I. John Russell, M.D. article from the Journal of Musculoskeletal Pain.
SUMMARY. Objectives: To describe new findings regarding metabolic and nociceptive processes in fibromyalgia (FS) patients.
Findings: The earlier observation of elevated substance P in the cerebrospinal fluid (CSF) of FS patients has been confirmed with the documentation of levels which approach correlation with the painful symptoms. Newly recognized biochemical abnormalities include elevated CSF kynurenine, low serum serotonin, and low red cell NADP, NADH, and ATP.
Conclusions: These findings support the hypothesis that FS is a metabolic disorder with demonstratable biochemical abnormalities potentially capable of explaining the clinical symptoms. It is hoped that the recognition of these abnormalities will facilitate the development of more specific and more effective therapies.
KEYWORDS. Fibromyalgia, pathogenesis, serotonin, substance P, kynurenine.
INTRODUCTION
Fibromyalgia syndrome (FS) is a chronic, painful, musculoskeletal disorder of unknown etiology. A growing body of epidemiological evidence has shown it to be relatively common, occurring in up to 2.5% of the general population. At that rate, it would be about twice as common as rheumatoid arthritis (RA). That comparison is relevant because the severity of the discomfort experienced by FS and RA patients is comparable. In addition, the ability of these two patient groups to perform specific work tasks is similarly impaired. FS patients sleep poorly, exhibiting non-restorative sleep physiology patterns relatively deficient in stage IV non-REM.
Efforts to develop rational therapy for FS symptoms have been hampered by a real lack of information about its etiology and pathogenesis. For many years it was believed that the problem was either psychological or that it resided in the skeletal muscles of affected individuals. Carefully controlled psychological and muscle histology studies have now largely discounted those ideas. Such revelations have indirectly provided support for the central metabolic hypothesis which predicts that an interpretative defect in the central nervous system (CNS) may be responsible for abnormal perception of pain in the absence of recognizable peripheral tissue injury.
The concept of metabolic pain enhancement in FS is not new but seems increasingly plausible. Moklofsky and colleagues proposed that serotonin might be deficient in FS because that neurotransmitter influences both deep restorative sleep and pain perception. It can influence pain perception at peripheral,, spinal and central nervous system levels. It is also known to modulate the function of substance P, particularly with reference to the interpretation of sensory stimuli.
Studies conducted by our group over the past 10 years in San Antonio have identified several very interesting biochemical abnormalities in cerebrospinal fluid and blood samples obtained from patients with FS. It is beyond the scope of this review to detail all of those findings, so the aim will be to focus on a small number of key new concepts which provide a basis for a better understanding of the biochemical defects in this disorder.
CEREBROSPINAL FLUID
Our group had predicted that the pathogenesis of fibromyalgia must involve one or more abnormalities in the processes by which the central nervous system interprets incoming pain signals. It seemed logical, therefore, to examine the metabolic status of cerebrospinal fluid (CSF) samples as a biological fluid which enjoys intimate contact with brain tissue and reflects the biochemical activities of the brain.
CSF Substance P
A prior study conducted by Vaeroy et al., in Norway found substance P (SP) to be 3-fold elevated in FS CSF compared with NC CSF, but did not document its relationship to symptoms. We conducted a similar study in San Antonio, collecting and analyzing CSF from 32 FS and 30 healthy NC. The findings were very similar to those reported by the Norweigan group. We found an average SP concentration of 42.79 plus or minus 14.94 fmole/ml in FS CSF and only 16.30 plus or minus 6.04 in CSF from NC. Those differences were statistically highly significant (Wilcoxon P<0.001). It was interesting to note that ethnicity had some influence on the Sp levels among FS patinets. SP in 12 Causasians FS (49.9 plus or minus 3.3) was higher (P = 0.02) than in 19 Hispanics FS (37.9 plus or minus 3.4). SP levels were not different on the basis of gender despite the fact that FS is so predominantly a female disorder (9 females affected for each male affected). There was clearly no relationship of SF SP concentration to age. As dramatic as the differences in CSF SP were between FS and NC, it was surprising that the correlations with clinical measures of pain were not stronger than were found. A borderline significant correlation was observed with the average tender point threshold (TPA) (R = 0.32, P = 0.07) and the total tender point severity index (TPI) (R = -0.30, P = -0.10), but there was no relationship between SP and the FS patients' subjective self-assessment of pain severity on a visual analog scale.
It seems likely that the elevated SP levels in the CSF of FS patients do indicate something about the pathogenesis of FS but the findings also suggest that it is not acting alone. Rather, the data mandates a "second hit" biochemical abnormality to explain the constellation of symptoms so characteristic of FS patients. It is still possible, as we earlier predicted, that the other abnormality is related to serotonin, a metabolite of tryptophan.
CSF Tryptophan
The same CSF samples collected from 32 well-characterized FS patients and 30 NC were studied for tryptophan concnetrations. The mean CSF tryptophan concentration in the FS patients was found to be numerically lower than the mean for NC CSF, but the differences were not statistically significant. Tryptophan can be metabolized to serotonin or can take an alternate
metabolic route by metabolism through the kynurenine pathway. The question then, was whether the serotonin or the kynurenine pathways of tryptophan metabolism would be found to evidence abnormalities in the CSF of FS patients.
CSF Serotonin
Serotinin is a potent, often inhibitory, neuromodulator which is effective locally at the interneural sites of its release. As a result, its levels in CSF are so low that no laboratory has yet devised methodology to accurately measure it. Serotonin's metabolic procudt, 5-hydroxyindole acetic acid (5HIAA), was measured because it is stable and is thought to accurately reflect the rate of serotonin turnover in brain tissue. Two different studies were conducted to answer this question and both gave essentially the same answer. The first study was performed on SCF from 17 FS patients collected in Norway and transported to San Antonio for examination along with CSF from 8 NC and 7 CSF from patients with RA. The second study involved CSF collected in San Antonio from FS and NC volunteers. In both experiments, the mean concentration of 5HIAA in CSF from FS patients was numerically lower than normal but in neither case was it significantly so. Whether that should steer investigation away from serotonin or simply indicate that the localized processes may be poorly documented by the CSF metabolite concentration will rquire more study. The question of serotonin involvement cannot be dismissed however, because it it clearly abnormal in the serum of FS patients and that abnormality has been shown to correlate with the painful symptoms of FS.
CSF Kynuernine
The kynurenine pathway was also examined in the CSF. For that purpose, we used the same 32 FS and 30 healthy NC CSF samples which had previously been examined for the concentrations of substance P. The results indicated that the metabolic conversion of tryptophan to kynurenine, catalyzed by the enzyme indole-2,3-dioxygenase (IDO) was abnormal in FS. Despite the lower mean concentration of tryptophan (the substrate for IDO) in FS patientsk the CSF concentration of kynurenine (the reaction's product) was significantly higher in the CSF of FS patients than in the CSF of NC. The next enzyme in the sequence is kynurenine-3-monooxygenqase (KMO) which catalyzes the conversion of kynurenine to kynurenine-3-hydroxylase. Surprisingly, the concentration of kynurenine-3-oxygenase was significantly lower than normal in FS CSF.
The explanation of these findings is not yet fully known. One possibility is that a single chemical messenger is responsible for both changes. The enzyme IDO, is known to be inductible by interfron-gamma (IFN-gamma). The activity of the enzyme KMO is dependent upon the concentration of its cofactor nicotinamide adenine dinucleotide phosphate (NADP). Interferon is also capable of inducing an intracellular depletion of NADP by inducing nucleotide transferase activity.
RBC NUCLEOTIDES
A series of experiiments with FS and NC red blood cells (RBC) were initiated to examine intracellular biochemistry. The specific purpose was to determinne whether the abnormal kynurenine metabolism, particularly the "bottleneck" at the KMO which requires NADP as a cofactor, might be explained by an intracellular deficiency of NADP. We examined the concentrations of nucleotides in the RBC from the venous blood of 27 demographically-matched FS and NC subjects (21), RBC nucleotides were extracted by the method of Stocchi et al., fractionated by high-performance liquid chromatography using an ion-paired, reverse-phase, gradient ellution technique and detected by optical density at 260 mu. The Nucleotides identified and quantified included AMP, ADP, ATP, NAD, NADH, NADP, and NADPH. The concentrations of several nucleotides were numerically low in FS RBC, but NADP (P - 0.05), NADH (P = 0.02), and ATP (P = 0.005) were significantly lower than normal. Within the FS patient group, the NADP concentrations correlated highly with the ATP concentration (R = 0.73, P < 0.001), and ATP correlated with the subject's age (P + 0.006), but none of the nucleotides concentrations ocrrelated with the patients' perception of their pain, with their perceived severity of fatigue or with the semi-objective measures of pain severity such as the TPI or the TPA. There were no important differences by gender and the only ethnic difference of note was that NADPH in the Caucasian FS patients, who exhibited a greater severity of pain and morning stiffness, was only half that of the Hispanic FS patients (P = 0.001).
IFN-GAMMA?
The abnormalities in tryptophan metabolism implied by these data are consistent with the literature reports of alterations inducible by elevated levels of IFN-gamma. In fact, several recognized biohemical abnormalities in FS seem to fit that concept. A diversion of tryptophan to kynurenine would decrease the level of tryptophan for serotonin synthesis so the level of serotonin would fall, making pain perception and deep sleep more problematic. Tyrptophan is an essential amino acid whose availability is also required for protein synthesis, so functions which depend on replacement proteins, such as skeletal muscle repair, might suffer. The problem nos is to determine whether IFN-gamma really is elevated in FS and, if so, to conceive why it might be elevated in the CNS of these pqtients. In the blood, IFN-gamma is thought to be released in response to viral infections or other inducers of inflammation. Since FS is not considered to be an inflammatory disorder, it is not clear why IFN-gamma should be present in excess. In additin, the white blood cells known to produce IFN-gamma in response to cytokine signals are not likely to be present in the CNS.
IFN AND SP
A search of the medical literature for a CNS connection between substnce P and IFN was without much success. There is published evidence to indicate that substance P can enhance production of IL-2 ty T lymphocytes. Increased IFN production by those cells parallels that increase in IL-2 production. Production of much IFN in the CNS, however, would require larger numbers of leukocytes than are normally present in the brain tissue.
One could speculate that there is a CNS stromal cell capable of producing IFN in response to substance P. For example, there is evidence that IL-1 can be produced by astrocytes, and that substance P stimulates that process by a calcium ion-dependent mechanism. Conversely, substance P production in neutral tissue is responsive to IL-1 and IL-1 down-regulates the numbers of receptors for substance P on astrocytes.
We have not yet identified any published accounts of IFN production in the CNS except in the presence of a bacterial or viral infection. Anothr potential alternative is that IFN is produced elsewhere, in response to a signal other than substance P, and then effects elsewhere and the products (e.g., kynurenine) might cross the blood brain barrier where we measured them in the CSF. That could explain the biochemical changes we have observed in other compartments in FS patients, i.e.,, the low tryptohan in serum and CSF,, low serotonin in serum, high kynurenine in CSF, low 3-hydroxykynurenine in CSF, and low NADP in peripheral red cells. The abnormality in tryptophan transport described by Yunas et al. could result from competition by kkynurenine for transport with the energy-requiring large amino acid receptor at the level of the blood brain transport.
5HT AND SP
Data from 2 laboratories, involving subjects from different ethnic backgrounds, have provided convincing evidence that substance P is increased in the CSF of FS patients. The cause and mechanism are still unknown, but considerations must include a hereditary or congenital defect, toxin exposure, tissue trauma or defective regulation. Substance P is normally released in the spinal cord by afferent neurons in response to painful peripheral stimuli. It is then involved in the early stages of nociceptive signaling, so an excess of substance P could amplify the perception of pain.
We suspect that some of the other biochemical abnormalities we have recently identified in FS patients (low tryptophan, low serum serotonin, increased CSF kynurenine, lower than normal adenine nucleotides in RBC may be epiphenomena. On the other hand, they probaly are pointing to critical abnormalities, as yet undetected, which are responsible for the array of symptoms experienced by FS patients. For that reason, a direct link between substance P and a tryptophan metabolite such as serotonin would be of interest.
There is evidence from a rat model that the nociceptive effects of SP are counteracted by 5HT agonists. The relqtionship between SP and 5HT is more complex than that, however. In the brain 5HT seems to augment the synthesis of SP while blockade of 5HT receptors in the brain decreases SP production. Conversely, a decrease in brain 5HT which causes a decrease in brain SP results in an increase in the spinal cord SP content. These data indicate that brain SP can be lower than normal while cord SP is relatively increased. Our own finding of increased lumbar CSF SP in FS would be consistent with that model and would further predict lower than normal 5HT effect in the brain.
The apparent dichotomy can be explained as follows: Brain SP seems to serve as a central down-regulator of nociception in the spinal cord, working through cord-level 5HT and opiates. SP release in the spinal cord induces nociception while in the brain it initiates retrograde anti-nociception down in the cord. Brain 5HT levels seem to be positively related to brain SP levels and inversely with cord SP levels. Therefore, a hypothetical model of relative central 5HT deficiency would predict lower normal central SP, but an excess of SP at the cord level resulting in excessive nociception which is inadequately down-regulated by deficient cord 5HT.
Substance P may also influence the function of the hypothalamic-pititutary-adrenal axis but dysregulation by 5HT seems even more likely. Perhaps one or both are responsible for some of the abnormalities FS patients exhibit in corticotrophin releasing hormone (CRH) and adrenal corticotrophic hormonne (ACTH) production and growth hormone (via IGF1) levels in FS.
WORKING MODEL
Our current working model is deceptively simple but it has the potential to explain many of the observed biochemical abnormalities and the exaggeration of pain perception in FS. It is perceived as a step on the way to a more complete understanding of FS pathogenesis partly because it predicts a large number of very testable subhypotheses. It is as follows:
Abnormal levels of or a disproportion of serotonin and substance ) in the brain and the spinal cord result in a variety of neuroendocrine, biothythm, and nociceptive abornmalities which explain the signs and symptoms of fibromyalgia syndrome. In the spinal cord they cause an exaggerated nociception from fairly normal stimuli (allodynia). In the brain, a relative deficiency of serotonin could be responsible for the sleep disturbances and the neuroendocrine abnormalities. In the abdomen, they culd cause the dysfunctional bowel synptoms experienced by FS patients. Finally, a relative deficiency of tryptophan for protein synthesis, in an environment deficient of IFG1 and DHEAS could be responsible for the intolerance to and delayed recovery from physical exercise among FS patients.