Carter et al (2003) investigated the haptoglobin (Hp) polymorphism and its influence on iron metabolism. They concluded that Hp type neither influences iron status in normal subjects nor predicts clinical presentation of hereditary haemochromatosis in South Wales, in contrast to our findings in Belgian populations (Langlois et al, 2000; Van Vlierberghe et al, 2001).In randomly selected control subjects, in blood donors homozygous for the HFE C282Y mutation, and in first-time blood donors lacking the HFE mutations, Carter et al (2003) found no influence of Hp type on transferrin saturation and serum ferritin concentration. The authors provided no ready explanation for the discrepancy between their results and ours (Langlois et al, 2000). Because of the limited number of individuals, their study may have not had the statistical power needed to detect the influence of Hp type on these serum iron indices. Our group investigated and needed a large cohort to clearly demonstrate the higher serum iron indices in Hp 2-2 individuals.It is possible that the conflicting results may reflect geographical differences between study populations. It is of interest, in this regard, that no relationship between serum iron indices and Hp type could be found in reports from Africa (Kasvosve et al, 2002) and southern California (Beutler et al, 2002).More importantly, potential confounders such as blood donation, intake of iron supplements, alcohol abuse and inflammation were excluded in our study (Langlois et al, 2000). Moreover, we found that the influence of Hp type on serum iron status was exclusively present in males aged 18-50 years. Table I presents serum ferritin concentrations in 359 healthy Caucasians aged 51-89 years, from the region of Flanders (Belgium), selected according to previously described exclusion criteria (Langlois et al, 2000). In contrast to our data from younger males, we were unable to demonstrate an influence of Hp type on serum ferritin concentration. This discrepancy between age groups might be attributable to different underlying conditions that may additionally affect iron status in ageing men.Haptoglobin 2-2 complexed with haemoglobin exhibits a higher affinity for the CD163 receptor than the other phenotypes (Kristiansen et al, 2001), which contributes to haem iron retention in monocyte macrophages as evidenced by a higher cytosolic l-ferritin content (Langlois et al, 2000). In the short term, the impact of the haptoglobin pathway is relatively small compared with major iron homeostasis pathways. In younger males, the Hp type-mediated variation of serum ferritin concentration occurred within the generally accepted reference ranges (Langlois et al, 2000). However, a long-term contribution to iron accumulation may have clinical consequences, e.g. in the insidious evolution towards haemochromatosis.In patients presenting clinically with haemochromatosis, who were homozygous for C282Y, Carter et al (2003) found no influence of Hp polymorphism on transferrin saturation and serum ferritin. Unfortunate...

Currently available protocols for induction of warfarin anticoagulation employ initial doses of 10 mg and are best suited to in-patient use. However, with the increasing number of elderly patients with atrial fibrillation requiring anticoagulation, there is a need for a less intense regimen which could be used for out-patients. We have established such a regimen and report on its prospective evaluation in 3 7 patients referred for out-patient initiation of warfarin, and a non-randomized comparison with 37 in-patients, with similar diagnoses, commenced on a traditional warfarin protocol. After exclusion of five patients on amiodarone, all of whom experienced supratherapeutic International Normalized Ratio (INR) results, the new out-patient regimen, employing an initial 5 mg dose, resulted in a lower maximum INR during the first 21 d therapy (median 2.9 v 4.0; P = 0.0001) and fewer INRs >4.5 (2/36 v 9/33) compared to the traditional 10 mg regimen. Time to reach stable anticoagulation was similar with each regimen; however, the 5 mg regimen gave a more accurate prediction of maintenance dose (correlation coefficient for predicted versus actual maintenance dose, r = 0.985). In comparison to a traditional 10 mg protocol, the proposed 5 mg warfarin induction regimen proved both safer and more reliable for initiation of prophylactic anticoagulation in patients with atrial fibrillation.

Summary:Patients undergoing allogeneic bone marrow transplants (BMT) are often malnourished prior to commencing the procedure. They face intensive treatment with often marked nutritional consequences. There is no consensus on the optimal nutritional management of these patients. Elective parenteral nutrition (PN), beginning pre-transplant irrespective of the patients nutritional status, or the use of 'salvage' PN, beginning during the post-transplant period if the patient fails to maintain nutritional status with oral diet, have been used. Enteral nutrition may benefit the patient by maintaining nutritional support throughout the transplant period, avoiding the complications and expense of PN and possibly, by using specific diets, protecting the gastrointestinal tract from the effects of chemoradiation. However, naso-gastric feeding during a transplant is not without risks, including the safe insertion of a tube in patients with mucositis and pan-cytopenias, tube displacement by vomiting and aspiration from gastro-oesophageal reflux. An alternative approach is to use nasojejunal (NJ) feeding tubes which are associated with less risk of loss due to vomiting and less risk of aspiration. We report a pilot study of 15 allogeneic BMT patients who had elective NJ feeding initiated before conditioning therapy irrespective of perceived nutritional compromise. This was well tolerated and feasible with a motivated nutritional team. Bone Marrow Transplantation (2001) 28, 1135-1139. Keywords: allogeneic bone marrow transplant; enteral feeding; naso-jejunal feeding tubes; pilot study Patients undergoing BMT are often under-nourished due to previous chemotherapy and the underlying disease process. They face a long hospital stay and intensive conditioning therapy, which can have marked nutritional consequences. Subsequent complications such as infections, veno-occlus-

Stenotrophomonas maltophilia (previously Pseudomonas maltophilia and Xanthomonas maltophilia) is an increasing problem as an opportunistic pathogen in immunocompromised patients such as those with acute leukaemia. Infection is difficult to treat and eradicate due to its inherent and quickly acquired resistance to many antibiotics. Presentation with unusual cutaneous lesions is not uncommon and can present diagnostic difficulties. We present two cases which highlight the classical features of this infection in leukaemic patients, and discuss the need for early antibiotic treatment with regimens including cotrimoxazole, and the consideration of prompt central venous catheter removal.

Fig 1.Time course of leucocyte and promyelocyte (blast) counts in peripheral blood. Day 1 was the data of readministration for relapse. BM ¼ bone marrow aspiration. Blast ¼ myeloblast and promyelocyte counts.