Preoperative cryoneurolysis for peri- and postoperative pain in total knee arthroplasty: a systematic review and pooled analysis

ABSTRACT

Background

Cryoneurolysis is seeing increased implementation in total knee arthroplasty (TKA), providing postoperative analgesia through preoperative cryoneurolysis of peripheral nerves. Thus, we are conducting this systematic review and meta-analysis to determine cryoneurolysis' efficacy in providing analgesia and improvements in recovery times for patients undergoing TKA.

Methods

We conducted a comprehensive search using five databases including PubMed, Cochrane Library, Web of Science, Embase, and Scopus. The screened articles were assessed for inclusion of primary outcomes: pain scores before nerve cryoablation, peri- and post-operative pain scores in TKA, length of pain relief, patient demographics, and type of pain score employed. The Higgins I² test served to discern the degree of heterogeneity between included studies. Cohen’s d was utilized to interpret the pooled effect size of the studies.

Results

Our search yielded six articles that met our inclusion criteria. The overall effect size illustrated a d = 1.468 (95% CI: 1.084–1.851; p < 0.001) for the Visual Analog Score (VAS), Numerical Rating Score (NRS), and Patient-Reported Outcomes Measurement Information System (PROMIS) pain subscore. Each individual study possessed a large effect size.

Conclusion

Our comprehensive review and meta-analysis indicated that cryoneurolysis provides effective pain management, providing TKA patients with significant pain reduction for 6–12 weeks post-surgery. Furthermore, improvements were observed in patient-centered outcomes: mobility, return to routine activities, and overall satisfaction. Future studies should be conducted to determine the long-term efficacy of cryoneurolysis and the appropriate timing and duration of cryoneurolysis to maximize its analgesic abilities. 

Protocolregistration

www.crd.york.ac.uk/prospero identifier is CRD42024542005

Plain Language Summary

Cryoneurolysis is a procedure in which a nerve is cut by utilizing extremely cold temperatures. By cutting the nerve that serves to conduct pain signals from the knee, this procedure can potentially serve to provide sustained pain relief during and after total knee arthroplasties. Additionally, this procedure has been shown to reduce the amount of opioid pain medications needed by patients after surgery and improve their physical rehabilitation.

1. Introduction

Total knee arthroplasty (TKA) is a common orthopedic surgery aiming to improve functional status and quality of life through replacement of some, or the entire, knee joint. Osteoarthritis (OA) is the most prevalent pathology leading to surgical intervention when conservative therapies (Non-steroidal anti-inflammatories [NSAIDS], intra-articular injections, and physical therapy) have failed [1]. TKA is considered a routine procedure; however, it is commonly associated with significant pre- and postoperative pain [2,3]. Regarding peri- and postoperative analgesia, the current literature supports a multimodal approach; previously studied modalities include intrathecal morphine, peripheral nerve blocks, gabapentinoids, opioids, acetaminophen, NSAIDS, epidural and perineural catheters [4–8].

Despite the multimodal approach, patients continue to suffer from significant pain throughout their postoperative period. The need for novel approaches is necessitated by growing medical concern of reducing opioid use post-TKA for its effects related to dependence and adverse outcomes such as increased rates of stiffness and revision arthroplasty [9,10].

Cryoneurolysis is a novel treatment approach to pain management, as it induces nerve ablation with extreme cold rather than conventional thermal ablations. Cryoneurolysis is achieved by the administration of nitrous oxide or carbon dioxide to the needle tip and subsequently the peripheral nerve for ablation [11]. Cryoanalgesia, a result of cryoneurolysis, is made possible through the induction of Wallerian degeneration of the nerve fiber axons and myelin distal to the site of the injury, which is reversible at a rate of axonal regrowth at one to two millimeters per day [12,13].

During this period of nerve ablation and regrowth, patients often do not experience pain in the targeted region, which in some cases provided pain relief for up to 6 months [14]. Some devices, such as the handheld Iovera device, operate at an ideal temperature of −88℃, providing analgesia without causing permanent damage as more extreme temperatures are associated with long-term injury to surrounding vasculature and neuronal tissue [15].

When applied to knee pathologies, cryoneurolysis has shown promise. Cryoneurolysis targeting the infrapatellar branch of the saphenous nerve (IBPSN) has been a common target for knee pain associated with OA [16–18]. The goal of this technique is to reduce anterior knee pain, restore function of the joint, and provide an improvement to the quality of life of the patient. Due to this evidence, employment for nerve cryoablation for pain management in TKA appears promising. Adequate analgesia for total knee replacement would allow for early mobilization, early restoration of functional status, and subsequently higher patient satisfaction [19,20]. Cryoneurolysis is a newer modality and its utility in pain management, specifically perioperatively, is still largely unexplored when juxtaposed to thermal ablations and various other nerve blocks.

The primary objective of this systematic review and pooled-analysis aims to determine the efficacy of cryoneurolysis in total knee arthroplasty for management of peri- and postoperative pain. Secondary variables of interest included duration of analgesia, nerve targeted, and cryoneurolysis technique.

2. Methods

A rigorous systematic review and meta-analysis on outcomes following cryoneurolysis for total knee arthroplasty was conducted employing the PRISMA 2020 guidelines [21] (Figure 1). This systematic review was registered through PROSPERO (Reference Number: CRD42024542005).

Figure 1.

PRISMA 2020 flow diagram of article inclusion.

2.1. Inclusion criteria

To be included in analysis, articles must contain primary clinical data on the use of cryoneurolysis for the peri- and post-operative analgesia in patients undergoing total knee arthroplasty. Randomized controlled trials, single-arm studies, case series, and retrospective cohort studies were included in analysis. Outcomes of interest were pre-cryoneurolysis pain levels, peri- and post-TKA pain scores, recovery time, and patient demographics. Studies without continuous data, or data reported as categorical or non-numerical data points that were considered incompatible with meta-analytics were not included. Studies that had usable data, but require a degree of transformation to be used were included, with their data points converted to effect size and standard error and then included in meta-analytics.

2.2. Exclusion criteria

Articles were not excluded based on language of publication, date of publication, or country of origin. Article formats that were excluded from analysis were single-patient case reports, articles without full-text available after extensive searching and requests via inter-library loans or directly to the author, studies without statistical analysis that cannot be utilized, or studies without properly reported statistics on pain score. Articles without a cryoneurolysis group, or those who received cryoablation of nerves, but did not undergo TKA, were excluded. Studies that included additional experimental treatment modalities in addition to cryoneurolysis were excluded in order to isolate their effects.

2.3. Information sources and search strategy

This systematic review and meta-analysis employed five medical databases (PubMed, Cochrane Library, Web of Science, Embase, and Scopus) to search for articles on the employment of cryoneurolysis for peri- and post-operative analgesia for TKA. The search string utilized was formed using MeSH terms, boolean operators, and associated terms with variables of interest which may not appear in MeSH. The search string utilized was: (“Cryoablation” OR “cryoneurolysis” OR “cryoanalgesia” OR “iovera”) AND (“total knee replacement” OR “total knee replacements” OR “total knee arthroplasty” OR “total knee arthroplasties” OR “total knee replacement arthroplasty” OR “total knee replacement arthroplasties”). To ensure an exhaustive investigation into the databases, manual investigation of article references and cited-by articles was conducted to ensure no missed articles. The initial article search was conducted on 4/30/24.

To control for duplicate articles, resultant studies from initial querying were identified and duplicates were resolved using Rayyan.ai online software. Following the resolution of duplicates, two independent and trained reviewers (BG and HB) manually sorted through the resultant articles to ensure no duplicates made it past resolution.

2.4. Study selection

Following resolution of duplicates, title and abstract appraisal was conducted for relevant inclusion/exclusion criteria. Article appraisal was carried out by two trained and independent reviewers (HB and BG) with a third reviewer brought in to break any ties (MK). Following a thorough title and abstract review, full-text appraisal for inclusion and relevant data was conducted by the reviewers. Studies deemed to fit all inclusion criteria, and having appropriate and relevant statistical reporting of outcomes of interest were then subject to data extraction. Authors of studies that did not have an available full text or statistics that were incomparable for statistical analysis were contacted individually and given 4 weeks to respond before articles were excluded.

2.5. Data collection

Data were extracted and collected into a separate data extraction form if deemed relevant, properly reported, and meeting inclusion. The primary outcomes extracted from the literature include pain scores before nerve cryoablation, peri- and post-operative pain scores in TKA, recovery time, patient demographics, and type of pain score employed.

2.6. Data analysis

Data extracted were then subsequently uploaded to the Statistical Package for Social Science (SPSS) version 21 (Armonk, NY: IBM Corp) for further evaluation. Studies were first analyzed for degree of heterogeneity. If heterogeneity was under Cochrane’s threshold of 50%, a fixed-effects meta-analysis would be conducted. If heterogeneity breaches the 50% threshold, then a random-effects meta-analysis would be conducted with the focus on the pooled (summary) effect size. If studies do not report data as continuous data, but did report a measure of effect size that could be converted to Cohen’s d for analysis, then they were converted accordingly. To control for any moderate-large degree of heterogeneity, post-hoc leave-out analysis will be conducted.

2.7. Certainty of evidence and risk of bias assessment

Articles included in the analysis were then scored for certainty of evidence utilizing the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) criteria [22]. The GRADE criteria and explanation for any decisions to downgrade the quality of evidence are detailed in the summary of findings chart, which was adopted from GRADEpro GDT (Table 1) [23].

Table 1.

Effect size for each individual study.

Study	Effect size	Std. error	Z	Sig. (2-tailed)	95% confidence interval		Weight	Weight (%)
					Lower	Upper
Pacira Pharmaceuticals	0.910	0.3500	2.600	0.009	0.0224	1.596	8.163	13.9
Dasa et al	1.922	0.2426	7.923	<0.001	1.446	2.397	16.993	28.9
Mihalko et al	1.160	0.3800	3.053	0.002	0.415	1.905	6.925	11.8
Urban et al	1.930	0.2662	7.250	<0.001	1.408	2.452	14.112	24.0
Roth	1.379	0.5168	2.669	0.008	0.366	2.392	3.745	6.4
Jennewine	1.166	0.3350	3.480	<0.001	0.509	1.822	8.911	15.1

Bias in included articles was assessed using metrics designed for their specific study type. To encompass all included study types, ROBINS-I was employed [24,25]. A description of the domains and the respective determinations for each study is included in Figures 2 and 3, which were generated using robvis [26]. This process is employed to determine the generalizability of the evidence provided by the analysis.

Figure 2.

ROBINS-I summary plot [26].

Figure 3.

ROBINS-I traffic-light plot [26].

3. Results

The objective of this systematic review and meta-analysis was to determine the efficacy of cryoneurolysis in total knee arthroplasty for management of peri- and postoperative pain. Secondary variables of interest included nerve ablated, technique, and duration of pain relief. Initial database querying resulted in 108 articles, of which 47 were unique and not duplicates. From these articles, 47 underwent title and abstract appraisal with 12 articles undergoing subsequent full-text appraisal. Six articles were included in the analysis with six being excluded. Of the included articles, three articles were retrospective cohort studies [10,27,28], two articles were randomized controlled trials [29,30], and one article was a case series (Supplementary Table S1) [31]. As of April 2024, 365 patients met our inclusion criteria and underwent cryoneurolysis analgesia prior to TKA. The average follow-up was 9 weeks.

3.1. Excluded articles

Of the articles excluded, two articles were excluded for being a single-patient case report that was unable to be included in statistical analysis [32,33], and the remaining four were excluded due to either different outcomes of interest, being the wrong study type (safety article, conference abstract, description of procedure) or unusable/unclear data reporting [34–37]. Corresponding authors of studies with unclear data reporting or full texts that were not available were contacted requesting clarification; after a month of no responses their study was then excluded.

3.2. Effect of intervention

The Higgins I² test was employed to discern the degree of heterogeneity between included studies. As a result of the I² = 51%, a random-effects model was employed for data analysis. However, due to the dichotomous nature of included study design, pooled-analysis was forgone in favor of meta-analysis (Supplemental Table 2).

When comparing pre-surgical pain levels to post-TKA pain status post cryoneurolysis, all of the included studies showed large effect sizes. Of the included studies, pain scores included: Visual Analog Score (VAS), Numerical Rating Score (NRS), Patient-Reported Outcomes Measurement Information System (PROMIS) pain subscore. The meta-analysis using the random-effects model depicts an effect size of Cohen’s d = 1.468 (95% CI: 1.084–1.851; p < 0.001). This effect can occur in two important epochs; the included studies report significant analgesia immediately following cryoneurolysis procedure, which was sustained until 6 or 12 weeks post-total knee arthroplasty (or end of study).

3.3. Leave-out analysis

In efforts to control for the moderate degree of heterogeneity, a repeated analysis was conducted removing one of the studies based on design or elevated risk of bias. Due to the included studies not containing a high degree of bias, the decision was made to repeat analysis without the case series. The overall effect size of remaining included articles without the case series included depicts d = 1.470 (95% CI: 1.039–1.901; p < 0.001). As a result of this repeated analysis, there was no significant change in effect size juxtaposed to including the case series (d = 1.470 vs. 1.468 respectively).

3.4. Certainty of evidence and risk of bias assessment

GRADE analysis revealed an overall moderate certainty of evidence (Table 1). Indirect comparisons across varying study designs lowered the certainty of evidence. Other factors that are taken into account for upgrading quality of evidence (i.e., large effect size, dose–response gradient, and plausible confounders that would have reduced effect size) were not applicable.

The Risk of Bias assessment revealed a low risk of bias in all included studies (Figures 2, 3). Each study included in this analysis was found to have a low risk of bias across all domains (Table 2).

Table 2.

Summary of findings of GRADE assessment of quality of evidence.

Participants (studies)	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Overall certainty of evidence
Efficacy of preoperative cryoneurolysis in total knee arthroplasty for management of peri- and postoperative pain
365 (3 retrospective cohort, 2 randomized controlled trials, 1 case series)	Not serious	Not serious	Serious	Not serious	None	Moderate
Explanation			Indirect comparisons

4. Discussion

In this systematic review and meta-analysis, we critically investigated the potential benefits of preoperative cryoneurolysis for sustained analgesia in the peri- and postoperative period. Our initial query yielded 108 articles, with 6 studies meeting our inclusion criteria. From our analysis, we can see that cryoneurolysis is a potentially valuable preoperative intervention for perioperative analgesia. Moreover, all studies depicted a sustained analgesic effect ranging from 6 to 12 weeks.

4.1. Clinical implications

The results of the study depict a high (d = 1.468) Cohen’s D value, denoting significant clinical effect in the reduction of peri- and post-operative pain. Cohen’s D value 0.8 or greater is indicative of a large effect of intervention. Due to elevated heterogeneity (I² = 51%), a Leave-Out analysis was conducted post-hoc to control for studies that may have contained a variation in study design or elevated bias. This post-hoc evaluation showed a minute increase in effect-size (d = 1.470), strengthening the results which may have been slightly weakened by the degree of heterogeneity.

While many prior studies have investigated cryoneurolysis for reduction of pain in knee osteoarthritis, employment as a perioperative analgesic in total knee arthroplasty has only begun to be explored. In addition to serving as a potential analgesic for perioperative pain, cryoneurolysis may also serve to help surgeons reduce opioid burden. In a retrospective analysis of a global data registry, researchers determined that preoperative cryoneurolysis helped to significantly decrease opioid burden post-operatively in opioid naive patients, while also decreasing sleep disturbances [38]. Moreover, cryoablation of the superficial genicular nerves prior to total knee arthroplasty has been determined to be a safe adjunct to TKA, and offers a similar profile of minor and major complications as other more commonly employed preoperative procedures [34]. Additionally, when undergoing physical rehabilitation postoperatively, patients who undergo cryoneurolysis before surgery were shown to also have a decreased overall morphine milli-equivalent (MME) requirement, increased range of motion, and improved patient-reported outcome measures when compared to those who did not [36]. Overall, cryoneurolysis of the superficial genicular nerves offers significant promise as an alternative analgesic modality, especially those with whom opioids should be used sparingly.

4.2. Future directions

Due to the fairly newer application of cryoablation as a perioperative analgesic, larger randomized-controlled prospective studies should be conducted to explore non-inferiority juxtaposed to modern standard of care. Cryoneurolysis should also be explored in a wide range of populations due to the nature of total knee arthroplasties, as the majority of literature currently focuses on elderly adults with advanced osteoarthritis. Exploring this modality in other orthopedic surgical applications would benefit the literature by understanding the limitations and potential applications of this intervention. Widening the potential for application of cryoablation through population and procedure will allow for further employment of this newer modality. Furthermore, it is necessary to investigate the effects of cryoneurolysis in this context over long-term follow-up periods, as the follow-up period in this study was limited to an average of 9 weeks. This would allow physicians to make informed decisions about options for the management of chronic pain following TKA.

There are barriers to the implementation of cryoneurolysis in clinical practice. Lack of standardized protocols, procedural training requirements, cost and insurance reimbursement, and integration into existing perioperative workflows currently prevent the adoption of cryoneurolysis. These practical challenges underscore the need for continued high-quality, multicenter trials.

4.3. Limitations

The present study contains limitations. Chief among these limitations is the small number of studies included in the analysis. Due to the newer application of cryoablation as a perioperative analgesic, large randomized-controlled trials have yet to be conducted, and those that have often contain limited sample sizes, limiting the generalizability of results. Furthermore, this study has a moderate degree of heterogeneity, and although leave-one-out analysis was conducted post-hoc, the dichotomous nature of the studies included can potentially lead to decreased strength of evidence. In terms of the outcomes of interest of this study, the differences in structure, sensitivity, and applicability of pain scales limit the strength of direct comparison between NRS, VAS, and PROMIS. However, it is important to note that pain is an inherently subjective and individualized metric; thus, tools to measure pain should be chosen based on specific patient characteristics. Finally, this paper had to exclude a fair number of studies due to incomplete or incompatible data. Although corresponding authors were contacted and given time to respond, exclusion of potentially useful data which was unable to be ascertained at the time of this writing could potentially weaken the strength of results.

5. Conclusion

An exhaustive and methodical systematic review and meta-analysis was conducted. As a result, it is determined that pre-operative cryoneurolysis significantly reduces pain in the peri- and postoperative periods following total knee arthroplasty. This observation illustrates its potential application as a pain management technique, suggesting that it may reduce the need for alternative pain management modalities. Further high-quality, randomized controlled trials are required to solidify cryoneurolysis’ place in the TKA pain management algorithm.

Funding Statement

This paper was not funded.

Article highlights

• Although more commonly employed for knee osteoarthritis, this article explores the potential efficacy of this modality as a perioperative analgesic.

• Preoperative cryoablation of the superficial genicular nerves can potentially lead to sustained pain relief in the perioperative and postoperative period.

• Preoperative cryoneurolysis can help to improve rehabilitation and functional outcomes after the procedure.

• Some studies have shown to decrease total opioid consumption in the perioperative period following cryoablation and total knee arthroplasty.

Author contributions

Brandon Goodwin is responsible for the overall development, conduction, and penning of the manuscript. Hanna Brancaccio was responsible for assisting with article selection and GRADE criteria. Mitchell Kaplan assisted in the authoring of this manuscript and article selection. Valerie Rome assisted in topic development and penning of the discussion. Sameer Shah was responsible for assisting in article retrieval and manuscript writing. Sweta Mukhopadhyay and Seungkyu Park were responsible for manuscript writing. Gilbert Siu was in charge of overall management and oversight of the manuscript.

Disclosure statement

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17581869.2025.2522063