Craniotomy Pain: Trying to Do Better

At one time, it was taught that the pain accompanying intracranial surgery was minimal and, when present, dangerous to treat. As a consequence, analgesic therapy for this group of patients has generally been modest.1 It is now known that pain from intracranial surgery is more similar to than different from that caused by other surgical procedures, and it is becoming acceptable to treat it. With traditionally limited analgesic therapy, 69% of patients undergoing craniotomy report some period of pain that is moderate to severe (≥4/10) on the first postoperative day, and 48% of patients experience this level of pain during some portion of the second postoperative day.2 Patients who were dissatisfied with analgesic therapy also reported significantly greater pain, a finding that is intuitive but infrequently observed in analgesic trials,3 and may reflect active withholding of analgesics in accordance with traditional teaching about intracranial surgery. As with other types of surgery, patient-controlled analgesia (PCA) now seems effective after craniotomy and demonstrates the extent that patients seek pain relief beyond that offered, even generously, on an “as needed” (PRN) basis.4 Similar to other surgical procedures for which procedure-specific pain syndromes are now well recognized,5 long-term pain often accompanies craniotomy. Persistent pain after infratentorial procedures is well known, common, and frequently severe.6–8 In a study in which the majority of procedures were supratentorial, 56% of patients reported persistent pain 2 mo after surgery, with half of those experiencing neuropathic symptoms.9 In that study, the incidence of persistent pain was reduced several-fold by long-lasting local anesthetic infiltration of the surgical site at the conclusion of surgery.9 Thus, after intracranial surgery, perioperative pain is meaningful and may persist, but relatively simple and familiar tools are effective. Given the progress delineated above in characterizing and treating pain from intracranial surgery, what is the impetus for further improvement and where do the opportunities lie for doing so? The answers reside with the rather unique anesthetic requirements of these patients. Most anesthetic plans for intracranial surgery seek a brisk and sustained return of consciousness, minimally impeded by residual anesthetic drugs and postoperative analgesics. Furthermore, apart from the opioid-related side effects such as sedation, miosis, nausea, and vomiting that could mask incipient intracranial catastrophe, perioperative medication should not unfavorably modulate intracranial physiology. Concern exists10 about the respiratory-depressant effects of opioids which, by promoting increases in carbon dioxide tensions, could lead to increases in cerebral blood flow and increased intracranial pressure. Therefore, the development of multimodal analgesic strategies that maximize analgesia while limiting the dose of any 1 drug so as to reduce or eliminate meaningful side effects such as sedation and respiratory depression seem to be particularly desirous for patients undergoing intracranial surgery. Although generally administered for surgical considerations, the anticonvulsants11,12 and glucocorticoids13 can also reduce perioperative pain and opioid consumption and can serve as components of a multimodal analgesic strategy. It was this opportunity that was exploited in the study by Türe et al.14 in this issue of Anesthesia & Analgesia. In their study, patients received 300 mg/day of phenytoin or 1200 mg/day of gabapentin for 1 wk before surgery, the day of surgery, and postoperatively well beyond the study period. Dexamethasone administration was initiated 1 day before surgery and continued beyond the study period in all patients. After induction, patients’ lungs were ventilated with air and oxygen, and anesthesia was maintained with infusions of propofol and remifentanil titrated via processed electroencephalography. PCA infusions of morphine sulfate were used to treat postoperative pain. Patients in the gabapentin group were somewhat slower to emerge from anesthesia than those in the phenytoin group and were significantly more sedated during the first few postoperative hours. However, they reported a small but significant decrease in pain compared with the phenytoin group during the first hour after surgery and consumed one-third less morphine throughout the 48-h study period. Two postoperative imaging studies were performed in the gabapentin group and none in the phenytoin group because of concern generated by the delay between completion of surgery and emergence. One patient in the phenytoin group and none in the gabapentin group experienced postoperative seizures. Pain after 48 h was not reported. Overall, after the first few postoperative hours, patients in both groups were relatively unsedated despite receiving PCA infusions of morphine sulfate sufficient to render their pain as generally mild (≤3 on a 0–10 scale). These results demonstrate that a multimodal analgesic regimen incorporating a gabapentinoid could reduce pain to a small extent and opiate consumption more so. However, the initial sedating effects, at least for the dosages used by Türe et al., may be limiting. The study by Türe et al. indicates that gabapentinoids could serve as a component of a multimodal analgesic regimen for craniotomy. Gabapentinoids now occupy an important niche in the therapy of chronic pain and a growing, but less-defined one, for acute pain.11,12,15 Studies of patients undergoing a variety of surgical procedures have demonstrated how a single preoperative dose of gabapentin can reduce postoperative pain and opioid consumption, although few have demonstrated a concomitant reduction in opioid-related side effects.15 In these studies, typical preoperative doses of gabapentin have been 1200 mg administered orally 1–2 h before surgery. However, doses as low as 300 mg can decrease acute pain and opioid consumption compared with placebo, with no further benefits from doses larger than 600 mg according to 1 study of perioperative analgesia after lumbar discectomy.16 Although the incidence of side effects escalated with dose in that study, somnolence was not associated with any group. Less is known about the role of perioperative gabapentinoids in reducing pain that persists after surgery, although several studies in which gabapentin was administered for several postoperative days demonstrated favorable longer-term outcomes.15,17 It is, therefore, worth speculating whether the subjects in the study by Türe et al., who were maintained for many months on gabapentin because of its anticonvulsant properties, also experienced reductions in persistent and neuropathic pain. Other aspects of anticonvulsants in general and gabapentinoids in particular guide the interpretation of the study by Türe et al. Although developed as an anticonvulsant, gabapentin currently plays its largest role as an analgesic. However, it is not the only anticonvulsant with analgesic properties,12,18 and even phenytoin may have some analgesic effect.19 Consequently, the control group, as Türe et al. acknowledge, may have benefited from phenytoin administration. More generally, this raises the issue of whether newer anticonvulsants, such as levetiracetam,18 that might eventually supplant phenytoin might also exhibit favorable analgesic properties. Furthermore, if perioperative anticonvulsant therapy can decrease persistent pain after surgery, perhaps the better-recognized persistence of pain after infratentorial procedures6–8 reflects, in part, the decreased likelihood of administering anticonvulsants for such surgery. An additional consideration with gabapentin administration is its ability to attenuate opioid-induced hyperalgesia.20 This attribute of gabapentin could account for a portion of the opioid sparing reported by Türe et al. and may be an important issue for neurosurgical procedures, which are frequently performed with opioid-based anesthetic techniques. Similarly, nitrous oxide, likely through N-methyl-d-aspartate antagonism, can also decrease injury-induced hyperalgesia and opioid-induced hyperalgesia.21 Because none of the patients in the study by Türe et al. received nitrous oxide, it is unknown whether the benefits of gabapentin administration would have been observed had patients in their study also received nitrous oxide. Overall, Türe et al. have performed an important service by demonstrating that anticonvulsants administered during neurosurgical procedures can measurably affect perioperative pain and opioid analgesic use. Although studies demonstrate that single preoperative doses of these drugs can enhance postoperative analgesia when used for other types of procedures, it remains unknown whether the more chronic administration of anticonvulsants that is common in neurosurgical patients can reduce persistent postsurgical pain and the development of neuropathic symptoms after intracranial surgery. The delayed emergence and sedation in the immediate postoperative period reported by Türe et al. seem to proscribe the primary use of gabapentin as an anticonvulsant to reap secondary analgesic benefits. However, the role of a single, perhaps smaller, preoperative dose of a gabapentinoid, particularly when anticonvulsants would not otherwise be administered, has not yet been determined. Furthermore, as with other analgesic studies of perioperative anticonvulsants,15 Türe et al. provided little evidence that the reduction in opioid consumption was associated with any reduction in opioid-related side effects, one of the stated goals of multimodal analgesic regimens. Perhaps reduction of opioid consumption sufficient to reduce opioid-related side effects might be possible only with additional analgesic modalities such as acetaminophen administration and preoperative wound infiltration/scalp block.2,9 However, similar to gabapentin,15 reductions in opioid consumption caused by acetaminophen administration22 did not reduce opioid-related side effects for other types of surgery. Finally, as other anticonvulsants supplant phenytoin use, it may be important to recognize their contribution to perioperative analgesia. Overall, the information provided by Türe et al. and the questions it poses will help to make analgesic therapy for craniotomy somewhat better.