兽研[内容请查看原文]
引言
细菌性角膜炎是一种威胁视力的疾病，若不及时有效治疗，可导致剧烈疼痛、角膜穿孔和永久性视力损害。对于中度至重度细菌性角膜炎的犬，标准治疗方案依赖于频繁局部给予抗菌药物，急性期通常每1–2小时给药一次[1]。尽管在药理学上是合理的，但此类方案对宠物主人和患者的依从性构成了重大挑战[2, 3]。给药频率、给药难度、患者性情以及宠物主人的限制因素均对依从性产生负面影响。特别是，每日给药超过四次的治疗方案始终与依从性显著下降相关[4]。
除依从性外，若干药代动力学和生理因素固有地限制了局部药物在眼表的有效性。由于眨眼、鼻泪管引流和反射性流泪，眼用溶液在角膜前滞留时间较短[5, 6, 7]。在犬中，泪膜药物浓度可迅速下降；例如，局部给予醋酸泼尼松龙后，其浓度在1分钟内下降约45%，15分钟内下降超过95%[8]。这些动力学特征限制了即使采用密集局部给药时有效抗菌药物暴露的持续时间。
由于局部抗菌治疗的局限性，包括患者对频繁给药的耐受性差、宠物主人的依从性以及生理屏障，长效肠外抗菌药物已被探索作为潜在的辅助治疗。全身给药药物的泪膜药代动力学已在有限数量的分子中进行了跨物种描述，包括多西环素、米诺环素、普拉多沙星、泛昔洛韦代谢物、伏立康唑、泼尼松、褪黑素和大麻二酚[9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]。然而，持续全身暴露并不一定转化为治疗相关的泪液浓度；值得注意的是，延长释放的头孢噻呋未能达到超过常见犬角膜病原体MIC的泪液水平[20]。此外，全身抗菌治疗可能引起胃肠道不良反应，包括腹泻和菌群失调，且可能在治疗期后持续存在[21, 22, 23]。因此，抗菌药物管理原则强调局部治疗仍是细菌性角膜炎管理的基石，抗菌药物选择应尽可能依据细胞学、培养和药敏试验指导[24]。相应地，任何考虑全身给予抗菌药物治疗眼表疾病的情况都必须限于经过仔细筛选的病例，而非作为局部治疗的經驗性替代方案。
头孢维星是一种长效、可注射的第三代头孢菌素，获批用于犬。皮下给药后，头孢维星表现出完全生物利用度、延长的表观消除半衰期和高血浆蛋白结合率（> 96%）[25]。尽管高蛋白结合限制了游离血浆浓度，但在白蛋白含量较低的外血管腔隙中，尤其是在炎症条件下，未结合部分可能比例更高[25]。头孢维星在体外对许多临床相关的葡萄球菌和链球菌 species 表现出活性[26]；然而，在大多数地区，它通常不被视为一线抗菌药物，其超说明书使用可能受当地抗菌药物管理建议和处方法规的约束。
因此，本研究的目的为：(i) 表征单次皮下注射后犬泪膜中头孢维星的浓度；(ii) 利用药理学结膜炎模型来近似模拟临床相关的血-泪屏障破坏；以及 (iii) 确定含白蛋白条件下常见犬眼部病原体的MIC分布，并将这些值与泪液暴露进行比较。总体而言，这些目的旨在评估头孢维星在犬细菌性角膜炎中高度选择性辅助使用（而非常规治疗）的药理学合理性。
[内容请查看原文]
材料与方法
[内容请查看原文]
动物
招募了来自爱荷华州立大学研究犬群的六只比格犬（三只去势雄性，三只绝育雌性；约2.5岁；7.7–10.6 kg）。根据委员会认证的眼科专家（M.A.K.）进行的完整体格和眼科检查，包括全血细胞计数、血清生化、尿液分析、STT-1、荧光素染色、回弹式眼压测量、裂隙灯生物显微镜检查和间接检眼镜检查，所有犬均被判定为健康。犬只群养，并在整个研究期间佩戴伊丽莎白项圈。每日进行体格检查和生命体征评估。
[内容请查看原文]
研究设计
[内容请查看原文]
预试验
一项初步研究评估了结膜炎对泪液中头孢维星浓度的影响。一只1岁去势雄性金毛寻回犬，一眼给予局部组胺，对侧眼给予人工泪液（对照）。将组胺二盐酸盐粉末与1.4%聚乙烯醇润滑滴眼液混合制备无菌1%组胺溶液。选择该浓度是为了诱导中度、自限性结膜炎，作为血-泪屏障破坏的模型[27, 28, 29]。在头孢维星给药前20分钟及每次泪液采样时间点前各给予1滴。头孢维星钠（8 mg/kg）皮下注射。使用Schirmer试纸置于下外侧结膜穹窿采集泪液样本，直至吸收至20 mm。采样时间点为：抗生素给药后0、0.25、0.5、0.75、1、1.5、2、4、8、12、24、36、48、72、120、144、168、192、216、240、288和336小时。血浆样本在0、168和336小时采集。所有样本储存于−80°C。
[内容请查看原文]
实验研究
基于初步研究结果，全部六只比格犬在头孢维星给药前20分钟及每次泪液采样时间点前双眼均给予1%组胺滴眼液。犬只接受单次皮下注射头孢维星钠（8 mg/kg）。泪液和血浆采集方法与初步研究一致，采样时间表相同。
[内容请查看原文]
泪液中头孢维星定量
头孢维星分析标准品（Zoetis）和头孢泊肟内标（Sigma）分别以1和5 μg/mL的浓度配制于50/50乙腈/水混合液中。头孢维星储备液稀释至100、50、10和1 μg/mL用于标准曲线和质量控制样本。校准标准品和QC样本配制于混合空白犬泪液中，空白犬泪液使用聚乙烯醇缩醛海绵从健康犬采集[30]。提取前，将18 μL校准品/QC施加于空白Schirmer试纸上。样本用1 mL含头孢泊肟的甲醇/乙腈/水（25:25:50）提取，冷藏、混匀、离心，上清液转移至自动进样瓶。使用Vanquish Flex LC系统 coupled to Exploris 120 Orbitrap质谱仪的LC–MS进行定量。
[内容请查看原文]
最低抑菌浓度
评估了从犬角膜溃疡中获得的10株临床分离菌各：假中间葡萄球菌、犬链球菌和铜绿假单胞菌（总计n = 30）。在含白蛋白条件下进行微量肉汤稀释以模拟泪膜蛋白质环境[31]。头孢维星进行系列浓度测试（稀释调整后最终孔范围160–20 480 ng/mL），平板按CLSI对每种微生物的建议进行孵育。使用数字MIC观察系统评估生长。
[内容请查看原文]
数据分析
MIC50和MIC90分别定义为抑制50%和90%分离菌生长的最低头孢维星浓度。配对t检验比较初步研究中对照眼和结膜炎眼之间的泪液头孢维星浓度。在实验研究中比较每个时间点右眼与左眼的泪液浓度；由于未检测到差异（p ≥ 0.10），双侧浓度取平均值进行分析。非房室PK分析使用PKanalix（2019R1），采用线性-对数梯形法则估算AUC。统计学显著性设定为p < 0.05。
[内容请查看原文]
结果
在初步实验中，除基线（t = 0）外，所有采样时间点泪液中均检测到头孢维星。对照眼泪液头孢维星浓度范围为0至183 ng/mL，结膜炎眼为0至14 190 ng/mL（p < 0.001；表1），对应各时间点1.5至430倍的差异。血浆头孢维星浓度范围为33.5至51.8 μg/mL。在六只比格犬中，至336小时的所有时间点泪液样本中均可检测到头孢维星。个体犬泪液浓度范围为68至73 034 ng/mL，各时间点平均泪液浓度见图1。平均±SD PK参数为：C max 27.2 ± 27.7 μg/mL；T max 61 ± 83 h；AUC 2734 ± 2563 μg/mL·h；t 1/2 64 ± 46 h。平均血浆浓度范围为3.6至24.8 μg/mL，实测C max为23.0 μg/mL。
TABLE 1. Tear film cefovecin concentrations in a pilot experiment involving a single dog with histamine‐induced conjunctivitis in one eye and a contralateral healthy control eye following a single subcutaneous injection (8 mg/kg).
| Time (hours) | Cefovecin tear concentrations (ng/mL) | ||
|---|---|---|---|
| Control eye | Conjunctivitis eye | Fold difference | |
| 0 | 0 | 0 | 0 |
| 0.25 | 3.7 | 5.7 | 1.5 |
| 0.5 | 5.3 | 680 | 128.3 |
| 0.75 | 37 | 242 | 6.5 |
| 1 | 52 | 140 | 2.7 |
| 1.5 | 135 | 1081 | 8.0 |
| 2 | 110 | 221 | 2.0 |
| 4 | 38 | 3663 | 96.4 |
| 8 | 80 | 3581 | 44.8 |
| 12 | 68 | 819 | 12 |
| 24 | 90 | 429 | 4.8 |
| 36 | 33 | 14 190 | 430 |
| 48 | 183 | 82 | 0.4 |
| 72 | 33 | 628 | 19.0 |
| 120 | 28 | 93 | 3.3 |
| 144 | 21 | 69 | 3.3 |
| 168 | 30 | 120 | 4.0 |
| 192 | 17 | 56 | 3.3 |
| 216 | 25 | 124 | 5.0 |
| 240 | 17 | 53 | 3.1 |
| 288 | 23 | 75 | 3.3 |
| 336 | 12 | 45 | 3.8 |
[图：六只犬单次皮下注射头孢维星（8 mg/kg）后不同时间点泪膜药物浓度的平均值±标准差]
犬链球菌的MIC范围为0.32至0.64 μg/mL（MIC50 = MIC90 = 0.64 μg/mL）。在完整的336小时研究期间，泪液中头孢维星的平均浓度始终高于0.64 μg/mL。假中间葡萄球菌的MIC范围为5.12至> 20.5 μg/mL（MIC50 = 5.12 μg/mL；MIC90 = 10.24 μg/mL）。泪液平均浓度在8至24小时之间以及48至288小时之间超过MIC50，在48至120小时之间超过MIC90。所有铜绿假单胞菌分离株在所有测试的头孢维星浓度下均能生长（MIC > 20.48 μg/mL），且各时间点的泪液平均浓度均低于该水平。
计算了游离头孢维星浓度高于MIC的时间百分比（fT > MIC）。对于犬链球菌，fT > MIC为100%。对于假中间葡萄球菌，fT > MIC50为76.2%，而fT > MIC90为21.4%。对于铜绿假单胞菌，fT > MIC为0%。
[内容请查看原文]
讨论
本研究表明，在组胺诱导的结膜炎条件下，单次皮下注射头孢维星（8 mg/kg）可在至少336小时内产生持续、可定量的泪膜浓度。在整个采样期间，泪液暴露量均超过犬链球菌的MIC90，并在48至120小时之间较有限的窗口期内超过假中间葡萄球菌的MIC90，同时始终低于铜绿假单胞菌的MIC阈值。这些发现与先前关于缓释头孢噻呋的研究形成对比，后者泪液浓度未能达到任何常见犬眼部病原体的MIC目标[20]。
对于β-内酰胺类抗生素，药效学疗效的最佳预测指标是游离药物浓度高于MIC的时间分数（fT > MIC）[32]。在本研究中，泪液平均浓度在犬链球菌MIC90阈值下达到100% fT > MIC，支持对该病原体的持续药理学覆盖。相比之下，对于假中间葡萄球菌，泪液平均浓度在研究期间的大部分时间内超过MIC50，但仅在21.4%的给药间隔内超过MIC90。由于MIC90值更好地反映敏感性分布的上限范围，并为经验性覆盖提供更大的信心，假中间葡萄球菌有限的fT > MIC90削弱了对其一致临床疗效的预期。
本研究结果必须在抗菌药物管理的框架内明确解读。证明持续的泪膜暴露并不证明常规或经验性使用头孢维星治疗犬角膜溃疡是合理的。相反，不加区分地使用长效、极其重要的抗菌药物有促进抗菌药物耐药性的风险，并损害负责任的眼科诊疗。在细菌性角膜炎背景下考虑肠外头孢维星时，应仅限于经过仔细选择的病例，依据细胞学、培养和药敏试验结果，且仅在预期病原谱与头孢维星活性相符时使用[24]。头孢维星的长作用持续时间进一步加剧了管理方面的担忧，因为一旦给药，不适当的选择无法轻易逆转。
泪液头孢维星浓度的显著个体间变异进一步限制了仅从平均浓度-时间曲线推断的强度。即使泪液平均浓度超过MIC阈值，最低浓度也经常低于这些目标，表明并非所有犬都达到等效的眼表暴露。此外，泪液浓度-时间曲线呈非线性，其特征为早期下降随后出现延迟的峰值浓度和逐渐消除。其他物种中已有全身给药后类似非线性泪液药代动力学的报道，这可能反映药物从血浆经结膜血管和/或泪腺贡献向泪液隔室的时间依赖性分布[9, 33]。
预试验中观察到的结膜炎对泪液头孢维星浓度的显著影响突出了血-泪屏障完整性的重要性。该屏障的破坏允许富含白蛋白的血浆蛋白渗漏进入泪膜，促进高蛋白结合药物如头孢维星向眼表转运[25, 27, 31]。使用白蛋白补充培养基进行MIC测试是本研究的 methodological 优势，因为标准药敏试验可能低估蛋白结合对抗菌活性的影响。此处使用的白蛋白浓度接近炎症犬泪液中报道的浓度，从而提高MIC估计的生理相关性[31]。
应承认若干局限性。研究人群由单一品种的年轻健康犬组成，限制了对老年犬、短头品种以及自然发生角膜炎或全身性疾病动物的外推。结膜炎在整个研究期间通过药物维持；在临床环境中，炎症预期会随时间消退，随着血-泪屏障恢复，可能降低头孢维星的泪液渗透。此外，虽然Schirmer试纸采样广泛用于泪液药代动力学研究，但反射性流泪可能稀释药物浓度，尽管先前工作表明泪液流速对全身给药药物的测量浓度影响可能有限[12]。
总之，肠外头孢维星可能提供延长的泪液暴露，在频繁局部给药不可行的情况下，可作为对敏感犬链球菌和某些敏感葡萄球菌分离株的辅助治疗而有益。然而，不应期望头孢维星替代局部抗菌治疗，也不应期望其提供对铜绿假单胞菌的可靠覆盖。对于假中间葡萄球菌，观察到的暴露窗口可能对部分犬和分离株具有意义，但个体间变异性和有限的fT > MIC90强调需要谨慎，并进一步强调头孢维星不应经验性或随意使用。
未来研究应评估自然发生细菌性角膜炎犬中头孢维星的泪液药代动力学，纳入动态炎症轨迹，并整合PK/PD建模以估计跨MIC分布的目标达成概率。此类工作对于确定肠外头孢维星是否以及在哪些严格定义的场景中可以负责任地纳入多模式治疗策略，同时保持抗菌药物管理，至关重要。
作者贡献
Jessica L. Payne：研究。David J. Borts：写作——审阅与编辑，形式分析，方法学。Meredith R. McClure：方法学，研究，写作——原稿，可视化。Lionel Sebbag：形式分析，概念化，方法学，写作——审阅与编辑，可视化。Melissa A. Kubai：概念化，研究，写作——审阅与编辑，方法学，资源。Rachel A. Allbaugh：写作——审阅与编辑。
资金
资金由爱荷华州立大学兽医临床科学系（ISU VCS）激励资助（SG2705634，负责人Kubai博士）和ISU VCS启动资金（Kubai博士）提供。
披露
作者未使用AI生成手稿的任何部分。
伦理声明
本研究符合ARVO眼科与视觉研究中动物使用声明，并获得爱荷华州立大学机构动物护理与使用委员会批准（IACUC 22-049）。
利益冲突
作者声明无利益冲突。
致谢
作者感谢Zoetis提供分析标准品，Chloe Stinman完成研究的药效学部分，VDL的Laura Burns和Dwayne Schrunk对样本分析的贡献，Dipak Kumar Sahoo提供实验室空间，以及捐献空白泪液的学生拥有的犬。
数据可用性声明
支持本研究结果的数据可应合理要求从通讯作者处获得。
参考文献
[内容请查看原文]
关联数据
数据可用性声明
支持本研究结果的数据可应合理要求从通讯作者处获得。
查看英文原文▾
[内容请查看原文]
Introduction
Bacterial keratitis is a vision‐threatening condition that can result in severe pain, corneal perforation, and permanent visual impairment if not treated promptly and effectively. In dogs with moderate to severe bacterial keratitis, the standard of care relies on frequent topical antimicrobial administration, often at intervals of 1–2 h during the acute phase [1]. Although pharmacologically justified, such regimens pose substantial challenges to owner and patient compliance [2, 3]. Dosing frequency, difficulty of administration, patient temperament, and owner constraints all negatively influence adherence. In particular, treatment regimens exceeding four daily doses are consistently associated with a marked decline in adherence [4].
Beyond adherence, several pharmacokinetic and physiological factors inherently limit topical drug effectiveness at the ocular surface. Ophthalmic solutions have short precorneal residence times due to blinking, nasolacrimal drainage, and reflex tearing [5, 6, 7]. In dogs, tear film drug concentrations can decline rapidly; for example, topical prednisolone acetate concentrations decrease by approximately 45% within 1 min and by more than 95% within 15 min of administration [8]. These dynamics constrain the duration of meaningful antimicrobial exposure even with intensive topical dosing.
Due to the limitations of topical antimicrobial therapy, including poor patient tolerance of frequent dosing, owner compliance, and physiological barriers, long‐acting parenteral antimicrobials have been explored as potential adjunctive therapies. Tear film pharmacokinetics of systemically administered drugs have been described for a limited number of molecules across species, including doxycycline, minocycline, pradofloxacin, famciclovir metabolites, voriconazole, prednisone, melatonin, and cannabidiol [9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]. However, sustained systemic exposure does not necessarily translate into therapeutically relevant tear concentrations; notably, extended‐release ceftiofur failed to achieve tear levels exceeding the MICs of common canine corneal pathogens [20]. Furthermore, systemic antimicrobial therapy may induce gastrointestinal adverse effects, including diarrhea and dysbiosis that can persist beyond the treatment period [21, 22, 23]. Therefore, antimicrobial stewardship principles emphasize that topical therapy remains the cornerstone of bacterial keratitis management, and antimicrobial selection should be guided by cytology, culture, and susceptibility testing whenever feasible [24]. Accordingly, any consideration of systemically administered antimicrobials for ocular surface disease must be restricted to carefully selected cases rather than used as an empiric alternative to topical therapy.
Cefovecin is a long‐acting, injectable, third‐generation cephalosporin approved for use in dogs. Following subcutaneous administration, cefovecin exhibits complete bioavailability, a prolonged apparent elimination half‐life, and high plasma protein binding (> 96%) [25]. Although high protein binding limits free plasma concentrations, the unbound fraction may be proportionally greater in extravascular compartments with lower albumin content, particularly under inflammatory conditions [25]. Cefovecin demonstrates in vitro activity against many clinically relevant Staphylococcus and Streptococcus species [26]; however, it is generally not considered a first‐line antimicrobial in most regions, and its off‐label use may be subject to local antimicrobial stewardship recommendations and prescribing regulations.
The objectives of this study were therefore: (i) to characterize cefovecin concentrations in the canine tear film following a single subcutaneous injection; (ii) to leverage a pharmacologic conjunctivitis model to approximate clinically relevant disruption of the blood‐tear barrier; and (iii) to determine MIC distributions for common canine ocular pathogens under albumin‐containing conditions and compare these values with tear exposure. Collectively, these aims were designed to assess the pharmacologic plausibility of highly selective adjunctive use, rather than routine treatment, of cefovecin in dogs with bacterial keratitis.
[内容请查看原文]
Materials and Methods
[内容请查看原文]
Animals
Six Beagle dogs from the Iowa State University research colony were enrolled (three castrated males, three spayed females; ~2.5 years old; 7.7–10.6 kg). All dogs were deemed healthy based on a complete physical and ophthalmic examination by a board‐certified ophthalmologist (M.A.K.), including CBC, serum chemistry, urinalysis, STT‐1, fluorescein staining, rebound tonometry, slit‐lamp biomicroscopy, and indirect ophthalmoscopy. Dogs were group‐housed and wore Elizabethan collars throughout the study. Daily physical examinations and vital sign assessments were performed.
[内容请查看原文]
Study Design
[内容请查看原文]
Pilot Study
A pilot study assessed the impact of conjunctivitis on tear cefovecin concentrations. A one‐year‐old, castrated male Golden Retriever received topical histamine in one eye and artificial tears in the contralateral eye (control). A sterile 1% histamine solution was prepared by combining histamine dihydrochloride powder with 1.4% polyvinyl alcohol lubricating drops. This concentration was selected to induce moderate, self‐resolving conjunctivitis as a model of blood‐tear barrier disruption [27, 28, 29]. One drop was administered 20 min before cefovecin dosing and before each tear sampling time point. Cefovecin sodium (8 mg/kg) was administered subcutaneously. Tear samples were collected with Schirmer strips placed in the ventrolateral conjunctival fornix until wicking reached 20 mm. Sampling occurred at the following time points: 0, 0.25, 0.5, 0.75, 1, 1.5, 2, 4, 8, 12, 24, 36, 48, 72, 120, 144, 168, 192, 216, 240, 288, and 336 h post‐antibiotic administration. Plasma samples were collected at 0, 168, and 336 h. All samples were stored at −80°C.
[内容请查看原文]
Experimental Study
Based on pilot findings, all six Beagles received 1% histamine ophthalmic solution in both eyes 20 min before cefovecin administration and before each tear sampling time point. Dogs received a single subcutaneous injection of cefovecin sodium (8 mg/kg). Tear and plasma collection mirrored the pilot approach and sampling schedule.
[内容请查看原文]
Cefovecin Quantitation in Tears
Cefovecin analytical standard (Zoetis) and cefpodoxime internal standard (Sigma) were prepared at 1 and 5 μg/mL, respectively, in a 50/50 acetonitrile/water mixture. The stock cefovecin solution was diluted to 100, 50, 10, and 1 μg/mL for the standard curve and quality control samples. Calibration standards and QC samples were prepared in pooled blank canine tears, which were collected from healthy dogs using polyvinyl acetyl sponges [30]. Prior to extraction, 18 μL of calibrant/QC was applied to blank Schirmer strips. Samples were extracted with 1 mL of methanol/acetonitrile/water (25:25:50) containing cefpodoxime, refrigerated, mixed, centrifuged, and the supernatant was transferred to autosampler vials. Quantitation was performed with LC–MS using a Vanquish Flex LC system coupled to an Exploris 120 Orbitrap mass spectrometer.
[内容请查看原文]
Minimum Inhibitory Concentrations
Ten clinical isolates each of Staphylococcus pseudintermedius , Streptococcus canis , and Pseudomonas aeruginosa (n = 30 total) obtained from canine corneal ulcers were evaluated. Microbroth dilution was performed in albumin‐containing conditions to approximate tear film protein environments [31]. Cefovecin was tested across serial concentrations (final well range 160–20 480 ng/mL after dilution adjustments), and plates were incubated per CLSI recommendations for each organism. Growth was assessed using a digital MIC viewing system.
[内容请查看原文]
Data Analysis
MIC50 and MIC90 were defined as the lowest cefovecin concentrations inhibiting growth in 50% and 90% of isolates, respectively. Paired t‐tests compared tear cefovecin concentrations between control and conjunctivitis eyes in the pilot study. Right versus left eye tear concentrations were compared at each time point in the experimental study; since differences were not detected (p ≥ 0.10), bilateral concentrations were averaged for analysis. Noncompartmental PK analysis used PKanalix (2019R1) with the linear‐log trapezoidal rule for AUC estimation. Statistical significance was set at p < 0.05.
[内容请查看原文]
Results
In the pilot experiment, cefovecin was detected in tears at all sampling times except baseline (t = 0). Tear cefovecin concentrations ranged from 0 to 183 ng/mL in the control eye and from 0 to 14 190 ng/mL in the conjunctivitis eye (p < 0.001; Table 1), corresponding to 1.5‐ to 430‐fold differences across time points. Plasma cefovecin concentrations ranged from 33.5 to 51.8 μg/mL. Across the six Beagles, cefovecin was detectable in tear samples at all time points through 336 h. Individual dog tear concentrations ranged from 68 to 73 034 ng/mL with mean tear concentrations at each time point depicted in Figure 1. Mean ± SD PK parameters were: C max 27.2 ± 27.7 μg/mL; T max 61 ± 83 h; AUC 2734 ± 2563 μg/mL·h; and t 1/2 64 ± 46 h. Average plasma concentrations ranged from 3.6 to 24.8 μg/mL, with a measured C max of 23.0 μg/mL.
TABLE 1. Tear film cefovecin concentrations in a pilot experiment involving a single dog with histamine‐induced conjunctivitis in one eye and a contralateral healthy control eye following a single subcutaneous injection (8 mg/kg).
| Time (hours) | Cefovecin tear concentrations (ng/mL) | ||
|---|---|---|---|
| Control eye | Conjunctivitis eye | Fold difference | |
| 0 | 0 | 0 | 0 |
| 0.25 | 3.7 | 5.7 | 1.5 |
| 0.5 | 5.3 | 680 | 128.3 |
| 0.75 | 37 | 242 | 6.5 |
| 1 | 52 | 140 | 2.7 |
| 1.5 | 135 | 1081 | 8.0 |
| 2 | 110 | 221 | 2.0 |
| 4 | 38 | 3663 | 96.4 |
| 8 | 80 | 3581 | 44.8 |
| 12 | 68 | 819 | 12 |
| 24 | 90 | 429 | 4.8 |
| 36 | 33 | 14 190 | 430 |
| 48 | 183 | 82 | 0.4 |
| 72 | 33 | 628 | 19.0 |
| 120 | 28 | 93 | 3.3 |
| 144 | 21 | 69 | 3.3 |
| 168 | 30 | 120 | 4.0 |
| 192 | 17 | 56 | 3.3 |
| 216 | 25 | 124 | 5.0 |
| 240 | 17 | 53 | 3.1 |
| 288 | 23 | 75 | 3.3 |
| 336 | 12 | 45 | 3.8 |
[图：Mean + SD tear film concentrations of cefovecin over time following a single subcutaneous injection (8 mg/kg) in six dog]
MICs for Streptococcus canis ranged from 0.32 to 0.64 μg/mL (MIC50 = MIC90 = 0.64 μg/mL). Mean tear cefovecin concentrations remained above 0.64 μg/mL for the full 336‐h study period. MICs for Staphylococcus pseudintermedius ranged from 5.12 to > 20.5 μg/mL (MIC50 = 5.12 μg/mL; MIC90 = 10.24 μg/mL). Mean tear concentrations exceeded the MIC50 between 8 and 24 h and again between 48 and 288 h, and exceeded the MIC90 between 48 and 120 h. All Pseudomonas aeruginosa isolates grew at all tested cefovecin concentrations (MICs > 20.48 μg/mL), and mean tear concentrations remained below this level at all time points.
Percentage of time that free cefovecin concentrations remained above MIC (fT > MIC) was calculated. For S. canis , fT > MIC was 100%. For S. pseudintermedius , fT > MIC50 was 76.2%, whereas fT > MIC90 was 21.4%. For P. aeruginosa , fT > MIC was 0%.
[内容请查看原文]
Discussion
This study demonstrates that a single subcutaneous dose of cefovecin (8 mg/kg) produces sustained, quantifiable tear film concentrations for at least 336 h under conditions of histamine‐induced conjunctivitis. Tear exposure exceeded the MIC90 for Streptococcus canis throughout the entire sampling period and exceeded the MIC90 for Staphylococcus pseudintermedius for a more limited window between 48 and 120 h, while remaining consistently below the MIC thresholds for Pseudomonas aeruginosa . These findings contrast with prior work on extended‐release ceftiofur, in which tear concentrations failed to reach MIC targets for any common canine ocular pathogens [20].
For β‐lactam antibiotics, pharmacodynamic efficacy is best predicted by the fraction of time that free drug concentrations exceed the MIC (fT > MIC) [32]. In the present study, mean tear concentrations achieved 100% fT > MIC at the MIC90 threshold for S. canis , supporting sustained pharmacologic coverage for this organism. In contrast, for S. pseudintermedius , mean tear concentrations exceeded the MIC50 for much of the study period but exceeded the MIC90 for only 21.4% of the dosing interval. Because MIC90 values better reflect the upper range of susceptibility distributions and provide greater confidence for empiric coverage, the limited fT > MIC90 observed for S. pseudintermedius tempers expectations for consistent clinical efficacy against this organism.
The present findings must be interpreted explicitly within the framework of antimicrobial stewardship. Demonstration of sustained tear film exposure does not justify routine or empiric use of cefovecin for canine corneal ulceration. On the contrary, indiscriminate administration of long‐acting, critically important antimicrobials risks promoting antimicrobial resistance and undermines responsible ophthalmic care. Any consideration of parenteral cefovecin in the context of bacterial keratitis should be restricted to carefully selected cases, informed by cytology, culture, and susceptibility testing, and only when the expected pathogen spectrum aligns with cefovecin activity [24]. The long duration of action of cefovecin further heightens stewardship concerns, as inappropriate selection cannot be readily reversed once the drug is administered.
Marked interindividual variability in tear cefovecin concentrations further limits the strength of inferences drawn from mean concentration‐time profiles alone. Even when mean tear concentrations exceeded MIC thresholds, minimum concentrations frequently fell below those targets, indicating that not all dogs achieved equivalent ocular surface exposure. In addition, the tear concentration‐time profile was nonlinear, characterized by an early decline followed by delayed peak concentrations and gradual elimination. Similar nonlinear tear pharmacokinetics following systemic drug administration have been reported in other species and may reflect a time‐dependent distribution from plasma into the tear compartment via conjunctival vasculature and/or lacrimal gland contribution [9, 33].
The pronounced effect of conjunctivitis on tear cefovecin concentrations observed in the pilot study highlights the importance of blood‐tear barrier integrity. Breakdown of this barrier permits leakage of albumin‐rich plasma proteins into the tear film, facilitating transport of highly protein‐bound drugs such as cefovecin to the ocular surface [25, 27, 31]. The use of albumin‐supplemented media for MIC testing represents a methodological strength of this study, as standard susceptibility assays may underestimate the impact of protein binding on antimicrobial activity. Albumin concentrations used here approximate those reported in inflamed canine tears, thereby improving the physiologic relevance of MIC estimates [31].
Several limitations should be acknowledged. The study population consisted of young, healthy dogs of a single breed, limiting extrapolation to older dogs, brachycephalic breeds, and animals with naturally occurring keratitis or systemic disease. Conjunctivitis was pharmacologically maintained throughout the study period; in clinical settings, resolution of inflammation would be expected over time, potentially reducing tear penetration of cefovecin as the blood–tear barrier recovers. Additionally, while Schirmer strip sampling is widely used for tear pharmacokinetic studies, reflex tearing may dilute drug concentrations, although prior work suggests that tear flow rate may have a limited impact on measured concentrations for systemically administered drugs [12].
In summary, parenteral cefovecin may provide prolonged tear exposure that could be beneficial as an adjunctive therapy for susceptible Streptococcus canis and some susceptible Staphylococcus isolates, in circumstances where frequent topical administration is not feasible. However, cefovecin should not be expected to replace topical antimicrobial therapy, nor to provide reliable coverage against Pseudomonas aeruginosa . For Staphylococcus pseudintermedius , the observed exposure window may be meaningful for a subset of dogs and isolates, but inter‐individual variability and limited fT > MIC90 emphasize the need for caution and reinforce that cefovecin should not be used empirically or indiscriminately.
Future studies should evaluate cefovecin tear pharmacokinetics in dogs with naturally occurring bacterial keratitis, incorporate dynamic inflammatory trajectories, and integrate PK/PD modeling to estimate the probability of target attainment across MIC distributions. Such work will be essential to define if and in which narrowly defined scenarios parenteral cefovecin can be responsibly incorporated into multimodal treatment strategies while preserving antimicrobial stewardship.
Author Contributions
Jessica L. Payne: investigation. David J. Borts: writing – review and editing, formal analysis, methodology. Meredith R. McClure: methodology, investigation, writing – original draft, visualization. Lionel Sebbag: formal analysis, conceptualization, methodology, writing – review and editing, visualization. Melissa A. Kubai: conceptualization, investigation, writing – review and editing, methodology, resources. Rachel A. Allbaugh: writing – review and editing.
Funding
Funding was provided by an Iowa State University Veterinary Clinical Sciences (ISU VCS) Incentive Grant (SG2705634 assignee Dr. Kubai) and ISU VCS Start‐up Funds (Dr. Kubai).
Disclosure
The authors have not used AI to generate any part of the manuscript.
Ethics Statement
This study complies with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and was approved by the Institutional Animal Care and Use Committee of Iowa State University (IACUC 22‐049).
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
The authors would like to acknowledge Zoetis for supplying analytical standards, Chloe Stinman for completing the pharmacodynamic portion of the study, Laura Burns and Dwayne Schrunk of the VDL for their contributions to sample analysis, Dipak Kumar Sahoo for use of his lab space, and the student‐owned canines who donated blank tears.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
[内容请查看原文]
Associated Data
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.