When cold complicates: the impact of cryoglobulins and pre-analytical conditions on monoclonal protein analysis

Abstract

Cryoprecipitation of monoclonal protein in two patients from separate hospitals led to falsely reduced concentrations. In one case, this was only later discovered because the hematology analyzer falsely flagged the cryoglobulins as abnormal platelets and lymphocytosis. This resulted in the generation of a peripheral blood smear in which the cryoprecipitation was microscopically identified. In both cases, initially unnoticed discrepancies existed between the total immunoglobulin and albumin concentrations derived from the general chemistry analyzer and those from serum protein electrophoresis analysis, which is what later alerted staff to the second case. These cases highlight the significant risk cryoglobulins pose in the diagnosis and accurate follow-up of plasma cell dyscrasias. We therefore propose several measures to improve laboratory detection and prevention of such discrepancies. These include; performing total immunoglobulin and albumin measurements alongside the protein electrophoresis analysis; implementation of a cross reference alarm in the laboratory information system for the discrepancies; consequent visual inspection of all samples prior to analysis; pre-warming of samples with a visible cryoprecipitate prior to serum protein electrophoresis; and the implementation of adequate operating procedures for handling samples from patients with known cryoglobulins to prevent pre-analytical loss caused by precipitation. Finally, we provide insight into the stability of immunoglobulins and monoclonal proteins at room temperature, which may be helpful to determine the optimal sample storage temperature in order to reduce the risk of unwanted cryoprecipitation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10238-026-02088-5.

Introduction

Cryoglobulinemia occurs when one or more subclasses of immunoglobulins precipitate at temperatures below 37 °C. These precipitates called cryoglobulins can be classified according to the Brouet criteria into three subtypes: Type I refers specifically to monoclonal immunoglobulins, type II to a combination of monoclonal and polyclonal immunoglobulins, and type III exclusively to polyclonal immunoglobulins [1–4]. There are a number of underlying conditions associated with each type. Hematological malignancies such as Multiple Myeloma, Monoclonal Gammopathy of Undetermined Significance (MGUS) and Lymphoplasmacytic Lymphoma (LPL) are frequently associated with type I, whereas type II and III are more often associated with chronic infections such as hepatitis B, hepatitis C and HIV, as well as autoimmune diseases [1]. While many patients with cryoglobulinemia are asymptomatic, there are multiple possible clinical manifestations with significant impact for patients. Type I is for example associated with hyperviscosity syndrome or Raynaud’s syndrome, whereas type II and III are more associated with vasculitis and glomerulonephritis [5–7]. Cryoglobulins have been extensively reviewed elsewhere [8–11].

Cryoglobulins have been reported to interfere with numerous laboratory tests such as the complete blood count and chemistry tests [12–21]. While some automated hematology analyzers are equipped with flags for the presence of cold agglutinins, there are currently no direct triggers for the presence of cryoglobulins, and immunoglobulin precipitation may also only occur after sample refrigeration. Laboratory tests which are infrequently performed and where samples are stored at 4 °C prior to analysis, are therefore particularly susceptible to unanticipated cryoprecipitation. Samples from patients with known cryoglobulinemia are therefore drawn in prewarmed tubes and are generally placed at 37 °C directly following phlebotomy.

Monoclonal proteins (M-proteins) are immunoglobulins produced by a monoclonal population of plasma cells. Being associated with both severe overt disease and subclinical abnormalities, quantifying the M-protein concentration and its immunoglobulin class is, alongside bone marrow analysis and other clinical features, important for disease classification as well as treatment and prognosis determination [22]. In this report we present two cases in which an M-protein was present but whereby the concentration was initially underestimated due to its cryogenic nature. We further present data demonstrating M-protein stability over a longer period at room temperature as a mitigation strategy for unexpected cryoprecipitation and suggest further preventative measures for such cases.

Case 1

An 83-year old man presented in 2023 to the internal medicine physician with lower back pain, a raised erythrocyte sedimentation rate (ESR), mild thrombocytopenia, normal CRP and a marginally abnormal glomerular filtration rate (GFR) (Table 1). Serum protein electrophoresis (SPE) and immunosubtraction showed an IgG-kappa M-protein concentration of 6 g/L, while the total IgG concentration was 36 g/L (Table 1). The albumin concentration as determined from the SPE analysis was 54.9 g/L while with the general chemistry analyzer this was 39 g/L. While the discrepancy between the M-protein concentration and the total IgG was not initially identified by the lab, it was noticed by the clinical team, who considered the raised total IgG to be polyclonal.

Table 1.

Laboratory results for case 1

Parameter	October 2023	9 November 2023	20 November 2023	July 2024	Reference Range
ESR	117	114	117	120	1–20 mm/hr
Hemoglobin	9.4	9.1	8.9	9.1	8.4–10.8 mmol/L
Leukocytes	4	4.9	5	4.3	4–11 × 10 9 /L
Thrombocytes	109	114	114	95	150–400 × 10 9 /L
MCHC	20.4	20.6	20.1	21.1	19–23 mmol/L
Microscopic Morphology	n.d.	Cryoglobulin like substrate	n.d.	n.d.	-
CRP	< 4	n.d.	< 4	< 4	< 10 mg/L
GFR	59	n.d.	57	61	> 90 ml/min/1.73m2
Total Protein	94	n.d.	95	97	57–82 g/L
Albumin (chemistry analyzer)	39	n.d.	n.d.	37	34–50 g/L
Albumin (SPE)	54.9	n.d.	42.1	42	31–47 g/L
Alfa-1-globulin	3.4	n.d.	3.0	9.2	2.0–6.0 g/L
Alfa-2-globulin	13.0	n.d.	9.5	7.6	5.0–11.00 g/L
M-protein type	IgG Kappa	n.d.	IgG Kappa	IgG Kappa	-
M-Protein conc.	6	n.d.	26	28	Not detectable g/L
IgG Chemistry analyzer	36.7	n.d.	n.d.	12.1	7–16 g/L
Cryoglobulin	n.d.	n.d.	Type I positive	Positive (known)	Not detectable

ESR, eryrtocyte sedimentation rate; MCHC, mean cell hemoglobin concentration; CRP, c−reactive protein; GFR, glomerular filtration rate; n.d., not determined

In an automated complete blood count performed a month later (Nov 2023), the hematology analyzer produced flags for an abnormal platelet distribution, giant platelets and lymphocytosis, despite a normal lymphocyte count. While hemocytometry had previously been performed, no flags were generated at that time. The flags resulted in a blood smear in which a background precipitation was seen which was brought to the attention of the laboratory specialist, who found it to be consistent with cryoglobulins (Fig. 1a-b). Reviewed analysis of the case revealed the chemistry discrepancies from October 2023 (Table 1) and cryoglobulin testing with repeat M-protein analysis was recommended to the hematologist. This revealed the presence of a type I cryoglobulin (IgG kappa) (Table 1) and prewarming of the sample provided a significantly higher M-protein peak compared to unwarmed electrophoresis results (Fig. 1b-c), with an M-protein concentration of 26 g/L (Table 1). There was no evidence for a mixed cryoglobulinemia and the cryoglobulin and the monoclonal IgG kappa are considered the same protein. While the total IgG was not tested on this occasion, the concentration in October was 36.7 g/L. This 10.7 g/L difference is considered likely to be due to an overestimation by the immunoturbidimetric method used, a known phenomenon for this analytical method [23, 24]. Therefore, the warmed M-protein and total IgG concentrations are considered consistent with each other. Interestingly, all non-cryogenic fractions of the protein spectrum (such as albumin and alpha globulins) are falsely raised in the unwarmed sample due to underestimation of the M-protein. This occurs because all protein spectrum fractions are calculated relative to the total protein concentration which is a fast routine chemistry analysis and therefore still largely includes the M-protein.

Fig. 1.

(a) Blood smear images showing the presence of a background substrate surrounding the leukocytes and erythrocytes, which was identified to be cryoglobulin. (b) Serum protein electrophoresis profile from the unwarmed sample and (c) the spectrum from the pre-warmed sample

In July 2024, there was a reversed discrepancy between the total IgG and IgG M-protein concentrations. This time, the sample had likely already precipitated prior to routine chemistry analysis leading to a falsely reduced IgG concentration, whereas for the M-protein analysis the sample was pre-warmed (Table 1). The patient declined further investigation, however, based on an M-protein concentration of < 34 g/L and the absence of other clinical diagnostic criteria, MGUS was considered the most likely diagnosis [22]. In retrospect, the patient had experienced cold hands and feet for years but without worsening of symptoms during colder temperatures. Dermatological or joint problems were absent and there were no evident signs of hyperviscosity.

Case 2

A 65-year old woman with a history of breast carcinoma, presented in 2021 with tiredness, persistent fever, an elevated ESR and anemia (Table 2). SPE revealed the presence of an IgM M-protein of 4 g/L for which a bone marrow puncture was performed. With an abnormal B cell population present in the bone marrow, a diagnosis was made of IgM-type lymphoplasmacytic lymphoma, also called Waldenström’s macroglobulinemia (WM) [22]. Considering the patient had no symptoms, a wait-and-see policy was adopted conform national guidelines [25].

Table 2.

Laboratory results for case 2

Parameter	April 2021	January 2022	October 2023	June 2024	Reference Range
ESR	79	n.d.	n.d.	n.d.	1–30 mm/hr
Hemoglobin	6.7	6.3	n.d.	6.6	7.5–10.0 mmol/L
Leukocytes	9.9	11.3	n.d.	12.3	4–10 × 10^9/L
Thrombocytes	n.d.	395	n.d.	427	150–400 × 10^9/L
GFR	> 90	81	n.d.	82	> 90 ml/min/1.73m2
Total Protein	79	87	n.d.	n.d.	60–80 g/L
M-protein type	IgM lambda	IgM lambda	n.d.		-
M-Protein conc.	4	5	n.d.	n.d.	Not detectable (g/L)
IgM chemistry analyzer	n.d.	38.1	n.d.	70.7	0.4–2.3 g/L
Cryoglobulin	n.d.	n.d.	Positive; Type I (IgM lambda)	positive	Not detectable

n.d, not determined

From 2022 onward the patient was monitored with blood tests every 3 to 4 months, which included a complete blood count, total IgM and M-protein analyses. There was a discrepancy between the total IgM concentration from the general chemistry analyzer (38 g/L) and the SPE M-protein concentration (IgM-lambda 5 g/L), which was too large to be explained by method differences (Table 2) [23, 24]. This discrepancy was not noted immediately by the clinical team or the laboratory, but caught attention at the end of 2022. In search for an explanation of this difference, several pre-analytical aspects were investigated during the next months. In January 2023, all samples were first stored at 4 °C prior to analysis. The total IgM and IgM M-protein concentrations from these samples were low and quite comparable (5.7 and 4 g/L respectively), which considering the discrepancy with the previous results, alerted the laboratory staff to the potential presence of cryoglobulins. Cryoglobulin analysis indeed identified a type I IgM-lambda cryoglobulin (Table 2). In addition, SPE performed on pre-warmed serum showed an M-protein peak of 22 g/L, compared with the previously quantified 4–5 g/L in cold serum samples. Interestingly, no difference was observed in total IgM levels in serum that was kept at 37 °C after blood withdrawal (according to the cryoglobulin protocol) or at room temperature, suggesting that this particular cryoglobulin only precipitated when the sample was stored at 4 °C and not at room temperature. From this point onward, the patient was monitored by measuring total IgM concentration. Due to worsening clinical and laboratory abnormalities, therapy was initiated in 2024. As in case 1, there were no evident signs of hyperviscosity.

Methods

General chemistry, hemocytometry, M-protein and cryoglobulin analysis

For case 1, general chemistry analysis was performed with the Atellica CH analyzer (Siemens Healthineers), hemocytometry with the XN-10 analyzer (Sysmex) and SPE with the Capillarys 3 OCTA (Sebia Capillarys). For case 2, samples were measured with the CobasPro system (Roche), XN-3000 analyzer (Sysmex) and SAS-1 and SAS-2 analyzers (Helena biosciences). The serum from both patients was removed from the corpuscular component and stored at 4 °C for up to a week prior to SPE.

Based on the in-house protocols for cryoglobulin testing, the analysis for case 1 was visual inspection of precipitation following 4 °C incubation and immunosubtraction of the serum at 37 °C and the supernatant from the 4 °C sample. Redissolving of the precipitate and returning the sample to 37 °C is not part of the routine protocol and is only performed when dubious results are obtained. Since this was not the case, the precipitate was not dissolved for further analysis and immunofixation was also not performed. Cryoglobulin analysis of case 2 included visual inspection of the cryoprecipitation following four days incubation at 4 °C, washing the cryoprecipitate, confirming that the cryoprecipitate dissolved again at 37 °C, and finally immunofixation electrophoresis (Sebia Hydrasys-2).

Immunoglobulin and M-protein stability at room temperature

10 samples in which IgG, IgM and IgA had been measured within the previous 24 h, were selected and removed from storage at 4 °C and allowed to acclimatize to room temperature for 2 h. The concentration of all 3 immunoglobulins was again determined and samples were left at room temperature, with repeat analysis on day 7 and 14. Analysis was performed with the Atellica CH analyzer (Siemens Healthineers).

Similarly, 9 samples in which M-protein analysis had been performed were selected, with the following isotypes; IgG (5x), IgM (3x) and IgA (1x). These samples had been collected over the previous 7 days and stored at 4 °C prior to analysis. At 1 day after analysis, the samples were removed from cold storage, sealed with parafilm and left at room temperature, with repeated analysis performed 7 and 14 days later using the Capillarys 3 OCTA.

Statistical analysis

To determine if immunoglobulin concentrations significantly changed with storage at room temperature, the percentage bias was determined for each immunoglobulin class compared to the initial concentration following blood collection. The bias was then compared to the reference change value (RCV) derived from the EFLM biological variance database [26]. The bias was additionally compared to the assay precision (intra-lab SD) as derived from our 2024 end of year report from the Dutch SKML external quality assurance program. These precision limits were also used to calculate the RCV.

Results

Considering both of the above cases, we questioned the feasibility of storing samples for M-protein analysis at room temperature instead of 4 °C. We first tested the stability of total immunoglobins IgG, IgM and IgA over a 14-day period with measurements at 2 h, 7 days and 14 days at room temperature. At 2 h, the bias was less than 2% for all three immunoglobulin types and within the precision limits of the assay (2.5%) (Fig. 2a-b). For IgG, the average bias of all 10 samples was 4.9% and 7.0% at days 7 and 14 respectively (Fig. 2a). IgA had a small positive bias of 2.8% and 3.3%. IgM showed in contrast a negative bias, with an average of −1.9% and − 10.6% at days 7 and 14 respectively (Fig. 2a). The maximum negative bias for IgM was − 18.0% for sample 2 at day 14 (Fig. 2a). For 8 of the samples, the negative RCV for IgM of −14.6% was not exceeded. The negative RCV for IgM was not exceeded at day 7.

Fig. 2.

(a) IgG, IgA and IgM concentration variances over time with the RCV and precision presented as grey and green dotted lines, respectively. (b) M-protein concentrations at initial measurement following 0-, 7- or 14-days storage at room temperature. Samples were stored up to 7 days at 4 °C prior to being placed at room temperature

Taking into consideration that M-proteins might behave differently compared to polyclonal immunoglobulins, we also tested the stability of these proteins at room temperature. For this, SPE analysis was repeated in samples following 7- and 14-day storage at room temperature. At day 7, the β2 peak was reduced and by day 14 completely gone. In contrast, the M-protein peaks remained visibly unchanged and could be identified correctly, with negligible differences in M-protein concentration. The average bias for all 9 samples on day 7 was − 5.8% and on day 14 − 6.0% (Fig. 2b). In the sample with the largest concentration difference, the concentration at day 14 was 59.1 g/L compared to an original concentration of 57.2 g/L. In contrast to the polyclonal immunoglobulins, the direction of the bias by the M-proteins could not be correlated to immunoglobulin class. 8 of the 9 samples showed a negative bias at day 7. One IgG M-protein showed a negative bias of −2.75% at day 7 and a positive bias of 0.55% at day 14, potentially due to evaporation. Another IgG M-protein showed a positive bias at both day 7 and 14.

Discussion

In the presented cases, the M-protein exhibited cryogenic features, leading to precipitation and under-quantification due to storage at 4 °C before analysis. This highlights a critical pre-analytical challenge in the accurate measurement of M-proteins, particularly considering analysis often does not occur on the same day as blood collection. Although in case 1 the diagnosis was not confirmed with bone marrow analysis, on the basis of the M-protein concentration, lack of anemia or significant renal dysfunction, and absence of other diagnostic criteria, a probable diagnosis of MGUS or smoldering multiple myeloma was made. While significant disease burden is absent in both entities, undetected underestimation of the M-protein could have impacted follow-up of disease progression. This also applies for case 2, where underestimation posed a risk of inaccurate follow-up of WM, a condition with potential for significant disease burden. Both cases emphasize the potential clinical impact of this issue, highlighting the need for increased vigilance in detecting potential cryoglobulin interference in M-protein quantification.

The inconsistency in hematology analyzer flagging between both cases was further notable. While case 1 triggered a flag on one occasion, case 2 did not, likely due to differences in pre-analytical processing time and the temperature sensitivity of the cryoprecipitate. This aligns with prior reports by Smit et al. and others, where abnormal scatter plots and visual cryoprecipitates were only evident with delayed testing [20, 27]. Even with delayed testing, cryoglobulins can present in different forms and sizes, making it difficult for the analyzer to detect all cases [27]. The ad-hoc nature of analyzer flagging due to heterogeneity in cryoglobulin characteristics and pre-analytical variables, limits their reliability as a screening tool for cryoglobulin interference.

Generally, it is recommended to store samples for immunoglobulin and M-protein analysis at 4 °C [28]. Sebia specifies a maximum sample storage time of 10 days for their SPE assays when samples are stored at 2–8 °C. Despite this, some assay manufacturers state an immunoglobulin stability of months at room temperature, with the exception of IgD and IgE [29]. With little available data over the true stability of M-proteins at room temperature, we set up experiments to test this alongside the stability of total immunoglobulins. Our data indeed indicate that immunoglobulins are stable at room temperature, with IgG and IgA not exceeding the RCV at day 14, and for IgM only 2 samples marginally exceeded the RCV at day 14. While negative biases were noted for M-protein concentrations at days 7 and 14, these were most significant in samples with low M-protein concentrations. While these data are a useful resource and help understand how M-protein stability differs to total immunoglobulins, laboratories need to consider their local procedures before implementing routine room temperature storage for M-protein analysis. For instance, altering storage temperature might alter samples logistics for other analytes determined from the same sample. Also, if an assay manufacturer specifies cold storage only in their instructions for use, altering this to room temperature may alter assay classification from routine to a laboratory developed test (LDT). Implementing an LDT usually requires more elaborate verification and quality control, depending on local laws and legislation. Nevertheless, despite these limitations, room temperature storage might be a promising mitigation strategy to prevent unwanted cryoprecipitation in some samples.

To mitigate the risk of falsely low M-protein concentrations, we primarily suggest the following strategies. Firstly, both SPE and total immunoglobulin levels should be included at both diagnosis and at each follow-up moment. Including an albumin measurement with the chemistry analyzer may also be considered. The laboratory should implement an algorithm that cross-references SPE results with the chemistry analyzer total immunoglobulin and the albumin results. This algorithm should generate an alarm for discrepant results. Laboratories, may consider prewarming of samples to 37 °C prior to analysis. Consequent visual inspection of all samples should be performed and labs should also establish clear operating procedures for handling samples from patients with known cryoglobulins, in order to minimize the risk of unintended cryoprecipitation. Ultimately, heightened awareness among clinicians and laboratory personnel is essential to ensure accurate interpretation and follow-up.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

AC, DR and MO prepared the manuscript and guided the experiments. DR, JS and MO were also involved in clarifying the discrepancies of both cases. DvA and MvUO carried out the experiments. AV and RV are the leading clinicians and also critically reviewed the manuscript. JRK and JJ provided helpful feedback on the experiments and critically reviewed the manuscript.

Funding

The authors have no funding sources to declare.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

The authors declare no competing interests.

Consent to publish

Written informed consent for publication was obtained from the patients.