Abnormal serum free light chain ratio predicts poor overall survival in mantle cell lymphoma

Authors


  • Presented in part as abstract (poster presentation) at: (a) British Society for Haematology 52nd annual scientific meeting (Glasgow, April 2012), (b) 11th International Conference on Malignant Lymphoma (Lugano, June 2011).

Correspondence: Dr Michelle Furtado, Department of Haematology, Combined Laboratories, Level 7, Derriford Hospital, Plymouth, Devon, PL6 8DH, UK.

E-mail: michelle.furtado@nhs.net

Summary

Serum free light Chain (sFLC) ratios have been correlated with survival outcomes in Hodgkin and non-Hodgkin lymphoma subtypes. This study was undertaken to investigate the prognostic impact of abnormal sFLC ratios in mantle cell lymphoma (MCL). two patient cohorts were analysed for sFLC parameters: a preliminary retrospective cohort and a uniformly treated cohort of 20 relapsed/refractory MCL patients, enrolled in a phase II clinical trial of single agent lenalidomide treatment. 52% of patients had an abnormality of one or more sFLC parameter (71% of the first cohort and 40% of the second cohort). In cohort two, a high baseline SFLC ratio correlated with poorer overall survival (OS) compared to a low/normal ratio (median OS: 1·4 months vs. 19 months respectively; P = 0·001). For patients presenting with an elevated sFLC ratio at trial entry a rise of >35% in the sFLC ratio correlated with disease progression and a sFLC ratio of >2× normal at trial entry correlated with aggressive disease. These data are the first to show a clear clinical correlation between sFLC ratios and survival outcomes in a uniformly treated cohort of MCL patients. We suggest that these markers may be useful in managing patients with MCL in the future.

The presence of a monoclonal immunoglobulin protein (M-protein) in the serum or urine of a patient is evidence of a monotypic B cell population, which may be of bone marrow or nodal origin (Pérez-Galán et al, 2011). M-proteins have been detected historically by serum protein electrophoresis (SPE) and isotyped by immunofixation electrophoresis (IFE) (Keren, 1999). Due to a lack of sensitivity, this screening algorithm may miss those diseases that produce only light chains and small M-proteins. Traditionally, assessment of Bence Jones protein (BJP) in the patient's urine was the only way to determine whether light chains were present – an insensitive method that may be affected by renal function and blood volume. Over the last decade, newly developed polyclonal immunoassays detecting free immunoglobulin light chains (Bradwell et al, 2001) in the serum have been validated (Kang et al, 2005) and are now routinely used in the assessment of plasma cell dyscrasias in the UK (Bird et al, 2009, 2010). As recommended by International Myeloma Working Group Guidelines (Dispenzieri et al, 2009), this assay provides diagnostic and monitoring information. Moreover, the predictive value of baseline serum FLCs for progression or prognosis extends to most monoclonal gammopathies (Dingli et al, 2006; Dispenzieri et al, 2006, 2008; Kyrtsonis et al, 2007; Itzykson et al, 2008).

The identification of M-protein production characteristic of B cell dyscrasia has routinely relied on serum and urine protein electrophoretic techniques. Whilst these are adequate for the detection of gross production, they may be insensitive in the detection of subtle M-protein production.

The serum free light chain (sFLC) assay (Freelite, The Binding Site, Edgbaston, Birmingham, UK) was developed using sheep polyclonal antisera as reagents. These antisera are specific for epitopes on free κ and λ light chains but not those bound to heavy chains as part of an intact immunoglobulin (Bradwell et al, 2001). Alongside the absolute levels of the individual κ and λ light chains, the ratio between the two (κ/λ) is used to assess monoclonality and monitor disease. Indeed the presence of an abnormal serum free light chain κ/λ ratio at diagnosis is a poor prognostic factor for plasma cell dyscrasias (Drayson et al, 2006; Kyrtsonis et al, 2007). The Freelite assay is a simple, quick automated test that provides higher sensitivity than IFE (detection level 3–4 mg/L compared to 100–150 mg/L for IFE) (Bradwell et al, 2001).

Role of sFLCs in plasma cell disorders and lymphomas (CLL/SLL, DLBCL, HL)

The increasing use of sFLC assays has allowed detection of immunoglobulin abnormalities in plasma cell disorders previously classified as non-secretory (Drayson et al, 2001) as well as other lymphoproliferative conditions, such as subtypes of Hodgkin lymphoma (HL), non-Hodgkin lymphoma (NHL) and chronic lymphocytic leukaemia (CLL) (Martin et al, 2007; Pratt et al, 2009; Pinto et al, 2010; Maurer et al, 2011a).

Martin et al (2007) noted 35% NHL/CLL patients had an M-protein detectable only by sFLC assay. Of the NHL subtypes, mantle cell lymphoma (MCL) and small lymphocytic lymphoma had the highest incidence of detectable sFLC (36% and 24%, respectively). Other groups have reported abnormal sFLC ratios in MCL, varying in incidence from 36 to 77% (Filippi et al, 2008; Pinto et al, 2008).

An abnormal sFLC ratio has been associated with a poorer prognosis in CLL (Pratt et al, 2009, 2011) and diffuse large B cell lymphoma (DLBCL) (Maurer et al, 2011a).

Mantle cell lymphoma

MCL is a B cell lymphoma comprising approximately 6% all B cell NHL (Swerdlow et al, 2008). It is considered one of the more aggressive lymphoid tumours. Despite general responsiveness to chemotherapy it characteristically has a poor overall survival (OS) due to frequent relapses and short remissions. The natural history of the disease results in patients receiving multiple lines of chemotherapy and being monitored closely for signs of relapsing disease. As well as full blood counts, monitoring traditionally includes radiological imaging and repeat bone marrow biopsies. Given the previous reports of abnormal sFLC ratios in MCL (Martin et al, 2007; Filippi et al, 2008; Pinto et al, 2008) we investigated sFLC in relapsed, refractory MCL patients to determine whether the routine addition of this non-invasive test would provide clinically valuable information.

By assessing serial sFLC data in a cohort of patients in a clinical trial, this is the first study to investigate the correlation of sFLC's with clinical behaviour.

Methods

34 patients with MCL treated in Plymouth were analysed. This comprised of two cohorts.

Preliminary cohort

With the emerging evidence of the presence of abnormal sFLC ratios in lymphoma subtypes, we retrospectively analysed frozen sera from patients on our MCL database to determine whether abnormal sFLC ratios could be detected at any time point during each patients disease and whether the sFLC ratios reflected clinical disease behaviour. All patients on the database were reviewed and were considered eligible for inclusion in this cohort if sequential sera (frozen at −80°C) were available. 14 patients were identified who fulfilled this criteria. Samples had been stored from consenting patients over a 5-year period (2002–2007) on an ‘ad hoc’ basis, corresponding with the development of significant clinical changes. All the samples were analysed using the serum Freelite assay. These initial 14 patients had received a wide range of chemotherapeutic regimens (Table 1). Our initial investigation was designed to assess whether there would be merit in collecting sFLC information from a more defined cohort of patients within a clinical trial.

Table 1. First cohort: disease status at the time of initial serum sample
PatientDisease statusaFLC κ (mg/L)FLC λ (mg/L)sFLC Ratio κ/λFLC (κ+λ)
  1. FLC, free light chain; sFLC, serum free light chain.

  2. a

    at time of first serum sample.

1Diagnosis. Pre-treatment21·0013·201·5934·20
2Relapsed. On 7th line treatment19·2030·300·6349·50
3Relapsed. Post-second line treatment19·9011·301·7631·20
4Relapsed. On 5th line therapy16·50144·000·11160·50
5Diagnosis. Pre-treatment109·0011·009·91120·00
6Pre-treatment (watch & wait)41·3043·800·9485·10
7On first line treatment44·80104·000·43148·80
8Diagnosis. Pre treatment24·0030·400·7954·40
9Diagnosis. Pre treatment47·5034·101·3981·60
10Relapsed. On second line treatment14·2025·800·5540·00
11On first line treatment11·6011·900·9723·50
12Relapsed. Pre-second line treatment3·75>7·2Unable to calculate>10·95
13Diagnosis. Pre-treatment7·9241·100·1949·02
14On first line treatment12·50115·000·11127·50

Second (clinical trial) cohort

Serum was collected in a uniformly treated cohort of 20 relapsed/refractory MCL patients enrolled in a phase II clinical trial of single agent (oral) lenalidomide (European Union Drug Regulating Authorities Clinical Trials No: 2007-005472-13) (Table 2). The trial drug administration was 25 mg lenalidomide for 6 months, then 15 mg lenalidomide until relapse or disease progression. Patients were fully consented for the storage and assessment of serum and ethical approval obtained.

Table 2. Second cohort (lenalidomide trial)
PatientAge at trial entry (years)Disease stageECOGLDH (iu/L)FLC κ (mg/L)FLC λ (mg/L)sFLC Ratio κ/λFLC (κ+λ)
  1. ECOG, Eastern Cooperative Oncology Group performance score; LDH, lactate dehydrogenase; FLC, free light chain; sFLC, serum free light chain.

PLY0015841784298·343·4837·34
PLY0024842 1·3932·20·0433·59
PLY003554177312·914·80·8727·7
PLY0046730 31·311·82·6543·1
PLY0056931 14·520·80·7035·3
PLY0074741 33·914·42·3548·3
PLY008644237824·119·51·2443·6
PLY00964412857·4311·20·6618·63
PLY0103940136910·770·20·1580·9
PLY0126240 13·711·11·2324·8
PLY013543042015·113·91·0929
PLY01476416606·496·11·0612·59
PLY015584031017·224·30·7141·5
PLY01672405017·2318·40·3925·63
PLY01852 0 12·516·90·7429·4
PLY019664028443291·53215·03330·53
PLY0216340 74·72·8326·4077·53
PLY02257428159·26170·5426·26
PLY0236840 7·577·191·0514·76
PLY0246541 1818·20·9936·2

Serum was frozen (−80°C) from all trial patients at set time points (pre-treatment, monthly to 6 months, 3 monthly thereafter) until disease progression, relapse or death.

Free light chain (Freelite) assay

All samples were analysed with the Freelite assay as previously described (Bradwell et al, 2001) at The Binding Site Laboratories, Birmingham, UK using the SPAplus analyser. Absolute sFLC κ and λ levels were recorded and sFLC κ/λ ratios calculated. As in previous studies (Martin et al, 2007; Pratt et al, 2009; Landgren et al, 2010; Yegin et al, 2010), abnormal absolute κ and λ levels and κ/λ ratios were defined according to published normal ranges (Katzmann et al, 2002). Thus a κ free light chain concentration above 19·4 mg/L, a λ free light chain concentration above 26·3 mg/L and a sFLC κ/λ ratio outside 0·26–1·65 was defined as abnormal.

Statistics

Survival data was analysed by Kaplan Meier, with log rank testing, using the statistical package for the social sciences (spss) Version 18.0. Statistical significance was defined as a P-value <0·05.

Results

Of the 34 patients investigated 52% (18/34) had an abnormality of one or more sFLC parameter.

First pilot study

In total, 10/14 (71%) MCL patients in the first cohort analysed had abnormal sFLC parameters that reflected disease behaviour–the sFLC ratio worsened with disease activity and improved when in remission. There were eight male and two female patients. Six patients had a concomitant ‘monoclonal’ sFLC abnormality (abnormal κ and/or λ plus an abnormal sFLC ratio) and four patients had a ‘polyclonal’ sFLC elevation (elevated absolute κ value with a normal sFLC ratio). In all of these 10 patients, the serum light chain levels and/or sFLC ratio correlated with disease behaviour during therapy (as described above), irrespective of the nature of that therapy.

For the majority of patients the association between sFLC parameters and disease behaviour was most marked with sFLC ratios, although in a subgroup of patients with normal sFLC ratios the absolute κ and λ levels reflected disease behaviour. In these patients an absolute κ and λ level >45 mg/L was associated with clinical deterioration. Figure 1 illustrates the findings in a representative patient.

Figure 1.

correlation of absolute κ+λ with disease behaviour in a representative patient from the first cohort. Dashed line: absolute κ+λ total, solid line: serum free light chain (sFLC) ratio (κ/λ).

Second cohort (clinical trial patients)

In the second cohort, 8 of the 20 patients (40% total) had a Freelite abnormality at trial entry. 7 had a ‘monoclonal’ sFLC abnormality (Fig 2). 1 further patient had a ‘polyclonal’ sFLC abnormality. No patients had ‘ratio-only abnormalities’.

Figure 2.

Patient data: κ vs. λ. Solid lines indicate upper and lower boundaries of normal range. Grey markers: individual patient data. FLC, free light chain.

Light chains and survival differences

A high sFLC ratio at the time of relapse and trial entry was significantly associated with a worse OS compared to a normal or low sFLC ratio (median OS: 1·4months vs. 19 months respectively; P = 0·0019 (Hazard Ratio [HR] 13·6, 95% confidence interval [CI]: 4·45–41·5); (Fig 3). There were not enough patients with a low sFLC ratio in this cohort to allow separate statistical analysis of this group.

Figure 3.

Overall Survival stratified by serum free light chain (sFLC) ratio.

The difference in survival remained significant if all patients with an abnormal sFLC ratio were pooled together (P = 0·001, HR 15·4, 95% CI: 5·2–45·9) (Fig 4) This echoes the results of previous investigators, which showed that an abnormal sFLC ratio (low or high) affects prognosis in other haematological disorders (Dispenzieri et al, 2008; Maurer et al, 2011a).

Figure 4.

Overall survival by normal/abnormal serum free light chain (sFLC) ratio.

Total κ+λ levels at the time of trial entry did not correlate with any significant differences in OS or progression-free survival (data not shown).

Monoclonal versus polyclonal abnormalities

Monoclonal abnormalities (Abnormal sFLC ratio)

Five of the seven patients with an abnormal sFLC ratio at trial entry had serial samples analysed. The normalization of sFLC ratios correlated with disease improvement (reduction in lymph node size and/or resolution of ‘B’ symptoms). Two patients with abnormal sFLC ratios at trial entry but no serial samples analysed were withdrawn from trial because of disease progression.

A sFLC ratio above 2× the upper limit of normal (i.e. sFLC κ/λ >=3·3) at the time of trial entry correlated with aggressive disease leading to significant disease progression or death within 2 months.

For patients with an elevated sFLC ratio at trial entry a rise in the sFLC ratio of >35% (using the presentation ratio as 0%) correlated with disease progression. This did not apply to patients with a low sFLC ratio at presentation as a rise in ratio acted to normalize the ratio and, as with the patients with elevated sFLC ratio at presentation, normalization of the ratio still correlated with disease improvement.

Polyclonal abnormalities (Normal sFLC ratio but polyclonal rises in immunoglobulins)

Of the patients with a normal sFLC ratio at the time of presentation, none subsequently developed abnormal sFLC ratios, although 4 did develop a polyclonal elevation of either the absolute κ or λ value. Polyclonal rises in immunoglobulins can be seen in many conditions including hepatic disease, chronic infections and rheumatological conditions; however, in our cohort of patients the absolute levels of the elevated light chains (κ+λ) did correlate with disease behaviour, despite the overall normal sFLC ratio (Figure S1, Tables SI, SII). This suggests that the polyclonal immunoglobulin was of tumour cell origin.

Discussion

MCL typically relapses multiple times, with increasingly chemo-refractory disease. A non-invasive method of assessing disease response to treatment, without recourse to computed tomography scanning/bone marrow biopsies, would be reassuring to both patients and clinicians.

It has been previously shown that a proportion of patients with MCL have a monoclonal immunoglobulin abnormality that is detectable using a sFLC ratio, although the proportion percentage has varied widely (Martin et al, 2007; Pinto et al, 2008; Maurer et al, 2011a). It is feasible that as a clonal B cell disorder, monoclonal immunoglobulin may be secreted by the tumour and therefore would reflect tumour bulk. Our hypothesis was that for patients with a detectable sFLC ratio, this could be used as an adjunct to non-invasively monitor tumour behaviour.

Overall, 52% of patients tested had an abnormality of their sFLC ratio. After the initial finding of a high proportion of abnormalities in the first cohort (71%), we then followed this sequentially in a uniformly treated patient population and found that 40% of patients tested had an abnormal sFLC ratio at the time of entry into a trial of single agent oral lenalidomide. This proportion is comparable to that found in previous studies (Martin et al, 2007; Maurer et al, 2011a). The patients investigated in this study all had relapsed refractory disease and had previously received a variety of chemotherapy regimens. All had active disease necessitating treatment at the time of serum collection.

This is the first study to analyse the correlation between sFLC ratios and MCL disease behaviour in a uniformly treated population of patients within a clinical trial. Although the number of patients is small, the results show a striking correlation between the sFLC ratio and MCL disease behaviour in the proportion of patients with an abnormal sFLC ratio at the time of presentation.

The sFLC ratio appears to mirror disease more closely if it is elevated at the time of presentation than if it is low. A low sFLC ratio indicates a λ monoclonal immunoglobulin elevation and in plasma cell dyscrasias this situation is more unusual than a κ monoclonal immunoglobulin elevation–the same may be true in MCL. This may be the reason that our cohort of patients did not include many patients with low sFLC ratios at presentation. It would be of interest to determine the incidence of λ monoclonal immunoglobulin elevation in MCL within the context of a prospectively collected trial and also to determine whether an initially low sFLC ratio would fall further with clinical deterioration – a situation that did not arise within our study.

Only 40% of the patients enrolled in the second (clinical trial) cohort had an abnormal sFLC ratio at presentation despite the fact that at the time of initial sampling all the patients had active disease requiring treatment. In these patients the abnormal sFLC ratio may reflect either the tumour bulk or the immune system trying to respond to disease. The two long-term responders (complete remission maintained for over 2 years on lenalidomide therapy) had normal ratios of sFLC.

As with CLL (Maurer et al, 2011b), the presence of a sFLC abnormality may, in the future, correlate with biological subtypes of MCL. Collecting Freelite data on all newly presenting MCL patients will help determine whether these assays can help detect disease progression in patients with initially indolent disease as well as those with classical MCL. The significant difference in OS seen between the patients with a normal and abnormal sFLC suggests that the sFLC ratio may have a role in the future as potential marker of aggressive disease. It is interesting to note that in diffuse large B cell lymphoma the absolute κ and λ levels were more predictive of outcome than the sFLC ratio (Maurer et al, 2011a) – the potential use of the sFLC ratio in disease monitoring may be applicable to certain NHL subtypes only.

Our results indicate that an elevated sFLC ratio of 2× the upper limit of normal at the time of presentation correlated with markedly aggressive disease and also that a rapid rise in the ratio of 35% or more indicated significant disease progression. Both these findings require validation in future studies, however they may potentially be useful non-invasive adjuncts to the assessment of patients.

The MCL international Prognostic Index (MIPI) score has only been validated for use at the time of initial diagnosis of MCL. Therefore, the details of its component parameters was not collected and the MIPI was not calculated for the patients included in this study as they all had relapsed disease. We therefore are unable to compare whether prognosis indicated by sFLC status correlates with prognosis from MIPI score or its components. Due to the small sample size we were unable to perform formal multivariate analysis on these results.

In clinical practice, it is common to evaluate patients on treatment with a combination of radiological imaging and bone marrow examination together with detection of new/increasing lymphadenopathy to detect relapses in disease. The finding that changing sFLC ratios may indicate relapsing disease before it is overtly clinically evident may allow investigations to be performed in a more timely manner and therapy instigated/changed before patients become unwell with disease.

There remain many questions regarding the role of the sFLC in MCL, however at present for those patients with a sFLC abnormality this has the potential (if validated in further studies) to become a non-invasive marker of disease activity in MCL.

Author contributions

MF performed the research, analysed the data and wrote the paper, SR designed the research study and critically reviewed the paper, NS performed the research, AL and SH contributed essential reagents, performed the research and critically reviewed the paper.

Competing interests

AL and SH are employed by The Binding Site Ltd. No other competing interests to declare.

Ancillary