Rheumatology International

, Volume 33, Issue 6, pp 1415–1418

Monoclonal anti-TNF antibodies can elevate hemoglobin level in patients with ankylosing spondylitis


    • Department of RheumatologyBakirköy Dr. Sadi Konuk Education and Research Hospital
  • Ayten Yazici
    • Department of RheumatologySakarya Education and Research Hospital
  • Mehmet Soy
    • Department of RheumatologyHisar Intercontinental Hospital
Original Article

DOI: 10.1007/s00296-012-2539-5

Cite this article as:
Bes, C., Yazici, A. & Soy, M. Rheumatol Int (2013) 33: 1415. doi:10.1007/s00296-012-2539-5


Anemia is one of the extra-articular findings of ankylosing spondylitis (AS), and anti-TNF therapy has been shown benefit in patients with anemia associated AS. In this study, we aimed to evaluate and compare the effects of biological and non-biological agents on hemoglobin levels in AS patients. One hundred consecutive patients who fulfilled ASAS criteria for AS were included in the study. Fifty-four of the patients treated with anti-TNF agents (20 patients treated with infliximab, 20 patients with adalimumab, and 14 patients with etanercept), and 46 patients treated with non-steroidal anti-inflammatory drugs and/or other disease modifying anti-rheumatic drugs. The C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), hemoglobin (HGB), hematocrit (HCT) counts, and BASDAI scores were compared before starting therapy and at 52 weeks. There was no statistically significant difference between patients about demographical data (age, sex) and disease age (p > 0.05 for all). Significant difference was determined between HGB, HCT, CRP, ESR, and BASDAI values before and after therapy (for infliximab p: 0.001; 0.000; 0.000; 0.000; 0.000, respectively, and for adalimumab p: 0.017; 0.03; 0.001; 0.002; 0.000, respectively). In etanercept group, there was no significant difference in HGB values, when compared with before starting therapy and at 52 weeks (p > 0.05). In the group of treated with non-biological agents, ESR values and BASDAİ scores showed distinctive improvement after 52 weeks of therapy, but was not a significant difference in hemoglobin and hematocrit values. Conclusion: Anti-TNF-alpha therapy with monoclonal antibodies (adalimumab and infliximab) did not only suppress disease activity but also provided a significant improvement in HGB levels. In the groups of treated with a TNF-alpha receptor antagonist (ETA) and non-biological agents, disease activity was suppressed, but there was not founded significant improvement in HGB levels after 52 weeks. Different outcomes of anti-TNF agents may be associated with their different effect mechanisms.


Ankylosing spondylitisAnti-tumor necrosis factor-alphaHemoglobin


Ankylosing spondylitis (AS) is a chronic, systemic, and inflammatory disease that primarily affects the sacroiliac joints and the axial skeleton. Other clinical manifestations include peripheral arthritis, enthesitis, and extra-articular organ involvement [1, 2]. Most common extra-articular manifestations include acute anterior uveitis, aortitis, conduction defects, pulmonary fibrosis, amyloidosis, and neurologic deficits [36]. Anemia is a frequent finding in chronically inflammatory processes and is also seen in AS patients. Pro-inflammatory cytokines, such as tumor necrosis factor-alpha (TNF-alpha) and interleukin-6 (IL-6), inhibiting erythropoiesis or malabsorption and/or gastrointestinal hemorrhage associated with non-steroidal anti-inflammatory drugs may be the reasons for anemia in AS [7, 8]. On the other hand, a recently determined protein called hepcidin is claimed to cause of anemia in chronic inflammatory diseases [9]. Infliximab (INF), adalimumab (ADA), and etanercept (ETA) are FDA certified TNF-alpha blockers, used in AS treatment. INF is a chimeric mouse–human monoclonal antibody, but ADA is a fully humanized monoclonal structured antibody, produced against TNF-alpha. Whereas ETA is a dimeric fusion protein of the TNF receptor linked to the Fc portion of human IgG1. ETA acts as a soluble TNF receptor to bind TNF. These biological agents promote patients’ life quality by suppressing inflammation and reducing disease activity [10]. There is no adequate publication in the literature about TNF-alpha associated anemia in AS patients. With this study, we investigated the effects of these biological agents (ETA, ADA, and INF) on hemoglobin levels.

Patients and methods

One hundred patients were included for this study from the rheumatology outpatient clinics. The diagnosis of definite ankylosing spondylitis was made according to ASAS criteria [11]. The number of patients taking INF, ADA, and ETA therapies (biological group) was 20, 20, and 14, respectively. Other 46 patients had been taking non-biological agents such as sulfasalazine, methotrexate, and/or non-steroidal anti-inflammatory agents (non-biological group: NBD). Patients who had history of gastrointestinal hemorrhage, iron deficiency related anemia, any malignancy or patients under anemia therapy were excluded. Hemoglobin levels, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), and disease activity (BASDAI) were evaluated at baseline and after 52 weeks of the therapy. The diagnosis of anemia was performed according to the criteria of World Health Organization (WHO); for men hemoglobin levels below 13 gr/dl and for women values below 12 gr/dl were accepted as anemia.

Statistical analysis

Median values and standard deviation of numeric data and frequency of quantities data were evaluated. Kolmogorov–Smirnov test was used to determine mean values of acquired data in order to detect their dispersion. The comparison of the parameters with normal distribution was performed with student t test. Mann–Whitney U test or Wilcoxon rank-sum test was used for rests. p value <0.05 was accepted statistically significant.


There was no significant difference between groups with respect to disease age, sex, and patient age (p > 0.05). The percentage of anemic patients in TNF-alpha group and non-biological therapy group were 35.18 % (19 of 54 patients) and 19.56 % (9 of 46 patients), respectively. After 52 weeks of the therapy, the number of anemic patients in biological group and non-biological group was 11 (20.37 %), 15 (32.60 %), respectively. After 52 weeks of the therapy, a significant improvement was ascertained in BASDAI, CRP, ESR values in both groups. In biological group, not only the disease activity was suppressed but also a significant increase in hemoglobin values was found (approximately 1 g/dl). Non-biological agents suppressed the disease activity but had no positive effect in hemoglobin values. Results are given in Table 1.
Table 1

Clinic and laboratory features of the patients


INF n:20

ADA n:20

ETA n:14

NBD n:46

Age (years)*

36.25 ± 11.46

33.45 ± 8.20

34.79 ± 6.80

34.66 ± 11.38.

Duration of disease (years)**

12.30 ± 9.67

10.75 ± 7.74

9.28 ± 2.48

10.01 ± 8.44

HGB: g/dl


13.025 ± 1.84

13.55 ± 1.29

13.10 ± 1.67

13.28 ± 1.20

 After treatment

14.39 ± 1.48

14.13 ± 1.29

13.33 ± 1.83

13.14 ± 1.40






HCT: %


38.7 ± 5.6

40.39 ± 3.87

39.44 ± 4.29

39.52 ± 3.16

 After treatment

42.3 ± 5.2

41.88 ± 4.42

41.88 ± 4.55

39.68 ± 3.86






CRP: mg/dl


4.47 ± 4.423

1.72 ± 1.20

3.25 ± 2.90

1.27 ± 1.14

 After treatment

0.77 ± 0.69

0.45 ± 0.44

0.59 ± 0.65

0.91 ± 0.98






ESR: mm/h


35.3 ± 28.6

22.65 ± 16.684

37.86 ± 27.76

20.98 ± 18

 After treatment

11.2 ± 8.7

6.15 ± 5.75

17.71 ± 14.64

14.26 ± 9.02.16








6.28 ± 0.99

6.45 ± 1.40

6.31 ± 1.91

3.71 ± 1.44

 After treatment

2.28 ± 0.80

1.99 ± 0.68

2.03 ± 0.99

2.06 ± 2.19






Bold values indicate statistical significance

HGB hemoglobin, HCT hematocrit, CRP C-reactive protein, ESR erythrocyte sedimentation rate, BASDAI: Bath Ankylosing Spondylitis Disease Activity Index

* p > 0.05

** p > 0.05


TNF-alpha blockers have been proven highly effective in improving the spinal symptoms and extra-spinal manifestations of AS. All three TNF-alpha blockers (INF, ADA, and ETA) have similar efficacy in the treatment for AS symptoms. However, the efficacy of TNF-alpha blockers varies in the treatment of extra-articular manifestations. For example, ETA does not appear to be effective in reducing the incidence of uveitis, while INF and ADA do appear effective [12]. In addition, INF and ADA are effective for Crohn’s disease, while ETA is not [13]. To date, there is no adequate data about the effectiveness of TNF-alpha blockers on the anemia associated with AS. Braun et al. [7] had shown a significant improvement in anemia, parallel to improvement in joint symptoms, with INF therapy. But they had not investigated the effects of other TNF-alpha blockers such as ADA, ETA, and non-biological agents on anemia, as their study was placebo-controlled. In our study, TNF-alpha blockers such as INF, ADA, and ETA, which are in use in our country, were compared with non-biological agents; in ADA and INF groups, a meaningful increase was determined after therapy of 1 year. Similar effect did not occur in ETA and NBD groups. Normally, parallel to the decrease in hemoglobin values, erythropoietin (EPO) level is expected to increase. But in chronic disease, anemia manifest increase is rare, which is explained with the inhibitory effects of cytokines such as IL-1 and TNF-alpha on EPO secretion. TNF-alpha is proved to block the effects of EPO on CD34 (+) hematopoietic stem/progenitor cells (HSPCs) [14]. This mechanism can explain the positive effects of TNF-alpha blockers on hemoglobin levels. Soluble TNF-alpha primarily binds to TNF-Rp55 receptor, where as transmembrane form (tmTNF) primarily binds to TNF-Rp75 receptor. Etanercept primarily blocks soluble TNF-alpha, so the tm-TNF pathway stays intact. INF and ADA have blocker effect both on soluble and on transmembrane TNF. Infliximab (INF) also binds tm-TNF with complement or antibody dependent way and precipitates cell lyses. INF furthermore provides reverse signal induction, in tm-TNF expressing cells, which results with apoptosis. Etanercept binds TNF-alpha and reversibly blocks its effects and does not cause cell death. INF is IgG1 structured TNF-alpha blocker, which causes caspase activation, free radical excretion, and apoptosis by reverse signal [15]. These effects can explain why ETA is inefficient in Crohn’s disease treatment and why ETA is ineffective to improving of anemia. Our study showed these monoclonal TNF-alpha blocker agents’ improvements mean hemoglobin levels in addition to their disease activity suppressing effect in AS, where as ETA and NBD did not have such effects. Based on this evidence, among TNF-alpha blockers, the monoclonal antibodies (INF or ADA) seem as more appropriate choice than ETA if chronic disease anemia is present.

Low number of patient is a limitation of this study. Our results need to confirmation with a greater group study.

Copyright information

© Springer-Verlag Berlin Heidelberg 2012