This retrospective observational study enrolled one hundred and fourteen patients with newly diagnosed WM/LPL between May 2001 and March 2021 from Shandong Provincial Hospital with informed consent; no personal identifiable information was recorded in this study. Diagnostic criteria for LPL in all cases were defined as a tumor of small lymphocytes showing evidence of plasmacytoid/plasma cell differentiation without any of the clinical, morphological, or immunophenotypic features of other lymphoproliferative disorders (18). The presence of IgM monoclonal protein associated with ≥10% clonal lymphoplasmacytic cells in bone marrow confirms the diagnosis of WM (32). Baseline demographic and clinical data concerning sex, age, and the International Staging System Waldenström Macroglobulinemia (ISSWM) score were collected from medical records and hospital registries at diagnosis. Eight patients were excluded due to lack of clinical information and laboratory examinations and death with no relation to WM/LPL. All procedures involving human participants in this study were approved by the Research Ethics Committee of the Shandong Provincial Hospital.

In this study, the thyroid complication group was divided according to the clinical diagnosis, from the hospital discharge registry system and electronic medical records (33). There are 13 cases enrolled in this group, including 8 cases with hypothyroidism, 2 cases with Graves’ disease, 1 case with hyperthyroidism, 1 case with thyroid cancer, and 1 case with Hashimoto’s disease but without abnormal thyroid function index. We analyzed 106 WM/LPL patients and collected parameters at diagnosis, including age at diagnosis, hemoglobin level, platelet count, serum albumin, β2-microglobulin, serum lactate dehydrogenase levels, free triiodothyronine (FT3), free tetraiodothyronine (FT), thyroid-stimulating hormone (TSH), anti-thyroglobulin antibodies (TG-Ab), antithyroid peroxidase autoantibody (anti-TPO), and thyrotropin receptor antibody. Subsequently, the survival difference between patients with and without thyroid complications was analyzed in WM/LPL patients. Based on ISSWM, an extended prediction model was established. A subsequent comparison of ISSWM and novel established index was performed. Finally, the role of thyroid-related genes in WM/LPL patients was further investigated.

The outcome of this study was disease progression or death. Disease progression was defined as ≥25% increase in serum IgM level from lowest nadir and/or progression in clinical features attributable to the disease according to IWWM-7 consensus (34). Patients who died of accidents or other medical conditions with no relation to WM/LPL were excluded from this study. Progression-free survival (PFS) referred to the time from the beginning of treatment to disease progression or death for any cause. Overall survival (OS) was measured from the date of diagnosis to the date of death, which was censored at the date of the last follow-up visit.

The dataset (GSE6691) is from the GEO database (https://www.ncbi.nlm.nih.gov/geo/), and the download data format is MINIML. The Limma package (version: 3.40.2) of the R software was used to study the differential expression of mRNAs. The adjusted p-value was analyzed to correct for false-positive results in GEO datasets. To further confirm the underlying function of potential targets, the data were analyzed by functional enrichment. Gene Ontology (GO) is a widely used tool for annotating genes with functions, especially molecular function (MF), biological pathways (BP), and cellular components (CC). To better understand the carcinogenesis of mRNA, the ClusterProfiler package (version: 3.18.0) in R was employed to analyze the GO function of potential targets. The box plot is implemented by the R software package ggplot2; the heat map is displayed by the R software package pheatmap.

The Chi-square test and Fisher’s precise test were used for the comparison of characteristics. The Kaplan–Meier method was used for the analysis of survival outcomes and the log-rank test was used for comparisons. Cox regression hazard analysis was also performed for univariate and multivariate analysis of OS and PFS. Receiver operating characteristic (ROC) curves and the Delong test were employed for the comparison of two different prognostic scoring systems. p < 0.05 was considered significant. All analyses were performed using SPSS software version 26.0 and Medcalc software.

The study population consisted of 114 WM/LPL patients, with 8 patients excluded because of inaccurate diagnosis and incomplete clinical information. The median age of the remaining 106 patients was 63 years (range, 14 to 86) at diagnosis, and patients over 65 years of age accounted for 41.5% of all patients. The frequency of B-symptoms was 29.2%, and 21.7% of patients showed involved extramedullary sites. More than half of the patients (63%) had a hemoglobin level lower than 11.5 g/dl, while only 21.7% of patients had thrombocytopenia. Almost all patients had an elevated IgM level, which ranged from 0.27 g/L to 126 g/L (median, 23.85 g/L). Among all patients, the most common complication was hypertension (18%), followed by diabetes mellitus (15%) and thyroid function abnormalities (12.3%). The positive MYD88-L265P fusion gene accounted for 73.5% of WM/LPL patients. A rituximab (anti-CD20 antibody)-based combination regimen was the first choice in 41.7% of WM/LPL patients, followed by a Bortezomib (proteasome inhibitor) combination regimen in 10.7%. The characteristics of the study population are summarized in Table 1 .

Over a median of 38 months of follow-up, there was lost-to-follow-up in 38 cases (37%), death in 8 cases (7%), and progression developed in 47 cases (45%). The median OS of all patients was not reached ( Figure 1A ) and the median PFS was 96 months ( Figure 1B ). Furthermore, patients with no progression, including complete response (CR), partial response (PR), and stable disease, showed better OS ( Figure 1C ) and PFS ( Figure 1D ) than patients with progressive disease.

In our study, a high incidence rate of thyroid complications was observed in WM/LPL patients. Given the abnormal IgM level as the most characteristic feature identified in WM/LPL patients’ sera, we wonder whether it might be related to thyroid complications. To investigate the potential association, we analyzed the correlation between thyroid complications and serum IgM levels. A relevant study showed that relative viscosity increased exponentially above a concentration of 30 g/L (3,000 mg/dl) (35). Moreover, it is reported that patients with IgM > 30 g/L are more likely to get symptomatic hyperviscosity (36). Thus, the patients were divided into low- and high-IgM groups with a cutoff of 30 g/L. Intriguingly, as Table 2 shows, a significant correlation between thyroid complications and relatively alleviated IgM levels was observed (p = 0.03). However, there was no statistically significant correlation between the thyroid function parameters and IgM level, including FT3 (p = 0.246), FT4 (p = 0.6), Anti-TG (p = 0.576), and TSH (p = 0.128). These results suggested the potential causal associations between thyroid complications and alleviated IgM levels, providing an answer for clinical problems. However, further studies with larger sample sizes were required to validate the relationship between the level of thyroid hormones and IgM.

To evaluate the effect of thyroid function abnormalities and IgM on disease presentation, we examined the clinical and biological baseline information of the enrolled patients. The demographic and clinical features of the enrolled patients are depicted in Table 3 . For clinical parameters, a significant dominance was identified in that patients without thyroid complications were more likely to get an overall response (CR+PR) to the initial treatment (p = 0.001). However, as to biological variables and laboratory findings, there were no indicators that presented statistical significance between WM/LPL patients with or without thyroid complications.

Moreover, patients with high levels of IgM more frequently presented with high levels of globulin (p = 0.001), elevated β2-MG (p = 0.033), and decreased hemoglobin with p-values around 0.001. There were no differences related to sex, age, presence of B-symptoms, extramedullary involvement, MYD88 L265P mutation, and response to treatment ( Table 4 ).

The present study further investigated the prognostic value of thyroid complications and IgM in patients with WM/LPL. Kaplan–Meier curves showed no difference in OS between different groups of IgM levels (median follow-up time: low-IgM group, 40 months vs. high-IgM group, 42 months). A similar result was observed in PFS, with a p-value of 0.93. As demonstrated in Figures 1E, F , the median OS for patients with thyroid function abnormalities was significantly shorter than those without thyroid function abnormalities (not reached vs. 91 months, p = 0.02). Consistently, the result also indicated a PFS of 30 months for patients with thyroid complications, while the median survival was 60 months for non-thyroid complications (p = 0.005).

Subsequently, clinical and laboratory variables including age, ISSWM risk group, presence of thyroid complications, levels of hemoglobin, platelet, lactate dehydrogenase (LDH), globulin, and β2-MG, together with thyroid function parameters and cytogenetic parameters of MYD88 L265P mutated status, were entered into the univariate and multivariate Cox regression analyses. In the univariate Cox regression model, age (HR 0.403, 95% CI 0.185–0.879; p = 0.022), ISSWM score (HR 1.059, 95% CI 0.409–2.741; p = 0.014), and thyroid complications (HR 0.229, 95% CI 0.106–0.491; p < 0.001) were significantly associated with an increased risk of progression. The multivariate Cox analysis showed that the impact of thyroid complications was maintained irrespective of age and ISSWM score, with an HR of 0.226 for thyroid complications (95% CI 0.084–0.608, p = 0.03, Table 5 ).

Among the WM/LPL patients enrolled in our study, 8 cases were diagnosed as hypothyroidism, the most common of thyroid complications. The remaining 93 cases without thyroid complications were categorized into two subgroups: subclinical hypothyroidism and normal thyroid function parameters. The diagnosis of subclinical hypothyroidism was based on thyroid function testing: raised serum TSH concentrations and normal serum thyroid hormone concentrations (37). As shown in Figures 2A, B , a significant difference in PFS and OS was identified among the three subgroups, with the p-value around 0.001 in PFS and 0.01 in OS (log-rank test across all three groups).

As mentioned above, thyroid complications were significantly related to ISSWM score (p = 0.014). Therefore, we constructed the subgroup survival analysis for different ISSWM risk groups in WM/LPL patients. According to ISSWM, the entire cohort was divided into two risk grades: a low–intermediate-risk group (ISSWM 0-2) and a high-risk group (ISSWM 3-5). In the low–intermediate-risk group ( Figures 2C, D ), hypothyroidism and subclinical hypothyroidism patients were more likely to have an evidently shorter PFS (p = 0.01) compared with those patients without hypothyroidism, while no significant difference in OS (p = 0.06) was identified among the three subgroups. Subsequently, an evident survival difference was observed between patients with subclinical hypothyroidism and those without hypothyroidism in the early stage of WM/LPL. On the other hand, regardless of OS or PFS, there was no significance between the high-risk stage patients with subclinical hypothyroidism and those without hypothyroidism, while the high-risk stage patients with hypothyroidism were more likely to have evidently shorter PFS (p < 0.01) and OS (p < 0.05) ( Figures 2E, F ). These results indicate that hypothyroidism and subclinical hypothyroidism may have diverse outcomes at different stages of the patients, and WM/LPL patients with subclinical hypothyroidism should be given equal concern as those with hypothyroidism at the early ISSWM stage since they might share an undesirable survival outcome.

Multivariate analysis showed that thyroid complications were the prognostic factor that independently predicted worse PFS in WM/LPL patients. Thus, adding the criterion of thyroid complications to ISSWM might improve its predictive capacity. Accordingly, a novel prognostic index (Th-ISS) can be generated by the sum of adding one point for thyroid complications to the original ISSWM. Considering the negative association between IgM levels and thyroid complications, thyroid complications concurrent with IgM ≤ 30 g/L will add two points to the original ISSWM. As a result, the Th-ISS index demonstrated a statistically significantly larger AUC compared with ISSWM alone in PFS evaluation (Th-ISS, AUC: 0.779, SE: 0.055, 95% CI: 0.667–0.867; ISSWM, AUC: 0.736, SE: 0.060, 95% CI: 0.620–0.831; p = 0.04) ( Figure 3A ). However, there is no evident significance in OS evaluation (Th-ISS, AUC: 0.733, SE: 0.091, 95% CI: 0.618–0.829; ISSWM, AUC: 0.652, SE: 0.087, 95% CI: 0.534–0.759; p = 0.12) ( Figure 3B ).

To further validate and confirm the prognostic capacity of Th-ISS, we firstly divided the cohort into two risk groups based on the original ISSWM: a low-intermediate risk group (ISSWM 0-2) and a high-risk group (ISSWM 3-5) ( Figures 3C, 3D ). However, only the PFS differed significantly among all risk groups according to ISSWM. For comparison, we next split the entire cohort into different risk grades based on the Th-ISS: a low-intermediate risk group (Th-ISS 0-3) and a high-risk group (Th-ISS 4-7). The PFS and OS differed significantly among all risk groups (p<0.001 and p<0.01, respectively) ( Figures 3E, F ). In pairwise comparison, an evident difference between OS and PFS was found between every pair of subgroups categorized by Th-ISS. Patients with high risk (Th-ISS 4-7) had worse survival. Thus, in conclusion, thyroid complications together with ISSWM can significantly differentiate each risk group from one another and improve the risk stratification of ISSWM.

The above findings prompted us to further investigate the role of thyroid-related genes in WM/LPL patients. To make it easier to demonstrate the interplay between thyroid-related genes and critical factors in WM/LPL, including MYD88 and CXCR4, we constructed a network of interactions among the genes ( Figure 4A ). After that, the significance of the interaction was assessed by the Pearson correlation coefficient. As shown in Figure 4B , the ordinate in the figure represents the expression distribution of MYD88 and CXCR4, and the horizontal axis represents the expression distribution of thyroid-related genes. Different colors represent correlation coefficients (in the diagram, red represents positive correlation and blue represents negative correlation). Color intensity indicates statistical significance. Statistically significant correlations between MYD88, CXCR4, and thyroid-related genes are shown in Figure 4C .

Subsequently, differential gene expression analysis between the WM/LPL and healthy donors was assessed. As shown in Figure 4D , a hierarchical cluster analysis was performed on differentially expressed genes between WM/LPL patients and healthy donors. We then constructed volcano plots using fold-change values and adjusted p. The red point in the plot represents the over-expressed mRNAs and the blue point indicates the under-expressed mRNAs with statistical significance. Thyroid-related genes with statistical significance were annotated in the graph ( Figure 4E ). Functional enrichment analysis of thyroid-related gene expression in WM/LPL microarray profiles was performed. Gene ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis indicated that thyroid-related gene was involved in response to peptide hormone and the PI3K-Akt signaling pathway ( Figure 4F ). The expression of thyroid-related genes between WM/LPL patients and healthy donors was shown using a box-violin plot. Five of them reached statistical significance, and one (TBL1X) reached marginal significance ( Figure 4G ).