Randomized Clinical Trial Of Preoperative Oral Versus Intravenous Iron In Anaemic Patients With Colorectal Cancer. Keeler BD, Simpson JA, Ng O, Padmanabhan H, Brookes MJ, Acheson AG on behalf of the IVICA Trial Group. British Journal of Surgery 2017;104:214-221

Presented by: Dr D. John


  • Bad things
    • Anaemia – Common – 40% of colorectal cancer patients
  • Adverse associations with blood transfusion
    • Dose independent factors – Expensive, scarce, infection, immunological reactions
    • Dose dependent factors – Biochemical derangements, post-operative infections, length of stay, cancer recurrence, mortality
  • Connected things
    • Pre-operative anaemia and peri-operative blood transfusion
  • Good things
    • Correcting pre-operative anaemia
  • Anaemia in colorectal cancer
    • Almost always due to chronic blood loss
    • Can be corrected with iron therapy
  • Pre-operative recommendations
    • FBC 6 weeks before surgery
    • Oral iron should be first line (ideally until 3 months after restoration of ‘normal’ Hb
      • Cannot be used post-operatively
    • Intravenous iron should be used for those intolerant to oral iron
      • Can be used post-operatively
    • Erythropoetin should not be used (not cost-effective)
  • Peri-operatively
    • Consider transfusion when Hb below 80g/l
    • Give transfusion when Hb below 70g/l
    • Clinical assessment trumps transfusion triggers
    • Transfusion should be by single units
    • Above applies to those with cardiovascular disease


  • Determine whether oral iron or intravenous iron is better:
    • At correcting pre-operative anaemia
    • At reducing the need for perioperative blood transfusion

Design and Setting

    • Ethical approval, trial registered
    • Multicentre randomised controlled trial
      • 7 UK sites
      • Transfusion practice in accordance with:
        • Joint UK Blood Transfusion and Tissue Transplantation Services Professional Advisory Committee Guidelines for Surgery


  • Patient selection
    • May 2012 – June 2014
    • Colorectal adenocarcinoma
    • Anaemia – ‘to 1g/dl [10g/l] below WHO definition of anaemia of <12g/dl [120g/l] for men and <11g/dl [110g/l] for women’
      • WHO definition of anaemia is 10g/l greater in both sexes – erroneous or poor expression
  • Exclusions
    • Clinical
      • Metastatic disease, pre-existing haematological disease, renal or hepatic failure, chemotherapy, iron contraindications – iron overload, allergy, symptomatic anaemia necessitating transfusion
    • Logistic
      • Children, prisoners, pregnant women, lactating women, women planning pregnancy, previous alimentary tract surgery (>50% of stomach or terminal ileum resected), urgent surgery, previous anaemia not due to colorectal cancer, incapable patients, planned blood donation for surgery, any other significant disease or disorder which may put the patient at risk or influence the trial result (investigator’s opinion), patient unable to comply with study’s requirements (investigator’s opinion), patient unwilling to disclose information to surgeon or GP,


  • Randomisation
    • Web-based system
    • Stratified by sex and age
    • Independent of the trial investigaotrs
  • Blinding
    • None

Randomised to:

  • Oral iron
    • Ferrous sulphate 200mg bd until surgery (BNF recommends tds for treatment of anaemia)
  • Intavenous iron
    • Ferinject dosed by Hb and weight (as advised)
    • If two doses required 7 day interval given

How did they go about things?

  • First consultation
    • At least 14 days prior to surgery
    • History, physical examination, randomisation done
    • Iron administration commenced on same day (pre-existing iron supplementation discontinued – investigator should exclude?)
    • If second intravenous dose required then done a week later
  • Peri-operative
    • Pre-operative review (‘around the time of surgery’)
    • Day 2 following surgery
    • Interval reviews until first post-operative clinic (2-3 months usually)
      • End of trial (or if unresectable at surgery, if dead at death, if lost at discharge)
  • At each visit
    • Blood transfusion requirements assessed
      • Volume, date and complications noted
      • Electronic blood bank database and case notes reviewed
    • Hb, ferritin, transferrin saturations
      • Pre-trial and day before surgery
  • Other data recorded
    • Operative blood loss (suction, swab weight), volume and type of intravenous fluid used intra-operatively documented at the time
    • Tumour location, size and grade fom histology report


  1. Primary Outcome – Reduction in perioperative blood transfusion
  2. Secondary outcome – Correction of preoperative anaemia


  • Population
    • Power calculations done – power of 90% of p<0.05
      • Assuming 1 unit difference between groups
      • 116 patients (accounting for drop-outs)
  • Statistics
    • Intention to treat
      • Non-Gaussian paired data used Wilcoxon signed rank
      • Non-Gaussin independent data used Mann-Whitney U
      • Gaussian data used Student’s t
      • Categorical data used Chi-squared
    • P<0.05 was cut off for significance
  • 116 patients
    • 55 to oral therapy
    • 61 to intravenous therapy
    • No demonstrable demographic difference
    • Median time between commencement of therapy and surgery was 21 days in both
    • 11 (6 oral, 5 intravenous) had surgery before 14 days of therapy (3 due to clinical reasons, 8 due to earlier date availability)
    • Median time between surgery and first out-patient appointment was 97 days for oral therapy and 87 for intravenous therpy
  • Adherence
    • Oral therapy by 91%
      • 3% (2 patients) reduced dose due to adverse reaction, 3% increased dose on clinical advice, 3% changed to ferrous fumarate
      • None converted to intravenous therapy
    • Intravenous therapy by 93%
      • 3.5% (2 patients) could not attend second appointment, 3.5% had surgery brought forward
      • 5.5% (3 patients) had post-infusion headache
      • 2% (1 patient) had a rash (oral antihistamines)
  • Cancelled surgery
    • 4 had surgery cancelled due to health decline
    • 1 died during anaesthesia
    • 1 unresectable
  • Blood transfusion (Primary outcome)
    • Pre-operative
      • Oral therapy – 2 patients
    • Intra-operative
      • Oral therapy – 6 patients
      • Intravenous therapy – 6 patients
        • No difference in number transfused (P=0.894)
      • No difference in volume transfused (P=0.863)
    • Post-operative
      • Oral therapy – 14 patients
      • Intravenous therapy – 10 patients
        • No difference in number transfused (P=0.470)
      • No difference in volume transfused (P=0.841)
  • Haemoglobin (Secondary outcome)
    • At recruitment – no significant difference
    • At surgery – significant difference
      • Median rise in oral therapy 5g/l (IQR -1.3 – 13.3), intravenous therapy 15.5g/l (9.3 – 25.8) (P<0.001)
      • Percentage still anaemic in oral therapy 90%, intravenous therapy 75% (P=0.048)
      • Percentage requiring iron therapy in oral therapy 54%, intravenous group 7% (P<0.001)
  • Other measures
    • At surgery
      • Median ferritin level in oral therapy 27.5mcg/l (IQR 17 – 51.5), intravenous therapy 558mcg/l (330-1085) (P<0.001)
      • Median transferrin saturation in oral therapy 9 (IQR 5 – 14), intravenous therapy 19 (16 – 29) (P<0.001)
    • At discharge
      • No difference in length of stay (6 days)
    • Mortality, morbidity, post-operative infection, grade of malignancy
      • No difference


  • Intravenous therapy better than oral therapy
  • No difference in transfusion
  • No difference in morbidity, mortality, length of stay
  • Intravenous iron not inferior
    • No significant adverse reaction


  • Randomised controlled trial
  • Multicentre
  • Each centre followed the same transfusion practice in accordance with the Joint UK Blood Transfusion and Tissue Transplantation Services Professional Advisory Committee Guidelines for Surgery
  • Asking a very relevant question


  • Admitted limitations
    • Lower transfusion rate than expected
      • Higher rate of laparoscopic surgery than previous studies (and consequent lower blood loss)
        • Potential type II error (incorrect acceptance of null hypothesis)
        • Inadequately powered
    • Inadequate duration of therapy prior to surgery
      • Most studies suggest 21 days of therapy is inadequate
      • Would adherence be lower with longer duration of oral therapy in clinical practice?
    • Conflicts of interest
      • Ferinject manufacturer donated drug to most centres and have provided grants, honoraria and travel expenses to many of the paper authors


  • IV iron probably better than oral
  • Treating anaemia important

Potential for Impact

  • Study not well conducted but may reflect clinical reality
  • Potential for bias significant
  • Probably underpowered