SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References

We evaluated the prognostic role of baseline levels of C-reactive protein (CRP) as well as CRP levels during conditioning in patients undergoing myeloablative allogeneic stem cell transplantation (SCT). Furthermore, we studied the impact of baseline clinical factors and conditioning regimens on CRP levels in the same period. We conducted a population-based retrospective study of 349 patients undergoing SCT at the National Danish SCT centre between January 2000 and January 2009. CRP levels increased significantly during the conditioning and peaked at day −3 before infusion of the graft. Elevated CRP was associated with older age, non-malignant disease, reduced pretransplant Karnofsky score and high-risk leukaemia. By univariate and multivariate analyses, increased CRP levels (>10 mg/l) before the start of treatment (day −7) and at the day of graft infusion (day 0) were associated with decreased overall survival [HR 1.35 (95%CL) (1.18–1.54); P < 0.0001] and increased treatment-related mortality [1.5 (1.24–1.82); P < 0.0001]. Similar findings were seen for mean CRP levels during the conditioning. CRP was not associated with risk of relapse or aGvHD in multivariate analysis. This study suggests that increased CRP levels before and during the conditioning are associated with baseline clinical factors and that elevated pretransplant CRP levels predict a poorer survival in SCT.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References

Allogeneic stem cell transplantation (SCT) is a treatment of high-risk leukaemia as well as non-malignant haematological diseases. The therapeutic effect of SCT depends partially on the cytoreductive effect of chemotherapy and total body irradiation given before the infusion of stem cells [1, 2]. This pretransplant conditioning regimen prevents graft rejection and reduces the number of remaining leukaemia cells. A contributing factor, securing lasting remission in patients transplanted for leukaemia, is the ‘graft-versus-leukaemia effect’ (GvL), mediated by the donor-derived T cells and NK cells. Although it is clear that the major toxic side effects of chemotherapy are initiated by a direct apoptotic and antimitotic effect primarily affecting rapidly dividing cells, the secondary downstream effects are complex and less well understood. Cytotoxic effects on the gastrointestinal tract (GI) are thought to play a key role in the pathogenesis of treatment-related complications in SCT due to translocation of bacterial products through leaking tight junctions in the GI mucosa. This influx of microbiologically derived products may lead to the induction of potentially harmful inflammatory responses, propagating a systemic inflammatory response [3]. The resulting pronounced local and systemic inflammation may cause further tissue damage and enhance alloreactivity.

C-reactive protein (CRP) is an acute phase protein produced by hepatocytes in response to IL-6 [4]. It is part of the innate immune system with both pro- and anti-inflammatory effects [5] and has proven to be a sensitive and reliable marker of acute inflammation in infectious as well as in non-infectious inflammatory disorders.

A number of previous studies have evaluated the prognostic value of CRP levels during the early engraftment period following infusion of donor cells [6-8], while only few studies have addressed the prognostic value of baseline CRP levels before the start of treatment [9-11]. In particular, very little is known regarding the prognostic significance of CRP responses during the pretransplant chemotherapy (the condition). Moreover, our insights into factors determining the height of the inflammatory response to chemotherapy are still incomplete.

In this retrospective study, we tested the hypothesis that the level of CRP at baseline (day −7) as well as the CRP levels during conditioning is associated with the outcome in SCT. Furthermore, we evaluated the impact of baseline clinical factors as well as differences in conditioning regimens on the level of CRP during the conditioning.

C-reactive protein levels at baseline were chosen as a time point for analysis, because variations in CRP at this stage may reflect individual differences in factors that influence the capability to mount an inflammatory response. These include differences in underlying diseases and the course of those, previous treatment and the impact of polymorphisms in immune response genes [12]. According to our previous studies [13-15], individuals with a high inflammatory response during chemotherapy are more prone to develop severe mucositis later in the course of transplantation and have an increased risk of developing treatment-related complications. Furthermore, the level of inflammatory mediators at the time of infusion of the graft may modulate the activity of alloreactive donor T cells, possibly influencing the risk of both acute graft-versus-host disease (aGvHD) and relapse.

We hypothesized (1) that patients with high baseline CRP levels will mount a higher inflammatory response in reaction to the pretransplant conditioning compared to patients with low CRP levels at baseline and (2) that higher CRP levels at baseline, at day 0 and mean CRP during the conditioning are associated with a poorer outcome.

Patients and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References

We included 349 patients (199 males, 150 females) undergoing myeloablative stem cell transplantation at the National Danish SCT centre in the period between January 2000 and January 2009. Patient characteristics based on prospectively recorded data to Center for International Blood and Marrow Transplantation Research (Table 1).

Table 1. Baseline characteristics of patients and transplantation in the period of 2000–2009
 Patients N (%)
  1. a

    Two patients received both BM and UCB.

Total349
Sex
Male199 (57)
Female150 (43)
Age [years, median (range)]26.7 (0.3–60.3)
Underlying disease
Acute lymphoblastic leukaemia92 (26)
Acute myeloid leukaemia106 (30)
Chronic myeloid leukaemia55 (16)
Myelodysplastic syndrome21 (6)
Other haematological malignancies14 (4)
Severe aplastic anaemia24 (7)
Other non-malignant diseases37 (11)
Standard-risk malignancies146 (42)
High-risk malignancies113 (32)
Unknown risk status28 (8)
Karnofsky >90224 (64)
Karnofsky ≤90121 (35)
Unknown Karnofsky4 (1)
Donor type
Matched sibling donor170 (49)
Matched unrelated donor179 (51)
Stem cell sourcea
Bone marrow (BM)227 (65)
Umbilical cord blood (UCB)3 (1)
Peripheral blood stem cells121 (35)
Conditioning regimen
Cyclophosphamide + busulfan97 (28)
Cyclophosphamide + total body irradiation168 (48)
Etoposide + total body irradiation63 (18)
Other regimes21 (6)
GvHD prophylaxis
Cyclosporine (CsA)44 (12.6)
Methotrexate (MTX)+CsA296 (84.8)
Other9 (3)

Patients were eligible if at least one CRP measurement was available during the conditioning period from day −7 to day 0. The mean number of qualified patients at each day of the pretransplant conditioning period was 283 (range: 227–334).

Donors were either matched sibling donors (n = 170) or matched unrelated donors (MUD) (n = 179). An unrelated donor was accepted if at least a 9/10 HLA match was achieved at high-resolution testing.

Underlying diseases leading to SCT were acute lymphoblastic leukaemia, ALL (n = 92), acute myeloid leukaemia, AML (n = 106), chronic myeloid leukaemia, CML (n = 55), myelodysplastic syndrome, MDS (n = 22), other haematological malignancies (n = 14), severe aplastic anaemia, SAA (n = 24) and other non-malignant diseases (n = 37).

Patients with a malignant diagnosis were stratified into risk groups. Patients with acute leukaemia (ALL/AML) in the first complete remission, myelodysplasia (MDS) or chronic myeloid leukaemia (CML) in first chronic phase were classified as standard-risk disease (n = 146). All other stages of malignancies were considered to be high-risk disease (n = 113).

The patients were transplanted with stem cells derived from bone marrow (n = 227), peripheral blood stem cells mobilized by granulocyte colony-stimulating factor (G-CSF) (n = 121) or umbilical cord blood (n = 3). Two patients received both bone marrow and umbilical cord blood stem cells.

All patients underwent a myeloablative conditioning regimen, although with varying intensities, including full intensity for patients with leukaemia as well as lower intensity for patients with severe aplastic anaemia and Fanconi anaemia. The majority of the patients received cyclophosphamide (CY) + total body irradiation (TBI) as conditioning regimen (n = 168). Others received either CY+ iv. busulfan (BU) (n = 97), etoposide (VP16) +TBI (n = 63) or other regimens (n = 21). Patients with an unrelated donor received three dosages of antithymocyte globulin (ATG) as part of the conditioning regimen.

GvHD prophylaxis was cyclosporine (CsA) alone (n = 44), methotrexate (MTX) + CsA (n = 296) or others (n = 9).

CRP measurement

C-reactive protein levels in plasma were measured using the turbidimetric assay CRPL3 on Hitachi 917 (January 2000–June 2003) and Modular P Modular (from June 2003) (normal range: 0–10 mg/l).

Ethics

The Danish Data Protection Agency and the Regional Ethics Committee approved the study.

Statistics

For day −7 and day 0, each patient was represented with one value. If two or more CRP values from the same patient were available at the same day, the first measurement was applied.

On average 6.5 CRP measurements (range: 2–8 measurements) were available from day −7 to day 0 from each patient. Based on these CRP values, the mean CRP for each patient was calculated to represent the levels of CRP during the conditioning period, and the mean value for each patient was used in the statistical analysis.

Spearman rank correlation test was used as a measure of association. Univariate analyses of baseline characteristics and CRP levels were tested using distribution free tests (Kruskal–Wallis or Mann–Whitney). Acute GVHD was analysed employing logistic regression analysis modelling the probability of aGVHD and presenting the odds ratio (OR).

Survival probabilities for time to event data were estimated using the Kaplan–Meier method in univariate analyses, and comparisons between strata were tested using log-rank test analysis. Time to relapse was analysed using a competing risks methodology with non-relapse mortality as a competing risk; treatment-related mortality (TRM) was analysed with relapse as a competing risk [16, 17].

Multivariate analyses of time to event data were performed using the Cox proportional hazards model. The endpoints were overall survival (OAS), TRM and time to relapse. Model assessment was performed using Martingale and Schoenfeld residuals.

C-reactive protein was log-transformed using log base 2, and accordingly, hazard ratios are for a twofold difference in CRP levels. Estimates are presented with 95% confidence limits. P-values <5% are considered significant unless otherwise stated. sas (version 9.2, SAS Institute, Cary, NC, USA) was used for statistical calculations.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References

CRP levels before and during the conditioning

C-reactive protein values during the pretransplant period are shown in Fig. 1. Prior to the start of conditioning (day −7), approximately 75% of the patients had a mean CRP level within the normal range (≤10 mg/l). During the conditioning period, a significant rise in CRP levels peaking at day −3 was seen (mean 48 mg/l; range: 0–306 mg/l, (P < 0.0001). This was followed by a gradual decline, although mean CRP at day 0 remained significantly elevated compared with baseline levels (P < 0.0001, Fig. 1).

image

Figure 1. Mean CRP levels by day during the pretransplant period. * indicates significant increase in CRP levels from day −7 to day −3 (P < 0.0001). # indicates significant increase in CRP levels from day −7 to day 0 (P < 0.0001). The bars represent 95% CL.

Download figure to PowerPoint

Associations between CRP levels and clinical factors

Univariate analysis

Associations between baseline parameters and CRP levels were analysed (Table 2). The following parameters showed significant associations with elevated CRP before and/or during the conditioning in univariate analysis, although with relatively small differences in medians: older age, non-malignant disease as cause of transplantation, reduced pretransplant Karnofsky score and high-risk leukaemia.

Table 2. Univariate analysis of factors associated with increase in CRP during the pretransplant period. Clinical baseline characteristics, transplant-related characteristics and CRP levels before the conditioning day −7, during the conditioning and at the end of conditioning before infusion of the graft were analysed
VariableCRP mg/l median (range) P-values
Day −7During conditioningaDay 0
  1. ATG, antithymoglobulin; TBI, total body irradiation; CY, cyclophosphamide; VP16, etoposide; BU, busulfan; aGvHD, acute graft-versus-host disease.

  2. a

    Data representing the conditioning period were calculated as the mean for each single patient and the median of those are shown for each subgroup).

Age groups (years)
0–5 versus3 (0–162)15 (0–258)12 (1–157)
6–12 versus3 (0–199)11 (0–199)16 (1–149)
13–18 versus5 (1–74)15 (0–242)14 (1–242)
19–39 versus5 (0.113)15 (0–306)20 (1–210)
≥405 (0–160)10 (0–208)15 (1–189)
 0.007<0.00010.017
Karnofsky score>90 versus3 (0–78)12 (0–285)15 (1–242)
Karnofsky score ≤906 (0–199)15 (0–306)20 (1–149)
 0.007<0.00010.109
Non-malignant versus3 (0–162)20 (0–306)24 (1–121)
Malignant4 (0–199)12 (0–285)15 (1–242)
  0.109 <0.0001 0.0049
High risk versus6 (0–199)17 (0–285)21 (1–189)
Standard risk3 (0–115)10 (0–242)11 (1–242)
 0.131<0.0001<0.0001
+ATG versus 27 (0–306)30 (2–242)
−ATG 8 (0–192)10 (1–101)
  <0.0001<0.0001
TBI+CY versus 14 (0–306)17 (1–242)
VP16+TBI versus 11 (0–215)8 (1–43)
BU+CY 11 (0–258)15 (–157)
  <0.0001<0.0001
+TBI versus 13 (0–306)15 (1–242)
−TBI 13 (0–258)16 (1–157)
  0.1390.069

Mean CRP levels during the conditioning and at day 0 were positively correlated with CRP levels at day −7 (= 0.36, < 0.0001 and r = 0.434, P < 0.001, respectively), indicating that patients with elevated CRP at baseline mounted a higher inflammatory response during the conditioning.

High CRP values at day 0 were associated with increased age, malignant diagnosis and high-risk disease (Table 2).

Associations between conditioning regimens and CRP during conditioning and at day 0 were analysed (Table 2). The following regimes were compared: TBI + CY, CY+ BU, TBI + VP16 and the use of ATG.

Elevated CRP levels during conditioning were associated with the use of TBI and CY. Moreover, the use of ATG was associated with a significant elevation of CRP (Table 2).

Multivariate analysis

In multivariate analysis, only reduced Karnofsky score remained associated with elevated CRP at day −7, while CRP levels at day 0 remained significantly associated with the use of ATG and type of chemotherapy. Further to this, multivariate analysis showed an approximately 40% increase in day 0 CRP for every twofold difference in day −7 CRP for the group not receiving ATG. For those receiving ATG, CRP levels were sixfold higher than for those not receiving ATG, and independent of CRP at day −7.

In addition, the types of conditioning revealed statistically significant association with elevated CRP in the multivariate analysis. In contrast, baseline patient and transplantation-related factors did not relate to the mean level of CRP during the conditioning in this analysis (data not shown).

CRP and clinical outcomes

Associations between clinical endpoints and CRP levels were evaluated by univariate and multivariate analyses. The examined endpoints were OAS, treatment-related mortality (TRM), time to relapse and aGvHD.

Overall survival was significantly reduced in patients with elevated CRP (>10 mg/l) at day −7, at day 0 and in patients with a mean CRP above the normal range in univariate analysis (Fig. 2). For TRM, a similar pattern was seen in univariate analysis with inferior outcome in patients with increased CRP (Fig. 3).

image

Figure 2. The upper panel shows the Kaplan–Meier estimates of survival probabilities for patients dichotomized by their CRP level (normal versus elevated) at day -7 (prior to start of conditioning). There were 191 patients with low CRP (61 deaths) and 72 patients with elevated CRP (36 deaths). The number of patients at risk in the group with CRP ≤10 mg/l at 24, 48 and 72 months was 141, 123 and 93, and for those with CRP> 10 mg/l, the numbers at risk were 43, 37 and 23, respectively. The lower panel shows the Kaplan–Meier estimates of survival probabilities for patients dichotomized by their CRP level (normal versus elevated) at day 0 (at the end of conditioning before infusion of the graft). There were 122 patients with low CRP (31 deaths) and 212 patients with elevated CRP (80 deaths). The number of patients at risk in the group with CRP ≤10 mg/l at 24, 48 and 72 months was 99, 88 and 62, and for those with CRP> 10 mg/l, the numbers at risk were 145, 123 and 95, respectively.

Download figure to PowerPoint

Furthermore, univariate analysis indicated increased frequency of relapse in patients with elevated CRP at day 0 and in patients with elevated mean CRP, but relapse was unrelated to CRP at day −7 (Fig. 4).

image

Figure 3. The upper panel shows the cumulative incidence estimates of TRM with relapse as a competing risk for patients dichotomized by their CRP level (normal versus elevated) at day −7 (prior to start of conditioning) (P = 0.005). The lower panel shows the cumulative incidence estimates of relapse with death as a competing risk for patients dichotomized by their CRP level (normal versus elevated) at day 0 at the end of conditioning before infusion of the graft (P = 0.026).

Download figure to PowerPoint

image

Figure 4. The upper panel shows the cumulative incidence estimates of relapse with death as a competing risk for patients dichotomized by their CRP level (normal versus elevated) at day −7 prior to start of conditioning (P = 0.31). The lower panel shows the cumulative incidence estimates of relapse with death as a competing risk for patients dichotomized by their CRP level (normal versus elevated) at day 0 at the end of conditioning before infusion of the graft (P = 0.008).

Download figure to PowerPoint

Multivariate analysis of OAS confirmed an association with CRP along with increasing recipient age, the use of ATG and malignant diagnosis (Table 3). Similar conclusions were obtained in separate analysis of malignant diseases (not shown). Variables tested for, but found insignificant in this model, were donor age, CMV mismatch, pretransplant Karnofsky score and type of chemotherapy.

Table 3. Cox regression analyses of factors associated with overall survival (OAS). CRP levels day −7, mean CRP during the conditioning and CRP day 0, were analysed versus malignant diagnoses, use of ATG and age of the recipient. CRP was log-transformed using log base 2, and therefore, hazard ratios for CRP are for a twofold difference in CRP levels
 Hazard ratio (95% confidence limits) P-values
Day -7 CRPDay 0 CRPMean CRP during conditioning
  1. ALL, acute lymphoblastic leukaemia; AML, acute myeloid leukaemia; CML, chronic myeloid leukaemia.

CRP log2

1.35 (1.18–1.54)

<0.0001

1.35 (1.19–1.53)

<0.0001

1.36 (1.18–1.56)

<0.0001

ALL

13.28 (3.07–57.41)

0.0005

6.46 (2.25–18.53)

0.0005

5.73 (2.02–16.24)

0.001

AML

10.22 (2.34–44.5)

0.002

4.69 (1.62–13.56)

0.0044

4.46 (1.55–12.84)

0.0056

CML

8.22 (1.77–38.19)

0.0072

3.13 (0.99–9.92)

0.00521

3.57 (1.16–11.05)

0.0271

Other malignancies

8.84 (1.82–42.86)

0.0068

5.09 (1.59–16.31)

0.0061

5.19 (1.62–16.69)

0.0057

ATG

2.26 (1.46–3.51)

0.0003

1.07 (0.98–1.18)

0.13

1.06 (0.96–1.16)

0.25

Recipient age in years

1.26 (1.07–3.51)

0.0048

1.21 (1.05–1.38)

0.0072

1.24 (1.09–1.42)

0.0012

In multivariate analysis, increased TRM was significantly associated with increased CRP as well as increased recipient age (Table 4).

Table 4. Multivariate analyses describing factors associated with treatment-related mortality (TRM). CRP at day −7, mean CRP during the conditioning and CRP day 0 were analysed versus donor match and the age of the recipient. CRP was log-transformed using log base 2, and therefore, hazard ratios are for a twofold difference in CRP levels
 Hazard ratio (95% confidence limits) P-values
Day −7Day 0Mean CRP during conditioning
  1. MUD, matched unrelated donor; SIB, matched sibling donor.

CRP log2

1.50 (1.24–1.82)

<0.0001

1.43 (1.18–1.72)

0.0002

1.36 (1.11–1.66)

0.0025

Donor match (MUD versus SIB)

1.15 (1.01–1.32)

0.0404

0.14430.2663
Recipient age in years

1.37 (1.09–1.72)

0.0077

1.41 (1.14–1.73)

0.0016

1.49 (1.21–1.83)

0.0001

For time to relapse, the multivariate analysis showed significant association with conditioning with VP16+ TBI, type of donor and diagnosis, while associations with CRP were insignificant (not shown).

The cause of treatment-related mortality in patients who had an elevated CRP at any of the three addressed stages of the pretransplant period included infection (n = 18), organ failure (incl. liver, cardiopulmonary and multi-organ failure) (n = 14), GvHD (n = 7) and unspecified non-relapse mortality (n = 14).

CRP and aGvHD

Two hundred and nine patients developed aGvHD, 181 with grades 1–2 and 28 with grades 3–4. Logistic regression analysis modelling the probability for aGVHD showed a non-significant association for CRP at day −7 (OR = 1.12, 95% CI: 0.95–1.30, P = 0.18). A similar result was seen for mean CRP during conditioning (OR = 1.10, 95% CI: 0.96–1.27, P = 0.16). Multivariate analyses did not demonstrate significant associations between aGvHD and CRP (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References

A number of studies have demonstrated an association between post-transplant CRP levels and outcome in SCT [7, 8, 18, 19]. In contrast, only few studies have addressed the prognostic value of pretransplant CRP levels, and to our knowledge, no published studies have addressed the prognostic significance of the course of CRP during the conditioning and at day 0.

Our results indicate that elevated CRP level at baseline, before the start of conditioning, is a strong prognostic factor for OAS. In agreement with our data, Pavlu et al. [9] reported increased TRM and poorer OAS in patients with elevated preconditioning CRP levels, but unaltered relapse rate in transplanted patients with CML. Artz et al. [20] and Remberger et al. [10] obtained comparable results in patients receiving reduced intensity regimes (RIC), and a recently published study by Sato et al. [11] showed increased TRM and inferior OAS in patients with elevated CRP undergoing myeloablative conditioning regimes.

These findings are also consistent with previous studies showing associations between outcome of SCT and genetic polymorphisms in the genes encoding IL-6, a main inducer of CRP production [21, 22].

The present study extends these studies by monitoring CRP values during the period of conditioning and at day 0. We observed a significant rise in CRP peaking at day −3, remaining elevated at day 0.

In multivariate analysis, reduced pretransplant Karnofsky score proved to be associated with elevated CRP. Although the difference in medians was relatively small and therefore of minor clinical interest, this finding indicates that factors determining the general physical state of an individual may influence baseline CRP levels.

In line with our hypothesis, we showed a positive correlation between CRP levels at day -7 and the levels of CRP after the conditioning, with even more pronounced elevation for patients receiving ATG.

CRP levels were related to TRM in accordance with previous studies, including studies by our group where a correlation between increased levels of inflammatory mediators during conditioning and the degree of mucositis later during the course of transplantation was demonstrated [9, 11, 13].

The intestinal epithelium is in constant interaction with the microbiological flora, and it is considered essential for the immunological homoeostasis that this interaction is kept in balance. However, destruction of tight junctions in the basal membrane may arise from the toxic effects of the conditioning, allowing bacteria and bacterial products to reach the submucosa and the circulation, leading to septicaemia, pronounced inflammation and increased risk of organ failure [23-25].

Several previous studies have supported the widely accepted hypothesis that induction of a systemic inflammatory response to chemotherapy is a central step in the pathogenesis of aGvHD. This hypothesis is, however, primarily based on studies in rodents, while evidence derived from studies in the human setting has been indirect, including studies showing associations between CRP rise during the post-transplant period and aGvHD. Thus, it is of interest that we found no association between pretransplant CRP levels and aGvHD, in agreement with a couple of previous studies [9, 10]. Likewise, we did not find any association between mean CRP levels during the conditioning and the risk of aGvHD. Consequently, the present study does not provide support for an induction of aGvHD through the primary wave of inflammation seen during the conditioning.

The findings of the present study are of interest for mainly two reasons. First, they suggest that measurements of CRP levels before the start of conditioning could be of benefit clinically in the risk stratification of patients to individualize the intensity of the conditioning chemotherapy and prophylaxis of complications.

Importantly, the present data further suggest that CRP levels during the conditioning and at day 0 are related to the poor survival due to increased TRM. Although risk stratification cannot be met by significant modifications in the treatment regimen at that stage, this observation is of interest because it stresses the need for the development of conditioning regimens with limited inflammation-inducing potential. Further, it suggests that the use of anti-inflammatory therapy during the conditioning period may help to reduce TRM.

Secondly, the study pinpoints baseline factors associated with increased CRP levels before, during and after the conditioning. This is in line with the hypothesis that increased induction of inflammation is a key element in the mechanisms by which various clinical and transplant-related factors affect the course of transplantation. Thus, the findings suggest an increased inflammatory response as a possible biological link between the well-established risk-factors in SCT such as leukaemia risk stage, Karnofsky score and age- and treatment-related complications and mortality.

The strengths of this study include the large population-based and clinically well-characterized cohort with long-term follow-up. There are, however, weaknesses related to the retrospective design with the inclusion of patients based on a variable number of CRP measurements for each of the three periods analysed.

In summary, increasing evidence indicates that preconditioning levels of CRP is an independent prognostic risk marker for patients candidating for SCT. The present study confirms these findings and extends them by pinpointing a strong association between OAS and CRP levels during the conditioning. These findings should be further evaluated in prospective studies.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References

This work has been supported by grants from The Danish Cancer Society and Tømrermester Holms Legat.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Patients and methods
  5. Results
  6. Discussion
  7. Acknowledgment
  8. Conflict of interests
  9. Financial disclosure
  10. References