CME Slides Forum - H.J. Lachmann

A joint Congress by ERA-EDTA and ISN

USE OF THE SERUM FREE LIGHT CHAIN ASSAY

Helen J. Lachmann, London, UK

Chair: Claudio Pozzi, Milan, Italy

Pierre Ronco, Paris, France

lachmann

Dr Helen J. Lachmann
UK National Amyloidosis Centre
University College London Medical School
London, United Kingdom

The paraprotein related renal diseases are a group of diseases which consist of myeloma cast nephropathy, AL-amyloidosis and light chain deposition disease. Much more rarely there’s also immunotactoid glomerulopathy and Fanconi syndrome. These diseases share a number of similarities: they typically affect older adults and the median age at presentation is 55-60 years; they’re the primary abnormalities in about 3% of native renal biopsies. Myeloma cast nephropathy is the commonest found in about 1% of native renal biopsies. AL-amyloidosis is found in slightly more renal biopsies (1.5%) because it is a less easy diagnosis so a greater proportion under go renal biopsy. Light chain deposition disease is found in about half a percent of renal biopsies in a British series. All these diseases cause proteinuric renal impairment with normal sized kidneys but they have differing histological appearances and different propensities to extrarenal involvement which to some extent is reflected in the mortality and morbidity associated with them. The co-existence of two or more paraprotein related renal lesions is rare but well recognized.

In myeloma cast nephropathy the monoclonal immunoglobulin light chains are deposited in a crystalline form, in AL-amyloidosis in a fibrillar form and in light chain deposition disease as granular non-organised deposit. In many respects it is worth about these diseases as a very similar group as the same principals are involved in their management.
All are complications of an underlying haematological disorder and the majority of these are complications of plasma cell dyscrasias which can be broadly grouped as myeloma, solitary plasmacytoma or monoclonal gammopathy of undetermined significance. In fact under current international definitions all the plasma cell disorders resulting in these diseases are myeloma because they result in end-organ damage away from the bone marrow. It is none the less worth considering their malignant potential because it makes a difference to the treatment that is going to be offered and to the prognosis given to the patients.
A minority of these patients will not have an underlying plasma cell disorder, but will have a lymphoproliferative disorder: a malignant B cell clone which can also make monoclonal immunoglobulin and the majority of these will be a lymphoplasmacytic lymphoma but both non-Hodgkin’s lymphoma and chronic lymphocytic leukaemia can result in these paraprotein renal related diseases.

What these clones all have in common is that they produce an excess of monoclonal immunoglobulin. These can be whole immunoglobulin or they can be monoclonal light chains alone. In health we have a 3 to 2 excess of kappa secreting cells but in fact in AL amyloidosis, we see an excess Lambda secreting clones.

The renal handling of immunoglobulin light chains is relevant to these diseases. The majority are absorbed by the PCT up to a capacity of about 10-30 g/day, any excess can be detected as Bence-Jones proteinuria. In these diseases deposition of the light chains can occur anywhere from the glomeruli, to the vessels, the interstitium and tubules. The presence of Bence-Jones proteinuria is not particularly useful in making the diagnosis in any of these patients with more subtle lesions.

The free light chain assay was developed by Jo Bradwell and a company called the Binding Site in the late 1990s and has had a major effect on our ability to manage these patients. They generated polyclonal sheep antibodies that can detect epitopes that are hidden in whole immunoglobulin but are revealed in free light chains. These assays are biological so they vary a bit from batch to batch but are standardised. They can be run on automated bench top nephelometers and they have a sensitivity of about 1 mg/l. This compares to sensitivities of conventional immunofixation of about 150 mg/l and electrophoresis of about 1 g/l.
So it is possible with these assays to detect free light chains within the normal range in all individuals and to detect very subtle elevations. The original work with this produced graphs which look a bit like this. So this is using free light chain assays to measure both serum Kappa and Lambda light chains and then to plot out to look at the ratio. You can see that normal individuals have low level production of both Kappa and Lambda with a ratio which sits on either side of 1.
Individuals without clones but with chronic renal impairment and end stage renal failure have higher levels of both because these light chains are predominantly cleared by the kidneys with a short half life but the ratio remains normal. This is very important in interpreting these results in patients with renal impairment where the absolute numbers are often not very significant. It’s vitally important to get both the Kappa and the Lambda so that the ratio can be interpreted. One of the major limitations of light chains is trying to interpret these results in patients with changing renal function where you end up often having to rely heavily on the ratio because absolute numbers may be rising reflecting a drop in creatinine clearance and glomerular filtration when the actual clonal load is falling. Subtracting the non clonal class from the clonal class may get around this problem.

What these results also showed is plotted here in purple are patients with Kappa myeloma showing extremely high levels of Kappa light chain production, up to grams/l. Patients with Lambda myeloma also have extremely high levels of production. Plotted in orange are patients with AL-amyloidosis and you’d expect the same sort of results in patients with light chain deposition disease showing that on the whole although not exclusively these are relatively more subtle clones producing lower levels of light chains which are just more inclined to get deposited. You can also see that there’s considerably more Lambda than Kappa showing what we already knew which is that although a majority of plasma cells are Kappa producing, in AL-amyloidosis there’s about a 2/3 Lambda bias.

So why do these light chains deposit in the tissues? The pathological effect of the deposited light chains are probably due to a combination of features. Some of them are the specific properties of the monoclonal immunoglobulin light chains themselves. This was first demonstrated very elegantly by Alan Solomon in 1991 who took concentrated BJP from humans with a number of light chain associated lesions, injected them into the peritoneum of mice and when the mice were sacrificed, they had the same renal lesions as the patients from whom the BJP had been taken. Showing that much of this is due to intrinsic properties of the light chains. We now know that there are often specific aminoacid substitutions within the variable regions. So for example in amino acid substitutions in the V6 Lambda region are very strongly associated with the tendency to deposit as amyloid. We also know that their concentration is extremely important. Myeloma cast nephropathy really does only occur in genuine myeloma. We don’t see it with rather lower grade clones. We also know that although the development of AL-amyloidosis and light chain deposition disease in myeloma is relatively unusual the relative risk is very much higher than for patients with lower grade monoclonal gammopathies showing a concentration effect. There is also clearly an effect from the local microenvironment. Some of this is shown by the evidence of recurrence of these lesions in renal transplants which can happen very early if patients don’t receive chemotherapy and exactly what’s going on there is clearly complex and not that well understood.

Well the major treatment is to reduce the supply of monoclonal FLCs and I’m only going to talk about chemotherapy and leave other modalities to other speakers. The aim is really to reduce the supply of the monoclonal FLCs as fast, as completely and for as long as possible without doing too much damage in the way of morbidity and mortality and combining this with support or replacement of compromised organ function. That can be very tricky because their renal function is extremely fragile.

So is there a role for chemotherapy in these diseases? This is data from AL-amyloidosis but it would work just as well for all the other diseases. This is the original paper from Robert Kyle from 1997 showing that patients treated with rather low grade chemotherapy, melphalan and prednisolone had significantly better survival than patients treated with colchicine which was as it happens is completely ineffective but at the time was the traditional treatment. Later data (this is taken from our centre and is non-selected data from a large cohort of 2500 patients) shows that patients who were given chemotherapy and respond to it have very much better survival than patients who don’t respond to chemotherapy demonstrating that effective chemotherapy has a major beneficial effect on survival in these patients.

So the rationale for treatment is that the organ dysfunction in these diseases is a complication of deposition of monoclonal FLCs. We also know that regression of deposits can occur if the supply of precursors is stopped or reduced. The aim of chemotherapy treatment is a rapid clonal response, a complete clonal response and a prolonged clonal response with minimal side effects. The difficult thing with these diseases is that many of these patients will have multiple organ involvement and will tolerate chemotherapy extremely poorly with a high treatment related mortality. There are still arguments as you will hear about the optimal treatment in terms of intensity, duration and regimes to use although the aims in terms of acheiving clonal responses are now not at all contentious.

This is a large cohort of patients from our database showing that in fact although in 52% of patients you can detect serum paraprotein we can only measure it in a third of patients. In the others we can detect it but only by immunofixation. In urine we can detect it in about 40% but that’s a very unreliable way of measuring clonal response because a large number of these patients will go into end stage renal failure and it’s very difficult to quantify proteinuria in the face of a changing creatinine clearance and change in proteinuria. But FLC assays allow us to serially detect an abnormal FLC ratio or absolute number and monitor it in almost 80% of these patients. So serum FLC assays have enormously improved our ability to measure and then monitor clones thereby allowing titration of chemotherapy.

This is original data from the first paper looking at light chains in AL-amyloidosis showing that patients whose light chains fell by more than 50% which in haematological terms would be regarded as a partial response had a 10-fold better survival than patients in whom light chains didn’t fall at all suggesting that serial monitoring of light chains was a very useful way of predicting survivors from non-survivors.

Is a complete response better than a partial response? This again is data from our centre and this from the Americans showing that patients with a complete response have an extremely good long-term survival compared to patients with a partial response and in this paper from Angelo Dispenzieri, you can see that it’s the light chain response which is important not the whole paraprotein response and patients who have a complete light chain response and a paraprotein partial light chain response have the same survival as patients with a complete response in all measured indices, whereas patients with a complete response in terms of their whole paraprotein but not their light chain response have much poorer survival.

Obviously this matters less in myeloma where they have a malignant clone and you wish to treat them to a complete response than in AL- amyloidosis or light chain deposition disease where one may be more prepared to tolerate an incomplete response and I’ll show you why for that in a moment. The other advantage of light chains is that they have a very short half-life. So provided patients have recovered form their immediate chemotherapy and have recovered their cell counts because obviously when they drop their cell counts they drop their production of light chains, you’ll see a light chain response within days. Whereas whole paraproteins are cleared by FC receptors and their clearance in fact slows during chemotherapy when there’s relatively fewer FC receptors and they have a half life of several weeks. So you get a much more rapid assessment of what’s going on in terms of chemotherapy responses.

This applies to AL-amyloidosis and light chain deposition disease but the question is do all patients need to achieve a complete response? You might say why don’t you want to achieve a complete response and the answer is that obviously one does but some of these patients are very frail and each time you give them more intense chemotherapy you’re risking a very serious morbidity as well as mortality as well as precipitating end stage renal failure. So the risks of prolonged chemotherapy in these patients are high and that has to be measured against their benefit.
In some of these patients with the more subtle light chain deposition diseases there is evidence of improved organ function in the face of a partial response and that evidence of improved organ function is strongly associated with better survival.
We have the advantage in London of being able to track AL- amyloidosis by means of a specific nuclear medicine scan shown here. This is a patient with AL-amyloidosis deposited in their bones, their liver and their spleen before they had any treatment. They were pretty frail so they had some rather weak treatment in the form of combination therapy with cyclophosphamide, thalidomide and dexamethasone and had a partial response but 6 months later after their partial response they had amyloid regression I hope I can persuade you that that liver is now no darker than spleen showing that there’s been amyloid regression from the liver. They had a drop in their proteinuria of 7g with a preserved eGFR. So this patient a partial response was adequate to allow recovery of renal function and given that they were rather frail you wouldn’t want to push it harder than that. So in some cases you can use the light chains with a partial response combined with measures of organ function to titrate chemotherapy aiming for maximum patient benefit against minimal risk although this obviously needs to be done on an individual basis.

We find SAP scans invaluable to monitor this but this is going to be combined more generally with other more subtle markers in the future. I can’t show you a renal marker as yet but there’s now good data for combing biomarkers in the form of cardiac function with light chains in order to predict outcome in these very complex diseases. The biomarker that is used in cardiac function is atrial NT-pro-BMP. What has been shown in AL-amyloidosis in a collaboration between us and the Italian group based in Pavia led by Professor Merlini is that patients with a partial response to chemotherapy who have an improvement in a biomarker of cardiac function have an extremely good outcome which matches those patients with a complete response. This is very useful in predicting what to do in patients with high risk disease. Patients who have no response do very badly and almost 80% of them are dead at 6 months follow up. Patients with a partial response and no BNP response also do quite badly with 38% dead. Patients with a complete response do very well, only 6% are dead at 6 months follow up. Patients with a partial response whose BNP falls suggesting that their cardiac function has got better (because the light chains are directly cardiac toxic) have the same good survival. This is here shown on a Kaplan-Meier plot suggesting that as we develop better biomarkers for looking at organ function and predicting outcome there will be the potential to combine measuring sensitive assays of what’s going on with the clone in the form of free light chains as well as other clonal markers with sensitive markers of what’s going on in terms of organ function and individually titrate patients treatment against their needs rather than simply going for a protocol where you can often risk quite high morbidity in end stage renal failure.

So I think that’s probably the future. I’m just going to finish by covering something rather different. One of the things I didn’t really talk about was the use of light chains in diagnosing these diseases. The reason for that is that you have to remember that monoclonal gammopathies are not uncommon and are quite a common incidental finding in renal clinic. Particularly as the patients who comes to renal clinic become older. Thisis Robert Kyle’s data from his series of patients seen near the Mayo Clinic and this was published in 2003. What this shows is the incidence of monoclonal gammopathy in the general population who turned up for routine blood tests. You can see that the incidence rises. It’s commoner in men than in women but by the time you look at 80 year old men 8% of them will have an incidental MGUS. Many of them will be picked up in renal clinic and they well have nothing to do with what’s going on with their renal problems. So one question is can you use these markers to predict which of these patients you need to worry about? And which patients you can ignore and don’t need to be referred on to the haematologists?

Why would you worry about them? Well the reason that you worry about them is there is a risk of progression of monoclonal gammopathy and the progression here is defined anything you don’t want to happen including progression to myeloma, progression to Waldenstrom’s macroglobulinemia progression to AL-amyloidosis or light chain deposition disease. There is a genuine risk of progression shown here and if we continued this graph onto 50 years it would continue rising and the risks accruing gradually over time.

None the less the majority of these patients who are elderly when they present with an MGUS will die of something completely unrelated and for the majority of these patients this is an incidental finding which doesn’t need monitoring. So can you select out those patients in whom you’re going to worry about from those in whom this is an incidental finding which particularly in the very elderly can be dismissed? There have been a number of attempts to stratify these risks. This stratification from a publication 3 years ago is based on the size of the M protein so if the M protein is more than 15 g/l that’s one risk factor. If it’s a non IgG paraprotein that’s a second risk factor and if the light chains are abnormal that’s a third risk factor. What they demonstrated was that almost 40% of patients had no risk factors and their risk of progression over 20 years is extremely low and these patients can essentially be reassured and won’t need much follow up. Patients with two risk factors who made up 20% of this cohort had a significantly higher 20% risk of progression in 20 years and patients with 3 risk factors had a 58% risk factor of progression and clearly need to be monitored very carefully and cannot be reassured that this is a benign disease.

So to finish in summary the paraprotein related renal diseases all present usually with proteinuria and renal impairment in older patients. The underlying haematological lesion is often subtle and may not be detected until after biopsy. For the majority of patients treatment is supportive combined with chemotherapy and other measures which will be covered later aimed at the underlying clonal disease. The use of the serum FLC assays is extremely helpful in diagnosis, monitoring treatment and in minimizing unnecessary treatment in patients who’ve already responded and recognizing ineffective therapies in patients who haven’t responded and need a change of treatment. Its use needs to be interpreted carefully in patients with changing renal function.
In the future combining FLC assays and other clonal markers with biomarkers may be the route by which more sophisticated tailoring of treatment is developed and in patients seen in renal clinic with incidental MGUS the use of light chain assays is in combination with other clonal markers can be used to select patients at high risk of progression and those who can be reassured that this is not going to be a major issue.

Chairman: Thank you very much Helen for this very clear presentation. We have time for one or two questions. Yes?

Question: Very clear presentation thank you. In patients presenting with unexplained kidney disease there’s quite good evidence that screening for myeloma is of no benefit. Would measurement of FLCs in unselected patients presenting to the renal clinic with CKD help to detect the people that need to be referred to the haematologist in the first place?

Dr. Lachmann: A small series from our clinic suggest that you’re going to be doing a lot of screening to pick up relatively few patients. Our biochemistry department certainly does not wish to support the cost outside trials. Our practice has been to do electrophoresis and look for BJP and then only look for light chains if those are abnormal or if the risk is considered very high. The difficulty is that the absolute FLC numbers can be quite high if they’ve gone into end stage renal failure and unless you measure both kappa and lambda and are confident in your interpretation of results there is risk of misleading false positives. So I think the number of patients in whom you’ll pick up unequivocally myeloma level results are quite low but not none. So it’s a useful test if your labs will provide it -- but I wouldn’t do it routinely.