Compared to non-Hispanic whites (NHWs), Black Americans have over a 2-fold higher incidence of multiple myeloma (MM). Based on data from the US Surveillance, Epidemiology, and End Results Program, the age-adjusted incidence rate of MM in Black individuals increased from 11 per 100 000 in 2000 to 14.6 per 100 000 in 2018, an average annual percent change (APC) of 1.7 [95% confidence interval (CI), 1.4 to 2.1]. Over the same time period, the corresponding rate in NHWs increased from 5.5 to 6.6 (average APC, 1.0; 95% CI: 0.4, 1.6). It is well established that virtually all cases of MM are preceded by the asymptomatic precursor state, monoclonal gammopathy of undetermined significance (MGUS), and individuals with MGUS have overall shorter survival than age- and sex-matched controls.¹ Rates of progression from MGUS to MM – about 1% per year¹ –do not appear to vary by race. Thus, the disparity in incidence of MM appears to be determined by a disparity in incidence of MGUS. Indeed, MGUS prevalence is higher in Black individuals than NHWs. Landgren et al. estimated the age-adjusted prevalence of MGUS in Black individuals ages 50 and older to be 3.7% (vs. 2.3% in NHWs) using data from the National Health and Nutritional Examination Survey (NHANES).² These estimates, however, were based on blood samples collected between 1988–1994 and 1999–2004. More recent nationwide prevalence estimates of MGUS are not available.

The Black Women's Health Study (BWHS) is a prospective cohort study that enrolled 59 000 self-identified Black women ages 21–69 in 1995 from across the US.³ The BWHS has followed participants for cancer and other outcomes through biennial questionnaires and linkage to 24 cancer registries in states in which 95% of participants live. Between 2014 and 2018, blood samples from 13 030 BWHS participants (approximately 25% of the surviving cohort) were collected and processed by Patient Service Centers at Quest Diagnostics (Madison, NJ). Participants who provided samples were 39 to 89 years old (median 58 years) at the time of blood draw. The samples are stored as serum, plasma, buffy coats, and red blood cells at the Boston University Molecular Genetics Core Laboratory in −80°C freezers. For the present analysis, 1707 participants ages 50–79 with available serum samples and no personal history of cancer were randomly selected for laboratory screening, with approximately equal number of individuals within 10-year age groups.

Samples were de-identified and all assays were performed and interpreted by individuals blinded to health, demographic, and other individual data. Total serum protein quantitation, serum protein electrophoresis (SPEP) by agarose gel, and immunofixation electrophoresis (IFE) were performed on all samples according to standard protocols at the Clinical Chemistry Laboratory at Boston Medical Center, a CLIA-certified laboratory. All samples were interpreted by one pathologist (YK). We also measured serum free light chains (sFLCs) by nephelometry using the Freelite® assay from Binding Site (Birmingham, UK); sFLCs were not measured for one individual.

We used diagnostic criteria from the International Myeloma Working Group to classify individuals as having MGUS, characterized by presence of a M protein <3 g/dl in the serum, consistent with the definition used in other cohorts.^{2, 4} A sFLC κ/λ ratio <0.26 or >1.65 is also an indication of MGUS and present in about 30–45% of MGUS patients; however, in the absence of data regarding renal function, light-chain MGUS diagnosis was determined based on SPEP/IFE only. We excluded one individual with a restricted band in the lambda region without heavy chain immunoglobulin expression whose sFLC λ was >100 mg/L and involved/uninvolved sFLC ratio was >100 – suggestive of myeloma – leaving an analytic cohort of 1706.

We used year 2000 US Census data to directly standardize crude prevalence estimates to the population distribution for women in 10-year age groups (50–59, 60–69, and 70–79) for comparison to previous literature. We tabulated immunoglobulin isotype and abnormal sFLC κ/λ ratio and calculated the median M protein concentration among cases, using half the limit of detection (LOD) when concentrations were below the LOD (<0.10 g/dl).

We identified 162 cases of MGUS (crude prevalence, 9.5%), 125 of which were identified by SPEP and confirmed by IFE and 37 of which were identified on IFE only. The overall age-adjusted prevalence was 9.0% (95% confidence interval: 7.6%, 10.4%). Prevalence of MGUS increased with increasing age, with a prevalence of 12.3% (9.3, 15.4) in women age > 70 (Table 1). Observed prevalence estimates in the BWHS were markedly higher than previously reported estimates in Black populations. In a community-based sample of 976 Black residents of North Carolina >70 years of age (1992–1993), MGUS prevalence was 8.4%.⁵ In the Southern Community Cohort Study (SCCS), 39 cases were detected among 1000 Black women ages 40–79 screened (3.9%).⁴ The only nationwide estimates are from the NHANES study mentioned previously; the prevalence of MGUS among Black women ages 50 and older was 3.4% (95% CI: 2.4, 4.7).²

Non-IgG isotype, abnormal sFLC ratio, and M protein ≥1.5 g/dl are predictors associated with increased risk of progression to MM or lymphoplasmacytic lymphoma.¹ IgG MGUS was observed in 114 cases (70%), IgA MGUS in 34 cases (21%) and IgM MGUS in 9 cases (5.6%). Three individuals presented with biclonal MGUS. Interestingly, we also observed one case of IgD MGUS, a rare subtype. One case was identified as a light-chain only based on SPEP/IFE (Table S1). The distribution of immunoglobulin heavy chain isotype is consistent with prior reports in Black populations.

Among the 162 MGUS cases, free-κ light chain values ranged from 7.3 mg/L to 508 mg/L (median, 21.2 mg/L), whereas free-λ light chain values ranged from 2.5 mg/L to 608.4 mg/L (median, 15.1 mg/L). Seventy-four cases (46%) presented with an abnormal sFLC κ/λ ratio (<0.26 or >1.65) (Table S1). M-spike was quantified in 106 cases (65% of all cases; 85% of cases with positive finding on SPEP). The median M protein concentration was 0.21 g/dl; the maximum M protein concentration was 1.46 g/dl (Table S1).

Among 1543 individuals with no evidence of MGUS on SPEP/IFE and available sFLC, free-κ light chain values ranged from 7.3 mg/L to 327.5 mg/L (median, 21.2 mg/L), whereas free-λ light chain values ranged from 3.2 mg/L to 146.5 mg/L (median, 15.0 mg/L). sFLC κ/λ ratios ranged from 0.21 to 18.1 (median, 1.45); 483 women (31%) had an abnormal sFLC κ/λ ratio (<0.26 or >1.65) (Table S1). We also present frequencies of abnormal sFLC κ/λ ratios for alternative reference ranges reported in the literature⁶ (Table S1) as well as the distribution of free-κ and free-λ light chain values among women without evidence of heavy-chain MGUS on SPEP/IFE in Table S2 and Figure S1.

A limitation of our study, as in previous studies, is the lack of serum creatinine levels, urinalysis, bone imaging information, bone marrow biopsies, or other clinical correlates of MGUS or myeloma. It is therefore possible that some individuals classified as MGUS cases in this study may have been experiencing a myeloma-defining event at the time of blood draw. Given lack of information on kidney function for participants in our study and consequent lack of defined reference range for sFLC κ/λ ratio in this population, we were unable to accurately determine how many women had light-chain MGUS. Our results suggest that previously published estimates of MGUS prevalence based on SPEP and IFE in US Black women may be underestimated. Reasons for the higher than expected prevalence are unknown, but could reflect temporal trends: NHANES samples were collected between 1988–1994 and 1999–2004, whereas BWHS samples were collected between 2014 and 2018. This possibility is supported by evidence that myeloma incidence has also increased over this time period. Because we performed IFE on all samples regardless of the finding on SPEP, we identified some MGUS cases that would have been missed in the NHANES prevalence study; however, even considering only cases with evidence of an M-protein on SPEP, our prevalence estimate was still higher than in previous cohorts. We note that participants in the BWHS are not a representative sample of Black women in the United States, and therefore our findings may not be generalizable to other populations. In particular, BWHS participants tend to have higher levels of education; however, socioeconomic status is not known to be associated with MGUS. Finally, it is also possible that our results could represent a chance finding. A better understanding of etiologic factors in MGUS and MM development is urgently needed to inform opportunities for risk reduction, particularly in high-risk populations.

与非西班牙裔白人 (NHW) 相比，美国黑人多发性骨髓瘤 (MM) 的发病率高出 2 倍多。根据美国监测、流行病学和最终结果计划的数据，黑人中 MM 的年龄调整发病率从 2000 年的每 10 万人中 11 例增加到 2018 年的每 10 万人中 14.6 例，平均年百分比变化 (APC) 1.7 [95% 置信区间 (CI)，1.4 至 2.1]。同一时期，NHW 的相应比率从 5.5 增加到 6.6（平均 APC，1.0；95% CI：0.4，1.6）。众所周知，几乎所有 MM 病例之前都有无症状的前体状态，即意义未明的单克隆丙种球蛋白病 (MGUS)，并且 MGUS 患者的总体生存期比年龄和性别匹配的对照者短。¹从 MGUS 进展为 MM 的速度——每年约 1% ^1——似乎不因种族而异。因此，MM 发病率的差异似乎是由 MGUS 发病率的差异决定的。事实上，黑人中的 MGUS 患病率高于 NHW。兰德格伦等人。根据国家健康和营养检查调查 (NHANES) 的数据，估计 50 岁及以上黑人中 MGUS 的年龄调整患病率为 3.7%（相对于 NHW 中的 2.3%）。²然而，这些估计是基于 1988-1994 年和 1999-2004 年期间采集的血液样本。目前尚无最新的全国 MGUS 患病率估计。

黑人女性健康研究 (BWHS) 是一项前瞻性队列研究，于 1995 年招募了来自美国各地的 59,000 名自我认定年龄为 21-69 岁的黑人女性。³ BWHS 通过每两年一次的问卷调查以及与 95% 参与者居住州的 24 个癌症登记处的联系来跟踪参与者的癌症和其他结果。2014 年至 2018 年间，Quest Diagnostics（新泽西州麦迪逊）的患者服务中心收集并处理了 13,030 名 BWHS 参与者（约占幸存队列的 25%）的血液样本。提供样本的参与者在抽血时年龄为 39 至 89 岁（中位年龄 58 岁）。样本以血清、血浆、血沉棕黄层和红细胞形式储存在波士顿大学分子遗传学核心实验室的 -80°C 冰箱中。在本次分析中，随机选择了 1707 名年龄在 50-79 岁之间、有可用血清样本且没有癌症个人史的参与者进行实验室筛查，10 岁年龄组中的人数大致相等。

样本被去识别化，所有检测均由对健康、人口统计和其他个人数据不知情的个人进行和解释。根据 CLIA 认证实验室波士顿医学中心临床化学实验室的标准方案，对所有样品进行总血清蛋白定量、琼脂糖凝胶血清蛋白电泳 (SPEP) 和免疫固定电泳 (IFE)。所有样本均由一名病理学家 (YK) 解读。我们还使用 Binding Site（英国伯明翰）的 Freelite® 测定法通过比浊法测量了血清游离轻链 (sFLC)；未测量某一个体的 sFLC。

我们使用国际骨髓瘤工作组的诊断标准将个体分类为 MGUS，其特征是血清中存在 <3 g/dl 的 M 蛋白，与其他队列中使用的定义一致。^{2, 4} A sFLC κ/λ 比值 <0.26 或 >1.65 也是 MGUS 的指征，约 30-45% 的 MGUS 患者存在；然而，在缺乏肾功能数据的情况下，轻链 MGUS 诊断仅根据 SPEP/IFE 确定。我们排除了 lambda 区域有限制带但不表达重链免疫球蛋白的个体，其 sFLC λ > 100 mg/L，且受累/未受累 sFLC 比率 > 100（提示骨髓瘤），留下 1706 名分析队列。

我们使用 2000 年美国人口普查数据来直接标准化 10 岁年龄组（50-59、60-69 和 70-79）女性人口分布的粗略患病率估计，以便与以前的文献进行比较。我们将免疫球蛋白同种型和异常 sFLC κ/λ 比率制成表格，并计算病例中的中位 M 蛋白浓度，当浓度低于 LOD (<0.10 g/dl) 时，使用检测限 (LOD) 的一半。

我们发现了 162 例 MGUS 病例（粗患病率，9.5%），其中 125 例由 SPEP 发现并由 IFE 确认，其中 37 例仅在 IFE 上发现。年龄调整后的总体患病率为 9.0%（95% 置信区间：7.6%、10.4%）。MGUS 的患病率随着年龄的增加而增加，70 岁以上女性的患病率为 12.3%（9.3，15.4）（表 1）。BWHS 中观察到的患病率估计值明显高于之前报告的黑人人群患病率估计值。在北卡罗来纳州 976 名 70 岁以上黑人居民的社区样本中（1992-1993 年），MGUS 患病率为 8.4%。⁵在南方社区队列研究 (SCCS) 中，在接受筛查的 1000 名 40-79 岁黑人女性中发现了 39 例病例 (3.9%)。⁴唯一的全国性估计来自前面提到的 NHANES 研究；50 岁及以上黑人女性中 MGUS 的患病率为 3.4%（95% CI：2.4，4.7）。²

非 IgG 同种型、异常 sFLC 比率和 M 蛋白≥1.5 g/dl 是与进展为 MM 或淋巴浆细胞淋巴瘤风险增加相关的预测因子。¹ IgG MGUS 114 例（70%），IgA MGUS 34 例（21%），IgM MGUS 9 例（5.6%）。三人患有双克隆 MGUS。有趣的是，我们还观察到一例 IgD MGUS，这是一种罕见的亚型。仅根据 SPEP/IFE（表 S1）将一个案例识别为轻链。免疫球蛋白重链同种型的分布与黑人群体中先前的报道一致。

在 162 例 MGUS 病例中，游离 κ 轻链值范围为 7.3 mg/L 至 508 mg/L（中位数为 21.2 mg/L），而游离 λ 轻链值范围为 2.5 mg/L 至 608.4 mg/L （中位数，15.1 毫克/升）。74 例 (46%) 出现异常 sFLC κ/λ 比值（<0.26 或 >1.65）（表 S1）。106 例 M-spike 被量化（占所有病例的 65%；SPEP 阳性结果占 85%）。中位 M 蛋白浓度为 0.21 g/dl；最大 M 蛋白浓度为 1.46 g/dl（表 S1）。

在 SPEP/IFE 和可用 sFLC 上没有 MGUS 证据的 1543 名个体中，游离 κ 轻链值范围为 7.3 mg/L 至 327.5 mg/L（中位数为 21.2 mg/L），而游离 λ 轻链值范围为从 3.2 毫克/升到 146.5 毫克/升（中位数，15.0 毫克/升）。sFLC κ/λ 比率范围为 0.21 至 18.1（中位数为 1.45）；483 名女性 (31%) 的 sFLC κ/λ 比率异常（<0.26 或 >1.65）（表 S1）。我们还提供了文献⁶（表 S1）中报道的替代参考范围的异常 sFLC κ/λ 比率的频率，以及无重链 MGUS 证据的女性中游离 κ 和游离 λ 轻链值的分布。表 S2 和图 S1 中的 SPEP/IFE。

与之前的研究一样，我们研究的局限性是缺乏血清肌酐水平、尿液分析、骨成像信息、骨髓活检或 MGUS 或骨髓瘤的其他临床相关信息。因此，本研究中被归类为 MGUS 病例的一些个体可能在抽血时经历了骨髓瘤定义事件。鉴于我们的研究参与者缺乏肾功能信息，并且因此缺乏该人群中 sFLC κ/λ 比率的明确参考范围，我们无法准确确定有多少女性患有轻链 MGUS。我们的结果表明，之前发表的基于 SPEP 和 IFE 对美国黑人女性 MGUS 患病率的估计可能被低估。患病率高于预期的原因尚不清楚，但可以反映时间趋势：NHANES 样本是在 1988-1994 年和 1999-2004 年之间收集的，而 BWHS 样本是在 2014 年至 2018 年之间收集的。这种可能性得到了骨髓瘤发病率也随时间变化的证据的支持。在此期间有所增加。因为无论 SPEP 的结果如何，我们都对所有样本进行了 IFE，因此我们发现了一些在 NHANES 患病率研究中可能被遗漏的 MGUS 病例；然而，即使仅考虑 SPEP 上有 M 蛋白证据的病例，我们的患病率估计值仍然高于之前的队列。我们注意到，BWHS 的参与者并不是美国黑人女性的代表性样本，因此我们的研究结果可能无法推广到其他人群。特别是，BWHS 参与者往往拥有较高的教育水平；然而，社会经济地位与 MGUS 之间的关系尚不清楚。最后，我们的结果也有可能代表一个偶然的发现。迫切需要更好地了解 MGUS 和 MM 发展的病因因素，以便为降低风险的机会提供信息，特别是在高风险人群中。