Mass spectrometry (MS) is becoming an accepted technique for clinical applications. It has already been shown to be a valuable technique for the characterisation of structural hemoglobin variants, and its applicability for screening purposes has been demonstrated by several research groups. It has not yet been implemented as an approved clinical laboratory technique to screen for hemoglobinopathies, especially the thalassemias. In order to replace existing methods the proposed approach has to provide improved technical, financial and medical benefits. For the purposes of antenatal/neonatal screening laboratories, all methods must at least detect HbS, HbC, HbD-Punjab, HbE, HbO-Arab, Hb Lepore, elevated HbF, and elevated HbA2 (to distinguish α and β thalassemia). Three mass spectrometry based approaches have been evaluated against a current standard method (cation exchange liquid chromatography, HPLC, and gel electrophoresis) in a clinical trial incorporating 2017 patient samples, in 12 batches. The MS methods used were:1.Measurement of the mass of intact globin chains and the relative proportion of β and δ globin chains. When any α or β variant was detected, this was followed by tandem MS of tryptic peptides to confirm identity.2.Ultra performance liquid chromatography+MS/MS experiments using pre-determined tryptic peptides to screen for clinically important structural variants.3.Mass measurement of intact globin chains followed by top down electron transfer dissociation (ETD) fragmentation of selected globin chains for confirmation of variant (no complex pre-analytical sample digestion/separation, or second stage analysis). Variant detection The clinical samples were all tested using methods 1 and 2. Method 3 has been used on a more limited sample set which, however, included all samples which were identified to have clinically significant variants. All methods agreed with the hospital lab results for >99.5% of samples. However, a number of additional observations could be made using the mass spectrometry data. The current clinical laboratory method was found to give false results on three samples (0.15%), two of which were identified by the existing approach as HbD-Punjab but were in fact characterised unambiguously by MS as HbG-Philadelphia. One sample was identified as containing elevated HbF by the existing method but was characterised as a β-chain variant using MS. In addition, seven patient samples were found by mass spectrometry to contain non-significant Hb variants which were not detected using the existing hospital method. HbA2 measurement (based on ratio of δ chains to ‘δ+β' chains) The correlation coefficient (R2) for HbA2 measurement between HPLC and MS in the 12 batches varied from 0.57-0.90 (method 1), and 0.72-0.90 (method 2, using the T5 tryptic digest peptide pair), but with significant systematic bias requiring either calibration to a standard or adjustment of the normal range. However, a cut-off between normal and abnormal was clearly evident, and there was complete agreement between HPLC and MS methods in categorisation of samples as beta-thalassemia trait. Method 3 has shown potential to measure HbA2levels but has not yet been tested on a large batch of clinical samples. The approaches differ in their instrumentation requirements, sample introduction, sample preparation and data interpretation. Key issues to be considered when selecting the most appropriate method to develop for future hospital use include cost, speed, sensitivity, selectivity, potential for automation and diagnostic information provided. Disclosures: Smith: Bruker UK Limited: Employment.

An analogue to digital converter architecture is presented. Statistical redundancy of signals such as audio allows the average number of cycles per sample to be reduced thanks to a local quantification scheme. The converter produces irregularly time spaced samples, without conversion error, at a higher speed than comparable architectures

Background: Haematohistopathologists require a sample of sufficient volume and quality to be both sensitive to recognising the presence of pathology and to make a specific diagnosis. An 8G trephine needle has a 3.43 mm internal diameter, larger than the 11G at 2.39 mm, which translantes to a 2x larger area. Not only can this provide substantially more material, it may allow for a more robust sample, less likely to suffer a crush artefact. However, the 8G needle can be intimidating, and there are concerns for patient tolerance and bleeding risk. There is no study comparing trephine gauge, so we performed a prospective observational study comparing these two sizes. Aims: A prospective study comparing use of 11G and 8G trephine needles on resultant sample quantity and quality, crush artefact, procedure ease, ease of sample extraction, haemostasis and diagnostic yield. Methods: Practitioners performing bone marrow biopsies at a large teaching hospital selected either 11G or 8G trephine needles for the procedure. They completed a questionnaire on the ease of performing the biopsy, whether several attempts were taken, if there was any prolonged bleeding, the ease of sample extraction and the approximate length. A haemato-histopathologist completed a questionnaire commenting on whether the sample was crushed, diagnostic, sufficient and the overall quality. These results were compared using the R numerical language. Results: Data was collected on 14 11G and 20 8G biopsy attempts, all using a 4" needle. The length of samples obtained were comparable (mean 2.04 vs 2.07 cm, 11G vs 8G). Ease of procedure appeared to be higher in 11G (86%) vs 8G (75%). However, ease of sample extraction was lower in 11G (79%) vs 8G (95%). A further attempt was required in 21% of 11G and 15% of 8G. Regular haemostasis was achieved less in 11G (86%) than in 8G (95%). Haemato-histopathologist assessment revealed slightly less diagnostic samples from the 11G (85%) than from 8G needles (90%), not including failed attempts. Crush artefact was seen in 23% of 11G vs 25% of 8G. Overall quality was comparable between 11G (mean 3.9 of max score of 5) and 8G (mean 4.0). Sufficient tissue quality was lower in 11G (79%) vs 8G (90%). Haemorrhage artefact was lower in 11G (8%) vs 8G (20%). Summary/Conclusion: Our continued study comparing 11G and 8G trephine needle suggests a similar overall sample quality between the needle sizes. With around double the area of sample, theoretically only half the length is needed to obtain the same volume of tissue. Despite this, operators were obtaining similar lengths with both biopsies. Some differences were seen, with the 11G needle holding advantages in the ease of taking the sample, and having less haemorrhage artefact. The 8G needle was easier to extract the sample from, had less repeated attempts, was diagnostic in more cases and more often provided sufficient tissue quantity. Haemostasis appeared to be slightly higher in the 8G biopsies, though this may be due to operator selection based on patient characteristics....