Elevated plasma vasopressin and normal cerebrospinal fluid angiotensin-converting enzyme in chronic pain disorder

Elevated Plasma Vasopressin and Normal Cerebrospinal Fluid Angiotensin-Converting Enzyme in Chronic Pain Disorder Kristian Wahlbeck, Markus Sundblom, Eija Kalso, Irma Tigerstedt, and Ranan Rim6n

The study was performed proceeding from the hypothesis that pain proneness in chronic pain disorder ( CPD) is a result of alterations in central mechanisms regulating pain sensations. To elucidate the function of the central renin-angiotensin system, the levels of angiotensinconverting enzyme (ACE) and arginine vasopressin (A VP) in cerebrospinal fluid (CSF) and peripheral blood were measured in 15 CPD patients and 19 healthy controls. Plasma A VP levels (p = .01) as well as the serum osmolality (p = .01) were significantly higher in the CPD group. No significant differences in CSF ACE levels were found. AVP is a stress-related peptide, but central antinociceptive effects have also been reported. Elevated plasma A VP levels possibly may constitute a response to chronic stress. © 1996 Society of Biological

Psychiatry Key Words: Pain disorder, angiotensin-converting enzyme, arginine vasopressin, cerebrospinal fluid, osmolality, cigarette smoking B I O L PSYCHIATRY

1996;40:994-999

Introduction Pain disorder is characterized by preoccupation with pain in the absence of any relevant organic pathology or pathophysiologic mechanisms. Psychological factors are judged to have an important role in the etiology of the disorder (American Psychiatric Association 1994).

From the Department of Psychiatry (KW, RR) and Department of Anesthesiology (MS, EK, IT), University of Helsinki. Helsinki, Finland. Address reprint requests to Kristian Wahlbeck, MD. Department of Psychiatry, Lappviksvagen, FIN-00180 Helsinki, Finland. Received March 6, 1995; revised October 23, 1995.

© 1996 Society of Biological Psychiatry.

Although there are several biological similarities between the chronic form of pain disorder (CPD) and depressive syndromes, e.g., low cerebrospinal fluid (CSF) levels of the serotonin metabolite 5-hydroxyindoleacetic acid and an increased frequency of serum cortisol nonsuppression to dexamethasone (Almay 1987), the pathophysiology of the CPD is still unresolved. The brain renin-angiotensin system (RAS) is a complex system of peptides regulating blood pressure and water homeostasis. The key enzyme in the RAS is angiotensinconverting enzyme (ACE, peptidyl-dipeptidase A, kininase II, E.C. 3.4.15.1), a dipeptidase with a broad substrate specificity. In the RAS, it converts angiotensin I into the 0006-3223/96/$15.00 SSDI 0006-3223(95)00577-3

Vasopressin and ACE in CPD

B1OLPSYCHIATRY

995

1996;40:994-999

octapeptide angiotensin II (AT II), which is further degraded to angiotensin III (AT III) and other fragments. AT II has a pressogenic effect on peripheral cardiovascular structures. Centrally, AT II is dipsogenic, it increases sodium intake, and it stimulates the release of arginine vasopressin (AVP) from the posterior pituitary (Wright and Harding 1992). Both AT II (Haulica et al 1986) and AT III (Yien et al 1993) seem to have antinociceptive effects in rats. A high number of AT II binding sites are present in the midbrain periaqueductal gray, where endogenous opioids produce analgesia, as well as in thalamus, the termination point for afferent pain information (Mendelsohn et al 1984; Gehlert et al 1986). Vasopressin seems to have a physiological antinocicepfive role in both the brain (Hasegawa et al 1987) and at the spinal level (Thurston et al 1992), but it also acts as a releasing factor for proopiomelanocortin (POMC), both in the brain and in the pituitary (Wiegant et al 1989). POMC is a precursor for both the endogenous opioid peptide endorphins and for the melanopeptides, which are endogenous antagonists of opioid peptides. ACE not only synthesizes AT II, but in vitro it also catalyzes degradation of a broad range of other natural peptide substrates. The substrates include, in the order of diminishing affinity, bradykinin (BK), neurotensin (NT), substance P (SP), luteinizing hormone-releasing hormone, and the endogenous opioids leu- and met-enkephalin (Terenius and Nyberg 1988; Welches et al 1993). Several of these peptides are neuromodulators or neuromediators in pain pathways (Carr and Lipkowski 1990). BK (Dray and Perkins 1993) and SP (Anonymous 1992) have been implicated to be hyperalgesic compounds in the regulation of peripheral nociceptive input. Endogenous opioids and NT seem to be potent analgesic agents centrally (Cart and Lipkowski 1990), and appear to play an important modulatory role at the spinal level, too (Zhang et al 1994). The net effect of ACE activity on nociception thus seems to be increased analgesia both in spinal and central regions. To our knowledge, there are no studies elucidating the function of the central RAS and the secretion of AVP in CPD patients, although alterations in CSF ACE have been observed in various other neuropsychiatric conditions (Oksanen et al 1985; Zubenko et al 1985; Beckmann et al 1984; Wahlbeck et al 1993). The authors hypothesized that a deficiency in the central RAS, which would result in a diminished activity of the antinociceptive processes regulated by AT II and AT III, and an incomplete degradation of nociceptive neuropeptides, would produce the pain experienced by CPD patients. The authors further hypothesized that the deficiency in the central RAS would result in lower CSF ACE and plasma

AVP levels in CPD patients than in asymptomatic controls.

Methods Thirty-three chronic pain disorder patients from the outpatient Pain Relief Unit of Helsinki University Central Hospital were asked to participate in the study by letter. The pain disorder diagnosis was made according to DSM-IV (American Psychiatric Association 1994). Subjects were under the age of 65 and free of somatic diseases, alcoholism, or drug abuse. Only CPD patients without any positive findings in extensive screenings by two experienced pain clinicians (E.K. and I.T.) for an organic etiology of the pain were included. Five patients with psychiatric diagnoses were excluded by psychological interviews as reported in an earlier paper (Sundblom et al 1994). Of the included 28 patients, 15 patients with a mean duration of pain of 11.0 -+ 9.3 years (range 4-37 years) agreed to participate in the study, 5 patients refused to take part in the study, 5 patients did not answer the letter, and 3 patients were not reached. The study group (n = 15) had pain all over the body (n = 5), in the pelvic (n = 4) or cranial (n = 3) regions, or in other localizations. Before entering the study, all medicated patients underwent a washout period of minimum 2 weeks or five times the half-life of the substances used, whichever was longer. Medications prior to washout consisted of antidepressant drugs (n = 5), nonsteroidal anti-inflammatory drugs (n = 5), benzodiazepines (n = 3), centrally acting muscle relaxants (n = 2), and neuroleptics (n = 1). The controls were unmedicated, normotensive, paid volunteers from the hospital staff recruited by an advertisement in the personnel paper. The same exclusion criteria were applied to both controls and patients. Psychiatric diagnoses of the control subjects were excluded by the Structured Clinical Interview for DSM-III-R (SCID) (Spitzer et al 1988) performed by a clinical psychiatrist (K.W.). Twenty subjects were included in the study, of which 1 subject was excluded after the psychiatric interview due to a personality disorder of mixed type. The series studied thus comprised 15 (7 women, 8 men) Caucasian outpatients with CPD and 19 (8 women, 11 men) Caucasian healthy controls. The CPD patients were 47.7 _+ 7.4 years of age and 169 +- 9 cm tall, while the mean age and height of controls were 32.4 + 10.6 years and 175 _+ 10 cm, respectively. The CPD group contained two smokers (data on 1 CPD patient missing); the control group contained six smokers. Informed consent was obtained from all subjects. The approval of the Ethics Committee of the Helsinki University Central Hospital was obtained.

996

BIOLPSYCHIATRY

K. Wahlbeck et al

1996;40:994-999

Both CPD patients and control subjects were admitted to a psychiatric hospital ward for the study. The lumbar puncture was performed in lateral recumbent position immediately after blood had been drawn in chilled tubes at 8 - 9 AM following a 12-hour fast and bed rest. Subjects were not permitted to smoke until sampling was performed. The CSF was collected in fractionated aliquots in chilled test tubes on ice without adding protease inhibitors. The CSF samples were immediately frozen at - 7 0 ° C until assayed. Only CSF samples without blood contamination and with normal cell count were accepted for assay. The blood samples were centrifuged at 1000 rpm at + 4 ° C and frozen at - 7 0 ° C immediately after separation. Urine samples were collected at 7 AM on the morning of the spinal tap. Ratings were made by Beck Depression Index (BDI) (Beck et al 1961), Quality of Life Index (QLI) (Spitzer et al 1981), and Symptom Checklist (SCL) (Derogatis et al 1973). Pain experience was measured using a 10-cm Visual Analogue Scale (VAS) (Guilford 1954). The A C E concentrations were determined with a sensitive inhibitor-binding assay (Fyhrquist et al 1984). A V P was measured with a radioimmunoassay (RIA) in routine use (Fyhrquist et al 1976). Osmolality was determined by the freezing point depression method (Gennari 1984) with a micro-osmometer (Model 3MO, Advanced Instruments Inc., Massachusetts). Descriptive statistical data are presented as means _+ standard deviations. Statistical probabilities (p) were calculated with unpaired t tests for normally distributed variables and with Mann-Whitney U tests for nonnormal variables. Normality assumptions were tested using Lilliefors' test. Correlations were calculated with Pearson's correlation coefficient (r) after ranking data. Parameter estimates and standard errors are reported in analyses of variance, which were performed with Systat for Windows® 5.01 (Systat Inc., Evanston, IL) software.

Results The CPD patients had significantly higher BDI (U = 224.5, p = .001) and SCL (U --- 212.5, p = .001) total scores, and lower quality of life scores (U = 59.5, p = .017) than the controls. The mean total scores in the CPD group were as follows: BDI 9.2 --- 7.2, SCL 57.4 - 39.7, and QLI 8.6 -+ 1.7. The CPD patients had significantly higher scores than the controls on all SCL subscales, except the anger subscale. When compared to the SCL scores of a Finnish normal population, the CPD patients differed primarily by high scores (2.35 --- 0.90) in the somatization dimension of SCL. Three out of 15 patients

Table 1. Plasma AVP, Serum Osmolality, and Serum and CSF ACE in CPD Patients and Healthy Controls

Plasma AVP (ng/L) Serum osmolality (mosm/kg H20) Serum ACE

CPD patients

Healthy controls

(n = 15)

(n = 19)

t value

df

p

5.1 _+ 1.6

3.5 -+ 1.7

2.6

31

.01

290.1 + 4.6

285.4 - 4.8

2.8

32

.01

27.3 --- 9.3

21.4 -4- 10.8

1.7

32

.10

0.45 + 0.11

0.39 -+ 0.12

1.7

32

.11

(U/L) CSF ACE

(u/L)

(20%) had a total BDI score of at least 17, which is a threshold value for moderate depression. The results of the assays are presented in Table 1. The CSF ACE did not correlate with serum ACE, CSF protein, age, sex, height, or the total rating scores or subscales on BDI, QLI, SCL, or VAS. Plasma AVP levels and serum osmolality were significantly higher in the CPD patients. Plasma A V P correlated with serum osmolality (r = 0.36, p = .04), but this correlation was lost when smoking subjects (n = 8) were analyzed separately (r = 0.07, p = .87). Plasma AVP did not significantly correlate with CSF ACE or with blood pressure. Urine osmolality did not differ between the groups, and there was no significant correlation between plasma A V P and urine osmolality. The groups differed significantly in age (t = 4.7, df = 32, p = .001) and height (t = 1.9, df = 31, p = .06) but not in smoking habits (×e = 2.605, df = 2, p = .27). Analysis of covariance for plasma AVP adjusted by age, height, and smoking habits was significant for the study group (parameter estimate for the difference = 1.67 --0.81, df = 27, p = .04). Smokers (n = 8) had significantly (t = 2.58, df = 30, p = .02) lower plasma AVP values (2.86 _ 1.27 vs. 4.63 ___ 1.78 ng/L) than nonsmokers (n - 24), a finding that was also significant when each study group was analyzed separately. The RIA for AVP had a detection limit of 0.15 ng/L and was unable to detect CSF A V P in all but 10 study subjects (2 patients and 8 controls). Thus, no reliable conclusions about differences in levels of CSF AVP can be made.

Discussion The absence of a correlation between CSF ACE and serum ACE or CSF protein indicates that the measured CSF ACE

Vasopressin and ACE in CPD

BIOLPSYCHIATRY

997

1996;40:994-999

was not from peripheral sources, but is produced within the blood-brain barrier. Since the CSF ACE levels may be altered both centrally and at the spinal level, no definitive conclusions about alterations in the central RAS in CPD can be drawn based on our data. A central hypofunction of the RAS can be masked by a spinal hyperfunction. There was a trend toward a negative correlation (r = -0.44, p = .07) between CSF ACE and diastolic blood pressure, which corroborates the physiological pressogenic role of the central RAS. Although the mean age and height differed between groups, analysis of covariance showed that difference in plasma AVP between groups was not an effect of these confounding factors or smoking habits. The two major physiological factors stimulating the release of renin and AVP are blood pressure and serum osmolality. Our study subjects were normotensive, but the CPD patients had a higher osmolality than the controls. Since the patients and the volunteers were subjected to the same standardized inpatient procedures prior to sampling, including 12-hour bed rest and fasting, the higher plasma AVP and serum osmolality in the CPD group reflect an alteration in the fluid homeostasis in CPD. The reasons for this alteration remain unclear. The secretion of AVP is regulated both by monoamine neurotransmitters (norepinephrine, dopamine, serotonin) and neuropeptides (AT II and endogenous opioids) (Keil et al 1984; Coiro and Chiodera 1991). The elevated osmolality may reflect an activated central renin-angiotensin system affecting salt intake, but the involvement of other regulatory systems cannot be excluded. Since the patients in this study had suffered from pain for several years, the elevated plasma AVP levels seem to be chronic, which leads to a down-regulation of the kidney target tissues. This renal escape could result in chronic hyperosmolality. Plasma AVP is of hypothalamic origin, whereas CSF AVP probably stems from extrahypothalamic vasopressinergic neurons and parvocellular hypothalamic neurons. Although AVP may have pain-modulating functions in the brain (de Wied et al 1993), and high doses of systemically administered AVP produce antinociception in rats (Berson and Bernston 1980), it seems unlikely that the elevated plasma AVP level in CPD constitutes a compensatory antinociceptive response. In rats, the synthesis of AVP has been showed to be enhanced by both acute (Herman and Sherman 1993) and chronic stress (de Goeij et al 1992a, 1992b). Thus, the elevated levels of plasma AVP in CPD

may reflect a chronic stress caused by the pain experience. On the other hand, no correlations between plasma AVP levels and perceived levels of pain could be detected. In depression, the levels of CSF AVP (Gold et al 1984; Gjerris et al 1985) and the plasma levels of the AVP carrier protein human neurophysin I (Laruelle et al 1990) have been reported to be decreased. Our finding of elevated levels of plasma AVP and serum osmolality points toward a different pathophysiology of CPD and depression. Our study shows the importance of taking smoking into account as a major confounding factor in AVP studies. We were able to show that smokers in both groups had significantly lower levels of plasma AVP. None of the above-mentioned studies on the CSF AVP secretion in depression reported smoking habits of the study subjects, which is a major bias of the studies since smoking is more frequent among psychiatric patients than in the general population. The lower plasma AVP levels in the smoking subjects were somewhat unexpected, since cigarette smoking is reported to increase plasma vasopressin concentrations (Waeber et al 1984). Since the subjects refrained from smoking for 12 hours before sampling, the decreased plasma AVP levels in the smoking subjects may reflect the abstinence state more than the effects of smoking. There was no correlation between plasma AVP and serum osmolality in the subgroup of smokers. This indicates that the normal osmoregulation of AVP secretion is disturbed in smokers refraining from smoking. The BDI and SCL scorings showed that the CPD patients are a heterogeneous group consisting of patients with and without depressive symptomatology. The nosology needs to be refined to differentiate the various subtypes of pain disorder. Until this has been done, it seems unlikely that any conclusive biological pathophysiology of the disease can be defined. In summary, these findings do not corroborate the original hypothesis of a central hypofunction in the RAS in CPD, but the fluid homeostasis and secretion of AVP seem to be dysregulated in CPD, which may reflect a chronic stress condition. This study was supportedby the Signe and Ane GyllenbergFoundation, Finska L~ikaresallskapet (K.W.), Perkl6ns Stiftelse (K.W.), and The Academy of Finland (E.K.).

)98

BIOL PSYCHIATRY

K. Wahlbeck et al

1996;40:994-999

References Almay BGL (1987): Patients with idiopathic pain syndromes. PhD Thesis. Umefi University. American Psychiatric Association (1994): Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association. Anonymous (1992): International Symposium on Substance P and Related Peptides: Pain, Inflammation, Visceral and CNS Functions. Proceedings. Shizuoka, Japan, November 3-6, 1992. Regul Pept 46:1-471. Beck AT, Ward CH, Mendelson M, Mock J, Erbaugh J (1961): An inventory for measuring depression. Arch Gen Psychiatry 4:561-571. Beckmann H, Saavedra JM, Gattaz WF (1984): Low angiotensin-converting enzyme activity (Kininase II) in cerebrospinal fluid of schizophrenics. Biol Psychiatry 19:679-684. Berson BS, Bernston GG (1980): Antinociceptive effect of intraventricular or systemic administration of vasopressin in the rat. Life Sci 26:455-459. =arr DB, Lipkowski AW (1990): Neuropeptides and pain. Agressologie 31 : 173-177. =oiro V, Chiodera P (1991): Naloxone increases the angiotensin II stimulated rise of vasopressin and oxytocin secretion in man. Neuroendocrinology 53:209-213. 3e Goeij DCE, Dijkstra H, Tilders FJH (1992a): Chronic psychosocial stress enhances vasopressin, but not corticotropinreleasing factor, in the external zone of the median eminence of male rats: Relationship to subordinate status. Endocrinology 131:847-853. ]e Goeij DCE, Jezova D, Tilders FJH (1992h): Repeated stress enhances vasopressin synthesis in corticotropin releasing factor neurons in the paraventricular nucleus. Brain Res 577:165-168. Derogatis LR, Lipman RS, Covi L (1973): SCL-90: An outpatient psychiatric rating scale--Preliminary report. Psychopharmacol Bull 9:13-28. ]e Wied D, Diamant M, Fodor M (1993): Central nervous system effects of the neurohypophyseal hormones and related peptides. Front Neuroendocrinol 14:251-302. Dray A, Perkins M (1993): Bradykinin and inflammatory pain. Trends Neurosci 16:99-104. 7yhrquist F, Wallenius M, Hollemans HJG (1976): Radioimmunoassay of vasopressin in unextracted plasma. Scand J Clin Lab Invest 36:841-847. Fyhrquist F, Tikkanen I, Gr6nhagen-Riska C, Hortling L, Hichens M (1984): Inhibitor binding assay for angiotensinconverting enzyme. Clin Chem 30:696-700. 3ehlert DR, Speth RC, Wamsley JK (1986): Distribution of [125I]angiotensin II binding sites in the rat brain: A quantitative autoradiographic study. Neuroscience 18:837-856. 3ennari FJ (1984): Serum osmolality. Use and limitations. N Engl J Med 310:102-105. 3jerris A, Hammer M, Vendsborg P, Christensen NJ, Rafaelsen OJ (1985): Cerebrospinal fluid vasopressin~hanges in depression. Br J Psychiatry 147:696-701. 3old PW, Ballenger JC, Robertson GL, et al (t984): Vasopressin in affective illness: Direct measurement, clinical trials and

response to hypertonic saline. In Post RM, Baltenger JC (eds), Neurobiology of Mood Disorders. Baltimore: Williams and Wilkins, pp 323-339. Guilford JP (1954): Psychometric Methods. New York: McGraw-Hill. Hasegawa Y, Kitani Y, Sugaya T, Nakajima S, Fujita T (1987): Mechanism of the antinociceptive effect by intracerebroventricular administration of arginin-vasopressin in the monkey. Pain 4(suppl):S51. Haulica I, Neamtu C, Stratone A, et al (1986): Evidence for the involvement of cerebral renin-angiotensin system (RAS) in stress analgesia. Pain 27:237-245. Herman JP, Sherman TG (1993): Acute stress upregulates vasopressin gene expression in parvocellular neurons of the hypothalamic paraventricular nucleus. Ann N Y Acad Sci 689:546 -549. Keil LC, Rosella-Dampman LM, Emmert S, Chee O, SummyLong JY (1984): Enkephalin inhibition of angiotensin-stimulated release of oxytocin and vasopressin. Brain Res 297: 329-336. Laruelle M, Seghers A, Goffinet S, Bouchez S, Legros JJ (1990): Plasmatic vasopressin neurophysin in depression: Basic levels and relations with HPA axis. Biol Psychiatry 27:12491263. Mendelsohn FAO, Quirion R, Saavedra JM, Aguilera G, Catt KJ (1984): Autoradiographic localization of angiotensin II receptors in rat brain. Proc Natl Acad Sci USA 81:1575-1579. Oksanen V, Fyhrquist F, Somer H, Gr6nhagen-Riska C (1985): Angiotensin converting enzyme in cerebrospinal fluid: A new assay. Neurology 35:1220-1223. Spitzer RL, Williams JBW, Gibbon M, First MB (1988): Structured Clinical Interview for DSM-lll-R--Patient Version (SCID-P 4/1/88). New York: Biometrics Research Department, New York State Psychiatric Institute. Spitzer WO, Dobson AJ, Hall J, et al (1981): Measuring quality of life of cancer patients. A concise QL-index for use by physicians. J Chron Dis 34:585-597. Sundblom DM, Halkonen S, Niemi-Pynttari J, Tigerstedt I (1994): Effect of spiritual healing on chronic idiopathic pain: A medical and psychological study. Clin J Pain 10:296-302. Terenius L, Nyberg F (1988): Neuropeptide-processing, -converting, and -inactivating enzymes in human cerebrospinal fluid. Int Rev Neurobiol 30:101-121. Thurston CL, Campbell IG, Culhane ES, Carstens E, Watkins LR (1992): Characterization of intrathecal vasopressin-induced antinociception, scratching behavior, and motor suppression. Peptides 13:17-25. Waeber B, Schaller MD, Nussberger J, Bussien JP, Hofbauer KG, Brunner HR (1984): Skin blood flow reduction induced by cigarette smoking: Role of vasopressin. Am J Physiol 247:H895-H901. Wahlbeck K, Rim6n R, Fyhrquist F (1993): Elevated angiotensin-converting enzyme (kininase II) in the cerebrospinal fluid of neuroleptic-treated schizophrenic patients. Schizophr Res 9:77-82. Welches WR, Brosnihan KB, Ferrario CM (1993): A comparison of the properties and enzymatic activities of three angiotensin

Vasopressin and ACE in CPD

processing enzymes: Angiotensin converting enzyme, prolyl endopeptidase and neutral endopeptidase 24.11. Life Sci 52:1461-1480. Wiegant VM, Sweep CGJ, Barna I, Veldhuis HD, De Wied D (1989): Involvement of vasopressin in the regulation of the brain proopiomelanocortin system. In Genazzani AR, Negri M (eds), Opioid Peptides in Biological Fluids. Carnforth: Parthenon, pp 25-32. Wright JW, Harding JW (1992): Regulatory role of brain angiotensins in the control of physiological and behavioral responses. Brain Res Rev 17:227-262. Yien HW, Chan JYH, Tsai HF, Lee TY, Chan SHH (1993):

B1OLPSYCHIATRY 1996;40:994-999

999

Participation of nucleus reticularis gigantocellularis in the antinociceptive effect of angiotensin III in the rat. Neurosci Lett 159:9-12. Zhang RX, Mi ZP, Qiao JT (1994): Changes of spinal substance P, calcitonin gene-related peptide, somatostatin, Met-enkephalin and neurotensin in rats in response to formalininduced pain. Regul Pept 51:25-32. Zubenko S, Volicer L, Direnfeld LK, Freeman M, Langlais PJ, Nixon RA (1985): Cerebrospinal fluid levels of angiotensinconverting enzyme in Alzheimer's disease, Parkinson's disease and progressive supranuclear palsy. Brain Res 328:215221.