Studies of immunopathogenesis behind IBD have recently identified the innate immune system as a major player in mucosal inflammation . The activated granulocytes and macrophages are important sources of pro-inflammatory cytokines [21, 25, 31]. Moreover, some of the pro-inflammatory monocytes are known to develop into locally infiltrating dendritic cells, which are professional antigen-presenting cells that maintain the mucosal inflammation by continuous stimulation of naïve T cells [41, 42]. In this context, the removal of these cells by leukocytapheresis provides us with a treatment strategy that targets major catalysts behind the intestinal inflammation. Furthermore, GMA (Adacolumn®) has been shown to improve clinical symptoms in patients with UC [22, 24, 43] and CD [44, 45]. GMA treatment has also been suggested to significantly reduce the dosage of prednisolone required to relieve symptoms in active UC-patients . A pilot study by Bresci et al even implies that GMA is more effective with less frequent and serious side effects than prednisolone in acute UC . Maiden and co-workers recently confirmed the efficacy of Adacolumn apheresis treatment in reducing relapse rates in UC as well as CD patients . However, to our knowledge the present study is the first to elucidate the effect of GMA-treatment in long-term follow-up including re-treatment at relapse.
The consecutive recruitment of patients with chronic inflammation resulted in a UC/CD ratio of 3/5 which may reflect the prevalence [18, 19]. The relatively small proportion of CD patients with small bowel involvement in this study, 3/25 (12%) compared to expected 48% in Stockholm County , may indicate a higher risk for CD patients with colonic inflammation to become chronic. However this reflection remains speculative within the scope of this report. An explanation to this recruitment discrepancy might also be a preference to refer patients with chronic inflammation localized to the small bowel or ileo-caecal region to limited surgery while trying GMA treatment on the colitis patients before colectomy.
This study includes a cohort of refractory patients who were re-treated with GMA upon relapse. We found that close to 100% of the patients who initially responded to a course of GMA also responded to subsequent GMA re-treatments. This may serve as a guide to select responders in the light of the high cost of this treatment.
In the present study including, 40 patients with chronically active IBD (15 with UC and 25 with CD) refractory to conventional medications, 34 patients (85%) responded to GMA. Further, 26 patients (65%) remained in clinical as well as endoscopic remission for an average of 14 months, ranging from two to 58 months. These remission values (65%) are in line with other studies, however with more homogenous patient selection [45, 50]. All 26 patients remained steroid-free during the follow-up period. Hence, GMA implicated substantial steroid-sparing effects. During the follow-up time, patients with relapsed disease were re-treated with GMA, resulting in high remission rates after the second to fourth time of GMA treatment. This finding indicated that maintenance therapy would be an interesting option but in lack of evidence at that time such a regime was not regarded as appropriate. Evidence for maintenance therapy has however recently been shown .
These encouraging results must be critically judged in the light of the open uncontrolled study design. Extracorporeal treatment may hold substantial placebo effects, even more pronounced than seen in similar patients groups . However, the patients included in this study were chronically inflamed and refractory to prior conventional treatments and almost all first-time responders who relapsed during the study period reached remission following re-treatment. The expected remission rate in one year in this sub-group of patients is unknown. The 65% clinical and endoscopic remission rate noted in the present group is in contrast to the placebo-controlled and randomised study  which was unsuccessful in reaching significant treatment effect. However, the placebo-controlled study included a more heterogeneous group of IBD-patients regarding disease patterns. In addition, the follow-up time only reached two to three weeks after completed GMA course. The discrepancy between the outcome of the current study and the aforementioned placebo-controlled study might possibly be explained by the presumption that chronically inflamed patients represent a more homogenous immunological phenotype regarding the importance of monocytes for driving the inflammation.
A total of 453 GMA procedures were performed and adequate venous access was achieved in all patients. Therefore, no exclusion of patients due to inadequate venous access was reported. Unlike other studies , we could overcome the problems with venous access partly by ensuring good peripheral blood flow and partly by using non-traumatising needles. The GMA treatment was well tolerated without serious side effect related to the apheresis procedure. Two patients developed transient headache, one person severe, and four patients showed transient fatigue. No patient discontinued the GMA treatment, which resulted in a compliance of 100%. HQoL were significantly increased in all sub-groups assessed between baseline and week 20. This is promising since the consistent active patients seem to reach lower quality of life scoring compared to other sub-groups . However, the dimensions "worry" and to a lesser extent even "general well being" in SHS increased assessed between the later time points measured. The impending fact that the close caring and weekly monitoring of these patients was reaching a final stage could partly explain these findings, in spite of the good contact availableness to the IBD nurses in continuation.
A total of 323 GMA sessions were given to the remission patients upon relapse and a total of 566 months in remission was achieved, equivalent to 1.75 months per apheresis column for patients in remission, which reflects a cost of 1150 $/remission month.
All 40 patients in the study suffered from chronic inflammation, defined as at least six months without remission in spite of conventional medication including immunosuppressors. Taking the pre-treatment clinical activity into account, each patient may be regarded as her or his own control. The continuous time period with active IBD prior to entry does not suggest that the GMA-induced remission correlates with the natural course of the disease in these chronically debilitated patients. Nonetheless, the observations in this relatively small cohort of IBD patients need to be further investigated in prospective, randomised controlled trials including larger cohorts of IBD-patients.
A sub-group of patients with IBD is characterised by a chronically activated immune response with a persistently inflamed mucosa. Defects in the innate immune response may be a major contributor to the onset of the inflammation, reflected by continuous activation of granulocytes and monocytes . These cells are crucial sources for the pro-inflammatory cytokines seen in the intestinal lesions , and many of the circulating monocytes are targeted for antigen-presenting functions at the inflammatory site. Hence, the removal of these leukocytes by extracorporeal selective adsorption may be an efficient approach to down-regulate the immune response in patients with limited effects from conventional medication.