Paradoxical Hyperalgesia of Supra-orbital and Infra-orbital Nerve Blocks when Anesthetizing the Trigeminal Nerves in TransSphenoidal Endonasal Pituitary Surgery: A Prospective, Randomized, Double-blinded Clinical Trial

Background: Pituitary neurosurgery executed via the transsphenoidal endonasal approach is commonly performed for pituitary adenomas. Reasons for prolonged hospital stay include postoperative migraine headache pain and protracted nausea with or without vomiting. Bilateral supercial trigeminal nerve blocks of the supra-orbital V1 and infra-orbital V2 (SION) nerves performed intra-operatively as a regional anesthetic adjunct to general anesthesia were hypothesized to reduce pain exposure and thus 6 hours post-operative morphine PCA (patient-controlled analgesia) use by patients. Methods: 49 patients, following induction of general anesthesia for their transsphenoidal surgery, were prospectively randomized in a double-blinded fashion to receive additional regional anesthesia as either a block (0.5% ropivacaine with epi 1:200,000) or placebo/sham (0.9% normal saline). The primary endpoint was the pain exposure and resulting systemic morphine PCA opioid consumption by the two groups in the rst 6-hours post-operatively. The secondary endpoints included (1) incidence of postoperative nausea and vomiting and (2) time to eligibility for PACU discharge. Results: Of the 49 patients that were enrolled, 3 patients were excluded due to protocol violations. Ultimately, there was no statistically signicant difference between morphine PCA use in the 6-hours post-operatively between the block and placebo/sham groups. There was, however, a slight visual tendency of the block group for higher pain scores, morphine use p=0.046, and delayed PACU discharge. False discovery rate corrected comparisons at each time point then revealed no statistically signicant difference between the two groups. There were no differences between the two groups for secondary endpoints. Conclusion: 6-hour post-operative migraine headache pain after endoscopic trans-sphenoidal pituitary surgery likely has a more complicated mechanism involving more than the supercial trigemino-vascular system and perhaps is neuro-modulated by other brain nuclei. (supra-orbital and infra-orbital nerve (SION)) study to compare the systemic post-operative pain medication requirements in patients having trans-sphenoidal endoscopic, endo-nasal pituitary surgery under general anesthesia. The study protocol is performed in accordance with the relevant guidelines and is included in detail below. The primary endpoint was the systemic morphine sulfate PCA (patient-controlled analgesia) opioid consumption by the 2 groups during the rst 6 hours post-operatively (regional + general anesthesia vs. general anesthesia alone). The secondary endpoints included: (1) incidence of PONV, (2) time to eligibility for PACU discharge, and (3) length of hospital stay/discharge. To our knowledge, this is the rst published clinical trial examining the use of SION block to control postoperative pain after endonasal pituitary surgery. The results from the two groups were compared using unpaired student t-test. P< 0.05 was considered to indicate a statistically signicant difference. has been shown that repetitive injection of local anesthetics in the distribution of the supra-orbital nerve (SON) V1 and infra-orbital nerve (ION) V2 can decrease the incidence of chronic idiopathic migraine headaches not related to surgery. 5 Thus, potentially local anesthetics injected in the distribution of the supra-orbital (V1) and infra-orbital (V2) nerves were postulated to pre-emptively prevent this sequence of events that ultimately lead to post-operative headache pain after trans-sphenoidal pituitary surgery. These study results showed that addition of supra-orbital and infra-orbital sensory trigeminal nerve blocks intra-operatively to general anesthesia during trans-sphenoidal, total endo-nasal pituitary surgery do not appear to reduce post-operative pain and thus 6-hour post-op morphine PCA consumption as compared to patients receiving only general anesthesia. In fact, the Placebo group had less postoperative pain and morphine use along with a shorter median PACU stay. Because the standard deviations in both groups were so wide, this effect was not clinically signicant. Possible explanations for this unexpected pain and morphine consumption post-op response in the Block group may include:


Introduction
Regional anesthesia combined with general anesthesia for minimally invasive procedures has become a common anesthetic technique. The goal of this approach is to reduce pain and promote a rapid recovery. 1 If regional anesthesia facilitates the administration of less systemic pain medications, patients could bene t in several ways: (1) reduced postoperative discomfort and opioid use, (2) decreased post-operative nausea and vomiting, (3) shortened PACU (post anesthesia care unit) stay time and (4) shortened time to discharge from hospital. A more streamlined, shorter and more pleasant hospital course for the patient could lead to increased patient satisfaction and potential economic gains to the hospital.
Pituitary tumor surgery is primarily executed through a minimally invasive approach through the nose and sinuses as an endoscopic procedure called the: "Trans-Sphenoidal Hypophysectomy". More recently, it is performed entirely endoscopically through the nose as the earlier modi cation, the "Caldwell Luc" approach, involved opening a bony window through a sub-labial oral (at the upper gum line under the top lip) approach leading to the maxillary sinus and then ultimately to the sellar cavity housing the pituitary tumor. Patients who have this endonasal endoscopic trans-sphenoidal pituitary surgery are usually same day admission patients who, following their operation, stay in hospital up to 24 hours post-operatively. Prolonged post-op admissions are often caused by persistent headaches or nausea with or without vomiting. 2,3,4,5 By administering regional anesthesia to the patient intra-operatively after induction of anesthesia, pain scores may be reduced 6 , PONV decreased, and PACU stay shortened. Bilateral infra-orbital nerve blocks (V2 maxillary) have been shown to provide adequate pain control following trans-sphenoidal pituitary surgery such that systemic opioid use can be reduced or eliminated, and the patient discharged earlier or same day of surgery. 7 The aim of our study was to investigate the effect of infra-orbital V2 and supra-orbital V1 nerve blocks to reduce pain scores and thus morphine opioid use in the rst 6 h post procedure. Secondary end points of the study were: (1) presence of PONV, (2) time to eligibility for PACU discharge, and (3) duration of hospital stay until discharge.

Page 3/13
This is a prospective, randomized, double-blinded sham controlled (preservative free normal saline) regional anesthetic (supraorbital and infra-orbital nerve (SION)) study to compare the systemic post-operative pain medication requirements in patients having trans-sphenoidal endoscopic, endo-nasal pituitary surgery under general anesthesia. The study protocol is performed in accordance with the relevant guidelines and is included in detail below. The primary endpoint was the systemic morphine sulfate PCA (patientcontrolled analgesia) opioid consumption by the 2 groups during the rst 6 hours post-operatively (regional + general anesthesia vs. general anesthesia alone). The secondary endpoints included: (1) incidence of PONV, (2) time to eligibility for PACU discharge, and (3) length of hospital stay/discharge. To our knowledge, this is the rst published clinical trial examining the use of SION block to control postoperative pain after endonasal pituitary surgery. The results from the two groups were compared using unpaired student t-test. P< 0.05 was considered to indicate a statistically signi cant difference.
Following Institutional Review Board approval at the University of California, San Francisco, USA (UCSF) study enrollment was offered to adult patients (male and female) (aged 18-65), having an ASA (American Society of Anesthesiologists) physical status I or II scheduled for same day admit endo-scopic endo-nasal trans-sphenoidal hypophysectomy to be performed by a single surgeon (SK). Additional inclusion criteria were: (1) elective pituitary surgery for tumors less than or equal to 2 cm diameter within the sella turcica and without cavernous sinus invasion, (2) both male or female patients, (3) English speaking patients providing informed consent, (4) not on chronic pre-op pain medications (non-narcotics) in the last week prior to surgery per patient report, (5) no opioid pain medications pre-op in the last month before surgery per patient report, (6) no abuse of recreational drugs (cocaine, methamphetamines, marijuana, opioids, heroin etc.…) per patient report and (7) no herbal medications for pain in the last 1 month prior to surgery per patient report. Urine or blood testing was not used to con rm opioid use pre-operatively. Exclusion criteria included: (1) patients less than 18 yo, (2) non-English speaking and unable to consent, (3) known allergy to ropivacaine, (4) chronic pain conditions, including idiopathic migraine as de ned by the ICHD (International Classi cation of Headache Disorders) II criteria, requiring the use of pain medications or (5) inability to comprehend or adhere to the study protocol (Chart 1).

Chart 1 Study Inclusion and Exclusion Criteria
Inclusion Criteria After written informed consent, the subjects were randomly assigned to 1 of 2 groups using randomization envelopes in 3 blocks of 20. A "triple mask design" was used: the subjects, healthcare providers (including the anesthesiologist in the operating room, anesthesiologist administering the SION blocks), randomization consent nurse and data collection nurse as well as the statistician were all blinded to the treatment groups assigned.
Intra-operative Block Protocol: The subjects assigned to Block group (B) would receive the SION blocks with 0.5% ropivacaine with epinephrine 1:200 000 following induction of general anesthesia and the Placebo group (P) would receive the same volume SION blocks using sterile, preservative free normal saline. Following general anesthetic induction and endo-tracheal intubation, supra-orbital and infra-orbital (SION) injections with study medication were performed. Supra-orbital nerve blocks were performed bilaterally using 3 cc (on each side for a total of 6 cc) of the study solution. A percutaneous technique was used to block both bilateral supra-orbital and supra-trochlear nerves. A 27 gauge, 1.5 in needle, attached to a 10-cc syringe was used. Then, the infra-orbital nerves were blocked intra-orally following retraction of the upper lip and using a 27-gauge, 2.54 cm needle attached to a 10-cc syringe inserted anterior to the ipsilateral canine tooth that runs along the maxilla. Here, 3 cc of study drug was placed bilaterally for a total of 6 cc. The needles were directed superiorly toward the infra-orbital foramen on the ipsi-lateral side as was described in the "Lynch" method. 8 Correct needle placement was con rmed by palpation of the needle tip near the infra-orbital foramen and by negative aspiration for blood. A wheel of study drug was palpated next to the nose under the inferior orbital rim. As mentioned earlier, the study drug was prepared by an independent physician who was not involved in the data acquisition, operating room care of the patient, or placement of the block.
Thus, a total of 12 cc of the study solution was placed per patient in the distribution of the supra-orbital branches of V1 (Trigeminal sensory nerve) and infra-orbital nerve V2 (Trigeminal sensory nerve).
The surgeons then administered standard topical intra-nasal anesthesia with lidocaine and epi soaked pledgets on the inner nasal mucosa. Less than 10 cc of 1% lidocaine with epi 1:200 000 was used to ensure hemostasis. Exposure of the pituitary gland was obtained trans-nasally with an endoscope via the sphenoid sinus and then via the sellar oor. Following removal of the sellar oor, the dura was opened, and the pituitary tumor was resected using a series of micro-dissectors and curettes. CSF leaks were then repaired, and fat was placed in the free intra-sellar space. The sellar oor was repaired and the sphenoid sinus and nostrils packed temporarily to close the sellar space. We removed the nasal packings just prior to extubation.
Following induction of anesthesia, the local injection study block drug/placebo was administered, and the patients were maintained with a balanced anesthetic technique using intra-venous agents, inhaled gas and a neuro-muscular relaxant. Des urane in O2/Air mixture with 80% inspired O2 was administered to provide anesthesia and to maintain the blood pressure within 20% of baseline.
Fentanyl was infused at 1 mcg/kg/hr from induction of anesthesia to end of surgery. Dexamethasone 10 mg IV or Hydrocortisone 100 mg IV was administered at the beginning of surgery as directed by the surgeon. An end-tidal CO2 of 30-35 mmHg was maintained. The patients were relaxed with rocuronium as needed to maintain 1-2 twitches by a NMT (neuro-muscular twitch) monitor. On emergence from anesthesia, ondansetron 4 mg IV was administered, and residual muscle relaxation was reversed with neostigmine 0.07 mg/kg and glycopyrrolate 0.01 mg/kg IV.
Post-operatively, a morphine PCA (1mg IV every 6 min lock out time) to a max of 10 mg IV morphine sulfate per hour, without a basal rate was administered. For pain ranked at >4/10, rescue pain medicine of morphine 1 mg IV was administered by the PACU recovery room nurses every 5-10 min until the pain was < 4/10 or acceptable to the patient. Blood pressure was maintained with a systolic BP< 140 mmHg using (labetalol, esmolol, nicardipine). Following intra-operative treatment of nausea with ondansetron and dexamethasone/hydrocortisone, the following escalating rescue algorithm was used for PONV: 1. Additional ondansetron 4 mg IV for 1 dose, then 2. prochlorperazine 5-10 mg IV, then 3. scopolamine patch 1.5 mg topical. 9 The primary outcome of the study was to compare the total amount of morphine (in mg) used post-operatively (for 6 hours) in the 2 study groups from T0-T6 time from PACU entry (T0) to 6 hours post PACU entry (T6). Study data was collected for 5 data time points: T0 (PACU entry), T1, T2, T4, T6 (1,2,4,6 hours after PACU entry, respectively). Pain scores (P0-P6) were collected using the Wong-Baker FACES Pain Rating Scale at times T0-T6 in the scale (0-10/10) with 10/10 being the worst imaginable pain. Nausea score N0-N6 was collected using the Baxter retching faces (BARF) nausea scale (0-10/10) with 10/10 being the worst nausea imaginable with or without vomiting. Other secondary data collected included Tr (total time in PACU from admission to "eligibility" for PACU discharge (using the UCSF Modi ed Aldrete Score), Th (total time from surgery start to eligibility for home discharge), Tn (time from surgery start to last dose of narcotic

Results
A total of 49 patients were randomized. Three of the randomized patients subsequently had to be disquali ed from the study due to protocol violations. There was no statistically signi cant difference between groups in demographic variables (Table 1).  There was no signi cant difference between groups in nausea exposure over time (Figure 3). Examining both groups combined, compared to PACU arrival, patients experienced signi cantly more nausea at 1 hour after PACU entry (mean difference 1.  This "sterile neuro-genic peri-vascular in ammation" can ultimately cause meningeal blood vessel vaso-dilation, increased cerebral blood ow and migraine type post-op headaches. 2,3,4 It has been shown that repetitive injection of local anesthetics in the distribution of the supra-orbital nerve (SON) V1 and infra-orbital nerve (ION) V2 can decrease the incidence of chronic idiopathic migraine headaches not related to surgery. 5 Thus, potentially local anesthetics injected in the distribution of the supra-orbital (V1) and infra-orbital (V2) nerves were postulated to pre-emptively prevent this sequence of events that ultimately lead to post-operative headache pain after trans-sphenoidal pituitary surgery.

Discussion
These study results showed that addition of supra-orbital and infra-orbital sensory trigeminal nerve blocks intra-operatively to general anesthesia during trans-sphenoidal, total endo-nasal pituitary surgery do not appear to reduce post-operative pain and thus 6-hour post-op morphine PCA consumption as compared to patients receiving only general anesthesia. In fact, the Placebo group had less postoperative pain and morphine use along with a shorter median PACU stay. Because the standard deviations in both groups were so wide, this effect was not clinically signi cant. Possible explanations for this unexpected pain and morphine consumption post-op response in the Block group may include: 1. It may be more di cult to accurately execute the supra-orbital V1 and infra-orbital nerve V2 blocks under general anesthesia than expected. The terminal sensory nerve anatomy of patients may vary and so the block, placed using landmarks, may not actually capture all the branches of the trigeminal nerves V1 and V2 that then contribute and eventually trigger the trigeminovascular effects leading to the post-op migraine headache. As the patients were under general anesthesia, the presence/absence and extent of block numbness was not evaluated prior to surgery. Post-op evaluation of this would have led to unblinding of the patient and data collection nurse, so this was not done. Thus, the two groups showed similar post-operative pain exposure data and morphine use.
2. It is possible that the administration of normal saline solution used in the placebo group may have some analgesic properties such that similar pain exposure and 6-hour post-op morphine use data resulted.
3. Most surprisingly/unexpectedly, there was a tendency for a paradoxically increased pain exposure, morphine usage and median time of PACU stay in the block group. One possible explanation for this could be the in uence of another mechanism in addition to the sensitization of the trigeminal ganglion in the production of post-op migraine pain. The spheno-palatine ganglion (SPG), a primarily parasympathetic ganglion, located in the pterygopalatine fossa (posterior to the middle nasal turbinate), is known to communicate directly with the trigeminal ganglion via several branches of the infra-orbital or maxillary V2 nerve (Figure 4.1).
Both the sphenopalatine and trigeminal nerve ganglia have been extensively studied in migraine headaches. 10,11,12 Perhaps, through some type of neuromodulation, the parasympathetic nervous system vasodilating effects of the SPG are enhanced if the trigeminal V1 and V2 sensory nerve blocks are performed. This may lead to more cerebral vasculature dilation and brain in ammatory mediator release (SP, CGRP, VIP) causing slight increase in migraine type post-op headache in the block group ( Figure 5.1,6.1). 11 A recent study 13 showed that bilateral SPG blocks as adjuvant to general anesthesia for endoscopic transsphenoidal pituitary surgery may decrease post-op pain. Blocking the SPG is more invasive and riskier 14 than the super cial trigeminal verve blocks of V1 and V2. Anesthetic pledgets, blind intra-nasal injections performed by the surgeons and specialized endoscopic guided catheters like the "SphenoCath" 12 performed by interventional radiologists under uoroscopy, can be used to administer this SPG block.
Perhaps future studies should look at randomized, controlled data using both the trigeminal nerve blocks combined with the sphenopalatine ganglion block as a multi-modal regional adjunct to general anesthesia for trans-sphenoidal pituitary resection.
As there appeared to be little difference in the pain exposure and morphine consumption post-op, the post-op nausea exposure over time did not show a signi cant difference between the Block and Placebo groups.
The post-operative time to discharge and time to last dose of narcotic from surgery start time for the Block and Placebo groups did not have statistically signi cant differences. Considering pain on discharge, again the Placebo group had less pain, although the nal median pain scores of 1/10 for Placebo and 2/10 for Block did not seem to have clinical signi cance. However, looking at the individual pain scores, the Block group had a much wider spread of pain data, showing some patients with signi cant pain on discharge.

Conclusions
In conclusion, the mechanism of the pain response to pituitary surgery via the total endo-nasal endoscopic technique appears to be more complex than expected. Likely, several nerves and ganglia interact via a complex mechanism of neuro-modulation to determine the nal post-op migraine headache intensity experienced by the patient. Simple peripheral sensory nerve supra-orbital V1 and infra-orbital V2 nerve blocks in addition to general anesthesia do not decrease pain exposure or 6-hour post-op morphine PCA consumption as compared to general anesthesia alone.

Declarations Ethics Approval and Consent to Participate
This study was approved on May 24, 2016 by the University of California, San Francisco, USA (UCSF) by the UCSF Institutional Review Board (IRB) and Independent Ethics Committee (IEC). All patients provided informed consent to participate in this trial.

Consent for Publication
All patient data collected has been deidenti ed in compliance with ethical standards. All manuscript authors have reviewed the nal manuscript and given consent for publication.

Availability of Data and Materials
All data generated or analyzed during this study are included in this published article and its supplementary information les. The corresponding author has all available raw data and can share on reasonable request.

Competing Interests
None declared

Funding and Acknowledgements
This study is an investigator-initiated trial that was funded by a Departmental Seed Grant. The authors wish to thank the University of California, San Francisco, USA, Department of Anesthesiology and Peri-Operative Care for a grant of $20,000.00 toward development and execution of this study. Figure 1 Cumulative Post-operative Intravenous Morphine Consumption in mg/kg   The Triggering of Pain via the Trigemino-Vascular Pathway