兽研引言
嗜血支原体属（又称血支原体）是对脊椎动物宿主红细胞具有趋向性的细菌，有可能在受感染宿主中引起溶血性贫血[1–3]。这些细菌属于支原体属，全球范围内有三种被认定为猫传染性贫血的病原体：猫血支原体、“Candidatus Mycoplasma haemominutum”和“Candidatus Mycoplasma turicensis”[4, 5]。
尽管存在一些地理差异，猫支原体感染的病原体呈全球分布（表1），在家猫、流浪猫和野猫中均有检出记录[6–10]。大多数研究报告“Candidatus Mycoplasma haemominutum”（“Ca. M. haemominutum”）的患病率最高，其次是猫血支原体（M. haemofelis），第三位是“Candidatus Mycoplasma turicensis”（“Ca. M. turicensis”）[11–14]。
表1. 猫支原体感染的国际患病率研究
同一宿主中多种病原体混合感染较为常见，因为这些病原体具有相似的风险因素和传播途径，其中猫血支原体与“Ca. M. haemominutum”的组合最为常见[14, 27–30]。
这些病原体也在多种野生猫科动物中有记录，无论是自由放养还是圈养的（狮子、老虎、猎豹、伊比利亚猞猁、豹、美洲豹、美洲狮等），其中“Ca. M. haemominutum”是最常见的种类，并有混合感染的记录，与家猫类似[31–35]。
通常在有户外活动机会的雄性猫中检出较高的感染患病率，其游走和领地行为被认为可能是与其他猫发生攻击性互动可能性增加以及接触节肢动物媒介机会增多原因[16, 36–39]。有时也注意到，被鉴定为猫免疫缺陷病毒（FIV）或猫白血病病毒（FeLV）阳性的动物与猫支原体属混合感染的风险较高，尤其以FIV为重，后者也与通过攻击性互动传播密切相关[7, 40–44]。
最常报告的临床症状通常是非特异性的，如嗜睡和厌食，但也包括反映潜在贫血严重程度的各种体征，如黏膜苍白、心动过速、脾肿大、黄疸、听诊时心脏杂音等[8, 45–47]。由于存在完全无症状的慢性携带者，有时未发现猫支原体感染与贫血之间的明显关系[48–50]。对于确实发生急性贫血的患者，通常为再生性贫血（大细胞性和低色素性），可能伴有红细胞大小不均、网织红细胞增多和多染性[21, 48, 51, 52]。血小板计数可能正常或略低于正常[53, 54]。还可能因肝脏缺氧导致肝酶升高，因溶血导致高胆红素血症，总蛋白轻度升高，或因多克隆丙种球蛋白病导致白蛋白/球蛋白比值改变[14, 55–57]。
猫血支原体的细胞学诊断是一种可及但灵敏度较低的方法，因为通常需要经验丰富的操作者才能观察到大量菌血症，且由于三种病原体形态相似而无法区分[5, 42, 53, 58]。相比之下，聚合酶链反应（PCR）在检测唾液、血液、尿液、粪便等不同类型样本中支原体属16S rRNA基因片段方面已证明具有高灵敏度和特异性[5, 59–61]。然而，所测定的血支原体载量并不直接与动物的临床状态相关，且有些病例尽管PCR阴性但仍持续血清阳性，因此推测在这些情况下血支原体以低于PCR检测限的水平持续存在[62–66]。另外，血清学技术是一种实用且便捷的诊断方法，得益于特异性抗体的存在，其滴度反映感染的进展，IgM滴度可区分急性感染和慢性感染[61, 67–69]。
本研究旨在评估HEV（兽医教学医院）患者中诊断的猫支原体属患病率，并比较阴性和近期感染猫在风险因素、临床症状和实验室参数方面的差异。同时还分析了可能影响阳性病例预后的因素。
材料与方法
研究设计与纳入标准
设计了一项分析性横断面研究，目标人群为2021年1月1日至2024年6月30日期间在HEV接受猫支原体感染筛查的猫患者。
将病例分类为急性猫支原体感染的标准是免疫球蛋白M（IgM）结果阳性。由于研究期间通过PCR方法筛查的病例较少，经PCR诊断的病例未被纳入。仅有免疫球蛋白G（IgG）阳性结果但无IgM信息的患者也被排除[56, 68, 70]。
在诊断血支原体感染时，并非所有病例都有完整的血常规和/或全面的生化分析可用，且部分临床记录无法完整收集信息，如疫苗接种状态、FIV/FeLV状态、体内外寄生虫预防频率等。对所获得的数据进行分析时，已意识到存在信息缺失的情况。
关注变量
基于先前关于猫血营养性支原体感染流行病学方面的研究[5, 39, 44, 71]，即水平传播和/或媒介传播的可能性，在收集每只猫生活方式的病史信息时格外谨慎，以评估其流行病学风险。收集了关于其来源（从繁育者处领养或街头救助）、是否生活在多猫家庭、户外活动情况（以评估媒介暴露或种间攻击）、疫苗接种和体外驱虫的规范性、FIV/FeLV状态，以及可从其医疗记录中获得的任何其他有用信息。在适用的情况下，记录任何既存状况如并发合并症或合并感染，以便后续评估其统计学价值。
另一个关注点是在诊断检查过程中盘点临床和分析发现，以便将围绕支原体属筛查时获得的任何关于临床症状（如精神沉郁、黏膜苍白、心动过速/呼吸急促、黄疸、发热等）以及常规全血细胞计数和生化面板的可用信息也记录下来用于分析。与支原体属筛查时间相隔过远的临床和分析数据被认为不适合进行统计分析。
最后，还记录了研究患者的临床结局和治疗细节，如实际开始治疗的病例（使用抗生素、皮质类固醇、输血等），以及其他猫被实施安乐死的情况。这些治疗细节，连同任何记录的合并症和合并感染，被认为是一个值得探索的有趣方面，以试图识别与病例临床结局相关的任何明显趋势。
病历检索
通过检索患者的临床记录收集患者信息，并整理到一个数据库中，包含其概况（性别、年龄、环境、生活方式、疫苗接种状态、体内外寄生虫预防、内科和手术史）。在核实每位患者的完整临床病史后，收集出现猫支原体病临床怀疑期间的血细胞计数值和生化分析结果。部分患者的全血细胞计数和生化分析与病原学诊断同时申请，但有些病例的血支原体筛查略有延迟，此类病例接受最大72小时的差异。除血细胞计数信息外，还收集了其他数据，如血尿素氮（BUN）、肌酐、天冬氨酸氨基转移酶（AST）、丙氨酸氨基转移酶（ALT）、碱性磷酸酶（ALP）、总蛋白、白蛋白、球蛋白、白蛋白/球蛋白比值和总胆红素。
还沿猫支原体病筛查的时间线提取了相关临床信息，如就诊期间观察到的临床症状（发热、黏膜苍白、呼吸急促/心动过速等）、所给予的治疗以及每个病例的结局。
随访信息追溯至每位患者最近的记录，部分患者返回复诊，而其他患者在医疗出院后无进一步记录。
提取数据
从临床记录中收集的数据被整理和分类到一个数据库中，总结于表2。
表2. 所收集变量的组织和分类
数据分析
使用Microsoft Office Excel（版本2409）、Jamovi（版本2.6.44.0）和IBM SPSS Statistics®（版本29）对收集的数据进行统计分析。所有执行的检验设定的统计学显著性值为p ≤ 0.05。
分类变量被分入适当的组别（有或无户外活动、IgM阳性或阴性、有或无黄疸等），而定量变量如全血细胞计数和生化值则根据外包实验室设定的参考范围分类为低于、高于或在生理参数范围内。
为计算患病率，将IgM阳性猫的数量除以样本总数，以百分比报告并附95%置信区间。定量变量如猫的体重和年龄以均值和标准差报告。
使用卡方检验研究IgM阳性结果与定性变量之间是否存在任何关系，在样本量不满足卡方检验要求的情况下使用Fisher精确检验。还确定了比值比（OR）及其95%置信区间（95% CI）。
分类变量被分为三个主要部分进行分析：IgM状态与可能风险因素之间的关系；IgM状态与临床症状或分析变化（全血细胞计数和生化参数）之间的关系；临床结局与可能预后因素（给予的药物、并发疾病等）之间的关系。
首先对每组研究变量进行单变量分析，随后选择p值<0.2的变量以评估其在构建多变量分析模型中的价值。此类分析评估一组自变量对因变量（通常称为“结局”）的同时效应，允许评估因果相关性并估计其对响应变量的影响程度[72, 73]。
在结局变量以定量方式表达的情况下，本应使用线性回归模型，但在本研究中，鉴于因变量本质上是二元的（IgM阳性或阴性；临床结局阳性或阴性），因此构建了逻辑回归模型[72, 74]。
结果
通过检索实验室数据库并应用上述纳入和排除标准，共识别出134例接受血巴尔通体或支原体属血清学检测的可行病例。由于同时筛查其他媒介传播病原体（VBPs）的病例数量较少（5例埃立克体阳性、2例立克次体阳性、无利什曼原虫阳性），无法分析其与猫支原体病的关系。同样，由于记录病例数量有限而未分析的其他变量包括：住院期间接受输血的病例、并发心脏病的患者、并发肾脏病的患者以及进行过Coombs试验的患者。
研究动物的特征
所选猫的平均年龄为7.4 ± 5岁。猫的平均体重为4.3 ± 1.3 kg。在134只猫中，74只鉴定为雄性，60只为雌性。大多数猫此前已绝育（114/134，85%）。大部分猫被描述为欧洲短毛猫（59/134，44%）或混种猫（55/134，41%）。还鉴定出其他较少见的品种（20/134，15%），如波斯猫、挪威森林猫、英国短毛猫、苏格兰折耳猫等。
大部分猫被描述为室内猫（103/134，76.9%），其余被归类为可接触户外的猫。一部分猫与其他宠物同住家中（84/134，62.7%）。一些动物被描述为此前从街上救助的流浪猫（37/134，27.6%）。
关于疫苗接种，大多数猫的疫苗接种状态为最新（80/122，65.6%），但体内（81/103，78.6%）和体外（78/104，75%）寄生虫预防习惯不一致。在研究病例中，31例（23%）发现外寄生虫，其中跳蚤是最常见的发现。体格检查时，没有猫出现提示近期攻击的伤口或其他病变。
关于研究病例的临床结局，部分猫（30/134，22.4%）死亡或应主人要求实施安乐死，而大多数（104/134，77.6%）实现了临床康复。
观察到的患病率
在HEV接受血支原体检测的134只猫中，32只IgM结果阳性，102只IgM结果阴性，因此观察到的患病率为23.8%（95% CI：0.169–0.320）。
血支原体感染与风险因素之间关系的分析
风险因素的单变量和多变量分析结果分别总结于表3和表4。单变量分析发现FIV阳性动物与IgM阳性状态之间存在统计学显著关联（OR为13.3，p < 0.001）。多变量分析证实了血支原体感染与FIV阳性动物的关联（调整OR为12.768，p = 0.05），以及血支原体感染与老年年龄组之间的负关联（调整OR为0.333，p = 0.043）。其他分析变量未发现统计学显著差异。
表3. 血支原体感染与风险因素关系的单变量分析
表4. 血支原体感染与风险因素关系的多变量逻辑回归分析
血支原体感染与临床/实验室参数之间关系的分析
血液学、生化和临床变量的单变量分析结果分别总结于表5、表6和表7。即使对单变量分析中显示p < 0.2的变量进行多变量分析后，也未发现统计学显著差异（表8）。然而，观察到IgM阳性动物与某些变量之间存在显著关联，如红细胞分布宽度（调整OR为2.588，p = 0.119）和淋巴结肿大（调整OR为7.31，p = 0.126）。相反，观察到猫支原体病与低ALT病例（调整OR为0.209，p = 0.157）和低蛋白血症病例（调整OR为0.374，p = 0.243）之间存在负关系。
表5. 血支原体感染与血液学参数关系的单变量分析
表6. 血支原体感染与生化指标关系的单因素分析
表7. 血支原体感染与临床症状关系的单因素分析
表8. 猫支原体病与临床/实验室指标关系的多因素逻辑回归分析
血支原体感染病例预后因素与结局关系的分析
预后因素的单因素和多因素分析结果总结于表9和表10。多因素分析显示，良好预后与使用皮质类固醇之间存在负相关（校正OR为0.068，p = 0.05）。虽然无统计学显著性，但单因素分析显示临床转归良好的动物与使用多西环素之间存在正相关趋势（OR为3.71，p = 0.355），而肝脏疾病（OR为0.12，p = 0.105）和肿瘤性疾病（OR为0.154，p = 0.292）等变量与良好临床结局呈负相关。
表9. 预后因素与临床结局关系的单因素分析
表10. 预后因素与临床结局关系的多因素逻辑回归分析
讨论
本研究中观察到的患病率为23.8%，与葡萄牙其他研究中的患病率值相对相似，分别为20.4%[75]、22.1%[76]和27.1%[11]。研究样本包括多种病例，从常规术前分析中仅诊断出轻度贫血或血小板减少症的表面健康猫，到因严重贫血就诊并在临床检查中被诊断的临床患病猫。
对危险因素与猫支原体属感染关系的分析发现，与FIV阳性动物存在统计学显著关系。其他研究也发现了这一关系，并提出逆转录病毒的免疫抑制作用使其他病原体得以感染，或者观察到的这种关联是相似传播途径的结果[42, 77, 78]。
还观察到猫支原体属与老年年龄组之间的关联，老年猫似乎较少为IgM阳性病例。猫支原体病由于累积暴露风险在成年动物中往往更为流行，但也有研究认为年轻猫更可能受影响，因为它们往往表现出更明显的临床症状[53, 55, 79]。与老年猫的这种负相关可能是由于这些动物获得近期感染性疾病的倾向较低，因为与生活方式更活跃的年轻猫相比，它们较少外出或参与促进血支原体传播的攻击性互动，尽管不同结果可能受各研究年龄范围定义方式的影响[44, 45, 80]。
尽管其他研究认为性别、生殖状态、生活方式或猫的来源等“典型”危险因素具有相关性，但收集的数据未能确定感染与这些因素之间的关系[61, 71, 79, 81]。猫血支原体与体外寄生虫驱虫剂使用或外寄生虫存在之间缺乏相关性，表明城市地区猫之间这些病原体的传播不限于节肢动物媒介的活动[34, 44, 82]。
在媒介和移除媒介的猫宿主中同时检测到血支原体遗传物质，证实了它们通过吸血活动摄入支原体属的能力[5, 10, 66, 83, 84]，但并不意味着它们仍具有传播给后续脊椎动物宿主的活力[14, 34, 85]。然而，猫支原体病病例与同时携带外寄生虫或未接受充分体外驱虫的猫之间经常观察到的关系，可以作为考虑存在媒介传播途径的理由[10, 17, 81, 86]。
另一方面，也有猫支原体病病例记录于未感染外寄生虫或甚至无法接触户外的猫[87, 88]，尽管进行了多次尝试，媒介实验传播研究提供了不确定的结果[83, 89, 90]。尽管猫支原体病病例与感染外寄生虫的动物之间存在明显关联，但这种关系可能仅代表更多接触户外的猫群体，因此更多地暴露于节肢动物媒介或其他健康状况未知的猫[91]。此外，最近有证据表明犬血支原体在 isolated 犬群中传播，所有这些犬均接受了适当的体外驱虫，表明血支原体传播并非严格依赖于节肢动物媒介的存在和活动[92]。
对急性猫支原体感染对临床、血液学和生化变量影响的单变量分析未发现统计学显著关联。然而，多变量分析发现感染病例与红细胞大小不均或淋巴结病的存在之间有一些正向趋势，这些是非特异性疾病体征，其他作者也已发现[47, 48, 57]。尽管无统计学显著性，但观察到感染动物与低ALT或总蛋白降低之间存在负相关，这可能反映了急性猫支原体感染病例中报道的肝酶升高和高蛋白血症[56, 57]。
与其他研究类似，未检测到血支原体感染与贫血相关全血细胞计数异常之间的明显关联[9, 66, 79]。这种与临床和分析体征的微弱关系可以用每只猫的个体反应与感染物种固有特征之间的复杂动态来解释，这可能转化为非特异性和短暂性的有限症状[21, 61, 93]。这些稀疏结果的另一个可能原因可能是较高比例的猫感染了致病潜力较低的物种，如“Ca. M. haemominutum”和“Ca. M. turicensis”，尽管由于分析的回顾性性质无法区分这三个物种[11–14]。
一些动物表现出有效反应，红细胞压积升高并自发清除病原体，而其他猫需要治疗干预以防止进展为致命性溶血性贫血，伴代谢性酸中毒和心血管衰竭[45, 66, 91, 94–96]。多西环素或马波沙星等抗生素已被证明可有效降低血支原体载量并有助于加速恢复，但药物治疗不能可靠地完全清除病原体[46, 69, 97]。
对能够影响临床结局的因素分析仅在使用皮质类固醇时具有统计学显著性，但观察到使用多西环素与恢复可能性之间存在明显正向趋势。同时，临床恢复与伴随疾病（肝胆或肿瘤性质）的存在呈负向趋势。有利预后与使用皮质类固醇之间的负相关关系可由存在更复杂的溶血性贫血病例来解释，这些病例对 antibiotic 给药无反应，导致在治疗方案中加入糖皮质激素，尽管其在支原体病病例中的有效性仍有争议[61, 98]。
阳性临床结局与肝胆或肿瘤疾病的负相关也是预期结果，因为伴随疾病的存在通常被认为是负性预后因素[45, 61, 65]。另一方面，多西环素的使用与有利的临床转归显示出明显趋势，证明了其有效性并确认了其作为猫支原体病初始治疗首选的地位[5, 47, 69]。
这项预后因素分析仅关注被认为需要治疗的IgM阳性病例，但应注意有一部分IgM阴性但IgG阳性的患者也在诊断后接受了抗生素治疗。鉴于这些病例伴有其他显著的临床状况，可能对研究的预后因素构成混杂效应，这些“慢性”病例未被纳入，以尝试进行聚焦于急性支原体病结局的分析。
本研究受到一些局限性阻碍，主要困难是缺乏关于影响每位患者的支原体物种的信息，以及无法区分共感染病例与单一感染病例。另一个困难是缺乏标准化，源于使用不同实验室进行的分析，这些实验室除设备不同外还使用不同的参考值。由于研究的回顾性性质，在收集临床病例时，患者记录已提供病例的进展和结局，但记录中的空白导致数据缺失，如品种、疫苗接种状态、体内外寄生虫预防频率和所用药物、猫的来源等。
结论
本研究结果证实，在里斯本一家兽医医院接受治疗的猫科患者中存在嗜血支原体属。这些病原体在表面上健康的慢性携带者以及明显贫血的猫中均有存在，这证实了这些病原体复杂的临床和流行病学动态。血清学或分子生物学阳性结果并不总是意味着活动性疾病，因此必须结合患者的全身状况、临床症状、血液学和/或生化异常，以及可能触发疾病的同时存在或未来出现的因素进行综合评估。引起猫支原体病的细菌存在于猫群中并通过尚未明确界定的传播途径循环，且给予适当的抗生素疗程并不能保证其清除。因此，鉴于对其人畜共患潜力的认识尚不充分，对于与人类密切接触的动物，应维持对这些病原体的充分监测。
查看英文原文▾
Introduction
Haemotropic Mycoplasma spp. (also known as haemoplasmas) are bacteria that exhibit tropism for vertebrate host erythrocytes, with the potential to cause hemolytic anemia in affected hosts [1–3]. These bacteria belong to the Mycoplasma genus, with three species recognized worldwide as agents of feline infectious anemia: Mycoplasma haemofelis, “Candidatus Mycoplasma haemominutum”, and “Candidatus Mycoplasma turicensis” [4, 5].
Despite some geographical variations, the agents of feline mycoplasmosis have a worldwide distribution (Table 1), with records of their detection in domestic, stray, and wild cats [6–10]. Most studies report a higher prevalence of “Candidatus Mycoplasma haemominutum” (“Ca. M. haemominutum”), followed by Mycoplasma haemofelis (M. haemofelis), and in third place “Candidatus Mycoplasma turicensis” (“Ca. M. turicensis”) [11–14].
Table 1. International prevalence studies of feline mycoplasmosis
| Country | Year | n | M. haemofelis | “ Ca . M. haemominutum” | “ Ca . M. turicensis” | Mycoplasma spp. |
|---|---|---|---|---|---|---|
| Iran [ 15 ] | 2024 | 361 | 2.2% | 10.5% | 0.6% | 15.5% |
| Spain [ 16 ] | 2023 | 332 | 10.5% | 18.4% | 3.3% | 22.9% |
| China [ 17 ] | 2021 | 668 | 0.9% | 3.4% | 1.2% | 4.9% |
| Thailand [ 18 ] | 2020 | 231 | 3.9% | 22.9% | 0.86% | 27.7% |
| Italy [ 13 ] | 2020 | 958 | 1.5% | 9.9% | 0.2% | 11.6% |
| Brazil [ 19 ] | 2018 | 200 | 12% | 12.5% | 3% | 35.5% |
| Cyprus [ 20 ] | 2017 | 174 | 7.5% | 20.7% | 6.9% | 26.4% |
| Chile [ 21 ] | 2016 | 384 | 4.4% | 7.8% | 2% | 15.1% |
| Turkey [ 22 ] | 2016 | 384 | 9.9% | 17.7% | 0.8% | 19.3% |
| Portugal [ 11 ] | 2015 | 236 | 14.4% | 17.8% | 5.9% | 27.1% |
| Albania [ 23 ] | 2014 | 146 | 10.3% | 21.9% | 5.5% | 30.8% |
| South Africa [ 24 ] | 2012 | 102 | 3.9% | 21.6% | 0% | 25.5% |
| Japan [ 25 ] | 2010 | 1770 | 2.4% | 15.8% | 2.7% | 26.4% |
| South Korea [ 26 ] | 2007 | 331 | 4.2% | 10.3% | 0% | 19.9% |
Coinfection by more than one agent in the same host is frequent, due to these agents having similar risk factors and transmission routes, with M. haemofelis paired with “Ca. M. haemominutum” being the most common combination [14, 27–30].
These agents are also documented in many species of wild felids, whether free-ranging or kept in captivity (lions, tigers, cheetahs, Iberian lynxes, leopards, jaguars, cougars, etc.), with “Ca. M. haemominutum” being the most prevalent species and records of mixed infections, similar to domestic cats [31–35].
A higher prevalence of infection is usually identified in male cats with access to the outdoors, with their ambulatory and territorial behavior being pointed out as a possible cause for the increased likelihood of aggressive interactions with other cats and greater exposure to arthropod vectors [16, 36–39]. It has also been noted on occasions that animals identified as positive for feline immunodeficiency virus (FIV) or feline leukemia virus (FeLV) have a higher risk of co-infection with feline Mycoplasma spp., with special emphasis on FIV, which is also strongly associated with transmission through aggressive interactions [7, 40–44].
The most commonly reported clinical signs are usually nonspecific such as lethargy and anorexia, but also include an assortment of signs that reflect the severity of the underlying anemia, such as pale mucous membranes, tachycardia, splenomegaly, jaundice, heart murmurs on auscultation, among others [8, 45–47]. Occasionally, no obvious relationship between feline mycoplasmosis and anemia is found due to the existence of completely asymptomatic chronic carriers [48–50]. For patients who do develop acute anemia, it is usually regenerative (macrocytic and hypochromic), and may be accompanied by anisocytosis, reticulocytosis, and polychromasia [21, 48, 51, 52]. Platelet counts may be normal or slightly subnormal [53, 54]. There may also be increases in liver enzymes due to hepatic hypoxia, hyperbilirubinemia due to hemolysis, some increase in total protein, or albumin/globulin ratio shifts due to polyclonal gammopathy [14, 55–57].
Cytological diagnosis of feline haemoplasmas is an accessible but low sensitivity method, as it usually requires exuberant bacteremia to be visualized by experienced operators and does not allow differentiation of the three species due to their morphological similarities [5, 42, 53, 58]. In comparison, polymerase chain reaction (PCR) has demonstrated high sensitivity and specificity for detecting fragments of the 16 S rRNA gene of Mycoplasma spp. in different types of samples such as saliva, blood, urine, feces, among others [5, 59–61]. However, the amount of haemoplasma determined does not directly correlate with the animal’s clinical status, and there are cases that remain persistently seropositive despite being negative by PCR, so there’s a presumed persistence of haemoplasmas at levels below the PCR detection limit in these situations [62–66]. Alternatively, serological techniques are a practical and convenient diagnostic method, thanks to the existence of specific antibodies whose titers reflect the progression of the infection, with IgM titers allowing the distinction between acute and chronic infections [61, 67–69].
The aim of this study was to evaluate the prevalence of feline Mycoplasma spp. diagnosed in patients at the HEV (Veterinary Teaching Hospital), as well as to compare negative and recently infected cats in terms of risk factors, clinical signs and laboratory parameters. Factors likely to influence the prognosis of positive cases were also analyzed.
Materials and methods
Study design and inclusion criteria
An analytical cross-sectional study was devised with the target population set as HEV feline patients who underwent feline mycoplasmosis screening, between January 1st, 2021, and June 30th, 2024.
The criterion used to classify a case as an acute feline mycoplasmosis infection was the presence of a positive Immunoglobulin M (IgM) result. Cases diagnosed by PCR were not included due to a shortage of cases screened by this method during the study period. Patients who only had a positive Immunoglobulin G (IgG) result, but no IgM information were also excluded [56, 68, 70].
Not all cases had a complete blood count and/or extensive biochemical analyses available at the time of haemoplasma infection diagnosis, and some clinical records did not allow for the complete collection of information such as vaccination status, FIV/FeLV status, frequency of internal and external parasite prophylaxis, among others. The data obtained were analyzed with the awareness that there is missing information.
Variables of interest
Based on previous studies that dwelled onto the epidemiological aspects of feline haemotropic Mycoplasma infection [5, 39, 44, 71], namely the possibility of horizontal and/or vector borne transmission, extra care was taken when collecting anamnesis information regarding each cat’s lifestyle to evaluate their epidemiological risk. Profile information was collected regarding their origins (adopted from breeder or street rescue), whether they lived in multicat households, outdoor access (to assess exposure to vectors or interspecies aggression), fastidiousness of vaccination and external parasiticide application, FIV/FeLV status, as well as any other useful information that could be obtained from their medical records. When applicable, any pre-existing conditions such as concomitant co-morbidities or co-infections were logged to assess their statistical usefulness later.
Another point of interest was to inventory clinical and analytical findings during the diagnostic work-up so any available information regarding clinical signs (such as depression, pale mucous membranes, tachycardia/tachypnea, jaundice, fever, etc.) and routine complete blood count and biochemistry panels performed around the time of Mycoplasma spp. screening were also recorded for analysis. Clinical and analytical data that were chronologically too distant from the moment of Mycoplasma spp. screening were considered unfit for statistical analysis.
Finally, study patients’ clinical outcomes and therapy details were also documented, such as cases that actually started treatment (with antibiotics, corticosteroids, blood transfusions, etc.) while other cats were euthanized. These treatment details, along with any logged co-morbidities and co-infections, were thought to be an interesting aspect to explore in an attempt to identify any visible trend with the clinical outcome of the case.
Medical record search
Patient information was collected by retrieving their clinical records and organized into a database containing their profile (sex, age, environment, lifestyle, vaccination status, internal and external parasite prophylaxis, medical and surgical history). After verifying the complete clinical history of each patient, blood count values and biochemical analyses were collected for the period in which clinical suspicion of feline mycoplasmosis arose. Part of the patients had complete blood count and biochemistry analyses requested simultaneously with etiological diagnosis, but some cases underwent haemoplasma screening with a slight delay, in which cases a maximum difference of 72 h was accepted. In addition to blood count information, other data were collected such as blood urea nitrogen (BUN), creatinine, aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), total protein, albumin, globulins, albumin/globulin ratio, and total bilirubin.
Relevant clinical information was also extracted along the timeline of feline mycoplasmosis screening, such as clinical signs observed during consultation (fever, pale mucous membranes, tachypnea/tachycardia, etc.), administered treatment, and outcome of each case.
Follow-up information was scoured until the most recent entries for each patient, with some patients returning for reevaluation while others had no further entries after medical discharge.
Data extracted
The data collected from clinical records were organized and sorted into a database, summarized in Table 2.
Table 2. Organization and classification of collected variables
| Variable | Classification | Reference range |
|---|---|---|
| IgM indirect immunofluorescence | Positive or negative result | |
| IgG indirect immunofluorescence | Positive or negative result | |
| Sex | Male or female | |
| Reproductive status | Intact or neutered | |
| Season of the year (Time of the year when feline mycoplasmosis screening was performed) | Winter | December 21st to March 20th |
| Spring | March 21st to June 20th | |
| Summer | June 21st to September 22nd | |
| Autumn | September 23rd to December 20th | |
| Breed | Pure or mixed breed | |
| Age range | Junior | < 1 year old |
| Adult | 1–10 years old | |
| Senior | > 10 years old | |
| Age | Expressed as years | |
| Other Vector Borne Pathogens (VBPs) | Presence or absence of other diagnosed VBPs ( Ehrlichia canis , Rickettsia conorii , Leishmania infantum , etc.) | |
| Ectoparasites | If ectoparasites such as fleas and/or ticks were found during the physical examination | |
| Internal parasiticide | Regular or inconsistent use of internal parasiticides | |
| External parasiticide | Regular or inconsistent use of external parasiticides | |
| FIV status | Positive or negative for feline immunodeficiency retrovirus | |
| FeLV status | Positive or negative for feline leukemia retrovirus | |
| Vaccination status | Up-to-date vaccinations or if vaccinations are overdue | |
| Living environment | With or without outdoor access | |
| Rescued stray | Sorted based on whether they were previously stray cats rescued from the street or not | |
| Number of cohabitants | Whether it’s the only pet at home or a household with multiple pets | |
| Body weight | Body weight in kg | |
| Hematocrit | Presence or absence of anemia | 24–45% |
| Reticulocytes | Classification of anemia as regenerative or non-regenerative | 3000–50,000/µL; considered regenerative if > 42,000/µL |
| Mean Corpuscular Volume (MCV) | Classification of anemia as macrocytic, normocytic, or microcytic | 39–52 fL |
| Mean Corpuscular Hemoglobin (MCH) | Classification of anemia as normochromic or hypochromic | 12.5–17.5 pg |
| Red Cell Distribution Width (RDW) | Presence or absence of anisocytosis | 17–23% |
| Leukocytes | Leukocytosis, leucopenia or normal count | 5.5–17.02 × 10³/mm³ |
| Lymphocytes | Lymphocytosis, lymphopenia, or normal count | 1.5–7 × 10³/mm³ |
| Monocytes | Monocytosis or normal count | 0–1400/mm³ |
| Neutrophils | Neutrophilia, neutropenia or normal count | 2.5–12.8 × 10³/mm³ |
| Eosinophils | Eosinophilia or normal count | 0–1.5 × 10³/mm³ |
| Platelets | Thrombocytopenia or normal count | 150–500 × 10³/mm³ |
| BUN | High values or values within the normal range | 30–65 mg/dL |
| Creatinine | High values or values within the normal range | 1.1–2.3 mg/dL |
| AST | High values or values within the normal range | 8–35 U/L |
| ALT | Low, high or values within the normal range | 33–119 U/L |
| ALP | High values or values within the normal range | 7–68 U/L |
| Total protein | Hyperproteinemia, hypoproteinemia or normal values | 6.4–8.5 g/dL |
| Albumin | Hypoalbuminemia or normal values | 3.1–4.3 g/dL |
| Globulins | Abnormal values or within the reference range | 2.8–5.1 g/dL |
| Albumin/globulin ratio | Abnormal values or within the reference range | 0.45–1.3 |
| Total bilirubin | Abnormal values or within the reference range | 0–0.4 mg/dL |
| Coagulation tests | Abnormal values or within the reference range | Prothrombin time: 7–10 s Thrombin time: 11.6–15.7 s Partial thromboplastin time: 8.6–12.9 s |
| Coombs test | Positive or negative result | |
| Clinical signs | Parameters recorded during physical examination such as prostration, hyporexia, fever, hypothermia, pale mucous membranes, signs of jaundice, abnormalities detected by cardiopulmonary auscultation, evident tachypnea or tachycardia, etc. | |
| Non-vectorial diseases | Variable that includes other concomitant diseases/medical-surgical history that were not the reason for feline Mycoplasma screening (dermatological, renal, hepatobiliary, oncological disease, etc.) | |
| Treatment | If targeted antibiotic therapy was needed for feline mycoplasmosis | |
| Doxycycline | If there was a need to institute doxycycline therapy | |
| Marbofloxacin | If there was a need to institute therapy with marbofloxacin | |
| Corticosteroids | If there was need to combine corticosteroid therapy with antibiotic therapy | |
| Blood transfusion | If transfusion with packed red blood cells or fresh frozen plasma was necessary | |
| Outcome | Whether the clinical case was solved or the patient died/was euthanized during the course of feline mycoplasmosis | |
Data analysis
Statistical analysis of the collected data was performed using Microsoft Office Excel (version 2409), Jamovi (version 2.6.44.0), and IBM SPSS Statistics® (version 29). The established statistical significance value was p ≤ 0.05 for all tests performed.
Categorical variables were sorted into appropriate groups (with or without outdoor access, IgM positive or negative, with or without jaundice, etc.) while quantitative variables such as complete blood count and biochemistry values were classified as below, above or within physiological parameters according to the outsourced laboratory’s set reference ranges.
To calculate the prevalence, the number of IgM-positive cats was divided by the total number of the sample, reported as a percentage with a 95% confidence interval. Quantitative variables such as weight and age of the cats were reported as mean and standard deviation.
The existence of any relationship between a positive IgM result and qualitative variables was studied using the Chi-Square test, or Fisher’s exact test in cases where sample size does not meet the requirements of the Chi-Square test. Odds Ratio (OR) with a 95% confidence interval (95% CI) was also determined.
Categorical variables were divided into three main sections for analysis: relationship between IgM status and possible risk factors; relationship between IgM status and clinical signs or analytical changes (complete blood count and biochemistry parameters); relationship between clinical outcome and possible prognostic factors (administered medication, concomitant diseases, etc.).
A first approach univariate analysis was performed on each group of variables under study, and variables that demonstrated a p-value < 0.2 were subsequently selected to assess their value in the construction of multivariate analysis models. This type of analysis assesses the simultaneous effect of a set of independent variables on a dependent variable (often designated as the “outcome”), allowing the assessment of causal correlations and estimating their degree of effect on the response variable [72, 73].
In situations where the outcome variable is expressed quantitatively, a linear regression model would have been used, but in the present study, given that the dependent variable is binary in nature (IgM positive or negative; positive or negative clinical outcome), logistic regression models were constructed [72, 74].
Results
By searching the laboratory database and applying the aforementioned inclusion and exclusion criteria, it was possible to identify 134 viable cases that underwent serology testing for Haemobartonella or Mycoplasma spp. Due to the reduced number of cases that were simultaneously screened for other vector-borne pathogens (VBPs) (5 positive for Ehrlichia, 2 positive for Rickettsia, and none positive for Leishmania), it was not possible to analyze their relationship with feline mycoplasmosis. Similarly, other variables that were not analyzed due to the limited number of recorded cases included: cases that received blood transfusion during their hospitalization, patients with concurrent cardiac disease, patients with concurrent kidney disease, and patients that had a Coombs test performed.
Characterization of the study animals
The average age of the selected cats was 7.4 ± 5 years. The average weight of the cats was 4.3 ± 1.3 kg. Of the 134 cats, 74 were identified as male and 60 as female. Most cats had been previously neutered (114/134, 85%). A large proportion of the cats were described as European Shorthair (59/134, 44%) or mixed breed (55/134, 41%). Other less frequent breeds (20/134, 15%) were also identified, such as Persian, Norwegian Forest Cats, British Shorthair, Scottish Fold, etc.
A large proportion of the cats were described as indoor cats (103/134, 76.9%) while the remainder were classified as cats with access to the outdoors. A portion of the cats lived with other pets at home (84/134, 62.7%). Some of the animals were described as previous strays rescued from the street (37/134, 27.6%).
Regarding vaccination, most cats had up to date vaccination status (80/122, 65.6%), but inconsistent internal (81/103, 78.6%) and external (78/104, 75%) parasite prophylaxis habits. Ectoparasites were found in 31 (23%) of the cases studied, with fleas being the most frequent finding. Upon physical examination, none of the cats presented wounds or other lesions that would suggest recent aggression.
Regarding clinical outcome of the study cases, some of the cats (30/134, 22.4%) passed away or were euthanized upon owner request while majority (104/134, 77.6%) of them achieved clinical recovery.
Observed prevalence
Of the 134 cats tested for haemoplasma at the HEV, 32 had positive IgM results, while 102 had negative IgM results, therefore the observed prevalence was 23.8% (95% CI: 0.169–0.320).
Analysis of the relationship between haemoplasma infection and risk factors
The results of univariate and multivariate analysis of risk factors were summarized in Tables 3 and 4, respectively. Univariate analysis identified a statistically significant association between FIV-positive animals and positive IgM status (OR of 13.3, p < 0.001). Multivariate analysis confirmed this association of haemoplasma infection with FIV-positive animals (adjusted OR of 12.768, p = 0.05), as well as a negative association between haemoplasma infection and the senior age group (adjusted OR of 0.333, p = 0.043). No statistically significant differences were identified for other analyzed variables.
Table 3. Univariate analysis of the relationship between haemoplasma infection and risk factors
| Risk factor | n | IgM negative (% row) | IgM positive (% row) | OR | 95% CI | p |
|---|---|---|---|---|---|---|
| Sex | 134 | |||||
| Male | 74 | 57 (77) | 17 (23) | 1.118 | 0.504–2.479 | 0.784 |
| Female | 60 | 45 (75) | 15 (25) | |||
| Reproductive status | 134 | |||||
| Neutered | 114 | 88 (77.2) | 26 (22.8) | 1.45 | 0.507–4.15 | 0.486 |
| Intact | 20 | 14 (70) | 6 (30) | |||
| Season | 134 | |||||
| Winter | 33 | 23 (69.7) | 10 (30.3) | 0.754 | ||
| Spring | 42 | 33 (78.6) | 9 (21.4) | |||
| Summer | 35 | 28 (80) | 7 (20) | |||
| Autumn | 24 | 18 (75) | 6 (25) | |||
| Breed | 134 | |||||
| Pure breed | 81 | 63 (77.8) | 18 (22.2) | 1.256 | 0.562–2.809 | 0.578 |
| Mixed breed | 53 | 39 (73.6) | 14 (26.4) | |||
| Age range (F) | 134 | |||||
| Junior | 7 | 6 (85.7) | 1 (14.3) | 0.14* | ||
| Adult | 80 | 56 (70) | 24 (30) | |||
| Senior | 47 | 40 (85.1) | 7 (14.9) | |||
| Ectoparasites | 134 | |||||
| No | 103 | 76 (73.8) | 27 (26.2) | 0.541 | 0.189–1.55 | 0.248 |
| Yes | 31 | 26 (83.9) | 5 (16.1) | |||
| Internal parasiticide (F) | 103 | |||||
| Regular | 22 | 18 (81.8) | 4 (18.2) | 1.379 | 0.416–4.575 | 0.775 |
| Inconsistent | 81 | 62 (76.5) | 19 (23.5) | |||
| External parasiticide | 104 | |||||
| Regular | 26 | 19 (73.1) | 7 (26.9) | 0.756 | 0.273–2.097 | 0.591 |
| Inconsistent | 78 | 61 (78.2) | 17 (21.8) | |||
| FIV (F) | 124 | |||||
| Negative | 115 | 91 (79.1) | 24 (20.9) | 13.3 | 2.59–68 | < 0.001* |
| Positive | 9 | 2 (22.2) | 7 (77.8) | |||
| FeLV (F) | 124 | |||||
| Negative | 118 | 88 (74.6) | 30 (25.4) | 0.587 | 0.066–5.22 | 1 |
| Positive | 6 | 5 (83.3) | 1 (16.7) | |||
| Vaccination | 122 | |||||
| Regular | 80 | 63 (78.8) | 17 (21.3) | 1.158 | 0.479–2.819 | 0.746 |
| Inconsistent | 42 | 32 (76.2) | 10 (23.8) | |||
| Living environment | 134 | |||||
| Indoor | 103 | 80 (77.7) | 23 (22.3) | 1.423 | 0.576–3.512 | 0.443 |
| Outdoor | 31 | 22 (71) | 9 (29) | |||
| Rescued stray | 134 | |||||
| No | 97 | 77 (79.4) | 20 (20.6) | 1.85 | 0.793–4.31 | 0.152* |
| Yes | 37 | 25 (67.6) | 12 (32.4) | |||
| Cohabitants | 134 | |||||
| Sole pet | 50 | 36 (72) | 14 (28) | 0.701 | 0.313–1.573 | 0.388 |
| More pets | 84 | 66 (78.6) | 18 (21.4) |
Table 4. Multivariate logistic regression analysis of the relationship between haemoplasma infection and risk factors
| Risk factor | OR | 95% CI | p |
|---|---|---|---|
| Age range | |||
| Junior - Adult | 0.256 | 0.021–3.12 | 0.285 |
| Adult - Senior | 0.333 | 0.115–0.964 | 0.043 |
| FIV | 12.768 | 2.167–75.225 | 0.005 |
| Rescued stray | 1.519 | 0.567–4.067 | 0.406 |
Analysis of the relationship between haemoplasma infection and clinical/laboratory parameters
The results of univariate analysis of hematological, biochemical, and clinical variables are summarized in Tables 5, 6, and 7, respectively. No statistically significant differences were identified even after multivariate analysis of variables that showed p < 0.2 in the univariate analysis (Table 8). However, a prominent association was observed between IgM-positive animals and certain variables, such as red blood cell distribution width (adjusted OR of 2.588, p = 0.119) and lymphadenopathy (adjusted OR of 7.31, p = 0.126). Conversely, a negative relationship was observed between feline mycoplasmosis and cases with low ALT (adjusted OR of 0.209, p = 0.157) and cases of hypoproteinemia (adjusted OR of 0.374, p = 0.243).
Table 5. Univariate analysis of the relationship between haemoplasma infection and hematological parameters
| Hematological parameter | n | IgM negative (% row) | IgM positive (% row) | OR | 95% CI | p |
|---|---|---|---|---|---|---|
| Hematocrit | 130 | |||||
| Normal | 74 | 57 (77) | 17 (23) | 1.01 | 0.445–2.31 | 0.974 |
| Anemia | 56 | 43 (76.8) | 13 (23.2) | |||
| Reticulocytes (F) | 74 | |||||
| Regenerative | 10 | 8 (80) | 2 (20) | 1.818 | 0.354–9.346 | 0.713 |
| Non-regenerative | 64 | 44 (68.8) | 20 (31.3) | |||
| MCV (F) | 130 | |||||
| Normocytic | 98 | 78 (79.6) | 20 (20.4) | 0.386 | ||
| Macrocytic | 14 | 10 (71.4) | 4 (28.6) | |||
| Microcytic | 18 | 12 (66.7) | 6 (33.3) | |||
| MCH (F) | 130 | |||||
| Normochromic | 119 | 92 (77.3) | 27 (22.7) | 1.278 | 0.317–5.153 | 0.715 |
| Hypochromic | 11 | 8 (72.7) | 3 (27.3) | |||
| RDW | 116 | |||||
| Normal | 57 | 48 (84.2) | 9 (15.8) | 2.159 | 0.871–5.352 | 0.093* |
| Anisocytosis | 59 | 42 (71.2) | 17 (28.8) | |||
| Leukocytes | 130 | |||||
| Normal | 82 | 66 (80.5) | 16 (19.5) | 0.227 | ||
| Leukocytosis | 23 | 18 (78.3) | 5 (21.7) | |||
| Leukopenia | 25 | 16 (64) | 9 (36) | |||
| Lymphocytes (F) | 130 | |||||
| Normal | 61 | 50 (82) | 11 (18) | 0.368 | ||
| Lymphocytosis | 5 | 4 (80) | 1 (20) | |||
| Lymphopenia | 64 | 46 (71.9) | 18 (28.1) | |||
| Monocytes (F) | 130 | |||||
| Normal | 123 | 96 (78) | 27 (22) | 2.667 | 0.562–12.648 | 0.2* |
| Monocytosis | 7 | 4 (57.1) | 3 (42.9) | |||
| Neutrophils (F) | 130 | |||||
| Normal | 90 | 70 (77.8) | 20 (22.2) | 0.551 | ||
| Neutrophilia | 25 | 20 (80) | 5 (20) | |||
| Neutropenia | 15 | 10 (66.7) | 5 (33.3) | |||
| Eosinophils (F) | 130 | |||||
| Normal | 128 | 99 (77.3) | 29 (22.7) | 3.414 | 0.207–56.281 | 0.41 |
| Eosinophilia | 2 | 1 (50) | 1 (50) | |||
| Platelets | 130 | |||||
| Normal | 99 | 76 (76.8) | 23 (23.2) | 0.964 | 0.368–2.52 | 0.94 |
| Thrombocytopenia | 31 | 24 (77.4) | 7 (22.6) |
Table 6. Univariate analysis of the relationship between haemoplasma infection and biochemical parameters
| Biochemical parameter | n | IgM negative (% row) | IgM positive (% row) | OR | 95% CI | p |
|---|---|---|---|---|---|---|
| BUN | 121 | |||||
| Normal | 83 | 61 (73.5) | 22 (26.5) | 0.52 | 0.191–1.412 | 0.195* |
| High | 38 | 32 (84.2) | 6 (15.8) | |||
| Creatinine (F) | 124 | |||||
| Normal | 106 | 81 (76.4) | 25 (23.6) | 0.926 | 0.279–3.068 | 1 |
| High | 18 | 14 (77.8) | 4 (22.2) | |||
| AST (F) | 22 | |||||
| Normal | 12 | 11 (91.7) | 1 (8.3) | 4.714 | 0.405–54.826 | 0.293 |
| High | 10 | 7 (70) | 3 (30) | |||
| ALT (F) | 121 | |||||
| Normal | 85 | 62 (72.9) | 23 (27.1) | 0.084* | ||
| Low | 20 | 19 (95) | 1 (5) | |||
| High | 16 | 13 (81.3) | 3 (18.8) | |||
| ALP | 91 | |||||
| Normal | 65 | 50 (76.9) | 15 (23.1) | 1.481 | 0.538–4.080 | 0.446 |
| High | 26 | 18 (69.2) | 8 (30.8) | |||
| Total protein | 99 | |||||
| Normal | 67 | 48 (71.6) | 19 (28.4) | 0.169* | ||
| Hypoproteinemia | 26 | 23 (88.5) | 3 (11.5) | |||
| Hyperproteinemia | 6 | 4 (66.7) | 2 (33.3) | |||
| Albumin | 117 | |||||
| Normal | 78 | 62 (79.5) | 16 (20.5) | 1.722 | 0.718–4.129 | 0.22 |
| Hypoalbuminemia | 39 | 27 (69.2) | 12 (30.8) | |||
| Globulins | 95 | |||||
| Normal | 74 | 55 (74.3) | 19 (25.7) | 0.905 | 0.292–2.805 | 0.862 |
| Abnormal | 21 | 16 (76.2) | 5 (23.8) | |||
| Albumin/globulin ratio (F) | 95 | |||||
| Normal | 86 | 62 (72.1) | 24 (27.9) | 0.721 | 0.632–0.822 | 0.106* |
| Abnormal | 9 | 9 (100) | 0 (0) | |||
| Bilirubin (F) | 42 | |||||
| Normal | 20 | 15 (75) | 5 (25) | 0.882 | 0.213–3.653 | 1 |
| Abnormal | 22 | 17 (77.3) | 5 (22.7) | |||
| Coagulation tests (F) | 10 | |||||
| Normal | 6 | 5 (83.3) | 1 (16.7) | 0.407 | 0.013–12.6 | 1 |
| Abnormal | 4 | 4 (100) | 0 (0) |
Table 7. Univariate analysis of the relationship between haemoplasma infection and clinical signs
| Clinical signs | n | IgM negative (% row) | IgM positive (% row) | OR | 95% CI | p |
|---|---|---|---|---|---|---|
| Prostration | 134 | |||||
| No | 61 | 48 (77.4) | 14 (22.6) | 1.14 | 0.514–2.54 | 0.743 |
| Yes | 72 | 54 (75) | 18 (25) | |||
| Mucous membranes | 134 | |||||
| Normal | 89 | 65 (73) | 24 (27) | 0.586 | 0.239–1.43 | 0.239 |
| Pale | 45 | 37 (82.2) | 8 (17.8) | |||
| Jaundice | 134 | |||||
| No | 111 | 84 (75.7) | 27 (24.3) | 0.864 | 0.293–2.55 | 0.791 |
| Yes | 23 | 18 (78.3) | 5 (21.7) | |||
| Weight loss (F) | 134 | |||||
| No | 116 | 89 (76.7) | 27 (23.3) | 1.27 | 0.415–3.88 | 0.767 |
| Yes | 18 | 13 (72.2) | 5 (27.8) | |||
| Fever | 134 | |||||
| No | 98 | 75 (76.5) | 23 (23.5) | 1.09 | 0.448–2.64 | 0.854 |
| Yes | 36 | 27 (75) | 9 (25) | |||
| Hypothermia (F) | 134 | |||||
| No | 125 | 95 (76) | 30 (24) | 0.905 | 0.178–4.59 | 1 |
| Yes | 9 | 7 (77.8) | 2 (22.2) | |||
| Cardiopulmonary auscultation | 134 | |||||
| Normal | 90 | 69 (76.7) | 21 (23.3) | 1.1 | 0.473–2.53 | 0.832 |
| Murmurs | 44 | 33 (75) | 11 (25) | |||
| Hyporexia | 134 | |||||
| No | 52 | 41 (78.8) | 11 (21.2) | 1.28 | 0.560–2.94 | 0.555 |
| Yes | 82 | 61 (74.4) | 21 (25.6) | |||
| Tachycardia (F) | 134 | |||||
| No | 121 | 94 (77.7) | 27 (22.3) | 2.18 | 0.658–7.2 | 0.301 |
| Yes | 13 | 8 (61.5) | 5 (38.5) | |||
| Tachypnea (F) | 134 | |||||
| No | 121 | 90 (74.4) | 31 (25.6) | 0.242 | 0.03–1.94 | 0.189* |
| Yes | 13 | 12 (92.3) | 1 (7.7) | |||
| Lymphadenopathy (F) | 134 | |||||
| No | 127 | 99 (78) | 28 (22) | 4.71 | 0.996–22.3 | 0.056* |
| Yes | 7 | 3 (42.9) | 4 (57.1) |
Table 8. Multivariate logistic regression analysis of the relationship between feline mycoplasmosis and clinical/laboratory parameters
| Variable | OR | 95% CI | p |
|---|---|---|---|
| RDW | 2.588 | 0.783–8.557 | 0.119 |
| Monocytes | 0.916 | 0.111–7.562 | 0.935 |
| BUN | 0.85 | 0.244–2.967 | 0.799 |
| ALT | |||
| Low - Normal | 0.209 | 0.024–1.826 | 0.157 |
| High - Normal | 0.809 | 0.168–3.892 | 0.792 |
| Total protein | |||
| Hypoproteinemia - Normal | 0.374 | 0.072–1.948 | 0.243 |
| Hyperproteinemia - Normal | 0.889 | 0.074–10.758 | 0.926 |
| Tachypnea | 0.4 | 0.041–3.912 | 0.431 |
| Lymphadenopathy | 7.31 | 0.572–93.47 | 0.126 |
Analysis of the relationship between prognostic factors and outcome of haemoplasma infection cases
The results of univariate and multivariate analysis of prognostic factors were summarized in Tables 9 and 10. Multivariate analysis revealed a negative correlation between a favorable prognosis and the use of corticosteroids (adjusted OR of 0.068, p = 0.05). Although not statistically significant, the univariate analysis showed a positive trend between animals with favorable clinical resolution and the use of doxycycline (OR of 3.71, p = 0.355), while variables such as the presence of hepatic disease (OR of 0.12, p = 0.105) and oncologic disease (OR of 0.154, p = 0.292) showed an inverse relationship with a positive clinical outcome.
Table 9. Univariate analysis of the relationship between prognostic factors and clinical outcome
| Prognostic factors | n | Euthanasia/death (% row) | Recovered (% row) | OR | 95% CI | p |
|---|---|---|---|---|---|---|
| Treatment (F) | 32 | |||||
| No | 4 | 1 (25) | 3 (75) | 2 | 0.164–24.3 | 0.512 |
| Yes | 28 | 4 (14.3) | 24 (85.7) | |||
| Doxycycline (F) | 32 | |||||
| No | 18 | 4 (22.2) | 14 (77.8) | 3.71 | 0.366–37.7 | 0.355 |
| Yes | 14 | 1 (7.1) | 13 (92.9) | |||
| Marbofloxacin (F) | 32 | |||||
| No | 16 | 1 (6.3) | 15 (93.8) | 0.2 | 0.02–2.03 | 0.333 |
| Yes | 16 | 4 (25) | 12 (75) | |||
| Corticosteroids (F) | 32 | |||||
| No | 27 | 3 (11.1) | 24 (88.9) | 0.188 | 0.022–1.62 | 0.163* |
| Yes | 5 | 2 (40) | 3 (60) | |||
| Dermatologic (F) | 32 | |||||
| No | 29 | 4 (13.8) | 25 (86.2) | 0.32 | 0.023–4.41 | 0.41 |
| Yes | 3 | 1 (33.3) | 2 (66.7) | |||
| Renal (F) | 32 | |||||
| No | 27 | 5 (18.5) | 22 (81.5) | 2.69 | 0.128–56.3 | 0.564 |
| Yes | 5 | 0 (0) | 5 (100) | |||
| Hepatic (F) | 32 | |||||
| No | 28 | 3 (10.7) | 25 (89.3) | 0.12 | 0.012–1.19 | 0.105* |
| Yes | 4 | 2 (50) | 2 (50) | |||
| Oncologic (F) | 32 | |||||
| No | 30 | 4 (13.3) | 26 (86.7) | 0.154 | 0.008–2.98 | 0.292 |
| Yes | 2 | 1 (50) | 1 (50) |
Table 10. Multivariate logistic regression analysis of the relationship between prognostic factors and clinical outcome
| Variable | OR | 95% CI | p |
|---|---|---|---|
| Corticosteroids | 0.068 | 0.005–1 | 0.05 |
| Hepatic | 22 | 1.333–362.84 | 0.137 |
Discussion
The observed prevalence in the present study was 23.8%, a value relatively similar to other studies undergone in Portugal with prevalence values of 20.4% [75], 22.1% [76] and 27.1% [11]. The study sample included a variety of cases, from apparently healthy cats that were diagnosed with only mild anemia or thrombocytopenia in routine pre-surgical analyses, to clinically ill cats that presented to the hospital in a state of severe anemia that were diagnosed during the clinical work-up.
Analysis of the relationship between risk factors and feline Mycoplasma spp. infection identified a statistically significant relationship with FIV-positive animals. This relationship was also identified by other studies, which have proposed that the retrovirus’ immunosuppressive effect enables infections by additional etiological agents or that this observed association is a result of similar transmission routes [42, 77, 78].
An association between feline Mycoplasma spp. and the senior age group has also been observed, with seniors appearing to be less likely to be IgM-positive cases. Feline mycoplasmosis tends to be more prevalent in adult animals due to the cumulative risk of exposure, but there are also studies that identified younger cats as more likely to be affected, as they tend to manifest more overt clinical signs [53, 55, 79]. This negative correlation with older cats may be due to a lower tendency for these animals to acquire recent infectious diseases as they’re less likely to leave home or engage in aggressive interactions that promote the transmission of haemoplasmas, compared to young cats who have a more active lifestyle, although different results could be conditioned by the way age ranges are defined for each study [44, 45, 80].
Despite being identified as relevant by other studies, collected data did not allow the identification of a relationship between infection and “typical” risk factors such as sex, reproductive status, lifestyle or origin of the cats [61, 71, 79, 81]. The absence of a correlation between feline haemoplasmas and the application of external parasiticides, or the presence of ectoparasites, suggests that transmission of these agents between cats in an urban area is not restricted to the activity of arthropod vectors [34, 44, 82].
The simultaneous detection of haemoplasma genetic material in vectors and feline hosts, from which they were removed, confirms their ability to ingest Mycoplasma spp. through their hematophagous activity [5, 10, 66, 83, 84], but does not imply that they are still viable to be transmitted to subsequent vertebrate hosts [14, 34, 85]. However, the frequently observed relationship between feline mycoplasmosis cases and cats that are also ectoparasite carriers, or that have not received adequate external parasiticides, can be a reason to consider the existence of vector-borne transmission [10, 17, 81, 86].
On the other hand, there are registered cases of feline mycoplasmosis in cats not infested with ectoparasites or cats that don’t even have access to the outdoors [87, 88], and although multiple attempts have been made, experimental transmission studies with vectors have provided inconclusive results [83, 89, 90]. Despite an apparent association of feline mycoplasmosis cases with animals infested with ectoparasites, this relationship may only represent a population of cats that have more access to the outdoors and are therefore more exposed to arthropod vectors or other cats of unknown health status [91]. Furthermore, there is recent evidence of canine haemoplasma transmission in an isolated group of dogs, all of which received proper ectoparasiticides, suggesting that haemoplasma transmission is not strictly dependent on the presence and activity of arthropod vectors [92].
Univariate analysis of the effect of acute feline mycoplasmosis infection on clinical, hematological, and biochemical variables found no statistically significant associations. However, multivariate analysis identified some positive trends between infected cases and the presence of anisocytosis or lymphadenopathy, nonspecific signs of disease that have also been identified by other authors [47, 48, 57]. Although not statistically significant, a negative correlation was observed between infected animals with low ALT or decreased total protein, which may be a mirror image of the increased liver enzymes and hyperproteinemia reported in acute cases of feline mycoplasmosis [56, 57].
Akin to other studies, no obvious association was detected between haemoplasma infection and complete blood count abnormalities associated with anemia [9, 66, 79]. This tenuous relationship with clinical and analytical signs can be explained by the complex dynamics between each cat’s individual response and inherent characteristics of the infecting species, which may translate into limited symptomatology that’s nonspecific and short-lived [21, 61, 93]. Another possible cause of these sparse results may be a higher proportion of cats infected by species with lower pathogenic potential such as “Ca. M. haemominutum” and “Ca. M. turicensis”, although it was not possible to distinguish between the three species due to the retrospective nature of the analysis [11–14].
Some animals show an effective response, with hematocrit rise and spontaneous elimination of the etiological agent, while other cats require therapeutic intervention to prevent progression to fatal hemolytic anemia, with metabolic acidosis and cardiovascular collapse [45, 66, 91, 94–96]. Antibiotics such as doxycycline or marbofloxacin have proven to be effective in reducing haemoplasma loads and help to hasten the recovery, but pharmacological therapy does not reliably allow for a complete elimination of the etiological agent [46, 69, 97].
The analysis of factors capable of affecting clinical outcome was only statistically significant with the use of corticosteroids, but a markedly positive trend was observed between the likelihood of recovery with the use of doxycycline. Meanwhile, clinical recovery registered a negative trend with the presence of concomitant diseases (of hepatobiliary or oncological nature). The negative relationship between a favorable prognosis and the use of corticosteroids is justified by the existence of more complex hemolytic anemia cases that didn’t respond to the administration of antibiotics, which led to glucocorticoids being added to the treatment plan even though their effectiveness in cases of mycoplasmosis is still debatable [61, 98].
The negative correlation of a positive clinical outcome with hepatobiliary or oncological disease is also an expected result as the presence of concomitant diseases is generally considered a negative prognostic factor [45, 61, 65]. On the other hand, the use of doxycycline has shown an obvious trend with favorable clinical resolution, proving its effectiveness and confirming its status as the first choice for an initial approach to feline mycoplasmosis [5, 47, 69].
This analysis of prognostic factors focused only on IgM-positive cases that were considered for treatment, but it should be noted that there was a portion of IgM-negative but IgG-positive patients who also received antibiotics after diagnosis. Seeing as these were cases with other prominent clinical conditions that may constitute a confounding effect on the studied prognostic factors, these “chronic” cases were not included as an attempt to perform an analysis focused on the outcome of acute mycoplasmosis.
This study was hampered by a few limitations, the main difficulty being the lack of information regarding the species of Mycoplasma affecting each patient and the inability to distinguish cases of co-infection from single infection cases. Another difficulty was the lack of standardization derived from the use of analyses performed by different laboratories, which, in addition to different equipment, use different reference values. Due to the study’s retrospective nature, as clinical cases were being collected, patient records already provided the case’s progress and outcome, but gaps in records resulted in missing data such as breed, vaccination status, internal and external parasite prophylaxis frequencies and applied drugs, origin of the cats, etc.
Conclusions
The results of this study confirm the presence of haemotropic Mycoplasma spp. in feline patients treated at a veterinary hospital in Lisbon. Their presence in apparently healthy chronic carriers but also clearly anemic cats confirms the complex clinical and epidemiological dynamics of these agents. A positive serological or molecular result doesn’t always imply active disease, so it must be evaluated with the patient’s general condition, clinical signs, hematological and/or biochemical abnormalities, as well as concomitant or future factors that may trigger disease. The bacteria responsible for feline mycoplasmosis are present and circulate in the feline population through transmission routes that are still not clearly defined, and their clearance is not guaranteed by administration of appropriate antibiotic courses. As such, adequate monitoring of these agents should be maintained in animals that live in close contact with humans, given the lack of knowledge regarding their zoonotic potential.