Case Presentation:
A 56-year-old female with a history of hiatal hernia and obesity presented as a referral to our clinic with recurrent hiatal hernia after hiatal hernia repair and Nissen fundoplication. Symptoms included nausea, vomiting, bloating, heartburn, and epigastric pain.

Past medical history included morbid obesity, hiatal hernia, oral herpes, anxiety, depression, gastroesophageal reflux disease (GERD), hypothyroidism, Sjogren’s disease, and systemic lupus erythematosus. Of note, the patient’s rheumatologic disease was well controlled on hydroxychloroquine, with a most recent erythrocyte sedimentation rate (before HZ reactivation) within normal limits and an anti-double-stranded DNA antibody level only slightly above the range of normal. Prior surgeries included incisional breast biopsy, laparoscopic hiatal hernia repair with Nissen fundoplication, and bilateral reduction mammaplasty. The patient denied tobacco use, occasionally used alcohol, endorsed minimal exercise, and used a self-created diet plan for weight loss. Medications included hydroxychloroquine, pramipexole, and rabeprazole. The patient had no prior history of coronary artery disease (CAD), angina, or family history of significant CAD.

Figure 1: Image of esophagogastroduodenoscopy retroflex view for the patient presented.
As shown above, the image is suggesting hernia recurrence and Nissen disruption.

Given the findings, we recommended a Nissen takedown, repair of the hiatal hernia, and subtotal gastrectomy with Roux-en-Y reconstruction. The patient was enrolled in our program, evaluated by dietician, social worker and referred to cardiology for risk stratification. The cardiologist noted she was a life-long non-smoker, had no family history of CAD, and denied symptoms of angina or dyspnea on exertion. Vital signs and physical exam were unremarkable. An echocardiogram showed an estimated ejection fraction of 55% without regional wall motion abnormalities. The valvular function was normal. Based on these findings, the cardiologist deemed the patient as low risk. This assessment was affirmed by the patient’s NSQIP perioperative risk score, which was estimated at 0.1% (Figure 2). The patient’s preoperative electrocardiogram (ECG) is shown in Figure 3.

When referring to risk of cardiac complications, the estimated risk for the patient was 0.1% where average risk was 0.4% meaning the patient was “below average” risk.

Figure 3: Image of the Patient’s Pre-operative Electrocardiogram

Before meeting the team for preoperative follow-up, the patient presented to the emergency department twice. The initial visit was two weeks before her procedure. She presented with a left-sided neck rash which was deemed to be shingles. She was started on valacyclovir and Silvadene cream. A week later, she represented with left-sided facial and neck pain. She was started on another course of antiviral medication. She completed the course of antiviral therapy, and her symptoms improved. Consideration was given to delaying surgery; however, given the patient’s improving symptoms and completion of her antiviral regimen, the decision was made to proceed with surgery.

On the day of surgery, the history and physical were reviewed with no changes noted. The patient’s shingles had almost resolved. She was taken to the operating room, and anaesthesia was induced with standard American Society of Anaesthesiology (ASA) monitors and average induction doses of medications. The anaesthesiologist noted a brief period of hypotension after induction, which was treated with phenylephrine. The patient’s intraoperative vital trends are shown in Figure 4. Intraoperative notes from approximately 30 minutes after induction noted transient ST-segment elevations in lead III. The blood pressure was stabilized, and ST segments improved. This episode was thought to be due to the reverse Trendelenburg position. Robotic-assisted Nissen takedown, subtotal gastrectomy with Roux-en-Y reconstruction, and repair of the paraesophageal hernia were completed with only 20 ml of blood loss. The total operative time was 3 hours and 29 minutes.

The patient was transported to the post-anesthesia care unit (PACU). However, postoperative ECG demonstrated ST-elevations in the inferolateral leads concerning STEMI (Figure 5). Cardiac enzyme elevations supported the diagnosis of MI, and a follow-up echocardiogram revealed inferior wall hypokinesis. Cardiac catheterization was deferred as perioperative heparinization would raise the risk of bleeding-related surgical complications. Instead, the patient was medically optimized with beta-blockers, a statin, and an angiotensin-converting enzyme (ACE) inhibitor. Antiplatelet therapy was added the night of surgery. The patient was discharged on postoperative day 4. Outpatient nuclear stress testing confirmed the presence of a sizeable inferolateral perfusion defect, and cardiac catheterization demonstrated a 100% proximal left circumflex stenosis, consistent with recent inferolateral STEMI. Diffuse 30% stenosis of the right coronary artery and 40-60% stenosis of the mid-left anterior descending artery was additionally noted, but these lesions were considered not to be “high grade” by the interventional cardiologist.

Figure 5: Image of the Patient’s Post-operative Electrocardiogram

Discussion
Preoperative cardiac risk stratification for this patient allows for evidence-based quantification of the perioperative risk of (MACE) according to the 2014 American College of Cardiology (ACC) and American Heart Association (AHA) Risk Stratification Guidelines. Standard risk-stratification calculators utilized to estimate perioperative risk have the benefit of being derived from large data sets and are either prospectively or externally validated to increase generalizability. Their accuracy comes from the c-statistic for each of these calculators which is reported to be between 0.7 and 0.8 [5,6]. The RCRI is a six-point index in which a score greater than or equal to 2 indicates a greater than 1 % risk of MACE [7]. This patient’s score of 1 (high-risk surgery) suggests the patient is at low risk for perioperative MACE. The NSQIP and Gupta Risk calculators are additional calculators derived from the NSQIP database, which incorporate both patient and surgical factors to calculate risk [8,9]. The patient’s perioperative risk of MACE was estimated to be between one (NSQIP) and four (Gupta) events per thousand cases, supportive of the assessment of the low likelihood of perioperative MACE (Figure 2, Figure 6). The fact that this patient suffered intraoperative STEMI may be explained by the fact that risk calculators may have underestimated her preoperative risk or that her risk may have been accurately estimated, but that she suffered the unfortunate 1-4 events per thousand cases suggested by preoperative risk assessment. However, an alternative and provocative hypothesis is that the patient’s recent HZ reactivation may have represented an additional unappreciated risk for perioperative MI. Viral migration from the neuron into the vasculature has previously been demonstrated, and the subsequent inflammatory response leading to vessel occlusion and ischemia provides a precise pathophysiologic mechanism for HZ-induced vascular events [10]. We speculate that the patient’s recent HZ reactivation combined with the expected intraoperative surgical stress response represented a “two-hit” mechanism to increase this patient’s risk of perioperative MI.