CRP, and IL-6, prospectively, in 350 patients after HSCT, conditioning of some patients included anti-thymocyte globulin (ATG). It is well known that ATG, which is a pool of heterologous antibodies against T-cells of direct form (opsonization and lysis via complement activation), causes early increase of TNF-α, CRP, IL-6 and PCT (few minutes after administration).[23,39]
Interleukin-6 (IL-6). Some studies show that IL-6 is a sensitive marker for acute infections in febrile neutropenia and it can be used as a marker of unfavorable results in this situation.[40-42]
The values of IL-6 (measured by the Enzyme-linked immunosorbent assay [ELISA] method) for febrile neutropenia vary in several studies. In the study by Steinmetz et al. (1995), the level of detection threshold was of 3 pg/mL, with intratest variations of < 15% (in the 3-100 pg/mL range) and < 20% for values > 100 pg/mL. A significant increase of IL-6 was considered as an increase higher than 20 pg/mL or 30% for levels < 100 pg/mL and ≥50 pg/mL or 20% for levels > 100 pg/mL. A cut-off value of IL-6 of 297 pg/mL presented sensitiveness of 62%, positive predictive value (PPV) of 50% and negative predictive value (NPV) of 70% (p = 0.016) for the identification of infection in neutropenic patients.
Interleukin-8 (IL-8). The rapid increase of IL-6 and IL-8, quickly detectable on the first days after the HSCT, may either be the mere reflection of the severity of the inflammatory response or it may indicate the deleterious synergic effect of these cytokines on tissues. In the study by Schots, the multivariate analysis showed that the increase of these cytokines, especially IL-8, was associated with a fatal result in HSCT. There is also a study that relates IL-8 to the adult respiratory distress syndrome.
The study by Prat et al. evaluated several biomarkers (PCT, CRP, neopterin, IL-6 and IL-8) in 61 adults (with 41 submitted to HSCT, and the other ones only under chemotherapy). IL-6 and IL-8 did not prove useful to prevent bloodstream infection before chemotherapy, on the first day of neutropenia nor subsequent days.
|Table 1. The most important studies with adult patients submitted to HSCT comparing use of biomarkers to predict infections.|
Studies of Inflammation Marker and Poor Prognosis (Death) Related to HSCT
An ideal death marker in the neutropenic population after HSCT is the one that indicates, at the onset of fever, or at most up to 24 hours after, the patients under potential risk of death transplant-related. There are few studies in literature to help the prediction of the prognosis of patients submitted to HSCT and the existing studies evaluate a small number of subjects. The following table shows the main studies that evaluated death predictors after HSCT (Table 2). The studies published about the prognosis in patients after HSCT are divided into allogeneic or autologous. Despite the advance in supporting treatment, the allogeneic HSCT still presents high toxicity, with mortality related to transplantation between 10-50% due to major complications during the first months after transplantation (infections, VOD, GVHD and pneumonitis).
|Table 2. The most important studies with adult patients submitted to HSCT comparing the use of biomarkers as predictors of death.|
In a prospective study of 296 patients submitted to HSCT at
our university hospital, IL-6, PCT and CRP were evaluated on the day of
confirmed neutropenia, on the day of the febrile event, 24 and 72 hours
after its onset, and 48 hours or 5 days, in case the fever,
persisted. The patients were evaluated after the discharge or death
within 30 days from the HSCT. Only CRP ≥120 mg/L was independently
associated with death. The other risk factors associated with death in
the multivariate analysis were: type of transplantation (allogeneic and
unrelated), bloodstream infection by Gram-negative, LDH≥390 IU/L and
urea ≥25 mg/dL. For allogeneic patients, only CRP≥120mg/L and
bloodstream infection by Gram- negative were risk factors for death. No
independent risk factor was evidenced for the subgroup of autologous
patients. We could not demonstrate the association between IL-6 and PCT
and death. Other authors demonstrated similar results, however, in
studies with a smaller number of cases and or without multivariate
analyzes or testing only one or two biomarkers.[21,22,35,37] Artz et
al. utilized multivariate analysis, however, with a smaller number
of patients (n = 112) and only analyzed two markers (CRP and IL-6) in
allogeneic HSCT. In their study, the median of CRP was 18 mg/L, and of
IL-6 was 78 pg/mL, showing that high levels of CRP, but not of IL-6,
before the conditioning for HSCT, were independent predictors of death
related to allogeneic transplantation. The study of Remberger et al.
(2010), despite including a large number patients, 299, was
retrospective, without multivariate analysis and, likewise the two
previous ones, evaluated only CRP in allogeneic HSCT before the
conditioning. The authors showed that mortality related to
transplantation was lower in patients with non-myeloablative regimens
with high CRP (67 vs 43%, p = 0.005, and 16 vs 30%, p = 0.036).
Finally, the study by Ortega et al. (2004) analyzed only CRP, did
not utilize multivariate analysis and the number of 100 (smaller than
that of the present study), in autologous and allogeneic patients, and
showed that CRP on the 5th day of
fever ≥ 16 mg/dL was associated with death, due to infectious causes
with a sensitivity of 100%, a specificity of 87%, PPV and NPV of 30 and
Schots et al. (2002) monitored the levels of CRP and other variables in 96 consecutive adults submitted to allogeneic HSCT, dosing from day 0 to +5. Only the high levels of CRP (> 50 mg/L) (p < 0.001) and type of allogeneic donor (p = 0.02) were the independent factors for mortality related to transplantation. The retrospective study of Remberger et al. (2010), evaluated the clinical impact of CRP in allogeneic HSCT patients, before the conditioning (n = 205) and after the not-myeloablative and myeloablative conditioning regimens (n = 299). There was a significant correlation between CRP levels in both groups with overall survival and mortality related to transplantation increased in individuals who had received non-myeloablative conditioning. Early death by infection (within 100 days) resulted in 25 and 8.7% of patients with high CRP with and without documented infection, respectively (p < 0.001 and p = 0.04), compared to patients with normal CRP, where early death by infection occurred in 2.3% of patients. Remberger et al. (2010) concluded that CRP may be utilized as a prognostic factor because it enables the identification of infection in earlier stages, as also observed in our results. In the studies by Pihusch et al., the serum levels of CRP (5.2 mg/dL vs 7.8 mg/dL, p < 0.001), IL-6 (147 ng/mL vs 494 ng/mL; p < 0.001) and PCT (2.1 ng/mL vs 4.0 ng/mL; p = 0.003) were high until week +2 in the group of patients who died of complications related to transplantation, compared to the group that survived.