Assessment of time from onset of first symptoms to the final diagnosis of multiple myeloma (MM) - possible risks and future solutions

Background: Most MM cells secrete an iIg and the corresponding FLC in excess. With its shorter half-life, FLC provides real-time information on tumor protein synthesis. In addition, immunoglobulin heavy chain/light chain analysis (HLC) is associated with a high sensitivity. Our study aimed to determine if early detection of response by FLC and HLC is predictive of outcome in patients with iIg MM. Methods: We prospectively studied 30 episodes of treatment in 24 patients with iIg MM measurable by serum protein electrophoresis (SPEP) ( 10 g/L) and FLC ( 100 mg/L). The M-spike, FLC (Freelite) and HLC (Hevylite) were measured before and after initiation of treatment weekly for the first 2-3 cycles. FLC and HLC assays were performed retrospectively by nephelometry and did not influence treatment. Patients were followed until next treatment or death. Response, based on IMWG criteria, was defined as at least partial response (PR). Criteria for HLC were the same as for FLC. Results: Median follow-up was 27.6 months (mos). Response by FLC preceded response by SPEP by a median of 3 weeks. By the end of cycle 1, PR was achieved in 23% of cases by SPEP, vs 63% by FLC. On cycle 2, median EFS for patients in PR by FLC compared to patients with stable disease (SD) was 17.5 vs 4.5 mos (p1⁄40.0039). After adjustment for age, line of treatment, ISS and cytogenetics, response by FLC on cycle 2 remained predictive of better EFS (HR 0.20 p1⁄40.008), in contrast to SPEP. On cycle 2, among patients with SD by SPEP, those with PR vs SD by FLC had a longer median EFS (17.5 vs 6.0 mos, p1⁄40.0441) and longer 1-year EFS (56 vs 22%) (figure 1). Among patients in PR by FLC, patients presenting with an unstable response (decrease in FLC by 50% followed by an unsustained increase in FLC by 25% and 25mg/L) had a significantly worse 1-year EFS compared to patients with a stable response by FLC (87.5% vs 30.8%, p1⁄40.0015). These paraprotein fluctuations within a cycle were not apparent with SPEP, reflecting the longer half-life of ilg. In our patients, HLC did not detect response earlier than FLC. Conclusion: Patients with ilg MM failing to respond by cycle 2 or with unstable response using FLC assay have significantly shorter EFS. However, addition of early HLC measurements is not more informative. Our results demonstrate that early FLC measurements after initiation of treatment are predictive of clinical outcome and could be used to better adjust therapy in patients to improve efficacy. PO-047 Assessment of time from onset of first symptoms to the final diagnosis of multiple myeloma (MM) possible risks and future solutions G. Graziani, J. Waldschmidt, G.W. Herget, M. Pandurevic, K. Henne, R. Selder, A.M. May, M. Moller, G. Ihorst, H. Reinhardt, W. Vach, J. Duyster, R. Wasch, M. Engelhardt Department of Medicine I Hematology and Oncology, Medical Center University of Freiburg (UKF); Comprehensive Cancer Center Freiburg e CCCF, Medical Center University of Freiburg; Department of Orthopedics and Trauma Surgery, Medical Center University of Freiburg; Department of Radiation Oncology, Medical Center University of Freiburg; Institute of Clinical Pathology, Medical Center University of Freiburg; Clinical Trials Unit, Medical Center University of Freiburg; Center for Medical Biometry and Medical Informatics, Medical Center University of Freiburg Introduction: The International Myeloma Working Group (IMWG) has recently updated the criteria for the diagnosis of MM, including definition of ultra-high-risk Smouldering Myeloma (SMM) and thereby revised the definition of active MM, which should be treated (Rajkumar et al. Lancet Oncol 2014). As a result, earlier diagnosis and treatment initiation of MM is proposed, albeit MM detection remains challenging due to the unspecificity of MM symptoms. Objective literature regarding the duration of diagnosis finding is lacking and possible risks which might lead to substantial latencies remain undetermined as yet. Methods: Here, 108 MM patients (pts) diagnosed between 1997-2014, treated within clinical trials at the Medical Center University of Freiburg and discussed in our MM tumor board, were meticulously analyzed via all medical records available. We assessed the duration and type of initial symptoms occurring until MM diagnosis and whether specific risks for possible delays could be determined. For the subsequent prospective analysis, we developed a concise MM-specific questionnaire to assess pts seen in our outpt clinic. Results: 102 of 108 pts showed symptoms before the diagnosis of MM. In 16 pts (15%), prior MGUS, SMM or solitary plasmocytoma were documented. The median time from first symptoms to the final MM diagnosis was 3 months (ms), albeit with large variation (0-120 ms). Pt frequencies diagnosed within 12ms are shown in Fig. 1A. Comparison of initial symptoms in pts diagnosed within 6 (n1⁄468) vs. >6ms (n1⁄434) did not reveal significant differences in the type of symptoms (Fig. 1B). Non-CRAB-related symptoms, like B-symptoms or infections, were more frequent with 35% and 25%, respectively, than previously reported (Engelhardt, Haematologica 2014). Of note, pts diagnosed in >6ms showed increased frequencies of light-chain-only MM, prior orthopedic or rheumatic comorbidities, residency in smaller communities and longer distance to tertiary centers compared to pts diagnosed within 6ms. Conclusions: The detection of possible risks for a delayed diagnosis finding is highly relevant, as our results show that 20% of MM pts experience latencies of 12ms. Differences in type of initial symptoms were not significant, whereas others, e.g. disease 15th International Myeloma Workshop, September 23-26, 2015 e107