Multiple Pharmacodynamics-pharmacokinetics (mPD-PK) Models And Their Application In Evaluation Of Analgesics

[Objectives]Chronic pain includes cancer pain and non-cancer pain which are caused by peripheral tissue damage, inflammation, nerve injury or cancer. Chronic pain has multiple symptoms in clinic including persistent spontaneous pain, primary and second hyperalgesia (allodynia) to thermal and mechanical stimuli, as well as mirror pain and hyperalgesia. Recently, it has been gradually noted that the mechanisms of the occurrence and persistence of pain might be mediated by muptiple signal transduction pathways along the pain pathways in both periphery and CNS, referred to as peripheral sensitization and central sensitization. These results suggest that gain of long-term and effective relief of chronic pain requires blockade of multiple molecular and cellular targets involved in the pathological processes. Therefore, strategies using multiple targets blockade might be promising and of particular significance in R&D of new analgesics.The key to screen new drugs or new clinical usage of old drugs is appropriate screening techniques or methods used. Pharmacodynamics-pharmacokinetics (PD-PK ) models are very important to evaluate the characteristics of a drug after administered into the body including effect-time course, drug concentration-time course, site concentration-effect (C-E), and site concentration-effect-time course (C-E-T). These parameters might reflect the relationship between ADME (absorption, distribution, metabolism and elimination) processes of a drug and its phamarcological effects. However, establishment of a multiple pharmacodynamics- pharmacokinetics (mPD-PK ) model would be much more useful in evaluation of the above parameters when considering that the PD-PK model might be changed depending upon different state that the patients are being. For an example, analgesics might play different pharmacological effects upon different pain states (physiological vs. pathological) and upon different stimulus-modalitiy evoked pain. We hypothesize that there should be a pain-state related difference or pain stimulus-modalitiy related difference in analgesia for analgesics. Thus we introduce a multiple pharmacodynamics - pharmacokinetics (mPD-PK ) model by which we try to highlight the significance of application of mPD - PK model in evaluation of the relationship between site concentration and pharmacological effects of a analgesic.To establish an appropriate mPD-PK model that could be used in evaluation of the relationship between site concentration and pharmacological effects of a analgesic, we studied the mPD-PK characteristics of morphine, a conventional clinical analgesic, for either pain stimulus-modalitiy related effects or pain-state related effects.[Methods]To see whether there are influences of pain stimulus-modalities on the PD-PK model of morphine, we selected two stimulus modalities including thermal stimulus and mechanical stimulus. To seek whether there are changes in PD-PK model under different pain states, we studied the mPD-PK chanracteristics of morphine under (1) physiological state with na?ve rats; (2) inflamed state with peripheral inflammation in rats; (3) neuropathic state with peripheral nerve injured rats. To rule out the effects of route of administration, oral gavage (intragastric, i.g.) and intraperitoneal (i.p.) were adopted for morphine.Quantification of pain response or hyperalgesia to thermal and mechanical stimulia) Paw withdrawal thermal latency (PWTL) was measured by counting the latency of the occurrence of paw withdrawal reflex after radiant heat stimulus.b) Paw withdrawal mechanical threshold (PWMT) was measured by the von Frey fiber bending force being able to produce 50% occurrence of paw withdrawal reflex. Models for peripheral inflammatory state or neuropathic state(1) Peripheral inflammatory pain state was produced by subcutaneous injection of bee venom (0.2 mg/50μl in saline) into the paw pad of one hind paw in rats (Chen et al., 1999b).(2) Neuropathic pain state was produced by transaction of tibial and peritoneal nerves with sural nerve remained intact in one side of hind paw of rats, referred to as spared nerve injury model (SNI, Costerd and Woolf, 2000).mPD-PK modeling of morphine(1) Blood or CSF samples were collected 5, 15, 30, 45, 60, 90, 120, 180, 240, 300 and 360 min after i.g. or i.p. administration of morphine, unbound morphine and conjugated (total-unbound) morphine were abstracted by water-bath (90 oC) and followed by the HPLC measurement.(2) The pharmacokinetic parameters were determined by using Auto-select model and real value with the pharmacokinetic computer program -3p87/3P97 or the new drugs Statistical Processor-21 (NDST-21, China) provided by Chinese Society of Mathematic Pharmacology. The data were compared by using WINNONLIN program from USA Pharsight Co. Statistical analyses were performed with SPSS 11.5 statistical software. All results were expressed by mean±SEM. Compareisions between groups in the each studies were performed using non-parametric Mann±Whitney U test. P value<0.05 was considered to be statistically significance.[Results]Part I: Characteristics of mPD-PK of Morphine against thermal pain response and mechanical pain response: Physiological pain state1. Difference in analgesic effect-time course of morphine across thermal and mechanical pain stimulus modalities in naive rat(1) Following both i.g. and i.p. administration, morphine produced an equal elevation of pain threshold over the baseline values in response to both radiant heat and von Frey filament stimuli applied to the rat hind paw compared with saline (Sal) control group, however, the effect-time course recording showed that morphine had differential anti-nociceptive effects in time course across difference stimulus modalities (thermal and mechanical) following both i.g. and i.p. route of administration. The steady-state of anti-mechanical pain effect lasted for 4 hours, while that of anti-thermal pain effect only lasted for 5-15 minutes. The area under the effect-time course (AUE) for morphine against mechnical pain response was 1-2 fold larger than against thermal pain response.(2) When comparing the Tmax, the onset of maximal effect of anti-mechanical pain response was about 5 min (for i.g.) or 15 min (for i.p.) slower than that of anti-thermal pain response.(3) This differential effect of morphine could also be reflected by the Emax, for which the % MPE on mechanical pain response was nearly 20% (for i.p.) or 40% (for i.g.) higher than that on thermal pain response.2. Pharmacokinetics of analgesia produced by morphine(1) Based upon the NDST-21 analysis software, the data for unbound morphine from both i.g. and i.p. groups were analyzed by one-compartment open model, while that for conjugated morphine were analyzed by two-compartment open model.(2) The values of area under the concentration-time course (AUC) of the estimated conjugated morphine was 4.4 (for i.g.) and 5.6 fold larger than the unbound form ( P<0.01) although there was also a significant difference between routes of i.p. and i.g. administrations. However, the Cmax values of both unbound and conjugated morphine were significant higher for i.p. than for i.g. administration.(3) The Tmax values showed that the time of peak concentration of conjugated morphine was 15 min (for i.g.) or 30 min (for i.p.) slower than that of unbound form.3. mPD-PK modeling of morphine: concentration-effect-time course (CET) hysteresis loop plottingTo look for mPD-PK modeling for differential effect-time course of morphine against thermal pain response and mechanical pain response, we adopted hysteresis loop plotting which can reflect well the concentration-effect-time course (CET) models. Through comparing hysteresis loop plots between unbound and conjugated morphine in both i.g. and i.p. administration groups, short anti-thermal pain effect of the conjugated morphine fell in a clockwise hysteresis model, namely the effect decreased quicker than plasma concentration, while long-lasting anti-mechanical pain effect of both the unbound and conjugated morphine fell in a counter-clockwise hysteresis model, namely the effect rose slowly and reached peak later but sustained longer than concentration (Perez-Urizar et al., 2000). As an except model, the temporal relationship between blood concentration of unbound morphine and the anti-thermal pain effect was highly correlated regardless of the administration routes used in the present study, namely the Cmax resulted in an Emax and vice versa.Summary: The results of Part I experiments suggest that different pain stimulus modalities cause changes in mPD-PK modeling and the different modes of blood CET hysteresis loop might be responsible for the differential antinociceptive effect-over-time across thermal and mechanical pain stimuli. The short-term and fast anti-thermal pain effect model (clockwise hysteresis model) reflects a high correlation between blood CET and AUE as well as between Emax/Tmax and Cmax/Tmax, while the long-term and slower anti-mechanical pain effect model (counter clockwise hysteresis model) reflects that there is no correlation between blood CET and AUE as well as between Emax/Tmax and Cmax/Tmax. This implicates that anti-thermal pain effect of morphine might be caused mainly in peripheral compartment, while anti-mechanical pain effect of morphine might be caused mainly in central compartment (across the blood-brain barrier).Part II: Characteristics of mPD-PK of morphine in production of differential anti-nociception against thermal and mechanical hyperalagisia: physiological and pathological pain states 1. Comparison of the effect-time course of morphine against thermal and mechanical hyperalagisia between physiological and pathological statesTo explore whether pain states also have influences upon the mPD characteristics, mPD of morphine following i.g. administration was analyzed in rats under na?ve, inflamed and neuropathic states.(1) Based upon the effect-time course of anti-thermal pain response, peripheral inflammation shortened the Tmax of morphine from 40.00±3.46 min to 22.50±5.61min(n=6).(2) Based upon the effect-time course of anti-mechanical pain response, peripheral inflammation and neuropathic pain states elongated the Tmax of morphine from 30 min for na?ve state to 45 min for inflamed state and 90 min for neuropathic pain state. Moreover, neuropathic pain state decreased AUE when compared with the na?ve and inflamed state and this result gave an explanation of the bad effect of morphine for the neuropathic pain treatment.2. PK characteristics of morphine under physiologica and pathological statesTo get understanding about the effects of pain states on the PK characteristics of morphine, we compared the concentration-time course curves and concentration-effect curves under na?ve, inflamed and neuropathic states. The results showed that: (1) the AUC of morphine under neuropathic pain state was greatly decreased than that of na?ve and inflamed states; (2) the Tmax of blood concentration of unbound morphine for neuropathic pain state (30 min) was significantly shortened than na?ve (60 min) and inflamed (45 min) states, while for that of conjugated morphine, inflamed state shortened the Tmax to 45 min from na?ve state (60 min) with the neuropathic pain state no change; (3) there was no significant difference in Cmax among the three pain states; (4) the concentration-effect curves for both unbound and conjugated morphine against thermal hyperalgesia were left-shifted to that of na?ve state by inflamed pain state, while those for both unbound and conjugated morphine against mechnical hyperalgesia were right-shifted to that of na?ve state by both inflamed and neuropathic pain states; (5) through comparing the values of EC50, EC50 value was decreased by inflamed state for anti-thermal response, while that values were increased by inflamed and neuropathic states for anti-mechanical response, implicating that stimulus modalities and pain states might cause changes in affinity or activity of unbound and conjugated morphine which leading to changes in concentration-effect of the drug. 3. mPD-PK modeling of morphine: comparison of the concentration-effect-time course (CET) hysteresis loop plotting between physiological and pathological statesTo get understanding about the effects of pain states on the mPD-PK characteristics of morphine, we compared the CET hysteresis loop plotting under na?ve, inflamed and neuropathic states. The results showed that there was no change in modes for CET between physiological and pathological states, namely the short-term and fast anti-thermal pain effect model was shown as clockwise hysteresis model, while the long-term and slower anti-mechanical pain effect model was shown as counter clockwise hysteresis model under inflamed and neuropathic pain states.Summary: The results of Part II experiments suggest that PD-PK modeling of morphine is not only influenced by pain stimulus modalities, but also by pain states. However, unlike the outcome of the effects of pain stimulus modalities which cause changes in CET modes, changes in pain states might cause changes in absorption, metabolim as well as ligand-receptor (unbound morphine-receptor or conjugated morphine-receptor) binding affinity or activities. For analgesic effects of morphine, both peripheral inflammatory and neuropathic pain states can result in increase in velocity of absorption and shortening of the Tmax of the unbound morphine, while only peripheral inflammatory pain states affect the process of metabolism (glucuronidation of morphine). The affinity of unbound morphine-receptor and conjugated morphine-receptor binding and/or activities might be influenced by both pain stimulus modalities and pain states.Part III: Compartmentation and action site of morphine for anti-thermal hyperalgesia and anti-mechanical hyperalgesia: mPD-PK modelingBased upon the results of Part I, we proposed that the anti-thermal pain effect of morphine might be caused mainly in peripheral compartment (blood-peripheral tissues), while the anti-mechanical pain effect of morphine might be caused mainly in central compartment (brain tissues across the blood-brain barrier). To explore compartmentation and action site of morphine for anti-thermal hyperalgesia and anti-mechanical hyperalgesia, we compared the characteristics of mPD-PK modeling of unbound and conjugated morphine between blood and CSF samples. This part of analysis was mainly based on the data obtained from inflamed rats with i.g. administration of morphine.1.Comparison of PD-PK characteristics of unbound and conjugated morphine between blood and CSF samplesSimilar to the blood samples, there were also two compartments for both unbound and conjugated morphine. The AUC for conjugated morphine was 2.5 fold larger that that of unbound morphine (Conjugated vs. unbound: 103813.18±190.92 vs. 42082.67±344.24). The Tmax for unbound morphine in CSF samples was 10 min faster than conjugated morphine (50.00±3.46 vs. 60.00 min), however, the Tmax for both of the two forms of morphine in CSF samples was 15 min slower than that in blood samples, suggesting a cross transportation of the drug through blood-brain barrier.Through comparisons of the concentration-effect (CE) curves and the EC50 values, the CE curves of unbound morphine for both blood and CSF samples were positioned left-ward to those of conjugated morphine and the unbound morphine was with lower EC50 values than conjugated form when evaluation of anti-thermal pain effects. However, when evaluating the anti-mechanical pain effects, the position of CE curve for conjugated morphine from CSF was changed into left-ward to the unbound morphine although there was no change for the CE curve position of unbound and conjugated morphine from blood samples, implicating pain stimulus modalities can also change the affinity and activities between ligand-receptor as well as CE curves of two forms of morphine in the central compartment.2. mPD-PK modeling of unbound and conjugated morphine from blood and CSF samples: comparison of the concentration-effect-time course (CET) hysteresis loop plottingTo further explore the action site of two forms of morphine for anti-thermal and anti-mechanical pain effects, mPD-PK modeling was compared between blood and CSF unbound and conjugated morphine by hysteresis loop plotting.For anti-thermal pain effects, short-term and fast effect CET mode (clockwise mode) for CSF unbound and conjugated morphine was similar to that for blood samples, however, the distinct difference was that the concentration Tmax for both unbound and conjugated morphine in CSF was dramatically delayed, namely, over the period between 60-120 min when two forms of morphine reached steady-state of peak concentration, the anti-thermal pain effects were decreased fast with time , implicating occurrence of a morphine tolerance.For anti-mechanical pain effects, long-term and slower effect CET mode (counter clockwise mode) for CSF unbound and conjugated morphine was also similar to that from blood, however, it is interesting to note that the steady-state of high concentration (300-450μg/L) of conjugated morphine from CSF was remained no change between 45-300 min which can underline the phenomenon of long-term and slower analgesic effect of morphine on mechanical pain response and hyperalgesia. The CET mode for CSF unbound morphine was without difference from that of blood.Through Gassian correlation regression analysis, it was shown that the anti-thermal pain effects of morphine were highly correlated to the concentration-time factor of blood and CSF unbound (r=0.9981/0.6218) and blood conjugated morphine (r=0.9909) but with no involvement of CSF conjugated morphine (r=0.4948), highly supporting that the anti-thermal pain effects of morphine are mainly determined by the blood unbound and conjugated morphine (peripheral compartment). In sharp contrast, the anti-mechanical pain effects of morphine were highly correlated to the concentration-time factor of CSF conjugated morphine (r=0.9992) and CSF and blood unbound (r=0.7509/0.5473) but with no involvement of blood conjugated morphine (r=0.4463), highly supporting that the anti-mechanical pain effects of morphine are mainly determined by the CSF conjugated and unbound morphine (central compartment).[Conclusions]1. Different pain stimulus modalities (thermal and mechanical) can influence characteristics of mPD-PK modeling of morphine which might be determined by the different CET modes and well reflected by CET hysteresis loop plotting. Namely, the short-term and fast anti-thermal pain effect is well reflected by a clockwise hysteresis model, while the long-term and slower anti-mechanical pain effect is well relected by a counter clockwise hysteresis model. 2. PD-PK modeling of morphine is not only influenced by pain stimulus modalities, but also by pain states. However, unlike the outcome of the effects of pain stimulus modalities which cause changes in CET modes, changes in pain states might cause changes in PK parameter as well as ligand-receptor (unbound morphine-receptor or conjugated morphine-receptor) binding affinity or activities. For analgesic effects of morphine, both peripheral inflammatory and neuropathic pain states can result in increase in velocity of absorption and shortening of the Tmax of the unbound morphine, while only peripheral inflammatory pain states affect the process of metabolism (glucuronidation of morphine). The affinity of unbound morphine-receptor and conjugated morphine-receptor binding and/or activities might be influenced by both pain stimulus modalities and pain states.3. The anti-thermal pain effects of morphine are mainly determined by the blood unbound and conjugated morphine (peripheral compartment); while the anti-mechanical pain effects of morphine are mainly determined by the CSF conjugated and unbound morphine (central compartment).4. This result first provides a line of evidence that pain stimlus modalities as well as pain states might influence the whole processes of absorption (A), distribution (D), metabolism (M) and elimination (E) of morphine in the body and the affinity and/or activities between two forms of morphine and their receptors. Thus application of mPD-PK modeling in screening and R&D of new analgesics is of particular importance and significance in both formation of new pharmaceutical theory and clinical pharmacotherapeutics.