Proposed Section Structure
This topic is best organized around clinical context, methods, comparative long-term outcomes, mechanistic interpretation, and practice implications. A logical structure is: Highlights; Background; Study Design and Target Trial Emulation; Key Results; Clinical and Biological Interpretation; Strengths and Limitations; Implications for Practice and Research; Funding and Registration; References.
Highlights
Across 2 real-world cohorts, cold knife conization (CKC) was associated with a lower risk of recurrent cervical lesions than loop electrosurgical excision procedure (LEEP).
In the Chinese multicenter cohort, CKC was also associated with higher HPV clearance at 3, 6, and 12 months.
The analysis used a target trial emulation framework with inverse probability of censoring weighting, an important methodological step to reduce bias in observational comparative effectiveness research.
The findings are clinically relevant but should be interpreted alongside unmeasured confounding, differences in procedure selection, and the absence of obstetric outcome data.
Background
Cervical intraepithelial neoplasia (CIN) and carcinoma in situ are common precursor lesions in the pathway to invasive cervical cancer. Excisional treatment is a cornerstone of management for high-grade lesions, particularly CIN2/3, because removal of the transformation zone can eradicate dysplastic tissue and reduce progression risk. Among the standard excisional options, CKC and LEEP are the most widely used.
These procedures differ technically and, potentially, biologically. CKC uses a scalpel to remove a cone-shaped portion of the cervix, often yielding a larger, less thermally damaged specimen and potentially more complete excision, especially for glandular extension or lesions extending into the endocervical canal. LEEP uses an electrically heated wire loop and is generally faster, less resource-intensive, and more common in many settings. It is often preferred because it can be performed efficiently and with lower perioperative burden. However, electrosurgical thermal artifact may complicate margin interpretation, and some clinicians have long suspected that recurrence or viral persistence may differ between procedures.
The key unresolved issue has been long-term comparative effectiveness. Short-term treatment success is only part of the story. Persistent high-risk human papillomavirus (HPV) infection after treatment is a clinically meaningful marker because viral persistence is strongly associated with residual or recurrent disease. A treatment that better promotes HPV clearance may plausibly reduce future lesion recurrence. Yet, robust long-term head-to-head data comparing CKC and LEEP have been limited, especially using modern causal inference methods.
Study Design and Methods
Yi and colleagues addressed this question using 2 observational cohorts analyzed within a target trial emulation framework. This design attempts to mimic the structure of a hypothetical randomized trial by clearly defining eligibility, treatment strategies, start of follow-up, outcomes, and analytic methods that reduce bias from censoring and treatment selection.
Cohort 1: Sweden
The first cohort was a nationwide Swedish cohort including 77,001 women who underwent cervical conization from January 1, 1997, to December 31, 2013. Women with CIN or carcinoma in situ were included. Of these, 75,497 underwent LEEP and 1,504 underwent CKC. Follow-up extended up to 22 years.
Cohort 2: Fujian Province, China
The second cohort was a multicenter hospital-based cohort from Fujian Province, China, covering October 1, 2013, to October 1, 2022. This cohort included 3,710 women treated with LEEP and 1,340 treated with CKC, with follow-up up to 9 years. Unlike the Swedish cohort, this cohort also enabled evaluation of HPV clearance at 3, 6, and 12 months.
Exposure, Outcomes, and Analysis
The exposure was CKC versus LEEP. The primary outcome was recurrence of cervical lesions. In cohort 2, secondary outcomes included HPV clearance at 3, 6, and 12 months. The investigators used inverse probability of censoring weighting to balance measured baseline covariates and weighted Cox proportional hazards models to estimate hazard ratios (HRs) and 95% CIs.
For clinicians, the methodological point matters. In nonrandomized data, the choice between CKC and LEEP is rarely arbitrary. It may depend on lesion severity, lesion location, transformation zone visibility, glandular involvement, colposcopic impression, patient age, fertility concerns, and institutional preference. Standard regression adjustment helps, but target trial emulation and censoring weights are intended to produce estimates closer to causal treatment effects, assuming the important confounders have been measured adequately.
Key Results
Long-term lesion recurrence
In the Swedish cohort, after 22 years of follow-up, recurrence occurred in 16,017 of 75,497 women treated with LEEP (21.2%) and 319 of 1,504 treated with CKC (21.2%). On raw percentages alone, the groups look identical. However, the weighted time-to-event analysis showed that CKC was associated with a significantly lower recurrence risk than LEEP, with an HR of 0.67 (95% CI, 0.65-0.68).
This apparent tension between similar crude cumulative percentages and a markedly lower adjusted hazard deserves attention. Hazard ratios reflect event timing over follow-up and the weighted comparison of patients after balancing measured covariates and censoring, not simply the final proportion with recurrence. The result suggests that after accounting for baseline differences and follow-up dynamics, CKC was associated with substantially lower recurrence risk over time.
In the Chinese cohort, recurrence occurred in 65 of 3,710 women treated with LEEP (1.8%) and 11 of 1,340 treated with CKC (0.8%) after up to 9 years of follow-up. The weighted HR was 0.41 (95% CI, 0.21-0.79), again favoring CKC. Although the absolute recurrence rates were lower than in the Swedish dataset, the direction and magnitude of effect were consistent, strengthening the overall signal.
HPV clearance
In cohort 2, HPV clearance after CKC was higher than after LEEP at 3, 6, and 12 months. The abstract does not provide the exact clearance percentages, but the direction of effect consistently favored CKC across all assessed intervals. This is clinically important because post-treatment HPV status is a powerful predictor of residual or recurrent high-grade disease and is increasingly integrated into risk-based surveillance pathways.
If CKC truly improves viral clearance, one plausible interpretation is that more complete removal of infected transformation-zone tissue reduces the local reservoir for persistent HPV infection. Another possibility is that CKC better clears occult endocervical extension that may be missed or incompletely treated with a shallower or smaller LEEP specimen in selected cases.
Clinical and Biological Interpretation
The findings are biologically plausible. CKC typically provides a larger and more intact specimen with clearer histopathologic margins and less electrosurgical artifact. In lesions extending into the endocervical canal, with glandular involvement, or when precise margin interpretation is especially important, CKC may offer oncologic advantages. More complete excision could translate into less residual dysplasia and lower HPV persistence.
That said, procedure selection is inherently tied to case complexity. In routine practice, CKC is often chosen for specific indications such as suspected adenocarcinoma in situ, unsatisfactory colposcopy, discordant cytology-histology, or concern about occult invasion. These indications can cut in opposite directions with respect to prognosis. Some patients selected for CKC may have higher-risk disease at baseline, while others may receive CKC precisely because it is expected to achieve clearer margins. This makes careful adjustment essential, but also explains why even sophisticated observational analyses cannot completely eliminate confounding.
The study also raises a practical question: should CKC be preferred more often for women with CIN or carcinoma in situ? The answer is not yet straightforward. On an oncologic endpoint basis, these data favor CKC. But procedure choice also depends on reproductive plans, surgical expertise, resource availability, anesthesia considerations, margin goals, and the balance between recurrence prevention and treatment-related harm.
Strengths of the Study
The article has several notable strengths. First, the sample size is unusually large, especially in the Swedish nationwide cohort. Second, the investigators assessed long-term outcomes, which is crucial because recurrent lesions can occur years after treatment. Third, the use of 2 independent cohorts from different health systems improves external credibility. Fourth, the target trial emulation approach is more rigorous than many conventional retrospective comparisons and aligns with current best practices in observational causal inference.
The inclusion of HPV clearance data in the Chinese cohort adds translational relevance. Recurrence is the hard clinical outcome, but virologic clearance may help explain why recurrence differs and may inform follow-up protocols.
Limitations and Cautions
Despite its strengths, the study remains observational. Residual confounding is the central limitation. Even with weighting, estimates are only as good as the variables captured and correctly modeled. Important factors such as excision depth, cone volume, margin status, HPV genotype, smoking, immunosuppression, prior treatment history, vaccination status, and pathology review consistency may not have been fully accounted for across both cohorts.
The two cohorts also differ substantially in calendar period, case mix, follow-up duration, screening environments, HPV epidemiology, and management practices. The Swedish cohort spans 1997 to 2013 with very long follow-up, whereas the Chinese cohort covers 2013 to 2022. Diagnostic technologies, HPV testing availability, and surveillance strategies likely changed over time. These differences do not invalidate the study, but they complicate direct numerical comparisons between cohorts.
The lack of detailed absolute HPV clearance figures in the abstract limits a more granular interpretation of the virologic benefit. Similarly, the abstract does not provide lesion-specific subgroup results, such as CIN2 versus CIN3 versus carcinoma in situ, nor margin-specific outcomes. These details would matter clinically because treatment effect may vary by disease severity and anatomy.
One of the most important omissions is the absence of obstetric and procedural safety outcomes. This is not a trivial gap. Excisional cervical treatments, particularly deeper or larger excisions, have been associated in prior literature with increased risk of preterm birth and other adverse obstetric outcomes. If CKC offers lower recurrence at the cost of greater reproductive morbidity, the net clinical decision may differ substantially for a nulliparous woman desiring future pregnancy versus a patient who has completed childbearing.
Finally, the striking precision of the Swedish estimate, HR 0.67 with a very narrow 95% CI of 0.65-0.68, suggests high statistical certainty but should not be mistaken for complete causal certainty. Precision reflects sample size and model behavior; it does not remove the possibility of systematic bias.
How These Findings Fit With Existing Practice
Current cervical precancer management guidelines emphasize risk-based management, margin assessment, and post-treatment surveillance, especially HPV-based testing. They do not generally declare a universal winner between CKC and LEEP for all patients because procedure choice depends on clinical context. CKC is often favored when specimen integrity and endocervical assessment are especially important, while LEEP remains widely used for many high-grade squamous lesions because it is effective, accessible, and efficient.
This study adds an important comparative effectiveness signal in favor of CKC for recurrence prevention and HPV clearance. For clinicians, the most reasonable interpretation is not that LEEP should be abandoned, but that the threshold for choosing CKC may deserve re-examination in selected higher-risk cases, particularly when complete excision and margin certainty are priorities. It also reinforces the importance of careful HPV-based post-treatment surveillance regardless of the excisional method used.
Implications for Practice and Research
For gynecologic surgeons and colposcopists, the study supports a more individualized discussion of excisional strategy. When counseling patients, especially those with extensive endocervical disease, recurrent CIN, or suspected glandular lesions, clinicians may reasonably discuss the possibility that CKC could reduce long-term recurrence risk. At the same time, reproductive implications and perioperative trade-offs must be part of the conversation.
For researchers, the next steps are clear. Comparative studies should report margin status, excision dimensions, HPV genotype-specific clearance, reproductive outcomes, and patient-reported outcomes. Pragmatic prospective registries or, if feasible, randomized trials in carefully selected populations would help determine whether the observed oncologic advantage of CKC persists after tighter control of confounding. Studies should also examine whether differences remain in the era of widespread HPV vaccination and primary HPV screening.
Another important question is whether the benefit is procedure-specific or specimen-size-specific. It is possible that recurrence is driven less by CKC as a technique than by the depth and completeness of excision. If so, optimizing LEEP technique in selected patients might narrow the observed gap.
Funding and ClinicalTrials.gov
The abstract provided does not report funding details. No ClinicalTrials.gov registration number is listed, which is expected for an observational cohort study rather than an interventional trial. Readers should consult the full JAMA Surgery article for complete disclosures, funding sources, and author conflict-of-interest statements.
Conclusion
In this large comparative effectiveness study spanning a nationwide Swedish cohort and a multicenter Chinese cohort, CKC was associated with lower long-term recurrence of cervical lesions than LEEP, and in one cohort, higher early HPV clearance as well. The consistency across cohorts and the use of target trial emulation make the findings noteworthy. Even so, the study should not be interpreted as a blanket mandate to replace LEEP with CKC. Procedure selection remains a nuanced clinical decision shaped by lesion characteristics, fertility goals, technical expertise, and the still-unanswered question of comparative obstetric harm. The main practical message is that CKC may offer an oncologic advantage in some settings, and that this possibility deserves stronger consideration in treatment planning and future research.
References
1. Yi H, Liu Y, Yin X, Huang J, Chen X, Zheng H, Liu S, Chen F, Hong Q, Sundquist J, Sundquist K, Chen J, Zheng X, Ji J. Long-Term Outcomes After Cervical Cold Knife Conization or Loop Electrosurgical Excision Procedure. JAMA Surgery. 2026-06-03. PMID: 42234447. https://pubmed.ncbi.nlm.nih.gov/42234447/
2. Perkins RB, Guido RS, Castle PE, et al. 2019 ASCCP Risk-Based Management Consensus Guidelines for abnormal cervical cancer screening tests and cancer precursors. Journal of Lower Genital Tract Disease. 2020;24(2):102-131. PMID: 32243307.
3. Kyrgiou M, Athanasiou A, Paraskevaidi M, et al. Adverse obstetric outcomes after local treatment for cervical preinvasive and early invasive disease according to cone depth: systematic review and meta-analysis. BMJ. 2016;354:i3633. PMID: 27469988.
4. Arbyn M, Redman CWE, Verdoodt F, et al. Incomplete excision of cervical precancer as a predictor of treatment failure: a systematic review and meta-analysis. Lancet Oncology. 2017;18(12):1665-1679. PMID: 29132818.
