Cardiovascular interventions using x-ray guidance are increasing in both volume and complexity. The long-term biological effects of chronic low-dose radiation exposure in operators performing these procedures are, however, largely unknown. Occupational safety limits are based on physical dosimetry only and do not consider individual biological sensitivity to radiation. We previously reported DNA damage in lymphocytes isolated from operators performing endovascular aortic repair (EVAR).¹ Expression of γ-H2AX and pATM (phosphorylated ataxia telangiectasia mutated), which are markers of acute DNA damage/repair, rose immediately after performing EVAR and normalized the following day. These markers, however, do not reflect the effects of chronic exposure, including chromosomal aberrations that may herald genomic instability and predisposition to malignancy. Here we report important findings pertaining to these aberrations in an international group of operators performing a large volume of complex endovascular interventions, including branched and fenestrated EVAR.

Peripheral blood was collected from endovascular operators (n=12; 11 male) and radiation-naïve general surgeons as controls (n=6; 5 male), all of whom gave informed consent. The study was approved by our institutional review committee (reference: 16/LO/1111). The median age of endovascular and control operators (50 [36–55] versus 47 [36–52], respectively; P=0.37) and years in practice (13.5 [3–20] versus 10.5 [5–16], respectively; P=0.19) were comparable. There were no intergroup differences in radiation exposures for personal health reasons, smoking history, or medications. Two endovascular operators had cancer, a squamous skin and renal lesion, both curatively treated at ages 49 and 16 years, respectively. Endovascular operators performed a median of 35 (20–100) standard EVARs and 70 (30–100) branched and fenestrated EVARs annually with a median annual personal radiation dose of 0.96 mSv (0.22–13.64) in the 3 years before blood sampling. All endovascular operators wore lead gowns and thyroid shields. Lead leg shields, headcaps, and goggles were used by 58%, 42%, and 92% of operators, respectively. Two operators used a ceiling-suspended radiation protection suit, and one used scatter radiation–absorbing drapes.

Giemsa-stained metaphase preparations were used to analyze the full complement of chromosomes in at least 3000 lymphocytes per operator (Figure [A]). Semiautomated scoring found a dicentric chromosome frequency of 0.11 (0.03–0.16) per 100 cells in the endovascular operators compared with 0.04 (0–0.06) in controls (P=0.002, Figure [B]). There was no correlation between age of operator and dicentric frequency (Pearson coefficient, r=0.04 [–0.44 to 0.50], P=0.876).

Figure. Chromosomal aberrations in exposed endovascular operators versus controls. A, Chromosome spread of a lymphocyte in metaphase, visualized by Giemsa staining, showing a dicentric chromosome (red arrow) and an acentric fragment (blue arrow). In this aberration, breaks in 2 chromosomes followed by an incorrect repair have resulted in the formation of a single chromosome containing 2 centromeres and a chromosome fragment containing no centromeres. B, Frequency of dicentric chromosomes per 100 cells in endovascular operators compared with radiation-naïve control operators (0.11 versus 0.04, respectively; *Mann-Whitney U test; P=0.002). C, Multiplex fluorescence in situ hybridization (m-FISH) demonstrating a chromosome spread of 22 pairs of autosomes and single X and Y chromosomes. A simple reciprocal translocation between chromosomes 3 and 8 is highlighted by the yellow arrows. D, An illustration depicting the formation of the reciprocal translocation seen in C, where ionizing radiation has caused breaks in chromosomes 3 and 8, which have then been repaired incorrectly such that a fragment of chromosome 8 is attached to chromosome 3 and vice versa. E, A complex, unstable, nontransmissible chromosome rearrangement visualized by m-FISH, with yellow arrows highlighting the chromosomes (1, 8, and 10) affected by breaks and incorrect repair. F, Illustration depicting the formation of the complex rearrangement seen in E, where ionizing radiation has caused breaks in 3 chromosomes followed by incorrect repair. The centromere-containing portion of chromosome 10 has attached to chromosome 8, and the acentric portion has attached to chromosome 1. Acentric portions of chromosomes 1 and 8 have also attached to form an acentric fragment. Lower case alphabetic labels denote part of chromosomes. G, A chromosome spread with aneuploidy visualized by m-FISH. Aneuploidy refers to the abnormal loss or gain of chromosomes within the cell. In this instance, the yellow arrows highlight the missing chromosomes 1 and 20. H, Bar chart showing the abnormal loss of each chromosome (aneuploidy) per 100 cells in endovascular operator samples (red) compared with the radiation-naïve controls (blue) after analyzing a total of >2000 cells by m-FISH; median of differences, 0.35; Wilcoxon signed-rank test; P=0.004.

More than 2000 lymphocytes from 9 operators (5 exposed, 4 control) were analyzed by multiplex fluorescence in situ hybridization using fluorescent probes hybridized to metaphase chromosomes. The frequency of unstable, complex exchanges that involve 3 or more breaks in 2 or more chromosomes (0.48 versus 0.24 per 100 cells, Mann-Whitney U test, P=0.32) and stable, reciprocal translocations (0.86 versus 0.59 per 100 cells, Mann-Whitney U test, P=0.38) trended higher in endovascular operators (Figure [C through F]). Stable exchanges can be passed on to subsequent cell generations during mitosis and are, therefore, particularly useful for monitoring cytogenetic effects of chronic radiation exposures. Aneuploidy, which refers to abnormal loss of chromosomes, was more frequent in radiation-exposed operators (Wilcoxon signed-rank test, P=0.004, Figure [G and H]), with a median difference of 0.35 per chromosome.

Dicentric chromosomes, formed by cleavage and incorrect repair of double-stranded DNA, indicate genomic instability and reflect radiation exposure during the lymphocyte’s lifespan, which is ≈3 years.² Their frequency increases proportionally to cumulative radiation exposures (<5Gy), allowing their use for biological assessment of chronic exposures.² The dicentric frequencies we observed fall below the threshold that allows reliable inference of effective exposure dose using current nomograms. Nevertheless, we found an almost 3-fold higher incidence of dicentrics in endovascular operators compared with radiation-naïve controls. The dicentric count in the latter group was comparable with that of the general population, which is generally quoted as ≈0.06 per 100 cells.² Our findings are corroborated by a recent report of higher dicentric frequency in interventional radiologists.³ These data highlight the need to investigate whether partial body irradiation to unshielded areas such as the legs may contribute to this level of chromosomal damage over time.¹

The chromosomal aberrations detected in the present study using multiplex fluorescence in situ hybridization, although not necessarily caused by occupational radiation exposure, are also associated with cancer.⁴ These increase the burden of genetic alterations that can cause defects in cell proliferation, induce proteotoxic stress, and promote tumorigenesis, but uncertainties remain around linking these actions to cancer risk. The effect of chronic low-dose occupational radiation exposure on the health of medical workers is uncertain and requires extensive epidemiological and mechanistic studies to inform.⁵ Our exploratory findings are hypothesis-generating and strengthen the case for larger-scale prospective studies that accurately record radiation doses to all body parts, capture health events, and relate these to cytogenetic markers of chronic exposure.

Data, materials, and methods will be made available to researchers through direct communication with the corresponding author.

The study was funded by the British Heart Foundation (FS/17/24/32596 to B.M).

Disclosures B.M. reports the following: consulting/advisory board for Cydar Medical; consulting for Philips; and advisory board, proctoring, and speaker’s fees for Cook. The other authors report no conflicts.

Circulation is available at www.ahajournals.org/journal/circ

For Sources of Funding and Disclosures, see page 1810.

使用 X 射线引导的心血管干预在数量和复杂性方面都在增加。然而，在执行这些程序的操作者中，慢性低剂量辐射暴露的长期生物效应在很大程度上是未知的。职业安全限值仅基于物理剂量测定，不考虑个体对辐射的生物敏感性。我们之前报道了从进行血管内主动脉修复 (EVAR) 的操作者中分离出的淋巴细胞中的 DNA 损伤。¹作为急性 DNA 损伤/修复的标志物 γ-H2AX 和 pATM（磷酸化共济失调毛细血管扩张症突变）的表达在进行 EVAR 后立即上升并在第二天恢复正常。然而，这些标志物并不能反映长期暴露的影响，包括可能预示基因组不稳定性和恶性肿瘤易感性的染色体畸变。在这里，我们报告了在执行大量复杂血管内介入治疗（包括分支和开窗 EVAR）的国际手术组中与这些异常有关的重要发现。

从血管内操作者（n = 12；11 名男性）和未接受放射治疗的普通外科医生（n = 6；5 名男性）收集外周血作为对照，所有这些人都知情同意。该研究得到了我们的机构审查委员会的批准（参考：16/LO/1111）。血管内操作者和对照组操作者的中位年龄（分别为 50 [36-55] 对 47 [36-52]；P = 0.37）和实践年数（分别为 13.5 [3-20] 对 10.5 [5-16]；磷=0.19) 具有可比性。由于个人健康原因、吸烟史或药物治疗，辐射暴露没有组间差异。两名血管内操作者患有癌症、鳞状皮肤和肾脏病变，分别在 49 岁和 16 岁时治愈。血管内操作者在采血前 3 年内每年平均进行 35 (20-100) 次标准 EVAR 和 70 (30-100) 次分支和开窗 EVAR，年平均个人辐射剂量为 0.96 mSv (0.22-13.64)。所有血管内操作者都穿着铅衣和甲状腺防护罩。分别有 58%、42% 和 92% 的操作员使用铅护腿罩、头罩和护目镜。两名操作员使用了悬挂在天花板上的辐射防护服，一名操作员使用了散射辐射吸收窗帘。

Giemsa 染色的中期制剂用于分析每个操作者至少 3000 个淋巴细胞中的完整染色体（图 [A]）。半自动评分发现，血管内操作者每 100 个细胞的双着丝粒染色体频率为 0.11 (0.03-0.16)，而对照组为 0.04 (0-0.06)（P = 0.002，图 [B]）。操作者的年龄与双着丝粒频率之间没有相关性（皮尔逊系数，r = 0.04 [–0.44 至 0.50]，P = 0.876）。

数字。 暴露的血管内操作者与对照组的染色体畸变。A，中期阶段淋巴细胞的染色体扩散，通过吉姆萨染色可见，显示双着丝粒染色体（红色箭头）和无着丝粒片段（蓝色箭头）。在这种畸变中，2 条染色体断裂，然后进行不正确的修复，导致形成一条含有 2 个着丝粒的染色体和一个不含着丝粒的染色体片段。B，与未接受过辐射的对照操作者相比，血管内操作者每 100 个细胞中双着丝粒染色体的频率（分别为 0.11 和 0.04；*Mann-Whitney U检验；P = 0.002）。C, 多重荧光原位杂交 (m-FISH) 显示 22 对常染色体和单个 X 和 Y 染色体的染色体分布。黄色箭头突出显示了 3 号和 8 号染色体之间的简单相互易位。D，描绘了在C中看到的相互易位形成的插图，其中电离辐射导致 3 号和 8 号染色体断裂，然后错误地修复，使得 8 号染色体的一个片段附着在 3 号染色体上，反之亦然。E，通过 m-FISH 观察到的复杂、不稳定、不可传播的染色体重排，黄色箭头突出显示受断裂和不正确修复影响的染色体（1、8 和 10）。F, 插图描绘了在E中看到的复杂重排的形成，其中电离辐射导致 3 条染色体断裂，随后出现不正确的修复。第 10 号染色体的含着丝粒部分与第 8 号染色体相连，无着丝粒部分与第 1 号染色体相连。第 1 号和第 8 号染色体的无着丝粒部分也已附着形成无着丝粒片段。小写字母标签表示染色体的一部分。G，通过 m-FISH 观察到的具有非整倍性扩散的染色体。非整倍性是指细胞内染色体的异常丢失或增加。在这种情况下，黄色箭头突出显示缺失的 1 号和 20 号染色体。H, 条形图显示在通过 m-FISH 分析总共 >2000 个细胞后，与未接受辐射的对照（蓝色）相比，血管内操作者样本（红色）中每 100 个细胞中每条染色体（非整倍体）的异常丢失；差异中位数，0.35；Wilcoxon 符号秩检验；P = 0.004。

使用与中期染色体杂交的荧光探针，通过多重荧光原位杂交分析了来自 9 名操作员（5 名暴露，4 名对照）的 2000 多个淋巴细胞。涉及 2 条或更多条染色体中 3 次或更多断裂的不稳定、复杂交换的频率（每 100 个细胞 0.48 对 0.24，Mann-Whitney U检验，P = 0.32）和稳定的相互易位（每 100 个细胞 0.86 对 0.59，Mann -惠特尼U检验，P= 0.38）在血管内操作者中呈上升趋势（图 [C 至 F]）。稳定的交换可以在有丝分裂期间传递给随后的细胞世代，因此对于监测慢性辐射暴露的细胞遗传学影响特别有用。非整倍体是指染色体的异常丢失，在辐射暴露的操作者中更为常见（Wilcoxon 符号秩检验，P = 0.004，图 [G 和 H]），每个染色体的中位数差异为 0.35。

双着丝粒染色体是由双链 DNA 的切割和不正确修复形成的，表明基因组不稳定性，并反映淋巴细胞寿命（约 3 年）期间的辐射暴露。²它们的频率与累积辐射照射（<5Gy）成正比增加，因此可以用于慢性照射的生物学评估。²我们观察到的双着丝粒频率低于允许使用当前列线图可靠推断有效暴露剂量的阈值。尽管如此，我们发现血管内操作者的双着丝粒发生率几乎是未接受放射治疗的对照组的 3 倍。后一组的双着丝粒数与普通人群相当，通常引用为 ≈0.06/100 个细胞。²我们的研究结果得到了近期介入放射科医师双着丝粒频率较高的报告的证实。³这些数据强调有必要调查对腿部等非屏蔽区域的部分身体照射是否会随着时间的推移导致这种水平的染色体损伤。¹

本研究中使用多重荧光原位杂交检测到的染色体畸变虽然不一定是由职业辐射暴露引起的，但也与癌症有关。⁴这些增加了可能导致细胞增殖缺陷、诱导蛋白毒性应激和促进肿瘤发生的遗传改变的负担，但将这些行为与癌症风险联系起来仍存在不确定性。慢性低剂量职业辐射暴露对医务人员健康的影响尚不确定，需要进行广泛的流行病学和机制研究才能提供信息。⁵我们的探索性发现产生了假设，并加强了大规模前瞻性研究的案例，这些研究准确记录了所有身体部位的辐射剂量，捕捉健康事件，并将这些与慢性暴露的细胞遗传学标记联系起来。

数据、材料和方法将通过与通讯作者的直接交流提供给研究人员。

该研究由英国心脏基金会（FS/17/24/32596 to BM）资助。

披露BM 报告以下内容： Cydar Medical 的咨询/顾问委员会；为飞利浦提供咨询；库克的顾问委员会、监考和演讲者费用。其他作者报告没有冲突。

流通可在 www.ahajournals.org/journal/circ

有关资金来源和披露信息，请参见第 1810 页。