When comparing serum and urine FLC results (e.g. serum Freelite results vs. urine electrophoresis) it is essential to ensure that the samples are from the same time point. If there is a significant time delay between the collection of serum and urine samples, even a few days, any observed difference may simply reflect response to treatment or disease progression because of the short serum half-life of FLCs (Chapter 3).
7.6.1. Monoclonal FLCs in urine, normal sFLCs
FLC analysis is generally more sensitive than urine electrophoresis for indicating the presence of monoclonal FLCs. However, this advantage is dependent upon efficient renal reabsorption of FLCs (Section 3.4). Small amounts of monoclonal FLCs have occasionally been identified in the urine of patients with normal sFLC ratios.
In a prospective screening study by Beetham et al.  monoclonal proteins were detected in 105 (22%) patients, 34 of whom had urinary Bence Jones protein (monoclonal FLCs). Of these 34 patients, eight had normal sFLC κ/λ ratios; however, seven were found to be positive for intact monoclonal immunoglobulins by SPE/sIFE and the remaining patient was considered to have a urine-only MGUS (<50 mg/L) of no apparent clinical consequence. These results support the findings of other studies, which recommend that SPE/sIFE and sFLC analysis can replace urine studies when screening for monoclonal gammopathy (Chapter 23). However, Beetham expressed some disquiet as to why monoclonal FLCs were present in urine when sFLC κ/λ ratios were normal . There is more than one mechanism whereby this may occur. One theory, proposed by Holding et al.  suggests that false positive results could be generated by the catabolism of intact immunoglobulins in urine. Although this theory remains unproven, it is known that renal podocytes express Fc receptors, which facilitate clearance of IgG that has traversed the glomerular filtration barrier into the urine (Section 3.5.3) , with subsequent separation of FLCs from intact IgG molecules. Consistent with this theory, Holding et al.  reported that the Freelite assay was able to measure urine FLCs (with values consistent with urine densitometry) in two patients with a normal κ/λ sFLC ratio and very low levels of urinary FLCs. Therefore, the Freelite assay is not missing epitopes in such cases.
A small proportion of AL patients may have trace amounts of urinary BJP but a normal sFLC ratio. Mead et al.  compared sFLC and urine IFE results in a cohort of 219 AL amyloidosis patients attending clinics at the National Amyloidosis Centre, London. Of these patients, 56 had abnormal sFLC ratios and monoclonal FLC detected in the urine; 52 had abnormal sFLC ratios but urine that was negative by IFE, and 16 had small monoclonal bands detected by uIFE but sFLC ratios within the normal range. Of this latter group, 12/16 had nephrotic-range proteinuria (>3 g/day), so saturation of protein reabsorption mehanisms by albumin and other proteins would explain the increased passage of FLCs into their urine (Chapter 3). Serum levels, in contrast, may not be raised sufficiently to produce abnormal κ/λ ratios. For the other 4/16 patients, other mechanisms must have been responsible. All four of these patients had sFLC ratios biased towards the tumour light chain (0.30, 0.34 and 0.49 for λ patients and 1.61 for the κ patient).
In a separate study by Palladini et al.  five of 115 (4%) AL amyloidosis patients had monoclonal bands detectable by uIFE but sFLC ratios within the normal range. Interestingly these five patients were all λ-type AL patients. This may reflect the higher proportion of λ AL patients with nephrotic-range proteinuria compared with κ patients .
7.6.2. Monoclonal FLCs detectable by sIFE but undetectable by sFLC immunoassay
On very rare occasions, sFLCs may be undetectable by immunoassay but detectable in the serum by IFE. In such cases, further investigation is always warranted. Possible explanations include: 1) antigen excess (Section 7.4); and 2) failure of the sFLC immunoassay to recognise a particular patient's FLC epitopes. This is highly unlikely. There has been only one reported case in light chain MM, in 2004 , where κ sFLCs were underestimated during a period of follow-up. The authors supplied Binding Site with urine from the patient, and purified FLCs were incorporated in the antigen pool used to produce subsequent batches of Freelite antisera .
7.6.3. No monoclonal proteins detectable by any routine laboratory method
For a small proportion of AL amyloidosis patients, no monoclonal FLCs are detected by any routine laboratory methods. In a large screening study by Katzmann et al., 11 of 581 (2%) AL patients were normal by sFLC analysis, sIFE and uIFE (Chapter 23) . Three possible explanations include: 1) some FLC molecules may have a high affinity for the amyloid deposits, resulting in circulating FLCs being rapidly removed; 2) in a similar manner, patients with extensive amyloid deposits might have a huge capacity for FLC removal. Any newly synthesised molecules would be cleared rapidly by a combination of binding to the amyloid mass and glomerular filtration, thereby preventing the accumulation of FLCs in serum; or 3) the amyloid is due to the deposition of a different protein .