Share this article on:

Oral Herpes Simplex Reactivation After Intrathecal Morphine: A Prospective Randomized Trial in an Obstetric Population

Section Editor(s): Birnbach, David J.Davies, Paul W. MD; Vallejo, Manuel C. MD; Shannon, Kelly T. MD; Amortegui, Antonio J. MD; Ramanathan, Sivam MD

doi: 10.1213/01.ANE.0000153013.34129.A7
Obstetric Anesthesia: Research Report

There is now evidence for an association between the use of epidural morphine and reactivation of herpes simplex labialis (HSL). There are no studies that definitively demonstrate the relationship between HSL reactivation and spinal intrathecal morphine. To investigate this relationship, we randomized and prospectively studied 100 obstetric patients with a history of HSL undergoing cesarean delivery under spinal anesthesia. One group received intrathecal morphine plus IV morphine via patient-controlled analgesia (ITM+PCA group) for postoperative analgesia, and a second group received only IV morphine via patient-controlled analgesia for postoperative analgesia (PCA-only group). Patients were followed for a 30-day period. In the ITM+PCA group 19 (38%) patients had HSL reactivation whereas eight (16.6%) had HSL reactivation in the morphine PCA-only group (P = 0.028). The incidence of pruritus in the ITM+PCA group was also more frequent in the early postoperative period. Our data show HSL reactivation in both the ITM+PCA group and PCA-only morphine group, with a more frequent incidence in the ITM+PCA group.

IMPLICATIONS: This study demonstrated a more frequent incidence of herpes simplex labialis reactivation and pruritus among patients receiving intrathecal morphine.

University of Pittsburgh School of Medicine, Department of Anesthesiology, Magee-Womens Hospital, Pittsburgh, Pennsylvania

Supported, in part, by a University of Pittsburgh Department of Anesthesiology Seed Grant.

Accepted for publication November 23, 2004.

Address correspondence to Manuel C. Vallejo, MD, Department of Anesthesiology, Magee-Womens Hospital, 300 Halket Street, Pittsburgh PA 15213. Address e-mail to

Intrathecal morphine (ITM) is widely used for postoperative analgesia after cesarean delivery. Common side effects include pruritus in the distribution of the trigeminal nerve, urinary retention, nausea, and, less frequently, delayed respiratory depression (1). It has been clearly demonstrated that there is a relationship between the use of epidural morphine and herpes simplex labialis (HSL) reactivation (1–5). However, there has not been a prospective randomized trial investigating the relationship between spinal ITM and HSL reactivation. The purpose of this study was to compare the incidence of HSL reactivation with morphine administered either intrathecally or parenterally. Furthermore, Crone et al. (1,2) proposed a relationship between epidural morphine, facial pruritus, and reactivation of HSL. We therefore evaluated the role of pruritus in relation to HSL recurrence under spinal anesthesia with ITM.

Back to Top | Article Outline


After approval by the local IRB, 100 ASA physical status I–II patients provided informed consent to participate in this study. Enrollment was limited to include only those patients with a positive history for HSL undergoing elective cesarean delivery. Only those patients describing recurrent lesions occurring in the same perioral site were included. Patients with an allergy to morphine, pregnancy-induced pruritus, current active perioral herpes lesions, diabetes mellitus, preeclampsia/eclampsia, or those with any immunomodulating diseases such as human immunodeficiency virus, myasthenia gravis, multiple sclerosis, and systemic lupus erythematosus were excluded.

All patients were seen in the preoperative evaluation clinic by an anesthesiologist and informed verbal and written study consent were obtained. In addition to obtaining the preoperative medical history and physical examination, the patient’s lips, mouth, and oropharynx were visually inspected to exclude the presence of active preoperative HSL lesions. Additionally, the anatomical location of past perioral lesions were ascertained from the patient history and documented for later comparison. Eight mL of peripheral maternal venous blood was obtained for herpes simplex virus (HSV) immunoglobulin (Ig)G antibody titer testing. HSV IgG antibody titer testing was performed using enzyme-linked immunoabsorbent assay using kits obtained from a commercial firm (Dako Corporation, Carpinteria, CA). Mouthwashes and swabs of the gingivobuccal folds were obtained and transported to the laboratory in 5 mL of Hank’s balanced salt solution. These samples were inoculated onto tissue culture layers of human fibroblasts and rabbit kidney cells (Biowhittaker, Walkerville, MD). The cultures were examined for 5 consecutive days. Specimens showing cytopathic effect were confirmed for the presence of HSV with immunoperoxidase using 3, 3 Diaminobenzidine as chromogen (Sigma Chemical Co, St. Louis, MO). All laboratory tests were done at the microbiology/virology laboratory at Magee-Womens Hospital.

Patients who met inclusion criteria were randomized by a computer-generated program into one of two treatment groups. Group 1 received spinal ITM plus IV morphine via patient-controlled analgesia (ITM+PCA) for postoperative analgesia. At our institution, patients who receive ITM are also are given IV PCA morphine as a backup for postoperative breakthrough pain. In Group 2, only IV morphine was used via PCA postoperatively without spinal intrathecal morphine (PCA-only group). Dosing variables for IV PCA in both groups were as follows: 4 mg initial loading dose, 1 mg bolus dose with a 6-min lockout interval, and a 4-h maximum dose of 30 mg.

On the day of surgery, all patients were given 30 mL of sodium citrate PO. They received 1200 mL of Ringer’s lactate solution IV before spinal anesthesia. A standardized spinal anesthetic consisting of hyperbaric bupivacaine (12–14 mg) plus preservative-free morphine (0.25 mg) was given in the ITM+PCA group and only bupivacaine (12–14 mg) was given in the PCA-only group. A bilateral sensory T4 dermatome level was recorded before incision. The uterus was tilted to the left with a wedge placed under the right hip.

After surgery, patients were followed for 3 postoperative days (POD) in the hospital by an independent blinded observer. The patient’s mouth and oropharyngeal mucosa were examined for the presence of any lesions. The size and location of the lesions were noted. Mouthwashes as well as swabs of lesions (if any) along with the swabs of the gingivobuccal folds were sent for culture. Patients were also asked whether they had pruritus and how severe it was. The severity of pruritus was assessed by using a Visual Analog Scale (VAS, 0–100, 0 = no pruritus, 100 = worst pruritus ever). Upon discharge, study patients were contacted by telephone on POD 10 and 30. During the telephone interview, patients were asked about any possible recurrence of HSL lesions and the condition of their baby. They were also asked if they used any topical treatment for their HSL lesion.

For sample size calculation, it was assumed that the incidence of reactivation in the ITM+ PCA group would be five times more frequent than that in the PCA-only group. At an α = 0.05 and β = 0.8, it was determined that 50 patients per group would be needed to detect this difference (6). Results were expressed as mean ± sd where applicable. Demographic data were analyzed using Student’s t-tests. Pruritus VAS scores were analyzed with the Mann-Whitney U-test. Relationships between the incidence and severity of pruritus and the likelihood of reactivation were analyzed using logistic regression. Frequency data were compared using χ2 analysis.

Back to Top | Article Outline


Two women assigned to the PCA-only group were excluded because of failed spinal anesthesia that required general anesthesia. Further data collection was suspended in these two patients. Their blood samples were not processed for HSV-IgG antibody titer. Hence, 50 patients in the ITM+PCA group and 48 patients in the PCA-only group were included in the analysis. There were no differences in height, weight, the number of HSL attacks per year, or in the number of previous oral, lip, and other facial HSL lesions (Table 1). Most patients reported lesions on the upper and lower lips (Table 1). The number of patients requiring PCA morphine and the total morphine dose used in the first 24 h was significantly more frequent in the PCA-only group compared with the ITM+PCA group (Table 1). No patient from the ITM+PCA group required an initial dose whereas all patients from the PCA-only group required an initial dose. Seventy patients from the original sample were found to be seropositive for HSV-1 (33 individuals from the ITM+PCA group and 37 from the PCA-only group; P = not significant).

Table 1

Table 1

Twenty-seven patients from both groups reported postoperative HSL lesions (19 individuals from the ITM+PCA group and 8 from the PCA-only group; P = 0.028) (Fig. 1). One patient from the ITM+PCA group reported extensive multiple lesions in the mouth and left side of the face starting 1 wk later for which she was placed on oral acyclovir therapy by her primary care physician. None of the other patients used any topical or oral treatment. HSL recurrences were reported at the same site in 93% of patients in both groups; in the remaining patients, an additional new site was reported. In both groups, no significant difference was noted with the HSL recurrence rate between IgG seropositive and IgG seronegative patients (Table 2). The mean IgG antibody level in those who reactivated was 90 ± 69 U compared with 80 ± 55 U in those who did not (P = not significant).

Figure 1

Figure 1

Table 2

Table 2

There was no difference in preoperative pruritus VAS scores between groups. However, postoperative VAS pruritus scores were higher in the ITM+PCA group in the postanesthesia care unit (PACU) and on POD1 with no significant differences on POD2 and POD3 (Table 3). The frequency and severity of pruritus was also increased in the ITM+PCA group (Table 3). Forty percent of patients who had pruritus in the PACU reactivated in the ITM+PCA group and 20% of patients in the PCA-only group reactivated (P = 0.16) (Table 4). When pruritus scores on POD1 were considered, 36% of patients had pruritus reactivated in the ITM+PCA group and only 22% had pruritus reactivated in the PCA-only group (P = 0.2) (Table 4). No statistically significant relationship could be demonstrated between the severity of pruritus and the incidence of recurrences in either group (Table 4).

Table 3

Table 3

Table 4

Table 4

Six women in the ITM+PCA group and two women in the PCA-only group had positive viral isolation before anesthesia (P = not significant) and one patient from each of these subgroups had a recurrence. In the ITM+PCA group, 6 patients reactivated before discharge from the hospital, 5 on the fourth day and the others between days 5 and 9. In the PCA-only group, 3 patients reactivated before hospital discharge and one reactivated on each of the days 4, 5, 7, 9, and 12. Of the 9 clinical recurrences in the hospital, 3 lesions were culture-positive, 2 from the ITM+PCA group and 1 from the PCA-only group. Mouthwashes obtained from three patients in the ITM+PCA group on POD1 were culture-positive and one of these patients had a clinical recurrence. Only one mouthwash obtained on POD1 was culture-positive without a clinical recurrence in the PCA-only group. The mean time to onset of a HSL lesion as identified by the examiner or self-reported by the patient was not statistically different between groups (Table 1).

Back to Top | Article Outline


Our data show that, regardless of the route of morphine administration (parenteral or neuraxial), HSL reactivation occurs in parturients. However, patients who received ITM+PCA experienced a more frequent reactivation compared with those who received PCA only.

Despite widespread agreement that epidural morphine increases the incidence of HSL (1–5), controversy remains regarding its effect when used intrathecally (7,8). This is the first study addressing the issue of ITM prospectively. Past reports have shown vari-ability between lesions diagnosed clinically and then confirmed by laboratory culture (2,5). In the studies by Crone et al. (2) and Boyle (5) only 43% and 65% of reactivated lesions were culture-positive, respectively. We do not know the real incidence of culture-positive recurrences in our study because 66% of recurrences in both groups occurred after discharge from the hospital. The incidence of viral shedding was infrequent in our study preoperatively and postoperatively. Crone et al. (2) reported a similar finding. A potential cause for the lack of reliability of laboratory culture findings may be in part attributable to deficiencies in specimen collection and transport before inoculation into tissue culture or lack of viral shedding from the healing encrusted lesions.

Pebody et al. (9) showed that seropositivity differs from country to country in Europe; rate ranges from 52% to 84%. Bünzli et al. (10) reported a seroprevalence rate of 80% in Switzerland. The seropositivity rate of 70% in our study is within this range. Even though all our patients gave a history of “cold sores,” only 71% were seropositive. We believe the difference between a positive history and seropositivity is that the so-called cold sores may have been nonspecific ulcerations unrelated to HSV-1 virus. Posavad et al. (11) showed that many patients who remained seronegative for HSV infection possessed specific T-cell immunity to the virus. Thus it appears likely that patients who possess significant immunity to HSV-1 do not seroconvert (11). In our study, seropositivity was not a specific predictor of HSL recurrence after surgery.

Pregnancy is an immunosuppressed state, which enables the mother to tolerate the fetal allograft (12). This, together with increased physical and emotional stress of surgery, may predispose to the increased reactivation rate of HSV. Stress, whether physical or emotional, plays an important role in HSV recurrence (3,13). Halford et al. (14) studied reactivation of HSV-1 virus in the trigeminal ganglion cells of mice latently infected with the virus. They found that heat stress predisposes to activation of HSV-1 virus (14). Halford et al. also found that signs of HSV-1 reactivations were hastened by the addition of dexamethasone to the trigeminal ganglion cultures (14). Peripheral sympathetic denervation alters immune function through activation of the hypothalamus-pituitary axis and increases serum corticosterone levels (15). Surgery is also associated with emotional and physical stress, which may further predispose to reactivation of HSL even in the absence of neuraxial opioids. Thus, reactivation can occur even in the absence of neuraxial morphine and, in fact, it can occur in response to surgical trauma (16). Other triggers include ultraviolet light, orofacial surgery, and plastic surgery on the face (3,17,18). The more frequent incidence of HSV reactivation seen with PCA morphine in our study may have been the result of methodological or population differences. It is not clear whether shorter acting opioids such as fentanyl can cause HSL reactivation in obstetrical patients and no such report exists (16).

Several antiviral drugs have recently become available (18). These may be administered topically, orally, or systemically. They have been shown to shorten the duration of the disease process, especially when started during the prodromal period. However, many recurrences occur without a significant prodromal period (18). Only one patient in our study received antiviral therapy for an extensive HSV lesion.

The HSV can remain dormant for a long time in the trigeminal ganglia and can undergo transformation into an infectious form. The mechanism by which neuroaxial morphine causes HSL reactivation is unclear. The fact that both facial pruritus and HSL reactivation affect the trigeminal nerve distribution has led many investigators to suggest that scratching causes skin damage predisposing to HSL reactivation (1,3,4). Although our data do not suggest a cause and effect relationship, we, like other authors, have demonstrated more pruritus with neuroaxial morphine (1–5). The increased trend towards itching in the ITM+PCA group did not attain statistical significance in our study. A larger study would perhaps clarify this issue.

The “scratch center” is believed to be located near the floor of the fourth ventricle at the lower end of the medulla oblongata (3). Neural inputs from this center and afferent impulses to the trigeminal nuclei allegedly convey the sensation of facial itch (3). Rostral spread of opioids can penetrate the scratch center or the trigeminal nuclei. Facial itch indicates a direct effect of morphine on the opioid receptors located at these sites and need not necessarily be a harbinger of HSL-reactivation (3). Intrathecal administration of morphine bypasses the blood-brain barrier and produces a cerebrospinal fluid concentration far in excess of that produced by IV administration (19). Consequently, any direct action of morphine on the trigeminal nuclei is likely to be more pronounced with intrathecal administration.

There are other theories of HSL reactivation that do not invoke itching as a predisposing factor. One theory is that morphine acts at a molecular level to stimulate replication of the herpes simplex viral genome in the infected trigeminal nuclei cells (3,5). Morphine is believed to cause modulation of neural function via opioid receptors on the trigeminal nuclear cell membrane, which are coupled to G-proteins (1–3). G-proteins affect membrane ion channel conductance and adenylate cyclase activity resulting in altered cyclic adenosine monophosphate dependent protein kinases (3). The altered protein kinases may be responsible for accelerated gene transcription and viral replication (3).

An alternative mechanism by which morphine may cause HSL reactivation is through its effect on human lymphocyte function (3). Patients often shed live viral particles into saliva but remain asymptomatic because the infected cells are removed by T cell-mediated mechanisms (3). When T cell-mediated immunity is impaired, infected cells are cleared more slowly and an active ulcerative lesion can develop despite the presence of humoral immunity (3). Morphine exposure reduces the expression of T-cell surface receptors responsible for resetting (3). This is an opioid receptor-mediated effect reversible by naloxone (3). The effective dose of morphine to produce this effect is thought to be 1000 times smaller when administered directly into the cerebral ventricles than when administered systemically (3).

In summary, there is a more frequent incidence of HSL reactivation and pruritus in parturients who receive ITM for cesarean delivery.

Back to Top | Article Outline


1. Crone LA, Conly JM, Clark KM, et al. Recurrent herpes simplex virus labialis and the use of epidural morphine in obstetric patients. Anesth Analg 1988;67:318–23.
2. Crone LA, Conly JM, Storgard C, et al. Herpes labialis in parturients receiving epidural morphine following cesarean section. Anesthesiology 1990;73:208–13.
3. Boyle RK. A review of anatomical and immunological links between epidural morphine and herpes simplex labialis in obstetric patients. Anaesth Intensive Care 1995;23:425–32.
4. Gieraerts R, Navalgund A, Vaes L, et al. Increased incidence of itching and herpes simplex in patients given epidural morphine after cesarean section. Anesth Analg 1987;66:321–4.
5. Boyle RK. Herpes simplex labialis after epidural or parenteral morphine: a randomized prospective trial in an Australian obstetric population. Anaesth Intensive Care 1995;23:433–7.
6. Glantz SA. Primer of biostatistics, 4th ed. New York: McGraw-Hill, 1977.
7. Abouleish E. Intrathecal morphine as a cause for herpes simplex reactivation should be scratched out. Anesthesiology 1991;75:919.
8. Pennant JH. Intrathecal morphine and reactivation of oral herpes simplex. Anesthesiology 1991;75:167.
9. Pebody RG, Andrews N, Brown D, et al. The seroepidemiology of herpes simplex virus type 1 and 2 in Europe. Sex Transm Infect 2004;80:185–91.
10. Bünzli D, Wietlisbach V, Barazzoni F, et al. Seroepidemiology of herpes simplex virus type 1 and 2 in western and southern Switzerland in adults aged 25–74 in 1992–93: a population-based study. BMC Infect Dis 2004;4:10.
11. Posavad CM, Wald A, Hosken N, et al. T cell immunity to herpes simplex viruses in seronegative subjects: silent infection or acquired immunity? J Immunol 2003;170:4380–8.
12. Somerset DA, Zheng Y, Kilby MD, et al. Normal human pregnancy is associated with an elevation in the immune suppressive CD25+ CD4+ regulatory T-cell subset. Immunology 2004;112:38–43.
13. Buske-Kirschbaum A, Geiben A, Wermke C, et al. Preliminary evidence for herpes labialis recurrence following experimentally induced disgust. Psychother Psychosom 2001;70:86–91.
14. Halford WP, Gebhardt BM, Carr DJ. Mechanisms of herpes simplex virus type 1 reactivation. J Virol 1996;70:5051–60.
15. Leo NA, Callahan TA, Bonneau RH. Peripheral sympathetic denervation alters both the primary and memory cellular immune responses to herpes simplex virus infection. Neuroimmunomodulation 1998;5:22–35.
16. Sinatra RS, Ayoub CM. Postoperative Analgesia: epidural and spinal techniques. In: Chestnut DH, ed. Obstetric anesthesia: principles and practice, 2nd ed. St. Louis: Mosby Inc, 1999:521–55.
17. Gilbert S. Improving the outcome of facial resurfacing-prevention of herpes simplex virus type 1 reactivation. J Antimicrob Chemother 2001;47:29–34.
18. Raborn GW, Grace MG. Recurrent herpes simplex labialis: selected therapeutic options. J Can Dent Assoc 2003;69:498–503.
19. Anderson HB, Christensen B, Findlay JW, Jansen JA. Pharmacokinetics of intravenous, intrathecal and epidural morphine and fentanyl in the goat. Acta Anaesthesiol Scand 1986;30:393–9.
© 2005 International Anesthesia Research Society