Parturients Infected with Human Immunodeficiency Virus and Regional Anesthesia : Clinical and Immunologic Response

With approval of our institutional Committee on Human Research and patient informed consent, we prospectively evaluated the obstetric course and anesthetic management of 30 patients infected with HIV-1 as part of the Bay Area Perinatal AIDS Center research project. Patients underwent extensive medical and laboratory evaluation before delivery and at least 6 months postpartum as part of a long-term study of pregnancy and HIV-1 disease. Follow-up visits were planned at 2, 4, 6, 9, and 12 months postpartum and then every 6 months thereafter. Infection with HIV-1 was confirmed by detection of antibodies to the viral protein by the enzyme-linked immunosorbent assay (DuPont, Biotech Research Laboratories, Rockville, MD) with positive results confirmed by Western Blot assay (Wironostika Teknika, Organon, OSS, The Netherlands) immunoblot tests. The immune function was monitored BY CD⁴sup +(helper/inducer) T-cell lymphocyte and DC⁸T-cell lymphocyte counts, critical modulators of the immune system, serum p24 antigen (p24), and beta²microglobulin evaluations. The beta²microglobulin levels in serum were determined by radioimmunoassay (Pharmacia, NJ), and the HIV-1 p24 antigen was detected by enzyme-linked immunosorbent assay. Routine blood analysis for leukocyte count and hematocrit were performed. To assess the severity or progression of the disease, patients were screened for use of antiviral drugs such as zidovudine (Azidothymidine) and chemoprophylaxis for Pneumocystis carinii. Disease status also was assessed by the occurrence of AIDS indicator diseases such as P. carinii pneumonia, Kaposi's sarcoma, candidiasis, cytomegalovirus disease, Mycobacterium avium, cervical dysplasia, cervical cancer, and/or oral hairy leukoplakia.

The anesthetic management was determined by the patient's request for elective pain relief or the necessity of surgical intervention requiring anesthesia. Decisions regarding anesthetic intervention were made in the usual clinical manner by the anesthesiologist in consultation with the obstetrician. Choices for pain relief were epidural or spinal anesthesia (regional anesthesia) or no anesthetic intervention. General anesthesia also was a choice for cesarean section. Patients with no anesthetic intervention had small amounts of opioid administered according to the obstetrician's orders or no analgesia. The newborn HIV infection rate was documented. After delivery, the patient was carefully monitored with special attention given to the immediate postpartum period and 4–6 months' follow-up for disease-related complications of any kind and their possible relationship to anesthesia. Continued extensive immunologic monitoring (CD⁴sup +, beta²microglobulin p24) was undertaken postpartum, as well as careful evaluation of the potential complicating infections related to AIDS. Neurologic symptoms, including unusual headaches, problems with coordination or balance, or numbing or tingling of arms, legs, or feet, were specifically surveyed (part of a follow-up questionnaire) more than 4–6 months postpartum.

Data were analyzed using the unpaired Student's t test when comparing the two groups, regional anesthesia group versus the no anesthetic intervention group. A P value of < 0.05 was considered significant. All values are expressed as the mean plus/minus SD.

The study groups, regional anesthesia (N = 18) versus no anesthetic intervention (N = 12) were similar with respect to maternal age, weight, height, parity, and history of intravenous drug abuse (Table 1). The patients were divided almost evenly with regard to parity and history of intravenous drug abuse. In the group who received regional anesthesia, 16 received epidural anethesia and two received spinal anesthesia, the latter for cesarean section. No patients received general anesthesia. There were three unintended dural punctures in the epidural group (18.8%) but no postdural puncture headaches in this subgroup or the spinal anesthesia group. There were no autologous epidural blood patches, either prophylactically or postpartum. There was a 20% cesarean section rate in the study patients, two of whom received spinal anesthesia; the remaining four received epidural anesthesia.

Immunologic values and blood counts before delivery and 4–6 months postpartum are compared in Table 2and Table 3. There were no significant differences between the groups, regional anesthesia and no intervention, or between predelivery and postpartum values within the two groups. The majority of patients had a CD⁴T-cell count greater than 200 cells/mm³both antepartum and postpartum. However, the mean CD⁴sup + T-cell count was approximately half the normal (no disease) cell count in both groups. The CD⁴sup + T-cell counts tended to remain stable and did not progress in most patients during the study period, and the p24 and beta²microglobulin values were not significantly changed.

There were no significant differences between groups in AIDS indicator diseases, but there were four cases of P. carinii pneumonia in the regional anesthesia group; two cases were diagnosed antepartum and two cases postpartum (Table 4). Five mothers in the regional anesthesia group and four in the no anesthesia group were given zidovudine therapy before delivery; zidovudine therapy was initiated postpartum in eight mothers in the regional anesthesia group and seven in the no intervention group, as indicated for a CD⁴sup + T-cell count of less than 500 cells/mm³. Oral hairy leukoplakia and persistent generalized lymphadenopathy were the most common indicator diseases or conditions. No cases of pulmonary mycobacterial disease or Kaposi's sarcoma were detected in the patients studied. Sixty-eight percent (n = 19) of the infants were HIV-1 negative with 21%(n = 6) having an indeterminate HIV status (probably negative) at this time, whereas 11%(n = 3) were HIV-1 positive.

Three dural punctures occurred when performing epidural anesthesia, but the epidural injections were placed successfully at another interspace without complications. Epidural catheters remained indwelling, as clinically required. There were no neurologic of infectious complications or any changes in the peripartum period or during the 4–6 month neurologic follow-up period to suggest that regional anesthesia had adverse sequelae. There were non superficial or deep infections noted at the catheter placement site. Only the visit at 4–6 months is reported here, but long-term follow-up of those patients who have continued in the study with numerous clinic visits reveals no unusual neurologic sequelae that would not be expected in the natural course of HIV-1 infection.

This study found no change in the disease progression or serologic markers in the parturient infected with HIV-1 that could be linked to the anesthetic management at delivery. The CD⁴sup + T-cell counts, a common indicator of disease progression (i.e., lower count, more advanced disease) remained relatively stable. Two other surrogate markers, the p24 antigen and beta²microglobulins, were evaluated. The p24 antigen is an HIV-1 core protein, and its presence in the blood has been associated with maternal-fetal viral transmission (vertical transmission) and possibly a more rapid progression of the disease. [10] beta²Microglobulin has been suggested as a predictor of disease progression because it indicates immune activation resulting from HIV-1 infection. Increased cord beta²microglobulin also may be linked to vertical HIV-1 transmission. [11] We measured only maternal beta²microglobulin levels in this study. Both of these surrogate markers remained stable when comparing the ante- and postpartum values. Whereas the use of p24 antigen, beta²microglobulins, and CD⁴sup + T-cell counts in some situations has been questioned with regard to their usefulness as indicators of disease progression, they are used frequently in many centers as clinical and/or research tools and, in this study, provide further evidence of the clinical stability of these patients in the peripartum period.

We found no neurologic or infectious problems in the immediate postpartum period or at the 4–6-month follow-up visits. This latter finding is in agreement with the related report of no sequelae related to performance of epidural blood patches in nonobstetric patients who are HIV-1-positive [12] and initial findings similar to ours. [13],** However, in a study of patients in whom epidural catheters were placed for chronic pain management, Du Pen et al. [9] found nine infections (six at the catheter-exit site, two deep-track infections, and one epidural infection) in patients with AIDS. These infections cleared after antibiotic therapy. Although the difference in findings may lie solely in the use of prolonged catheterization versus acute use of regional anesthesia, it may be that most of our obstetric AIDS patients were in the early stages of the disease (CD⁴sup + T-cell count > 200 cells/mm³). Indeed, one limitation of our study is that the majority of our patients had CD⁴sup + T-cell counts greater than 200 cells/mm³(largely Centers for Disease Control class A²or B²) and, thus, were relatively healthy. Therefore, our findings might not apply to patients with further advanced HIV-1 disease.

This is the first prospective study of the clinical and immunologic function of the HIV-1 positive parturient to include relevant evaluation of the effects of anesthetic management on the disease and to evaluate maternal outcome. Because individuals infected with HIV-1 have depressed immune function, there is obvious concern that anesthesia and surgery may further impair the patient's immune status with an increased risk of infection, extension of HIV-1 disease, or poor wound healing. There is a critical lack of information regarding the effects of anesthesia in these patients, and that which exists is often only contradictory medical opinions. [7,8,14].

However, there are data documenting isolated, transient immunologic consequences secondary to anesthesia and surgery in patients with normal immune systems, but this is rarely correlated with adverse clinical outcome. [15] Markovic et al. [16] demonstrated that halothane and isoflurane induced “anesthetized natural killer (NK cells) system,” which could lead to changes in clinical management as we better understand the mechanisms involved. [17] However, we did not measure CD⁴sup + T-cells or other surrogate markers frequently enough (repeated acute measurements within the days before and immediately after the anesthetic intervention) to see such transient effects. Despite this, our data indicate no real change over time in the function of the parturient's immune system or the disease process that were related to anesthesia in the parturient infected with HIV-1.

The results of the current study support the growing evidence that pregnancy has no obvious effect on the progression of HIV-1 disease in generally asymptomatic parturients. [18–20] The risk of vertical HIV-1 transmission appears to be the chief concern associated with pregnancy. Our newborn HIV-1 infection rate of approximately 11% is in marked contrast to early studies reporting rates as high as 73%, [15] and our reported lower rate may be decreased further by maternal zidovudine therapy. [21].

Of particular concern in the patient infected with HIV-1 is the extensive neurologic complications that may appear early in the disease and progress with time. [6,15] HIV-1 infects the central nervous system early in the course of the disease, and isolation of the virus from cerebrospinal fluid specimens of seropositive patients with and without neurologic symptoms has been reported. [22–23] A recent report would suggest that failure to obtain positive cerebrospinal fluid cultures from patients who are HIV-1-seropositive probably results from sampling error, not absence of the virus from the cerebrospinal fluid. [24] Thus, it is our view that the anesthesiologist performing regional anesthesia need not fear contamination of the cerebrospinal fluid with HIV-1, for it is present very early in the disease.

This prospective study of parturients infected with HIV-1 (n = 18) is too small to declare that regional anesthesia is “safe” in this patient population. Further, the relatively stable HIV-1 infections in our patients may not be representative of a larger group of parturients. However, in the patients studied, there were no neurologic or infectious complications related to the anesthetic or obstetric course. In the immediate postpartum period, the immune function measurements remained essentially unchanged, as did the severity of the disease. Thus, we believe these data, though limited, support the use of regional anesthesia in the parturient infected with HIV-1 after careful clinical evaluation and review of routine data.

The authors thank Sara Tarbox, for her administrative support, and Winifred von Ehrenburg, for her editorial assistance.

*Centers for Disease Control and Prevention: HIV/AIDS Surveillance Report. 5. No. 3. Atlanta. Centers for Disease Control and Prevention. 1993, p 6.

**Gershon RY, Williams DM, Berry AJ, Lindsay MK, Watkins C: The effect of anesthesia on the infected parturient (abstract). ANESTHESIOLOGY 77:A1037, 1992.