Effectiveness and safety of cryoablation therapy for early-stage lung cancer patients: a systematic review and meta-analysis Authors: Kun Zhang, Han Wu, Wentong Zhang, Tianlan Wang, Bo Huang, Lei Cao Published: 2026 ABSTRACT BackgroundThe effectiveness and safety of cryoablation therapy in early-stage lung cancer remains unclear now. Therefore, this meta-analysis aimed to further identify the effectiveness and safety of cryoablation therapy among early-stage lung cancer patients.MethodsPubMed, Chinese National Knowledge Infrastructure (CNKI) and Web of Science (WOS) databases were searched up to January 10, 2026. Endpoints assessing the effectiveness included the overall survival (OS), cancer-specific survival (CSS), progression-free survival (PFS), local recurrence and metastasis rate. And endpoints evaluating the safety of cryoablation therapy included the risk of hemorrhage, pneumothorax, pneumothorax with chest tube required and 30-day mortality. Additionally, subgroup analysis focusing on patients with stage IA was performed.ResultsSeven single-arm pilot studies were included with 353 patients. Most patients were stage IA, with a small proportion of stage IB and two studies reporting a few stage II cases (N0). For effectiveness, pooled results indicated that the 1-, 2-, 3- and 5-year OS was 99% [95% confidence interval (CI): 93–100%], 94% (95% CI: 81–100%), 90% (95% CI: 77–99%) and 67.85%. The overall, 1-, 2- and 3-year PFS was 89.4%, 93% (95% CI: 88–98%), 83% (95% CI: 78–88%) and 68% (95% CI: 54–81%) and 5-year CSS was 56.6%. The local recurrence and metastasis rates were 7% (95% CI: 2–14%) and 4% (95% CI: 0–14%). Regarding safety, the proportions of hemorrhage, pneumothorax, pneumothorax with chest tube required and 30-day mortality were 5% (95% CI: 0–15%), 25% (95% CI: 14–36%), 11% (95% CI: 3–22%) and 0%, respectively. Subgroup analysis for patients with stage IA manifested better effectiveness and safety.ConclusionsFor patients with early-stage lung cancer, cryoablation appears to be a feasible minimally invasive treatment option with acceptable short- to mid-term outcomes, particularly in patients with stage IA disease. Highlight box Key findings • Cryoablation demonstrates high effectiveness and favorable safety in patients with early-stage lung cancer, achieving a 3-year overall survival of 90% and progression-free survival of 68%. • The procedure is associated with low rates of local recurrence (7%) and distant metastasis (4%), along with an excellent safety profile (0% 30-day mortality), especially in patients with stage IA disease. What is known and what is new? • Surgical resection is the standard of care for early-stage lung cancer, but many patients are medically inoperable. Cryoablation has emerged as a promising minimally invasive alternative, yet comprehensive, systematic evidence evaluating its collective clinical value has been lacking. • This study is the first systematic review and meta-analysis to provide robust, pooled estimates of survival and complication rates for cryoablation in early-stage lung cancer, demonstrating its high efficacy and safety—particularly for stage IA patients. What is the implication, and what should change now? • Cryoablation is a highly feasible, safe, and effective option for early-stage lung cancer patients who are unfit for surgery, providing acceptable short- to mid-term oncologic outcomes. • Multidisciplinary teams should consider cryoablation as a viable alternative to other non-surgical modalities [e.g., stereotactic body radiotherapy (SBRT)] for selected cohorts. Future research should prioritize large-scale, prospective randomized controlled trials to directly compare cryoablation with surgery and SBRT. Introduction Lung cancer remains the most frequently diagnosed malignancy and the leading cause of cancer-related death worldwide, accounting for approximately one in five cancer deaths globally (1). With the increasing implementation of low-dose computed tomography screening programs, an increasing proportion of lung cancers are being detected at an early stage, particularly stage I disease (2). For patients with early-stage lung cancer, surgical resection remains the standard of care, with lobectomy or anatomically appropriate sublobar resection providing excellent long-term oncologic outcomes (3,4). Despite advances in surgical techniques and perioperative management, surgery is still associated with a certain degree of invasiveness and perioperative risk, which may limit its applicability in selected patient populations. Notably, a considerable proportion of patients with early-stage lung cancer are deemed medically inoperable or decline surgical treatment due to advanced age, impaired cardiopulmonary function, multiple comorbidities, or personal preferences. For these patients, non-surgical treatment strategies such as stereotactic body radiotherapy (SBRT) and systemic therapies, including targeted therapy in selected molecular subgroups, have emerged as alternative options (5,6). In parallel, image-guided percutaneous thermal ablation techniques have an established role as minimally invasive local therapies for selected patients, depending on lesion size and location, as recommended in the National Comprehensive Cancer Network (NCCN) guidelines. Among them, cryoablation has evolved substantially in recent years and has been widely applied in the management of various solid tumors, including hepatocellular carcinoma, renal cell carcinoma, bone metastases, and lung tumors, demonstrating favorable local tumor control and acceptable safety profiles (7-10). Compared with other thermal ablation modalities, cryoablation offers potential advantages such as real-time visualization of the ice ball, reduced procedure-related pain, and relative preservation of surrounding structures. In recent years, several investigators have explored the feasibility, effectiveness, and safety of cryoablation in patients with early-stage lung cancer, particularly in elderly or high-risk populations. However, the available evidence is largely derived from small-scale, single-arm cohort studies, and the reported outcomes vary substantially across studies. To date, a systematic synthesis of the existing evidence remains lacking, limiting the ability to draw robust conclusions regarding the clinical value of cryoablation in this setting. Therefore, the present meta-analysis was conducted to comprehensively evaluate the effectiveness and safety of cryoablation therapy for early-stage lung cancer patients, with the aim of providing a more reliable evidence base to inform clinical decision-making. We present this article in accordance with the PRISMA reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1-0304/rc) (11). Methods Literature search PubMed, Chinese National Knowledge Infrastructure (CNKI) and Web of Science (WOS) databases were searched from database inception up to January 10, 2026 with following terms: cryoablation, lung, pulmonary, tumor, cancer, carcinoma, neoplasm, early stage, stage I and stage IA. Detailed search strategy was as follows: (cryoablation) AND (lung OR pulmonary) AND (tumor OR cancer OR carcinoma OR neoplasm) AND (early stage OR stage I OR stage IA). MeSH terms and free texts were applied. Criteria for inclusion and exclusion Studies met following criteria were included: (I) patients were diagnosed with primary lung cancer; (II) the cryoablation was received due to the lung cancer; (III) the effectiveness representing as the survival, recurrence and metastasis rates or (and) safety representing as the risk of adverse events; (IV) full texts were available; (V) papers were published in English or Chinese. Studies met following criteria were excluded: (I) studies with the type of letter, editorial, case report, review or animal trial; (II) duplicated or overlapped data. Data extraction Following data were extracted from included studies: the first author, year, country, sample size, study design, age, tumor-node-metastasis (TNM) stage, image characteristic and endpoint. In this meta-analysis, endpoints for the effectiveness included the 1-year overall survival (OS), 2-year OS, 3-year OS, 5-year OS, 5-year cancer-specific survival (CSS), overall progression-free survival (PFS), 1-year PFS, 2-year PFS, 3-year PFS, local recurrence rate and metastasis rate. Safety endpoints for the safety included the incidence of hemorrhage, pneumothorax, pneumothorax with chest tube required and 30-day mortality. Risk of bias assessment The risk of bias of the included non-randomized studies was assessed using the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool. This tool evaluates potential bias across several domains, including confounding, selection of participants, classification of interventions, deviations from intended interventions, missing data, measurement of outcomes, and selection of reported results. Each domain was judged as low, moderate, serious, or critical risk of bias. Disagreements were resolved through discussion among the research team. Statistical analysis All statistical analyses were conducted using the online meta-analysis platform METAANALYSISONLINE (https://metaanalysisonline.com). As all outcomes were proportions, including survival rates, recurrence rates, metastasis rates, and adverse event incidences, a single-arm meta-analysis of proportions was performed. Considering the expected clinical and methodological heterogeneity across studies, as well as the limited number of included studies for each outcome (maximum of seven studies), a random-effects model was uniformly applied for all pooled analyses, regardless of the magnitude of statistical heterogeneity. Proportions were pooled after Freeman-Tukey double arcsine transformation to stabilize variances. Statistical heterogeneity was assessed using the Cochran Q test and quantified by the I2 statistic, with higher I2 values indicating greater heterogeneity. Subgroup analyses were conducted according to tumor stage, with a particular focus on patients with stage IA lung cancer. Due to the small number of included studies, sensitivity analyses and publication bias assessments were not performed, as funnel plot-based methods and formal statistical tests for publication bias are considered unreliable when fewer than ten studies are available. All pooled estimates were presented with 95% confidence intervals (CIs). Results Literature selection As shown in the Figure 1, 187 records were searched and 26 duplicated records were removed. After reviewing titles, abstracts and full texts, seven available studies were included (12-18). PRISMA flow diagram of meta-analysis. Basic characteristics All these seven studies were single-arm pilot studies and the overall sample size was 353 patients. All studies were from China or USA and most studies focused on elderly and stage IA patients. One study focused on patients with ground-glass nodular lung cancer. The results of the risk-of-bias assessment using the ROBINS-I tool are summarized in Table S1. Overall, most studies were judged to have a moderate risk of bias, mainly due to their observational single-arm design and potential confounding factors. No study was considered to have a critical risk of bias (Table 1). Table 1Basic characteristics of included studiesAuthorYearCountrySample sizeStudy designAge (years)TNM stageImage characteristicEndpoint Moore (12)2015USA45Single-arm pilot study74.8 [49–85]INREffectiveness and safetyXing (13)2016China28Single-arm pilot study≥70IA–IIBNREffectiveness and safetyLiu (14)2019China14Single-arm pilot study63±10.61IAGGOEffectiveness and safetyChen (15)2022China38Single-arm pilot study70.1±10.5IANREffectiveness and safetyMabud (16)2023USA19Single-arm pilot study≥80IA-IIBNREffectiveness and safetySänger (17)2023USA33Single-arm pilot study75 [68–83]IANREffectiveness and safetyRehman (18)2026USA170Single-arm pilot study≥50IANREffectiveness and safetyData are presented as mean or median with standard deviation or range. CI, confidence interval; GGO, ground-glass opacity; NR, not reported; TNM, tumor-node-metastasis. Basic characteristics of included studies Data are presented as mean or median with standard deviation or range. CI, confidence interval; GGO, ground-glass opacity; NR, not reported; TNM, tumor-node-metastasis. Efficacy of cryoablation therapy for early-stage lung cancer patients Pooled results across all included studies demonstrated post-cryoablation 1-year OS of 99% (95% CI: 93–100%; Figure 2A), 2-year OS of 94% (95% CI: 81–100%), 3-year OS of 90% (95% CI: 77–99%; Figure 2B), 5-year OS of 67.85%±15.3%, 5-year CSS of 56.6%±16.5%, overall PFS of 89.4%, 1-year PFS of 93% (95% CI: 88–98%; Figure 3A), 2-year PFS of 83% (95% CI: 78–88%; Figure 3B), 3-year PFS of 68% (95% CI: 54–81%; Figure 3C), local recurrence rate of 7% (95% CI: 2–14%; Figure 4A) and metastasis rate of 4% (95% CI: 0–14%; Figure 4B) (Table 2). Pooled analyses for the 1-year (A) and 3-year (B) overall survival in early-stage lung cancer patients undergoing the cryoablation. CI, confidence interval; IV, inverse variance. Pooled analyses for the 1-year (A), 2-year (B), and 3-year (C) progression-free survival in early-stage lung cancer patients undergoing the cryoablation. CI, confidence interval; IV, inverse variance. Pooled analyses for the local recurrence (A) and metastasis rates (B) in early-stage lung cancer patients undergoing the cryoablation. CI, confidence interval; IV, inverse variance. Table 2Pooled results for efficacyItems No. of studiesProportion (%)95% CI (%)I2 (%)P valueOverall 1-year OS59993–10074.30.004 2-year OS19481–100 3-year OS39077–9979.50.008 5-year OS167.85±15.3––– 5-year CSS156.6±16.5––– Overall PFS189.4––– 1-year PFS39388–9837.80.20 2-year PFS38378–880.00.69 3-year PFS26854–8162.00.11 Local recurrence572–1442.20.14 Metastasis rate540–1475.90.002Stage IA 1-year OS210099–1000.00.61 3-year OS194.790.8–98.8–– Overall PFS189.4––– 1-year PFS39388–9837.80.20 2-year PFS38378–880.00.69 3-year PFS26854–81620.11 Local recurrence372–155.30.35 Metastasis rate200–300.75Data are presented as mean ± standard deviation. CI, confidence interval; CSS, cancer-specific survival; OS, overall survival; PFS, progression-free survival. Pooled results for efficacy Data are presented as mean ± standard deviation. CI, confidence interval; CSS, cancer-specific survival; OS, overall survival; PFS, progression-free survival. For patients with stage IA lung cancer, the efficacy of cryoablation therapy for early-stage lung cancer patients was manifested as 1-year OS of 100% (95% CI: 99–100%; Figure S1A), 3-year OS of 94.7% (95% CI: 90.8–98.8%) and metastasis rate of 0% (95% CI: 0–3%; Figure S1B). And similar overall PFS of 89.4%, 1-year PFS of 93% (95% CI: 88–98%; Figure S1C), 2-year PFS of 83% (95% CI: 78–88%; Figure S1D), 3-year PFS of 68% (95% CI: 54–81%; Figure S1E) and local recurrence rate of 7% (95% CI: 2–15%; Figure S1F) were revealed (Table 2). Safety of cryoablation therapy for early-stage lung cancer patients Pooled results demonstrated that the risk of adverse events after cryoablation therapy was relatively low with, the hemorrhage incidence of 5% (95% CI: 0–15%; Figure 5A), pneumothorax incidence of 25% (95% CI: 14–36%; Figure 5B), pneumothorax with chest tube required incidence of 11% (95% CI: 3–22%; Figure 5C) and 30-day mortality of 0% (Table 3). Pooled analyses for the risk of hemorrhage (A), pneumothorax (B) and pneumothorax with chest tube required (C) in early-stage lung cancer patients undergoing the cryoablation. CI, confidence interval; IV, inverse variance. Table 3Pooled results of safetyItemsNo. of studiesProportion (%)95% confidence interval (%)I2 (%)P valueOverall Hemorrhage750–1587.9<0.001 Pneumothorax72514–3677.7<0.001 Pneumothorax with chest tube required6113–2283.8<0.001 30-day mortality70–––Stage IA Hemorrhage410–30.00.48 Pneumothorax41810–2855.90.08 Pneumothorax with chest tube required380–2488.3<0.001 30-day mortality40––– Pooled results of safety For patients with stage IA, lower incidences of hemorrhage (1%, 95% CI: 0–3%; Figure S2A), pneumothorax (18%, 95% CI: 10–28%; Figure S2B) and pneumothorax with chest tube required (8%, 95% CI: 0–24%; Figure S2C) were observed (Table 3). Discussion In this meta-analysis, we systematically evaluated the effectiveness and safety of cryoablation therapy in patients with early-stage lung cancer. The pooled results demonstrated favorable short- to mid-term survival outcomes, with high 1-, 2-, and 3-year OS and PFS, as well as low incidences of local recurrence and distant metastasis. In terms of safety, cryoablation was associated with an acceptable complication profile, with pneumothorax being the most common adverse event and no reported procedure-related 30-day mortality. Subgroup analyses further suggested that patients with stage IA disease may derive greater benefit from cryoablation, exhibiting superior efficacy outcomes and lower complication rates. Collectively, these findings indicate that cryoablation represents a feasible and relatively safe local treatment option for selected early-stage lung cancer patients, particularly those who are unsuitable for surgery, and may have meaningful clinical relevance in individualized treatment decision-making. Surgical resection remains the standard treatment for early-stage lung cancer, providing the most reliable long-term oncologic outcomes. Large surgical series have reported 5-year OS exceeding 70% for stage I non-small cell lung cancer following lobectomy or appropriate sublobar resection (19,20). However, surgery is associated with perioperative risks, especially in elderly patients or those with limited cardiopulmonary reserve. SBRT has emerged as the preferred non-surgical alternative for medically inoperable patients, with prospective studies reporting local control rates of 85–95% and 3-year OS ranging from 55% to 70% (5,21-23). In comparison, cryoablation appears to offer comparable short-term local control and survival outcomes based on available evidence, albeit derived mainly from single-arm studies. Regarding safety, surgery carries risks of postoperative complications such as pneumonia, prolonged air leak, and cardiopulmonary events, while SBRT is associated with radiation pneumonitis and chest wall toxicity in a subset of patients (24,25). Cryoablation, by contrast, is minimally invasive and primarily associated with procedure-related complications such as pneumothorax and mild hemorrhage, which are generally manageable. Nonetheless, the absence of direct comparative trials limits definitive conclusions regarding the relative effectiveness and safety of cryoablation versus surgery or SBRT. Cryoablation possesses several intrinsic advantages that support its clinical value in selected patient populations. The procedure is minimally invasive and can be performed percutaneously under image guidance, resulting in reduced procedural stress and faster recovery compared with surgery. The ability to visualize the ice ball in real time allows precise control of the ablation zone, potentially reducing damage to adjacent critical structures. Moreover, cryoablation is associated with less intraprocedural pain than heat-based ablation techniques and preserves surrounding lung parenchyma to a greater extent (26-28). These characteristics make cryoablation particularly suitable for elderly patients, individuals with poor pulmonary reserve, patients with significant comorbidities, or those who refuse surgical intervention. In addition, cryoablation may be advantageous for patients with small peripheral tumors, ground-glass nodules, or metachronous or multifocal lesions, in whom lung function preservation is a critical consideration. Thus, cryoablation may serve as an important component of a multidisciplinary, patient-centered treatment strategy for early-stage lung cancer. Despite these encouraging findings, several areas warrant further investigation. Future research should prioritize well-designed prospective studies and randomized or matched comparative trials to directly compare cryoablation with surgery and SBRT in terms of long-term survival, local control, and quality of life. Greater emphasis should also be placed on identifying optimal patient selection criteria, incorporating tumor size, imaging characteristics, location, and molecular features. Additionally, standardized reporting of procedural techniques, ablation protocols, and outcome definitions would enhance comparability across studies. Finally, longer follow-up durations are needed to clarify the durability of oncologic control following cryoablation and to better define its role within evolving treatment algorithms for early-stage lung cancer. Notably, although the present meta-analysis primarily targeted early-stage lung cancer, a small number of stage II patients were reported in two studies. In the study by Xing et al., the stage II cases were limited to patients without lymph node metastasis (N0), while in the study by Mabud et al., only a single stage II patient was included in the entire cohort (13). Given the extremely small proportion of stage II patients relative to the overall study population, their inclusion is unlikely to have materially influenced the pooled estimates. Nevertheless, to further reduce potential heterogeneity, we performed a subgroup analysis focusing on stage IA disease, which yielded consistent or even more favorable outcomes. Future studies with more detailed stage-specific reporting would help further clarify the role of cryoablation across different early-stage subgroups. Several limitations of this meta-analysis should be acknowledged. First, the overall sample size of included studies was small, and all eligible studies were conducted in China or USA, which may limit the generalizability of the pooled results to other populations and healthcare settings. Therefore, the conclusions of this study should be interpreted with caution. Second, all included studies were single-arm cohort studies, and direct comparisons between cryoablation and other established treatment modalities, such as surgical resection or SBRT, were lacking. As a result, the relative advantages and disadvantages of cryoablation compared with alternative treatments could not be adequately evaluated, highlighting the need for future comparative or randomized studies. Third, due to the absence of detailed individual-level data, subgroup analyses based on additional clinically relevant factors—such as imaging characteristics of pulmonary nodules or tumor location—could not be performed, which may have obscured potential sources of outcome variability. Four, substantial heterogeneity was observed across several pooled endpoints; however, owing to the limited number of available studies and insufficient reporting of potential confounders, further exploration of the underlying sources of heterogeneity was not feasible. Five, another methodological limitation relates to the pooling of fixed-time survival outcomes (e.g., 1-, 2-, and 3-year OS) across studies. Ideally, time-to-event analyses using hazard ratios would provide a more robust evaluation of survival outcomes. However, because the included studies were single-arm cohorts and most did not report hazard ratios or provide sufficient data for survival reconstruction, such analyses were not feasible. Therefore, the pooled fixed-time survival estimates should be interpreted with caution. In addition, most included studies did not report complications according to standardized grading systems (such as CTCAE or Clavien-Dindo classification), which precluded further analysis of complication severity. Conclusions The available evidence suggests that cryoablation is a feasible minimally invasive treatment option with acceptable short- to mid-term outcomes for selected patients with early-stage lung cancer, particularly those who are not suitable for surgical resection. However, further high-quality comparative studies are required to better define its role in clinical practice. Supplementary The article’s supplementary files as Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2026-1-0304/prf