We thank Prof. Björnsson for his thoughtful editorial 1 and would like to mention that his recent publication on drug-induced liver injury (DILI) due to antibiotic therapy has validated the importance of cefazolin as a cause of cholestatic liver injury and provided an estimate of its incidence, occurring in 1 in 6400 persons exposed 2. An important remaining question is why cefazolin has not been prominent in other established DILI registries 3, 4. The likely reason is that DILI from cefazolin is underrecognized, primarily because it is usually administered intra-operatively by the anesthesiologist while the patient is unconscious, without a formal prescription, and with documentation only in the surgical operative notes or anaesthesia records. Thus, when patients present with jaundice 1 to 3 weeks later, neither they nor their physicians may know of the exposure. Without knowing about the cefazolin injection, the liver injury may be mistakenly attributed to the general anaesthetic (such as propofol or sevoflurane), a local anaesthetic (bupivacaine), or postoperatively prescribed oral antibiotics (cephalexin), analgesics (tramadol), or anticoagulants (rivaroxaban) 5, 6. This phenomenon ("false attribution") underscores the need for heightened awareness among physicians of the potential for hepatotoxicity of cefazolin, particularly in the perioperative setting. The clinical phenotype of cefazolin DILI is similar to the cholestatic injury seen with other beta-lactam antibiotics, such as the penicillins and other cephalosporins 7. Notably, cefazolin shares the HLA-A*02:01 genetic association with amoxicillin and amoxicillin/clavulanic acid 8. Cefazolin-induced liver injury is usually mild, self-resolving, and not linked to deaths 7. Nevertheless, recognising cefazolin as the causative agent in cases of unexplained cholestatic liver injury is important because it helps avoid unnecessary, invasive, and expensive diagnostic procedures such as a liver biopsy, endoscopic retrograde cholangiopancreatography (ERCP), or endoscopic ultrasound (EUS), which may compound the iatrogenic harm. To better understand the frequency of DILI from cefazolin, a prospective study assessing liver tests in patients having minor outpatient surgery who receive a single dose of cefazolin would be informative and welcome. Liver tests at baseline and 1, 2, and 4 weeks after administration of cefazolin may help unveil cases that otherwise would not have been recognised. Such a study would help delineate incidence, severity, and guide prevention strategies. We hope that our paper, along with the accompanying editorial, raises awareness of cefazolin as a potential cause of DILI and encourages more accurate and inclusive identification in clinical practice and future registry data. We have indeed come a long way from the initial association, first noted by a plastic surgeon in a letter to the editor of a surgery journal, where concern was raised over the routine use of cefazolin as prophylaxis against postoperative infection in women undergoing augmentation mammaplasty 9. Raj Vuppalanchi: writing – original draft, writing – review and editing, conceptualization. Jay H. Hoofnagle: conceptualization, writing – review and editing, writing – original draft. Don C. Rockey: writing – review and editing. Herbert L. Bonkovsky: writing – review and editing. Yi-Ju Li: writing – review and editing. Dina Halegoua-DeMarzio: writing – review and editing. Robert J. Fontana: writing – review and editing. Christopher Koh: writing – review and editing. Joseph Odin: writing – review and editing. Huiman Barnhart: writing – review and editing. The authors declarations of personal and financial interests are unchanged from those in the original article 7. This article is linked to Vuppalanchi et al papers. To view these articles, visit https://doi.org/10.1111/apt.70284 and https://doi.org/10.1111/apt.70313. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
Vuppalanchi et al. (2025) studied this question.