Pectus excavatum (PE), characterized by depression of the sternum, is the most common chest wall deformity, with the majority of patients initially presenting between 11–15 years of age and a 4:1 male-to-female ratio.1,2 PE can have a substantial impact on a patient's quality of life, with severe deformities potentially resulting in cardiopulmonary impairment and psychological trauma. Contemporary surgical correction of pectus excavatum (SCOPE) is generally via modified Nuss procedure, which has become a prevalent operation in pediatric surgical practice.3 Per the Pediatric Health Information System (PHIS), from 2013–2019, over 7,000 patients underwent SCOPE, with hospitals that performed higher case volumes achieving lower odds of postoperative complications.4 Despite studies demonstrating highly successful clinical outcomes with SCOPE, the instantaneous reshaping of the chest wall results in significant pain, often requiring high opioid use.5, 6, 7, 8, 9 Days to weeks of intense postoperative analgesia are required, with pain likely proportional to advanced skeletal maturity.10

Proposed strategies for pain management following SCOPE are numerous and include thoracic epidurals (TE), intravenous patient-controlled analgesia (PCA), indwelling chest wall catheter infusions or elastomeric pain pumps, and local or regional nerve blocks in addition to other pharmacologic and non-pharmacologic methods.11

In 2016, the use of intercostal nerve cryoablation (INC) was first described as a modality of analgesia following SCOPE.12 Before its novel application for patients with PE, cryotherapy as a therapeutic modality had been performed for decades. In 1961, the first handheld cryoprobe for clinical application was invented by neurosurgeon Irving Cooper and engineer Arnold Lee, initially used by Cooper to treat an array of neurologic diseases.13 In 1967, the use of surgical cryoablation was established for cardiac conditions, such as drug-refractory atrial arrhythmias.13 In 1974, Nelson et al. was among the first to publish the application of cryoablation to the intercostal nerves for management of post-thoracotomy pain.14 While cryoanalgesia initially gained favor in the adult surgical community, TEs soon predominated, slowing the more widespread application of INC to other thoracic/chest wall operations and patient populations. In 2015, interest in INC was reignited following the Food and Drug Administration (FDA) approval of the first commercially available cryoablation probe (CryoICE®, AtriCure, Mason, OH) for peripheral nerve analgesia.12 The same year, Kim et al. described their thoracoscopic transthoracic cryoanalgesia technique for use during the Nuss procedure.15 Over time, INC has proven to be an effective perioperative pain management strategy following a variety of adult and pediatric thoracic operations (e.g., thoracotomy, rib fracture stabilization).16, 17, 18

INC is a surgeon-directed modality for analgesia that applies a handheld probe to supercool perineural tissue to −50°C to −70°C, inducing osmotic cellular damage and microvascular injury while sparing the endoneurium for gradual axonal regeneration.19, 20, 21, 22

To perform INC, first, the sternal defect is inspected and bilateral incisions along the anterolateral chest wall are made at the sites of Nuss bar entry/exit in the standard fashion. Single-lung ventilation is initiated. Insufflation and direct thoracoscopic visualization of the hemithorax are achieved. Through the same incision, the parietal pleura is again entered via a different rib space. The cryoprobe is then inserted into the chest cavity (Image A).

To assist with exposure and access to the ipsilateral chest wall, the cryoprobe is gently shaped extracorporeally; for example, bent to form a hook or curved to form a sine wave. Next, the target rib levels for INC are identified. Classically, intercostal nerves between T3 through T8 are deemed safe.23 Selection of the appropriate target rib levels must account for the myotomes and dermatomes of the thoracoabdominal cavity as well as the location of Nuss bar placement.24 Placement of the cryoprobe superior to T3 poses a risk of Horner's syndrome, and placement inferior to T8 could result in denervation diastasis of the obliques. Generally, 4–5 intercostal nerves (e.g., two intercostal levels superior and two intercostal levels inferior to bar entry) are cryoablated.

The cryoprobe is placed at minimum 4 cm lateral to the spine, to avoid injury to the sympathetic chain.18,23 The transition along the posterior chest wall from membranous to muscular pleura serves as a useful anatomic landmark approximating this distance. At this point, the cryoprobe is held in steady contact to the cephalad portion of the intercostal space at the presumed site of the neurovascular bundle (Image B.1). INC is initiated. The device cools to reach the effective temperature. Once this threshold is achieved, a digital timer is set for 1–2 min of freezing.23,25, 26, 27 Gross changes to the tissue are immediately appreciated during this cycle as ice crystals develop (Image B.2). The commercial cryoprobe manufacturer (CryoICE®, AtriCure, Mason, OH) recommends two minutes of contact between the probe and the target tissue. However, histologic evidence has demonstrated adequate axonal degeneration with a shorter, one-minute interval per level freeze, which has also been clinically validated.25 Once the freeze cycle is complete, the cryoprobe rapidly defrosts and is ready for application at the next rib level (Image B.3).

As INC is performed along the chest wall, care is taken to protect the lung parenchyma and skin from contact with the cryoprobe. As the cryoprobe is retracted through the incision to reach lower rib levels, a surgical sponge may be wrapped around the instrument to protect the skin.

This process is then repeated on the contralateral side.

Effective INC relies on the precise execution of this technique, including neurolocation, or the isolation of the nerve within the subcostal groove.27 Hardy et al. evaluated blind placement of the cryoprobe and the anatomic location of the proximal intercostal nerves in 30 cadavers, finding that only 17 % of subjects exhibited intercostal nerves in the classical subcostal position.28 Since the success of cryoanalgesia relies on the cryoprobe's proximity to the nerve, failures of this technique may result from anatomic variability and an inability to isolate the intercostal nerve. More recently, Talsma et al. proposed targeting the lateral cutaneous nerve and the collateral branch in addition to the main branch of the intercostal nerve.29 This approach requires cryoablation at a minimum of 16 sites, as the collateral branch lies immediately superior to the next inferior rib, necessitating cryoablation both above and below each chosen rib level. Percutaneous image-guided INC techniques have been described with promising success; however, to date, no published studies have compared outcomes to the more widely adopted intrathoracic approach.30,31

The favorable impact of INC on length of stay (LOS) and opioid usage after SCOPE has been well-established in the literature.23, 32, 33, 34, 35, 36, 37, 38 Numerous studies report an association between INC and a significant decrease in LOS relative to other analgesic modalities.27,38,39 Compared to TE, INC has been shown to reduce LOS by 2-3.5 days.40, 41, 42, 43 Most studies comparing INC to elastomeric pain pumps or regional catheters reported similar findings.38,39 The mode reported hospital LOS after SCOPE is one night when INC is used as an adjunct in a multimodal pain regimen. More recently, studies have shown that even same-day discharge SCOPE is feasible, achieving effective analgesia when INC is combined with peripheral nerve block.44

Like LOS, use of INC, relative to other analgesic modalities, is associated with a significant reduction in patient opioid requirement.27 In a large cohort study of 5,442 patients using PHIS, Linton et al. found that INC was associated with a median decrease of 80.8 total oral morphine equivalents (OME) compared to no INC.37 Similarly, Graves et al. conducted a randomized control trial (RCT) in 2019 (N=20) that found a mean decrease of 416 mg OME for patients who underwent INC compared to TE (p=0.0001).43 In the 2017 retrospective cohort study by Harbaugh et al., less opioids were administered at discharge in procedures with INC compared to TE (INC: 30 [30–40] doses; TE: 42 [40–60] doses; P = 0.005).45 Of note, those patients who underwent INC had a statistically significant higher Haller index, indicating a more severe deformity and subsequent correction. Compared to thoracic paravertebral catheters, Zeineddin et al. found that INC had a 19-fold and 5.6-fold reduction in average inpatient and total postoperative opioid use, respectively.25 Dekoneko et al. conducted a prospective study of INC compared to RCT results for TE with PCA, concluding that INC resulted in less time to pain control with oral medication.32 Some have postulated that the timing of opioid use during the perioperative period may differ by modality, with patients having undergone INC requiring opioids earlier in their admissions compared to those who received TEs; however, this remains debated in the literature, with higher level evidence favoring no difference.45

INC has not been found to be associated with any significant increase in adverse events, with no observed difference in readmission or reoperation.37

From a respiratory standpoint, INC has not been shown to worsen short-term pulmonary function, as measured by incentive spirometry.46 Some investigators have expressed concern that INC may contribute to complications such as post-operative late pneumothorax. A single center observational study (N=101) by Parrado et al. found no significant difference in late pneumothorax with INC compared to other modalities.47

Risk of neuropathy is commonly mentioned by critics of INC, though prospective studies on long-term pain outcomes after SCOPE with INC are limited. Fraser et al. conducted a prospective survey study of 110 patients who underwent INC where 25 % of respondents reported requiring intermittent pain medication at three months; however, this investigation was limited by a 44 % response rate and the absence of a control group.48 In the previously mentioned randomized control trial by Graves et al., there was no difference in mean pain score between patients who underwent INC versus TE up to one year, and no increased neuropathic pain in the INC group.43 The vast majority of patients (76.9–100 %) experienced complete return of chest wall sensation one year post-INC. A 2020 multi-institutional retrospective review by Zobel et al. screened for neuropathic symptoms using the validated Leeds Assessment of Neuropathic Symptoms and Signs and found that no patients less than 21 years of age experienced neuropathic pain; in addition, these patients had shorter mean time for resolution of numbness compared to their older counterparts (3.4 months versus 10.8, p=0.003).49

While analgesic alternatives, such as elastomeric or ON-Q (I-Flow, Kimberly-Clark, Irvine, CA) pumps, have been found to be associated with higher surgical site infection rates (8.3 % vs. 2.4 %, p<0.001), INC has not been shown to increase infection risk.50 Bundurant et al. showed that INC resulted in lower core body temperature but had no significant difference in wound infections or pneumonia.51 While the effect of INC on core body temperature has been debated, a direct rebuttal of this concern, published by Carter et al. in 2023, found that INC did not result in inadvertent perioperative hypothermia, or a core body temperature less than 36.0 degrees Celsius.52 The estimated basal metabolic rate of the average teenager (approximately 56.3 kg) is nearly 20 times greater than the flux from the cryoprobe; therefore, since a patient's innate thermoregulatory mechanisms would expectedly overcome the energy flux generated by the cryoprobe, the plausibility for a causal relationship between cryoablation and lower core body temperature after surgery is reduced.

Perhaps the largest tradeoff or barrier to the use of INC is the observed increase in surgical time.32 Estimates of the additional time required to perform INC vary, possibly a reflection of differences in technique, experience/familiarity with equipment, and institutional resources. For instance, in the matched cohort study of NSQIP-Pediatric patients by Arshad et al., INC did not significantly increase mean total operative time or mean total time under anesthesia (97.0 min +/− 26.4; 158.0 min +/− 36.6) compared to the control group (97.3 +/− 39.0; 167.8 +/− 45.8, respectively).33 Rettig et al. found that when TEs are placed in the operating room prior to incision, INC had a comparable overall operating room time.42 When applying the cryoprobe for one minute per rib level, Zeineddin et al. found surgical time increased, on average, by nine minutes when INC versus thoracic paravertebral catheters were performed (p<0.01).25 Graves et al. found mean operating room time was 46.5 min longer with INC relative to TE (p=0.0001).43 More recently, Perez Holguin et al. quoted a difference in mean operative time of 153 min with INC versus 89 min with TE (p<0.001).26 Breglio et al. reported a longer average operative time with INC (161 +/− 36 min) than with no block (105 +/− 21 min) or regional catheters (90 +/− 11 min, p<0.001).39 Overall, more prospective, multi-institutional studies are needed to ascertain the true time requirement of INC and to test interventions to improve efficiency.

Despite the reported increased operative time, INC after SCOPE has been found to decrease overall hospital costs.25 Hegde et al. estimated an 89 % probability of a reduction in total hospital costs associated with INC compared to no INC (RR: 0.9, correlation: 0.8–1.1).36 The cost savings with INC have been estimated between $2,000–$7,000 per patient. Rettig et al. reported increased initial costs associated with INC, with institutional costs estimated to be $2,400 per case.42 Nonetheless, when compared to TE, total hospital cost averaged less for INC ($18,336 v. $15,976, respectively; p<0.0005). Conversely, Perez Holguin et al. found increased total hospital costs with INC compared to TE, with notably higher median total costs for both groups ($24,742.5 (IQR 20,861.8–31,799.0) versus $21,621.9 (IQR 16,906.2–25,380.3)).26 For those studies that found increased costs, operating room supplies were the primary driver.53