All consecutive patients undergoing isolated CABG through left thoracotomy between July 2021 and July 2022 from a single surgeon at Lankenau Medical Center, PA, USA, were included in the study. Patients using the cryoICE™ device for nerve block were compared for baseline demographics and pre-operative characteristics with those that did not use the cryoICE™ device. During the first 5-months of the surgeon (GT) clinical practice, he did not use the cryoICE™. After that period, he introduced the cryoICE™ device in his clinical practice for all patients undergoing isolated left thoracotomy CABG. Patients were identified via operation codes in a digital operation registry, as well as from a centralized cardiac surgery database for all CABG operations. Clinical data were collected retrospectively from medical records. The underlying in-hospital outcomes were recorded from the charts and death certificate made out by the responsible physician. Follow-up of patients was performed at our outpatient’s clinic, from the hospital registry and by telephone call (IRB E-22-5266). Inclusion criteria were consecutive patients that underwent isolated CABG through left thoracotomy. Exclusion criteria were patients with a concomitant cardiac procedure and that did not use cryoICE™, cardiac surgery via full sternotomy, pregnancy, current use of opioids, myocardial infarction (MI) within 30 days of informed consent, history of substance abuse, chronic pain syndrome, and psychiatric, physical, or mental conditions that would interfere with pain assessment. A psychiatric evaluation was performed by a psychiatrist who made an overall assessment of the patients and provided insights on whether a patient was mentally suitable for this type of procedure.

At follow-up, we investigated in-hospital and follow post-operative pain in the post-operative period, with particular attention on skin hypersensitivity and persistence of skin numbness, utilizing visual-pain questionnaires.

The study protocol was approved by the Main Line Health Hospitals Institutional Review Board (IRB 45CFR164.512).

Primary outcome was grade of incisional pain and numbness. Secondary outcomes were all cause mortality, major adverse cardiovascular and cerebrovascular events (MACCE), nonfatal stroke, nonfatal MI, and repeat intervention. Pre-operative definitions and post-operative outcomes definitions followed the Society of Thoracic Surgeons definitions (Supplementary Document 1). Anesthesia management and covariates and exposures are included in Supplementary Document 2. A description of the cryoICE™, surgical indications, and complications according to the company’s suggestions are described in Supplementary Document 3. Patients’ questionnaire was added in Supplemental Document 4. Only 1 surgeon (GT) performed all the operations. A detailed description of the surgical procedure is added in Supplementary Document 2.

Anaesthetic management was similar in all patients, and included standard monitoring techniques (electrocardiography, central venous/pulmonary artery and arterial pressure monitoring, urinary output, and nasopharyngeal and urinary bladder temperature monitoring).

Standardized intravenous (IV) and oral opioids were used for pain management during the postoperative period as part of institutional program. Intravenous fentanyl was used in patients in the operating room. Ketorolac was also used in the operating room and in the management of post-operative pain. Oral oxycodone was used in the post-operative period for pain management. We used local injections of lidocaine at the surgical incision site in the operating room.

Cryo nerve block consists in temporarily blocking intercostal nerve conduction under each rib. After the procedure, the patient may experience a sensation of numbness in the incision site which translates in decreased pain. The probe is inserted in the thoracotomy incision, the intercostal nerve that innervates the surgical incision site is localized, and the cryosphere is placed adjacent to the intercostal bundle (nerve-artery-vein) close to the innermost intercostal muscle in a location proximal to the latero-cutaneous nerve branch, at a distance of at least 2 cm from the ganglia and 4 cm from the base of the spine. We maintain gentle probe pressure on the tissue throughout cryoablation period. Cryoablation is performed for 120 seconds at –65 °C at 2 levels above the incision, at the level of the incision and 2 levels below the lowest incision.

None of the authors have received funding related to this study and none of the authors are affiliated to the company which produces the Cryonerve device.

Rescue analgesia was defined as hydromorphone or oxycodone administration if pain intensity was higher than 5 (based on a scale from 1 to 10 where 1 is the lowest intensity of pain and 10 is the most severe pain), or if pain persisted after administration of acetaminophen and gabapentin. In the post-operative, 1 gr IV acetaminophen is administered for pain management every 4 hours with a maximum dose of 3 gr. If pain persists, oral gabapentin 300 mg is administered up to three times per day, after patient mental status assessment looking for signs of delirium. In case the pain persisted, hydromorphone 0.2 mg IV every 2 to 3 hours, as needed was administered up to a maximum of 3 mg per day. Next step, in case of continuous pain, was administration of oral oxycodone 5 mg as needed. In case creatinine clearance was 30 mL/min 50% of the usual oxycodone dose was used. In both study arms, standardized IV and oral opioids were used for pain management during the post-operative period as part of an institutional protocol (Supplementary Document 2). In our analysis, we collected and presented the cumulative, intra-operative and post-operative amount of opioid use. A second cohort undergoing the same surgical procedure from the same surgeon without cryoICE™ was used as a control group for in-hospital opioids use. The conversion rate of opioids to morphine milligram equivalent was done following the American Pain Society guidelines [7]. For simplicity we added the conversion table in Supplementary Document 2.

We evaluated each demographic and pre-operative variable as a risk factor for clinical outcomes of all-cause mortality, MACCE, stroke, MI, reoperation, and angina. Groups were compared by two-sample t-tests or Wilcoxon Rank Sum Test for continuous variables and chi-square test of independence for categorical variables. A propensity-adjusted matching was used via a multiple logistic regression with cryoICE™ as the dependent variable and all demographics and pre-operative variables added to the model. A 1:1 greedy nearest neighbor with no replacement match and caliper width of 0.2 produced two groups, with cryoICE™ (N = 60 patients) and non-cryoICE™ (N = 43 patients). Success of matching was assessed by computing the percent bias (similar to standardized mean difference) of each covariate with a cut-off of 10% to denote acceptable balance. Matched samples were compared with McNemar’s test and marginal homogeneity tests for categorical variables and matched paired t-tests and signed rank tests for continuous variables. This analysis allowed inclusion of patients who died within the first 30-days of surgery and was performed to account for possible immortal time-bias. All analyses were performed in Stata 17.0 (Statacorp, LLC. College Station, TX, USA).

Study patients were followed in the post-operative period and at hospital discharge. All patients were contacted after surgery by telephone to screen for hyperalgesia using a visual-pain questionnaire (Supplementary Document 4). In case of a positive response, the patient was asked to make an office visit for a more thorough allodynia assessment. The cotton tip applicator test was used, a test previously validated as a reproducible screening method to identify cutaneous allodynia [5].

There was a total of 103 patients undergoing isolated left thoracotomy CABG, of whom 60 patients received cryoICE™ and 43 patients did not. Mean follow-up period was 5.7 months.

After matching 14 pre-operative variables, diabetes, and chronic obstructive pulmonary disease (COPD) were significantly higher in the cryoICE™ cohort. The cryoICE™ cohort (Table 1) had a mean age of 69.6 $\pm$ 9.5 years and a mean STS score (%) of 1.4 $\pm$ 1.1 while, the non- cryoICE™ had a mean population age of 67.5 $\pm$ 10.1 years and a mean STS score was 1.2 $\pm$ 1.1%.

Intraoperatively (Table 2), there were no differences among the two groups. Both groups had off-pump CABG, and LITA anastomosis to the left anterior descending (LAD) artery was used in 100% of cases. The radial artery graft was used in 1 (1.7%) patient in the cryoICE™ cohort and 5 (11.9%) patients in the non-cryoICE™ one. In addition, there were 3 (5%) saphenous venous grafts anastomoses in the cryoICE™ cohort and 1 (2.4%) in the non-cryoICE™ cohort, while 51 (85%) patients and 36 (83.7%) were extubated in the OR in the cryoICE™ and cryoICE™ cohorts, respectively. There was no intra-operative death but there was a non-lethal stroke from which the patient fully recovered.

Postoperatively, there were no differences among groups. Ventilation time was 1.05 $\pm$ 3.42 vs 0.45 $\pm$ 1.28 hours, in the cryoICE™ versus cryoICE™ cohorts, respectively, while no patient had prolonged ventilation time ($\ge$24 hours). Total hospital length of stay (LOS) was 4.1 $\pm$ 2.6 vs 4.1 $\pm$ 1.9 days in the cryoICE™ versus non-cryoICE™ cohorts, respectively (Table 3). One patient had a reoperation for bleeding in the cryoICE™ group, and there was no in-hospital death.

At follow-up, there were no significant differences among the two groups (Table 4). Two patients (4.6%) had an all-cause death in the non-cryoICE™ cohort. One (1.7%) patient had a nonfatal MI that was medically treated. No patient had a cardiac death or surgical reintervention. MACCE was 1 (1.7%) in the cryoICE™ cohort and 2 (4.6%) in the non-cryoICE™. In addition, 1 (1.7%) patient experienced non-fatal stroke in the cryoICE™ cohort.

Pain questionnaire elicited important findings, including a mean value of incisional pain at hospital discharge of 1.5 (scale 0–10) and at follow-up of 0.69 (scale 0–10) (Table 5). The mean value of elicited pain with movement at hospital discharge was 1.15 (scale 0–10). Mean values of skin numbness at hospital discharge were 1 (scale 0–10) and 0.57 (scale 0–10) at follow-up. Mean values of skin hypersensitivity at hospital discharge were 1.1 (scale 0–10) and 0.9 (scale 0–10) at follow-up. Mean values of pain affecting the quality of sleep and pain at breathing were both 0.34 (scale 0–10) upon hospital discharge and follow-up. Most of the patients that experienced pain at hospital discharge and follow-up described the quality of pain as stabbing pain.

With respect to intra-operative drug use, including fentanyl, ketorolac and lidocaine, there were no significant differences among the two cohorts (Table 6). Postoperatively, hydromorphone IV use was significantly lower in the cryoICE™ vs non-cryoICE™ cohorts, 6.0 $\pm$ 7.0 vs 10.8 $\pm$ 14.1 mg (p = 0.025) (Table 6). Intraoperatively, there was an overall 27% increase of total opioid consumption while postoperatively there was a total reduction of 36% of opioid consumption in the cryoICE™ compared to non-cryoICE™ cohort. During the entire hospital LOS there was a 49% opioid consumption in the cryoICE™ compared to non-cryoICE™ cohort (Table 7).