Anesthesia and the HIV infected parturient: a retrospective study

Inlernationol Journal of Obsrerric Anesthesia (1997) 6, 7641 0 1997 Pearson Professional Ltd

ORIGINAL

ARTICLE

Anesthesia and the HIV-infected parturient: a retrospectivestudy R. Y. Gershon, D. Manning-Williams Department of Anesthesiology,Emory University School of M e d icine, Atlanta, Georgia, USA SUMMARY W e wished to determine whether immunosuppression and/or acceleration of human immunodeficiency virus (HIV)-associated disease is related to the m o d e of anesthesia in the HIV-infected parturient. Ninety-six known HIV-infected asymptomatic parturients who delivered between January 1990 and January 1992at Grady Memorial Hospital were reviewedfor pre-delivery health status, m o d e of delivery, m o d e of anesthesia and peripartumpost-partum complications. Statistics used x2 analysis. There were 36 regional anesthetics(28 epidural, 8 spinal), 11 general anesthetics,22 local anesthesia/intravenoussedation and 27 patients who did not receiveanesthesia.There were no differencesin the 2448 h and 4-6 week complication rates.None of the women studied developed neurologic sequelaefor 2 years after anesthesia.Thirty-one of the 96 women had CD4/CD8 T-cell lymphocyte data at second trimester and 2448 h post partum which showed no worsening of maternal immune status in women receiving regional anesthesia, local anesthesia/intravenoussedation or no anesthesia.W e believethat both regional and generalanesthesiacan be performed safely on the asymptomatic HIVinfected parturient. The choice of anesthesiashould be based on the usual obstetric and clinical considerations.

HIV infection is associatedwith a variety of neurologic disorders: acquired immunodeficiency syndrome (AIDS) dementia complex (ADC), acute or chronic non-bacterial m e n ingitis, chronic headaches,vacuolar degenerationof the spinal cord and polyneuropathy.” HIV-induced immunocompromise may render the patient susceptibleto intracranial toxoplasmotic mass lesions, central nervous system (CNS) lymphomas, cryptococcal m e n ingitis and cytomegalovirus retinitis.” Autopsy studies on patients who died of AIDS show a 7&80% incidence of neuropathologic changes.“,‘*The concern has therefore been raised regarding the risk of spinal and epidural anesthesia-relatedinfections and neuropathologic exacerbations.‘3 m ’5 There are presently no data to support the contention that either general or regional anesthesia affects HIV-related diseasein the parturient. Hughes et alI6 prospectively studied 30 HIV-infected parturients. Regional anesthesia was a d m inistered in 18 women, and 12 received small doses of opioids or no analgesia. Clinical outcome showed no infections, complications or neurologic changes in the peripartum period. Immunologic outcome data compared CD4 and CD8 pre-delivery levels to those 4-6 months post partum, from which they concluded that immune function was stable in the peripartum period. The purpose of this retrospective study was to determine whether immunosuppression and/or acceleration of

INTRODUCTION The immunomodulating effects of spinal and epidural anesthesia on the non-HIV infected population may be negligible, while general anesthesiamay be associated with a transient immunodepression.‘,’Evidence exists for an altered immune status during normal pregnancy, including a decreasein CD4 lymphocyte count3 and functiotP as well as natural killer (NK) cell activity.’ Impairment of the immune responsemay affect the course of viral infection during pregnancy.* Tonnesen & W a h lgreen’compared general anesthesia to epidural anesthesia in women undergoing total abdominal hysterectomy for benign disease.General anesthesia decreasedCD4 and B-cell numbers up to 3 days after operation, while patients in the epidural group remained immunologically intact. Concern exists for the possibility of further immune function compromise and peripartum complications in the HIV-infected parturient after general anesthesia.

Raphael Y. Gershon MD, Chief, Obstetric Anesthesiology,

Grady Memorial Hospital, Assistant Professor, Department of Anesthesiology, Dawn Manning-Williams MD, Assistant Professor, Emory University School of Medicine, Thomas K. Glenn Memorial Building, 69 Butler Street S E, Atlanta, Georgia 30303, USA, Tel: + 1404 616 8760/9655 Fax: + 1 404 616 9660. Correspondence to: Raphael Y. Gershon MD. 76

Anesthesia and the HIV-infected parturient

77

Anemia [hct < 30%) Intravenous Drug Abuse Venereal

Disease Hepatitis

Positive PPD Lymphadenopathy

$

,

, , ,

0

10

20

Parturients

,

,

,

,

30

40

50

(%) (n=96)

Fig. 1 Pre-delivery morbidity in the HIV-infected parturient. hct = hematocrit; PPD = purified protein derivative (of tuberculin).

HIV-associated disease relate to the mode of anesthesia in the HIV-infected parturient. MATERIALS

AND METHODS

Patients were selected retrospectively from participants in an ongoing longitudinal study on the relationship between pregnancy and HIV-related disease. Ninety-six known HIV positive (enzymelinked immunosorbent assay and Western immunoblot confirmed) parturients who delivered between January 1990 and January 1992 at Grady Memorial Hospital, and who gave informed consent, were included in this Institution Review Board approved study. Total lymphocyte, CD4 and CD8 cell counts were determined during the second trimester and 2448 h post partum. T-cells were quantitated with monoclonal antibodies (Cytostat Coulter Clone, Coulter Immunology, Hialeah, Florida). Cells were analyzed by indirect immunofluorescence in an EPICS PROFILE II (Coulter Electronics, Hialeah, Florida). Charts were subsequently reviewed using a prepared questionnaire for pre-delivery medical and obstetric demographics, mode of delivery, mode of anesthesia and peripartum complications. Peripartum complications were assessed by a senior obstetric resident. The 6-week follow-up was performed by a senior obstetric resident in the obstetric gynecology HIV clinic, after which the women were followed in the infectious disease HIV clinic. Statistical analysis used x2, and P I 0.05 was considered statistically significant.

RESULTS No patients were found to have symptomatic disease (Centers for Disease Control, group IV) at study onset. However, patients did present with various types of pre-delivery morbidity (Fig. 1). Over the course of pregnancy, four women were treated with aerosolized pentamidine. Seven women were treated with oral zidovudine (AZT). The types of anesthetics and drugs used are listed in Table 1. Clinical outcome Ninety-six records were reviewed. There were 75 spontaneous vaginal deliveries, 17 cesarean sections, 1 low forceps and 3 mid-second trimester abortions. Thirtysix of the women had regional anesthesia (28 epidural and 8 subarachnoid block). Eleven underwent general anesthesia, 22 local anesthesia with intravenous sedation, and 27 no anesthesia. Six-week post partum complications are listed in Table 2. ‘No anesthesia’ refers to no analgesic intervention peripartum. There were no 2448 h post-delivery complications. Chisquare analysis using a four by four contingency table (x2 = 6.8, d.f. = 9, P = 0.65) showed that the complication rate was independent of the anesthetic technique used. No neurologic sequelae were noted in any of the women up to 24 months post partum. Immunologic outcome Thirty-one of the ninety-six parturients had appropriately timed cell counts (first sample: second trimester,

78 International Journal of Obstetric Anesthesia Table 1. Anesthetic drug regimens (96 parturients)

Epidural (n = 28)

Spinal (n = 8)

Bupivacaine (92%) Bupivacaine (75%) Lidocaine (85%) Lidocaine (25%) Chlorprocaine (11%) Fentanyl(6%)

General (n = 11)

Paracervical nerve block Pudendal nerve block infiltration (n = 22)

Intravenous sedation (n = 22)

Thiopental Ketamine Lidocaine Succinylcholine Atracurium Morphine Fentanyl Midazolam Neostigmine Atropine Glycopyrrolate Isoflurane Ethrane Nitrous oxide

Lidocaine (100%)

Fentanyl Ketamine Midazolam Droperidol Nalbuphine Pethidine Promethazine

Percentage in parenthesis reflects patients in each subgroup. In none of the parturients did general anesthesia follow regional anesthesia. Anesthesia was conducted by different anesthesiologists, with varying drug regimens.

Table 2. Clinical outcome: Cweek post partum assessment (n = 96) n

Respiratory infection/bronchitis

Regional (36) (epidural28) (spinal 8) General (11) Local/intravenous sedation (22) No anesthesia (27)

Poor wound healing

Conversion to CDC group IV

2

2

1

2 1 2

1 1 0

1 0 1

Respiratory infections were non-pneumocystis carinii CDC: Centers for Disease Control and Prevention. CDC group IV: symptomatic AIDS.

second sample: 2448 h post partum). Twenty-five of the 3 1 underwent spontaneous vaginal deliveries, and 6 had cesarean sections. No statistically significant differences were found in pre- and post-anesthetic T-cell subpopulation ratios among regional, local anesthesia/intravenous sedation or no anesthesia groups (Fig. 2). Seven parturients had fewer than 400 CD4 cells/mm3 during the second trimester (epidural: 2, spinal: 2, general anesthesia: 1, no anesthesia: 2). The CD41CD8 ratios of these 7 parturients remained stable, but the absolute CD4 and CD8 cell counts declined in 4 of them (epidural: 2, spinal: 1, general anesthesia: 1). Five of the 31 parturients underwent a decline in their CD4/CD8 cell ratios greater than 0.2, 24-28 h post partum (epidural: 2, spinal: 1, local anesthesia/intravenous sedation: 1, no anesthesia: 1). In the patient with no anesthesia, the CD4/CD8 ratio declined from 0.90 to 0.40 within 24 h of spontaneous vaginal delivery; she was diagnosed with end-stage AIDS within one year post partum.

DISCUSSION

Immunologic outcome has been examined in non-HIV infected individuals following various anesthetic regimens.‘~2~‘7~20 Whether certain anesthetic agents are more likely than others to cause immunodepression, and consequently increase the risk of postoperative tissue and wound colonization and sepsis, remains an open question. Whether certain anesthetic drug regimens might cause further decrements in the CD4/CD8 cell ratios of immunologically susceptible HIV-infected patients, which could allow AIDS-related disease progression, is unknown. Because of marked variations in the preoperative immunologic profile, disease state and nutritional status of patients, it is difficult to assess immunologic outcome. Preoperative, intraoperative and postoperative managements differ greatly between studies as well. Preoperative events such as hemolysis, hemorrhage, burns and chronic infections may cause immunodepression. Immunologic changes

Anesthesia and the HIV-infected parturient

7

-

02nd -post

79

trimester delivery

CD4/CDS Ratios (values are mean * sem

Kegtonal (n=14)

General (n=l)

’ Local/IV Sedation (n=9) Type of Anesthesia

’ Anesthesia (n=7)

Fig. 2 CD4:CD8 ratios in the HIV-infected parturient.

brought about by intraoperative events such as hypothermia, blood transfusion or variable surgical trauma cause further difficulties in study analysis. Furthermore, most in vivo studies of immunocompetence test peripheral blood, which may reflect poorly ongoing immunologic activity at the tissue level.z Nevertheless, the consensus remains that some anesthetics depress cell-mediated immunity and if such an anesthetic-related immunodepression does occur, it is clinically subtle and transient.* In our study, we did not control for the agents used to perform general anesthesia, or for the type of local anesthetic administered for regional anesthesia. However, patients in our study were treated as any other parturient who delivered in that time, regardless of HIV status. Furthermore, we were not able to associate a possible transient immunodepression in 11 HIV-infected parturients given general anesthesia with a poor clinical outcome, although such outcome changes may be quite subtle. Pregnancy alters both cell- and humoral-mediated host immunity.” Several investigators have shown a decrease in CD4 cell counts during normal pregnancy, the nadir appearing in the third trimester, and usually returning to normal approximately 3 months post partum.4,’The CD8 cell count increases as pregnancy approaches term.2’This suggests that the parturient’s cell-mediated immune system is suppressed. The normal CD4 cell count in a healthy person is approximately 1000 cells/mm (1 x 109/L). CD4 cells are normally twice the number of CD8 cells. However, most HIV-infected parturients have CD4 cell counts lo-20% lower than seronegative controls, with an inversion of their CD4/CD8 cell ratios.23.?4 Pregnant women with CD4 cell counts less than 400 cells/mm’ have a greater chance of fetal viral transmission, while counts less than 300 cells/mm’ correlate with the

development of maternal complications.” Pregnancy impairs the immunologic response in general, and that to viral illnesses such as cytomegalovirus and rubella in particular.* It has been hypothesized that both the AIDS virus and AIDS-related diseases may progress more rapidly in the parturient.26 While it is unknown whether certain modes of delivery are more immunodepressing, the combination of pregnancy, HIV infection and general anesthesia could conceivably increase both morbidity and mortality. We found no difference in clinical outcome and peripartum complications in HIV-infected parturients undergoing general anesthesia compared to regional anesthesia, local anesthesia/intravenous sedation, or no anesthesia. The drugs used are agents commonly used for general anesthesia in the healthy parturient. CD4/CD8 cell ratios were measured during the second trimester and 2448 h post partum. While the sole general anesthetic/cesarean section patient with both appropriately timed blood samples did undergo a decline in her absolute CD4 and CD8 cell counts, the ratio remained stable and there were no peripartum complications. AIDS dementia complex is the most common central nervous system manifestation of HIV.” Memory loss and the inability to concentrate on complex tasks due to a shortened attention span are among the earliest signs of the complex. Up to two-thirds of HIV-infected patients eventually show signs of the complex. The AIDS virus is highly neurotropic and is capable of causing acute meningitides as well.‘? The virus has been implicated in spinal cord vacuolar degeneration with associated paraparesis, ataxia, incontinence and peripheral neuropathies.‘Ox”Several authors have therefore advised against the use of regional anesthesia’3m’in5 the HIV-infected patient. By the time the HIV-infected patient has progressed to

80 International Journal of Obstetric Anesthesia AIDS, the incidence of clinical neurologic disease has approached 40%.*’Chalmers et aP found a high incidence of the virus in otherwise healthy asymptomatic HIV seropositive men. It therefore seems that precisely due to the neurotropic nature of the AIDS virus, it is highly unlikely that a central nervous system HIV infection will be initiated by a small amount of viremic blood, possibly introduced from a regional anesthetic, in the already HIV-infected parturient. Whether epidural is safer than spinal anesthesia, because of the possible protective nature of the dura/arachnoid, is unknown. In our review, neither epidural nor spinal anesthesia were related to periparturn clinical or immunologic morbidity, compared to general anesthesia, local anesthesia/intravenous sedation, or no anesthesia. To date, no central nervous system morbidity has been attributed to epidural blood patches performed as treatment for post dural puncture headache in HIV-infected patients. It thus seems that both techniques appear equally safe.29m32 Ryhanen et aF3 studied the influence of elective cesarean section, using general and epidural anesthesia, on NK cell activity in healthy parturients and their newborns. A significant decrease in the NK cell activity was found only in the mothers given general anesthesia. The decrease in per cent cytotoxicity and NK cell function of the parturients induced by operative stress was prevented by epidural anesthesia. While regional anesthesia is often preferred in the parturient for many reasons, we reasoned that in the HIV-infected parturient with a greater risk of immunodepression, regional anesthesia is still the technique of choice. With this in mind, we find it interesting that the parturient whose CD4/CD8 cell ratio rapidly declined within 24 h after a normal spontaneous vaginal delivery, and who received no anesthesia, was diagnosed with end-stage AIDS 1 year post partum. We are unsure why those HIV-infected parturients who underwent the stress of labor without the possibly protective influence of epidural analgesia did not apparently fair worse either clinically or immunologically. Nor did our study patients in the regional anesthesia group seem to fare any better than the other groups. In this limited review, no differences were found in clinical outcome, nor in CD4/CDS ratios with any mode of anesthesia compared to no anesthesia. While it is inappropriate to make broad conclusions given the difficulty in detecting subtle clinical and immunologic changes in our limited retrospective data, our experience suggests that both regional and general anesthesia can be safely given to asymptomatic HIVinfected parturients, and that the choice of anesthesia should be based on the usual obstetric and clinical considerations.

ACKNOWLEDGEMENT Presented in part at the 1992 Annual Convention of the American Society of Anesthesiologists, New Orleans, Louisiana.

REFERENCES 1. Scannell K A. Surgery and human immunodeficiency virus disease. J Acquir Immune Defic Syndr 1989; 2: 43-53. 2. Stevenson G W, Hall S C, Rudnick S, Seleny F L, Stevenson H C. The effect of anesthetic agents on the human immune response. Anesthesiology 1990; 72: 542-552. 3. Sridama V, Pacini F, Yang S L, Moawad A, Reilly M, DeGroot L. Decreased levels of helper T-cells: a possible cause of immunodeficiency in pregnancy. N Engl J Med 1982; 307: 352-356. 4. Barnett M A, Learmoth R P, Pihl E, Wood E C. T-helper lymphocyte depression in early human pregnancy. J Reprod Immunol 1983; 5: 55-59. 5. Coulam C B, Silvertield J C, Kazmar R E, Fathman C G. T lymphocyte subsets during pregnancy in the menstrual cycle. Am J Reprod Immunol 1983; 4: 88-93. 6. Beer A E. Imnmnology of reproduction. In: Samter M, Talmaee D W. Frank M M. Austen K F, Clamn H N. eds. Immunological diseases. Boston: Little Brown, 1988: 329-360. I. Aksel S. Immunologic aspects of reproductive disease. JAMA 1992; 268: 2931-2933. 8. Gall S A. Maternal adjustments in the immune system in normal pregnancy. Clin Obstet Gynecol 1983; 26: 521-536. 9. Tonnesen E, Wahlgreen C. Influence of extradural and general anaesthesia on natural killer cell activity and lymphocyte subpopulations in patients undergoing hysterectomy. Br J Anaesth 1988; 60: 500-507. 10. Cornblath D R, McArthury J C, Kennedy P G E, Witte A S, Griffin J W. Inflammatory demyelinating peripheral neuropathies associated with human T-cell lymphotropic virus type III infection. Ann Neurol 1987; 21: 3240. 11. De Girolami U, Smith T W, Henin D, Hauw J J. Neuropathology of the acquired immunodeficiency syndrome. Arch Path01 Lab Med 1990; 114: 643-655. 12. Resnick L, Berger J R, Shapshak P, Tourtelotte W W. Early penetration of the blood-brain barrier by HIV Neurology 1988; 38: 9-14. 13. Greene E R. Acquired immunodeficiency syndrome: an overview for anesthesiologists. Anesth Analg 1986; 65: 10541058. 14. Greene E R. Spinal and epidural anesthesia in patients with the acquired immunodeficiency syndrome. Anesth Analg 1986; 65: 108991093. 15. Marx G F. AIDS and the Obstetric Anesthesiologist. Obstetric Anesthesia Digest 1986; 6: 289-326. 16. Hughes S C, Dailey P A, Landers D, Dattel B J, Crombleholme W R, Johnson J L. Parturients infected with human immunodeficiency virus and regional anesthesia. Anesthesiology 1995; 82: 32-37. 17. Thomson D A. Anesthesia and the immune system. J Burn Care Rehabil 1987; 8: 483487. 18. Finlayson D C. Immunologic changes in critically ill patients after anesthesia and surgery. Anesth Clin North Am 1989; 4: 883-895. 19. Shapiro H M, Grant I, Woinger M B. AIDS and the central nervous system: implications for the anesthesiologist. Anesthesiology 1994; 80: 187-200. 20. Moudgil G C, Singal D P, Gordon J. Comparitive effects of halothane and isoflurane anesthesia on perioperative lymphocyte function. Anesthesiology 1992; 77: A354. 21. Weinberg E D. Pregnancy-associated depression of cellmediated immunity. Rev Infect Dis 1984; 6: 814820. 22. Hirahara F, Gorai I, Tanaka K, Matsuzaki Y, Sumiyashi Y, Schiojima Y. Cellular immunity in pregnancy: subpopulations of T lymphocytes bearing Fc receptor for IgG and IgM in pregnant women. Clin Exp Immunol 1980; 41: 353-359.

Anesthesia and the HIV-infected parturient 23. Landesman S H. Human immunodeficiency virus infection in women: an overview. Semin Perinatol 1989; 13: 2-6. 24. Biggar R J, Pahwa S, Minkoff H. Immunosuppression in pregnant women infected with human immunodeficiency virus. Am J Obstet Gynecol 1989; 161: 1239-1244. 25. Minkoff H L, Henderson C E, Willoughby A. The influence of HIV serostatus on pregnancy outcome: results of prospective case control study. Society of Perinatal Obstetricians, 10th Meeting, Houston, 1990. 26. Chu S Y, Buehler J W, Berkelman R L. Impact of the human immunodeficiency virus epidemic on mortality in women of reproductive age, United States. JAMA 1990;264: 2255229. 27. So Y. Neurologic manifestations of AIDS. AIDS Clin Care 1989; 1: 3740. 28. Chalmers A C, April1 B S, Shephard H. Cerebrospinal fluid and human immunodeficiency virus. Findings in healthy,

29. 30. 3 1. 32. 33.

81

asymptomatic, seropositive men. Arch Intern Med 1990; 150: 153881540. Tom D J, Gulevich S J, Shapiro H M, Heaton R K, Grant J. Epidural blood patch in the HIV-positive patient. Anesthesiology 1992; 76: 943-947. Frame W A, Lightmann M W. Blood patch in the HIVpositive patient (letter). Anesthesiology 1990; 73: 1297. Bevacqua B K, Slucky A V Epidural blood patch in a patient with HIV infection. Anesthesiology 1991; 74: 952-953. Gibbons J J. Post-dural puncture headache in the HIV-positive patient (letter). Anesthesiology 1991; 74: 953. Ryhanen P, Jouppila R, Lanning M, Jouppila P, Hollmen A, Kouvalainen K. Natural killer cell activity after elective caesarean section under general and epidural anaesthesia in healthy parturients and their newborns. Gynecol Obstet Invest 1985; 19: 139-142.