Title of Invention

OPIOID DELIVERY SYSTEM

Abstract

An opiod formulation for pulmonary administration in the treatment of pain, a pulmonary drug delivery device containing, method of administering, kit containing, and uses of same. The formulation contains at least one rapid-onset opioid and preferably also contains a sustained-effect opioid to reduce the frequency of administration. The invention employs the side effects of the opioid formulation to permit patients to self-limit drug intake, thereby avoiding toxicity while achieving analgesia. A pharmacokinetic and pharmacodynamic model is employed to determine optimum drug formulations and optimum parameters for administration.

Full Text

Field of the invention This invention relates to pharmaceutical preparations and methods of administering same and, more particularly, to opioid based analgesics and a method for their administration. Background of the invention Opioids are among the oldest drugs in existence, and remain a mainstay of pain management. Opium, the original opioid, is derived from poppy plants. "Opiates" are natural derivatives of opium, and include morphine, methadone, and heroin. "Opioids" are a broader class of drugs, that includes opium, opiates, and synthetic drugs with the same pharmacological effect of opium. Commonly used synthetic opioids include meperidine, fentanyl, alfentanil, sufentanil, and remifentanil. Opioids are believed to exert their effects through binding of the mu receptor in the spinal cord and brain, and peripheral tissues. Binding at the mu receptor induces a wide variety of pharmacological effects, including therapeutic effects such as analgesia, effects which may be viewed as either side effects or therapeutic effects, depending on context, including sedation and decreased bowel motility, side effects such as nausea, vomiting, urinary retention, pruritis, ventilatory depression, addiction, and toxicity such as severe ventilatory depression, loss of consciousness and death. Opioids differ from each other in many ways, including their route of delivery, their physicochemical composition, their drug absorption rate, their pharmacokinetics, and their pharmacodynamics. Noninvasive routes of opioid delivery include oral, rectal, transdermal, transmucosal, and via inhalation. Invasive routes of opioid delivery include intravenous, intramuscular, epidural, spinal, and by injection into joints. When injected intravenously, some opioids quickly enter the brain and spinal cord and thus have a very rapid onset of drug effect (e.g., alfentanil and remifentanil), while others are absorbed slowly to the site of action and have very slow onset of drag effect (e.g., morphine). Similarly, for some opioids the drug effect is very short-lived, owing to very rapid metabolism (e.g., remifentanil), while other opioids may have very slow metabolism and prolonged effect (e.g., methadone). In terms of pharmacodynamics, the potency of opioids covers nearly 5 orders of magnitude, from extraordinarily potent opioids such as carfentanil and etorphine (both used to stun elephants) to relatively less potent drugs such as methadone and morphine. The equivalent potencies of opipids (measured as a "therapeutic equivalence ratio") are well established in the literature, and are often used when- changing a patient's treatment regimen from one opioid to another. Despite these differences, all opioids have the same potential to produce both profound levels of analgesia, and profound toxicity from hypoxia, which can be fatal. Because of the risk of hypoxia, physicians are reluctant to use appropriate dose|s of opioids to treat acute and chronic pain; As a result, hundreds of thousands of patients who could be provided better pain control receive inadequate doses of opioids. Conversely, even with an understandably cautious approach by the health care community to treatment of pain, every year, many patients die from opioid-induced ventilatory depression. Pain is highly variable and highly subjective. Different patients respond differently to opioids. As a result, different patients need different amounts of analgesia to treat their pain. As such, it has become desirable to allow patients to vary the amount of analgesic they receive. One attempt to better adjust opioid dosing in patients has been the introduction of "patient controlled analgesia" ("PCA") (Ballantyne JC, et al. Postoperative patientcontrolled analgesia: Meta-analyses of initial randomized control trials. J Clin Anesth 1993:5:182-193.) With the PCA system, the patient must be awake, and must activate a delivery mechanism to receive more opioid, before the drug is given. If the patient becomes overdosed from the opioid, then the patient will become unconscious and not request additional drug. In this manner, the PCA system uses a side effect of opioid, sedation, to limit the amount of opioid given. One problem with the PCA system ig that the drug is injected rapidly after the patient requests it (typically, the time frame of administration of drug is under 1 minute) and because the drug most frequently used in the PCA is morphine, a drag that is slowly transferred from the plasma to the si :e of action - this results in a delay between the patient request for drug and the analgesic effect of the drug. As a result of this delay, patients often request a second (or third) dose of the drug while the opioid effect level of the first injection is still rising. PCA systems include a "lockout" period (commonly 5 minutes), which helps prevent patients from administering more opioid while the opioid drug effect is still rising. Lockout periods are typically controlled, defined or programmed by the health care provider, and there have been many instances where user error or inadvertence in programming the lockout period have resulted in the death of the patient. The patient also often feels •frustrated by the lockout, as it diminishes the patients' control of dosing. Other disadvantages of the PCA include the invasive parenteral (intravenous) administration as well as the expensive infusion pumps thus restricting the use of the PCA to institutionalized patients. A second attempt to better adjust opioid dosing in patients is in the selfadministration of Nitrous Oxide during labour associated with childbirth. A nitrous ox|de mask is held to the face by the patient during contractions, and is released from the face when adequate analgesia is achieved. However, this mechanism is a titration to analgesic effect and not used as a safety mechanism, since overdosing on nitrous oxide using this system of administration is not a significant concern. Furthermore, nitrous oxide is a gas which requires a heavy steel tank for storage and a complex delivery system for administration. Therefore, the use of nitrous oxide is primarily restricted to the hospital environment and not for ambulatory patients. An additional potential problem with nitrous oxide relates to its low potency and thus the necessity of administering a high concentration (more than 50%) of nitrous oxide in oxygen with a potential of a hypoxic mixture. The current invention seeks to use two physiological responses of opioids: sedation and ventilatory depression, to limit the total dose of opioids that patients receive. la this manner, the invention seeks to increase safety of opioid drug delivery beyond what is currently accomplished with PCA or other existing opioid administration methods whereby only a single side effect is used to limit the exposure of patients to dangerously high levels of opioid drug effects. The invention also improves the use of sedation by removing the need for a "lockout" period, currently required in PCA systems, arid removing the frustration and user error possible therein. Summary of the invention Accordingly, the invention provides in accordance with a first aspect, an opioid formulation for use in a method of providing analgesia to a patient comprising the-steps of: continuously inhaling the formulation using a pulmonary drug delivery device to produce analgesia; and stopping inhalation when satisfactory analgesia is achieved OK at the onset of a side effect; wherein the formulation comprises an effective amount of at least one rapid-onset opioid; and a pharmaceutically acceptable carrier, the concentration arid type of each bpioid, and amount of and particle size of the formulation delivered from the device on each iimalation, being selected so that, during inhalation, analgesia is achieved before the onset of said side effect, and the onset of said side effect occurs before the onset of toxicity, and so that the maximum total opioid plasma concentration does not reach toxic levels, whereby the onset of said side effect can be used by the patient to terminate inhalation to avoid toxicity. The concentration of each opioid is such that the at least one opioid enters the patient's system in an incremental and gradual fashion so that analgesia is achieved, followed by the onset of side effects, and well in advance of toxicity. Any conventional diluent suitable for use in formulations for pulmonary administration may be used such as saline or sterile water. It will be appreciated that the concentration will be adjusted leased on other parameters, including the type of pulmonary delivery device that is used and, more particularly, the amount of formulation which is dispensed by the device on each inhalation and particle size of the formulation dispensed (expressed in terms of "mass medium aerodynamic diameter"). In order that the formulation is bioavailable (i.e. is deposited hi the lungs), the formulation should be dispensed by the pulmonary drug delivery device at a mass medium aerodynamic diameter of from 1 to 5 microns, or from 1 to 3 microris, or from 1.5 to 2 microns. To avoid toxicity following termination of inhalation, the maximum total opioid plasma concentration at the onset of side effect is preferably no less than 66% of the maximum total opioid plasma concentration, and more preferably no less than 80% of the maximum total opioid plasma concentration. The rapid-onset opioid may be chosen from fentanyl, alfentanil, sufentanil and remifentanil and is preferably fentanyl and alfentanil. When used as the sole rapid-onset opioid, the amount of sufentanil may range from 1.7 to 100 meg/ml and the amount of remifentanil may range from 3.3 to 3 300 meg/ml. In order to reduce the frequency of administration, the opioid formulation may comprise an effective amount of at least one sustained-effect opioid to provide sustained relief. It will be appreciated that the relative concentration of the sustained-effect opioid and rapid-onset opioid must be adjusted in order that the formulation may still achieve jhe same desired result, namely, to produce analgesia followed by at least one side effect before a toxic level of opioid in plasma is reached, and to avoid toxic levels being reajch after termination of inhalation. In mixed formulations contained rapid-onset and sustained-effect opioids, the rapid-onset opioid serves to limit the dose of both drujgs below a dose that would cause toxicity. The sustained-effect opioid may be chosen from morphine, morphine-6- glucuronide, methadone, hydromorphone, meperidine, an opioid encapsulated in a biocompatible carrier that delays release of the drug at the lung surface (as are known in the art), and a liposome encapsulated opioid (e.g.. liposomally encapsulated fentanyl). Morphine-6-glucuronide may be present hi a concentration of from 3.3 to 3300 mg/ml, methadone may be present in a concentration of from 0.3 to 33 mg/ml, hydromorphone may be present in a concentration of from 0.03 to 7 mg/ml, and meperidine may be present in a concentration of from 1.7 to 170 mg/ml. Depending on the identity of the opioids used in the formulation, the ratio of concentration of total sustained-effect opioid to concentration of total rapid-onset opioid will vary. A preferred formulation giving rise to both rapid-onset and sustained analgesic effect is one in which the opioids consist of fentanyl and liposomally encapsulated fentanyl. The ratio of concentration of liposomally encapsulated fentanyl to fentanyl may be from 1:2 to 6:1, or from 1:1 to 5:1, or from 2:1 to 4:1, or about 3:1. Furthermore, the total opioid concentration may be from 250 to 1500 meg/ml. The liposomally encapsulated fentanyl may be present in a concentration of from 3.3 to 3300, or from 250 to 1500, or from 267 to 330, or from 100 to 750 meg/ml. This total opioid concentration maybe about 500 meg/ml, the fentanyl concentration may be about 200 meg/ml and the liposomally encapsulated fentanyl concentration may be about 300 meg/ml. The concentration of fentanyl, and amount and particle size of the formulation delivered from the device, may be selected so that from 4 to 50, or from 10 to 20, or about 15 mcg/min., of fentanyl is deposited in the lungs during inhalation. The concentration of liposomally encapsulated fentanyl, and amount and particle size of the formulation delivered from the device, maybe selected so that from 5 to 150, or from 10 to 90, or from 15 to 60, or from 20 to 45 mcg/min of liposomally encapsulated fentanyl is deposited in the lungs during inhalation. In another embodiment, the opioids in1 the formulation may consist of alfefitanil and morphine. The alfentanil may be present in a concentration of from 300 to 6700 meg/ml. Furthermore, the concentration of alfentanil, and amount and particle size of the formulation delivered from the device, may be selected so that from 100 to 500, or about 250, mcg/min. of alfentanil is deposited in the lungs during inhalation. The morphine may be present in a concentration of from 650 to 13350 or from 0.3 tos33 mg/ml. As well, the concentration of morphine, and amount and particle size of the formulation delivered from the device, may be selected so that from 100 to 2000, or from 200 to 1000, or about 500, mcg/min of morphine is deposited in the lungs during inhalation. In accordance with a second aspect of the invention, there is provided a method of administering an opibid formulation to provide analgesia to a patient, comprising the steps of: continuously inhaling the formulation using a pulmonary drug delivery device to produce analgesia; and stopping inhalation when satisfactory analgesia is achieved or at the onset of a side effect; wherein the formulation comprises an effective amount of at least one rapid-onset opioid and a pharmaceutically acceptable carrier; the concentration and type of each opioid, and amount of and particle size of the formulation delivered from the device on each inhalation, being selected so that, during inhalation, analgesia is achieved befori the onset of said side effect, and the onset of said side effect occurs before the onset of toxicity, and so that the maximum total opioid plasma concentration does not reach toxic levels, whereby the onset of said side effect can be used by the patient to terminate inhalation to avoid toxicity. The pulmonary drag delivery device may comprise: a container containing said formulation; means for'forming the formulation into particles having a mass mediim aerodynamic diameter of from 1 to 5 microns for delivery to a patient; an outlet through which the formulation is dispensed; and means for dispensing said formulation through said outlet; wherein the device is adapted to dispense the formulation only through an exercise of conscious effort by the patient. The time to onset of the side effect (e.g. ventilatory depression and/or sedation) will vary from patient to patient and will depend on factors such as the opioids used, thoir concentrations, relative amounts, and particle size. For example, the onset may occur from 3-15 minutes following the start of inhalation, and the side effect may peak within 1-2 minutes after the patient stops inhaling the formulation (i.e. at the end of dose). Typically, the patient will inhale the formulation over a period of from about 2 to 20 minutes, at a normal inhalation rate, before the onset of side effects. Similarly, the tirie after the end of the dose by which the maximum plasma concentration is reached will also vary based on various parameters include the above-mentioned parameters. For examp e, the maximum plasma concentration may be reached within 0 to 5 minutes following tie end of dose. Normal inhalation rates are typically less than 20 and more often from 5 to 15 breaths per minute. In accordance with a third aspect of the invention, there is provided a pulmona]ry drug delivery device comprising: a container containing a formulation comprising an effective amount of at le^st one rapid-onset opioid and a pharmaceutically acceptable carrier; means for forming the formulation into particles having a mass medium aerodynamic diameter of from 1 to 5 microns for delivery to a patient; an outlet through which the formulation is dispensed; and means for dispensing said formulation through said outlet; wherein the device is adapted to dispense the formulation only through an exercise of conscious effort by the patient; and the concentration and type of each opioidj, and amount of and particle size of the formulation delivered from the device on each inhalation, is selected so that, during inhalation, analgesia is achieved before the onset of said side effect, arid the onset of said side effect occurs before the" onset of toxicity, and so that the maximum total opioid plasma concentration does not reach toxic levels, whereby the onset of said side effect can be used by the patient to terminate inhalation to avoid toxicity. Examples of devices which require a conscious effort by the patient to actuate include those having a button which must be manually depressed using a certain degree of force, and those which are breath actuated. The device may further comprise a mechanical or electrical delivery rate controlling means for limiting the,rate at which the formulation is dispensed to below a selected threshold such that the objects of the invention may be achieved^. The delivery rate controlling means may include a mechanism for measuring ventilatory depression and limit the rate of dispensation based on this information. Ventilatory depression) may be measured by the respiratory rate, inspiratory force or value of end tidal CO2 cf the patient. The device may weigh from 250 to 2500 grams and may be designed such that the weight is adjustable on a patient by patient basis. Furthermore, the outlet ma designed to be difficult to maintain in the patient's mouth if the patient is not conscious. For example, the outlet may comprise a fenestration which must be seal y be fully dby the lips of the patient in order for the formulation to be dispensed, m this way, continued administration of the formulation may be hampered by the onset of side effects such as sedation and ventilatory depression. The invention also provides an opioid administration kit, in accordance w^ith a fourth aspect, comprising: . a pulmonary drug delivery device comprising a container; means for forming a formulation contained in the container into particles having a mass medium aerodynamic diameter of from 1 to 5 microns for delivery to a patient; an outlet through which the formulation is dispensed; and means for dispensing said formulation through said outlet; wherein the device is adapted to dispense the formulation only through an exercise of conscious effort by the patient; and an opioid formulation contained in the container or for receipt by the contaiier, said formulation comprising an effective amount of at least one rapid-onset opioid acd a pharmaceutically acceptable carrier; the concentration and type of each opioid, ,aid amount of and particle size of the formulation delivered from the. device on each inhalation, being selected so that, during inhalation, analgesia is achieved before the or set of said side effect, and the onset of said side effect occurs before the onset of toxicity, £tnd so that the maximum total opioid plasma concentration does not reach toxic levols, whereby the onset of said side effect can be used by the patient to terminate inhalation! to avoid toxicity; and | instructions for using said kit comprising the steps of continuously inhaling the formulation using a pulmonary drug delivery device to produce analgesia; and stoppi ig inhalation when satisfactory analgesia is achieved or at the onset of a side effect. The 1 it might include an empty device which must be filled with the opioid formulation prior to administration. In such event, the instructions may include the step of pre-filling the container. " hi accordance with further aspects of the invention, mere is provided a use of the present formulation in providing analgesia to a patient and in the manufacture of la medicament for doing same. Useful drug formulations and parameters for administration according to present invention can be determined by the person skilled in the art based on kno pharmacological data as well as through pharmacokinetic and pharmacodynamil modeling as herein described. Such modeling is intended to ensure that analgesic effe is achieved before the onset of a side effect, and that the onset of the side effect occ well in advance of toxicity, and to ensure that once the patient stops inhaling the formulation, there will not be a continued rise in total opioid concentration in the plasm to toxic levels. Brief Description of the Drawings Figure 1 is a flow diagram which represents a computer simulation model for sedation. Figure 2 is a flow diagram which represents a computer simulation model for ventilatory depression. Figure 3 is a flow diagram which represents a computer simulation model for an inhalation device. : , ' ' • ' I Figure 4 is a flow diagram which represents a computer simulation model for the pharmacokinetic profiling of opioid as administered to a patient through a pulmonary route. Figure 5 is a flow diagram which represents the Stella™ computer simulation of the pharmacpkinetics of the administration of a single opioid. Figure 6 is a graph showing output of the Stella™ computer simulation of Figure 5 (ventilatory depression and sedation models disabled) expressed as a time course of quantity of opioid in the inhalation device and in the lung of the patient. Figure 7 is a graph showing the time course of ventilatory depression hi the Stella™ computer simulation of Figure 5 (ventilatory depression and sedation me dels disabled). Figure 8 is a graph showing the time course of quantity of opioid in the inhalation device and in the lung of the patient, in the Stella™ computer simulation of Figure 5 (ventilatory depression model enabled, sedation model disabled). Figure 9 is a graph showing the time course of ventilatory depression in the Stella™ computer simulation of Figure 5 (ventilatory depression model enabled, sedition model disabled). Figure 10 is a graph showing the time course of quantity of opioid in the inhalation device and in the lung of the patient, in the Stella™ computer simulatiojn of Figure 5 (ventilatory depression and sedation models enabled). Figure 11 is a graph showing the time course of ventilatory depression in the Stella™ computer simulation of Figure 5 (ventilatory depression and sedation models enabled). Figure 12 is a flow diagram which represents a computer simulation model foj: the administration of two opioids. Figure 13 is a flow diagram which represents the Stella™ computer simulation of the pharmacokinetics of the administration of two opioids. Figure 14 is a graph showing the output of Stella™ computer simulation of Figjire 13 expressed as a time course of total quantity of opioid in the inhalation device and] in the lung of the patient (ventilatory depression and sedation models enabled). ! • . I Figure 15 is a graph showing the time course of concentration of each opioid a|nd of total opioid at the effect site in the Stella™ computer simulation of Figure 13 (ventilatory depression and sedation models enabled). Figure 16 is a graph showing the time course of ventilatory depression during ajtid after delivery of opioids in the Stella™ computer simulation of Figure 13 (ventilatcjry depression and sedation models enabled). Figure 17 is a flow diagram which represents the StellaT M computer simulation of the pharmacokinetics of the administration of two opioids, where the two opioids beijig administered are alfentanil and morphine. 9 Figure 18 is a graph showing the time course of concentration of alfentanil, morphine, and combined opioid at the effect site in the Stella™ computer simulation Figure 17 (ventilatory depression and sedation models enabled). Figure 19 is a graph showing the time course of ventilatory depression in 1 Stella™ computer simulation of Figure 17 (ventilatory depression and sedation models enabled). Figure 20 is a graph showing maximum concentration of opioid in the plasma against end of dose concentration of opioid in the plasma of patients administered opioid. Figure 20A shows patients administered with a combination of fentanyl and liposomal y encapsulated fentanyl through a pulmonary route. Figure 20B shows patients administered with fentanyl intravenously. Figure 21 is a graph showing time to side/toxic effect versus tune to end of dos|e for the side effects and toxic effects of patients administered a combination of fentanyl and liposomally encapsulated fentanyl through a pulmonary route. Figure 22 is a table showing the statistical correlation of side effect to toxic effect. Detailed description of the invention In this application, the following terms have the following meanings: "Analgesic effect" or "analgesia" means the relief from pain resulting from action of a drug. "Drug delivery profile" means the concentration of the drug, over, time, at the of drug effect, as determined by the amount and rate of drug administered to the patient and the pharmacokinetics relating dose inhaled to concentration in the lungs, plasma and at the site of drug effect. the site "Hypoxia" is atoxic effect of opioid administration, and is defined in this application decrease in blood Oz concentration to less than 90% saturation. as a 'Ventilatory depression" means a decrease in the fate, tidal volume, and/or flow rate of air into the lungs. Ventilatory depression may manifest as dizziness, shortnej s of breath, or a slowing in rate of breathing. "Opioid induced ventilatory depression" mfers to ventilatory depression caused by the action of an opioid at a site of drug effect. "Sedation" means a decrease in attention, mental awareness, focus, and stale of consciousness caused by opioids, and manifests in a lack of physical strength (muscle fatigue), lack of voluntary activity, lethargy, drowsiness, and sleep. "Opioid-indiiced sedation" refers to sedation caused by the action of an opioid at a site of drug effect. "Rapid onset", when used to describe a drug formulation, means a formulation which has an' analgesic effect that rapidly follows the rise in plasma opioid concentration. A "rapid onset opioid" is an opioid that has an analgesic effect within 5 minutes of administration. "Sustained effect" means a formulation which has an analgesic effect that is sustained over several hours. A "sustained effect opioid" means an opioid that has analgesic effect that lasts over 2 hours. "Side effect" means an effect of an opioid that is not analgesic or toxic. For example, severe ventilatory depression is an example of opioid toxicity, while mild ventilatory depression and sedation are not considered signs of opioid toxicity, but are side effects of the opioid. "Site of effect" refers to a physical or hypothetical site of drag action within the patient. "Site of effecf' may be a compartment of the body, such as the brain, the liver, or the spleen, or it may be a theoretical and unknown location based on correlation and pharmacokinetic modeling. For example, it is known that opioids exert their analgesic actions, in part, in the substantia gelatinosa of the spinal cord, so this is a site of opioid analgesic effect. The concentration of opioid at the effect site may be determined by direct measurement, or through the use of pharmacokinetic and pharmacodynarnic modeling. "Effective amount" means the amount of drug needed to reach an analgesic effejct. "mass medium aerodynamic diameter" means the aerodynamic diameter j ah aerosol such that half of the cumulative mass of all particles is contained in particles wjith smaller (or larger) diameters and wherein the aerodynamic diameter is defined as ihe diameter of a unit-density sphere having the same gravitational settling velocity as particle being measured. "breathing rate" means the number of breaths taken per unit of time. 'Titration to effect" means administering an opioid until a satisfactory analgejsic effect is felt'by the patient, then ceasing administration of the opioid. 'Titration to side effect" means administering an opioid until a side effect is f^lt, then ceasing administration. The ceasing of administration may be voluntary (for example, by instructing patients to cease administration of the opioid when they start! to feel drowsy, dizzy or short of breath) or involuntary (for example, when patients are ao longer able to breathe effective dosages of opioid due to ventilatory depression or sedation). The terms "toxic", "toxicity", "toxic effect" and "opioid toxicity" refer to effe ;ts of opioids that place a patient at risk of death. For example, opioids commonly produce modest amounts of ventilatory depression that pose little risk to a patient. This is rot considered an example of opioid toxicity. However, severe Ventilatory depression poses the risk of hypoxia, loss of consciousness, and death. Thus, severe ventilatory depression is an example of opioid toxicity, while mild ventilatory depression is not considered a sign of opioid toxicity. The present invention is for use in patient self-administration of opioids. Tie invention utilizes the opioid's side effects to self-regulate the amount of opioid given tol a patient, thereby tailoring the dose to achieve the patient's analgesic requirements, while avoiding toxicity and death. The use of the invention begins with the patient's perception of pain. There are many modalities of treating mild to moderate pain, but opioids are the mainstay of treatment for severe pain. In response to the severe pain, either the patient or .the patient's • care provider open a 'prefilled vial of opioid in liquid solution, or, alternatively, in an emulsion. The liquid is added to a nebulizer. The nebulizer is then brought to the mouth, and is held there with the hand. The ' . ' • • ' ••,: '•:.' ;. . ' • ' , ' • . I . : • " ' . ' • • • • ' nebulizer is not attached to the face with straps, as this prevents the 'self-Uniting mechanism from working. With each breath, the nebulizer releases a small amount of the liquid opioid as an aerosol. The aerosol passes through the patient's mouth and into the trachea and Ipngs, where the aerosolized opioid is deposited. Throughout mis patent application, the nebulizer is also called an inhaler lor an aerosol pulmonary drug delivery device. An inhaler may refer to either a nebulizer or a nebulizer combined with a source of compressed air or oxygen, or any other aerosol generating device for the administration of drug by way of the lungs. An aerosol pulmonary drug delivery device refers to any device that allows the aerosolization of a substance for delivery into the lungs. Various nebulizer technologies are known and available in the art. The rate of onset of opioid drug effect is believed to be dictated by the spejd at which the opioid enters the lungs and the rate at which the opioid crosses the blood jrain barrier. Some opioids, such as alfentanil and remifentanil, cross the blood brain barrier very quickly, and thus produce very rapid onset of drug effect. Other opioids, such as morphine and morphine-6-glucuronide, cross the blood brain barrier very slowly, and thus produce a slow onset but sustained effect. As the opioid crosses the blood brain barrier, it starts to exert effects at the site of drug action. Although in some instances, patients may feel effects differently, typically, as concentration of opioid increases, the effects felt are analgesic effect, side effect, and toxic effect, in that order. Ventilatory depression is up and down regulated by the opposing actionb of opioids (which depress ventilation) and carbon dioxide (which increases ventilation). This occurs in a feedback loop as follows: initially the opioids will depress ventilation. Because the patient is not exhaling as much carbon dioxide, the level of carbon dioxide in the patient's blood will rise. As the carbon dioxide rises, it stimulates ventilation, partly offsetting the opioid-induced ventilatory depression. The opioid-induced. yentilatoiy depression must come on sufficiently rapidly so that it occurs as the patient is inhaling tt e opioid, thus serving to limit the amount of opioid inhaled. However, it must not come on so rapidly as to place the patient at risk of toxic effects before the carbon dioxide has had a chance to rise and offset the opioid induced ventilatory depression. The amount of opioid inhaled by the patient each minute is proportional to tt|e ventilation during that minute. As ventilation becomes depressed, the rate of opioijd delivery to the lungs is depressed proportionally. In this way, the rate of delivery :s slowed by ventilatory depression, decreasing the ability of the patient to self-administer a toxic dose of opioid. The slowed uptake of opioid from ventilatory depression creates the opportunity for complete cessation of drug delivery through the onset of sedation. As the opioids exert their analgesic effects, patients will become sedated, in part from the mitigation of their pain, in part due to the side effects of the opioids. As sedation develops in patients, it becomes difficult to hold the device to the mouth, maintain a sei with the lips, and breathe through the device to administer additional opioid. Instead, the patient begins to breathe through the nose, or through the mouth but around the mouthpiece of the nebulizer. With increasing sedation, the arm drops away from the airway, removing the device from the mouth. This dropping away of the arm may the encouraged to take place at a lower level of sedation by making the device deliberately heavy, or by adding a weight to the device. Weight of the device can be adjusted from patient to patient, depending on the individual patient's strength pre-sedation. Since the side effects of the opioids typically occur at lower opioid concentrations (as compared to the opioid toxic effects), a safer, patient self-limited opioid administration has been created through the pulmonary administration of an opioid (or a combination of opioids) at a rate sufficiently slow to allow for a time gap between th e onset of side effects and the onset of toxic effects. The rate must also be slow enough (is compared to the rate of onset of the opioid) to allow for the onset of side effects -while the dose is being administered. In a clinical study relating to this invention, healthy subjects were directed tio inhale a fentanyl formulation consisting of rapidly acting free fentanyl and sustained acting liposomal encapsulated fentanyl over 2-20 minutes. In this study, several subjects attempted to self-limit the dose and required external assistance to receive the entire c ose. Some subjects self-limited the dose because of opioid-induced ventilatory depression, with a decrease in ventilation rate reducing the amount of drug irihaled. Other subsets self-limited the dose because of sedation, and their inability to hold the device tc the mouth to continue inhaling fentanyl. Some subjects exhibited both side effects. The trial demonstrated that patient will, in fact, self-Limit fentanyl administration via the pulmonary route before a toxic level of fentanyl is administered, when 1) the drug is intended to be inhaled over a deliberately extended period of time (e.g. 2-20 minutes), 2) opioid induced ventilatory depression occurs while the drug is being given (and before a toxic dose is administered), and/or 3) sedation occurs while the drug is being given {and before a toxic dose is administered). We have found mat these factors can be controlled by designing the rate at which an opioid is given to a patient accordingly. Prefe-rably, the opioid formulation is administered over 2-20 minutes. The total amount of opioid administered over the 2-20 niinutes will depend on several factors, including the type of opioid or combination of opioids delivered, and the mass medium aerodynamic diameter (MMED) of the particles being inhaled. This administration period results hi a rate of onset to effect that is influenced by the rate of administration and affords the patient the ability to involuntarily self-limit the dose through the onset of ventilatory depression and sedation. We have found that, for an alfentanyymorphine combination drug, a range of 100-500 mcg/min of alfentanyl and 200-1000 mcg/min of morphine is optimal (measured as drag delivered to the lung of the patient ("systemically available drug")). For a free and liposomally encapsulated fentanyl formulation, we have found that the levels for systemically available drug to be optimum at 5-50 mcg/min of free fentanyl and 15-150 mcg/min of liposomally encapsulated fentanyl. We have found that nebulized particles with an MMED of 1-5 microns typically have bioavailability of about 20%, which means the optimum drug flow from the nebulizer should be between 25J250 mcg/min of free fentanyl and 75-750 mcg/min of microsomaUy encapsulated fentanyl. Our current preferred embodiment of the invention comprises a drug flow from the inhaler of 75 mcg/min of free fentanyl and 250 mcg/min of limposomally encapsulated fentanyl. For other opioid formulations, we expect that a therapeutically equivalent rate )f systemically available drug to have similar advantages. In order to prevent peaks of. opioid effect that are more-potent than the concentration at which patients stop taking the drug in a multiple opioid formulation wih at least one rapid-onset opioid and at least one sustained effect opioid, we expect that the ratio of sustained-effect opioid to rapid-onset opioid administered should be less than 1 in terms of therapeutic equivalent potency. Another factor affecting the rate of administration of opioid is the patient creating rate. We have found that a breathing rate of 10-15 breaths per minute (i.e. "normal" breathing rate) is preferred. Opioid response is highly individualized. This reflects, in part, varying levels painful stimulation. In the presence of very severe pain, very high doses of opioids can I administered without undue toxicity. Patients being administered chronic opioids requi higher doses to produce both the desired therapeutic effects and opioid toxicity. This al; reflects the development of tolerance to opioids. Physicians have sought improved meai of administering opioids in part because of the wide range of doses required to adequate tailor the opioid to the needs of individualized patients. With the described invention, patients who need large doses of opioids to provi analgesia can elect to administer either a larger volume of drug (inhaled over a long period of time), or can be offered a more concentrated solution of drug to be inhaled ov the expected 2-20 minutes. Either way, the opioid-induced ventilatory depression ar sedation will still attenuate, and eventually terminate, drug administration before tox doses are inhaled. Preferably, the patient will inhale the drug over a longer period of time Conversely, a patient who requkes only a small dose will experience the desired pain relief during inhalation. The patient can elect to not inhale additional drag. The patieit who unwisely continues to self-administer opioid despite obtaining the desired pain re] will experience Ventilatory depression and sedation, which will then either voluntar (according to the instructions given to the patient) or involuntarily'(cUie to the side effe themselves) attenuate and subsequently terminate drag administration before inhalation of a toxic dose of opioid. The patient is therefore empowered to self-titrate to analgesic effect, without a lockout period and with a lower risk of toxicity. The selection of opioid and opioid concentration (as disclosed above, or otherwise) for the device requires consideration of the time course of opioid absorption from the lung into the plasma, and the time course of opioid transfer from the plasma into ' . ' • ' . ' • • : / . .' j • • - ' . : • „ • ' •- the site of drug effect (e.g., the brain or spinal cord). ' " • • ' • • : • • ' ' _. '. * • • .- • • i Some opioids are associated with very rapid absorption from the lung into the systemic circulation. For example, the absorption of free fentanyl from the lung into the plasma is nearly instantaneous. This would likely be true of remifentanil, alfentanil. and sufentanil as well. The absorption of free fentanyl released from the lipospmal encapsulated fentanyl from the lung to the plasma is far slower. Some opioids are associated with very rapid transfer from the plasma to the site of drug effect. For example, peak alfentanil and remifentanil concentrations at the si;e of drug effect occur within 2 minutes of intravenous-injection. Other opioids are associated with very slow transfer from the plasma to the site of drug effect. For example, the peak drug effect from an intravenous dose of morphine maybe delayed by 10-15 minutes from the time of the injection. For the self limiting opioid delivery system to work, one of the opioids srjould have both rapid transfer from the lungs to the plasma, and rapid transfer from the plasma to the site of opioid drug effect. Fentanyl, alfentanil, sufentanil, and remifentanil all liave this characteristic (rapid onset). It may be that meperidine and methadone also have effect, but that is not presently known. Although it is possible to obtain the reqi parameters of the invention with a single opioid, we have found that combining the onset opioid with a slower acting, but sustained effect opioid gives a preferred resu the patient typically feels analgesic effect for longer periods of time with su combination. " If the desire is to maintain the opioid analgesic effect, then it may be necessi combine the rapid onset opioid with an opioid that has a slower onset, but sustc this aired apid t, as li a effect. Examples of such formulations include (1) a formulation of fentanyl and liposDmal encapsulated fentanyl, (2) a formulation of remifentanil, alfentanil, sufentanil, or fen;anyl in combination with morphine, and (3) a formulation of remifentanil, alfen;anil, sufentanil, or fentanyl in combination with methadone. Care must be taken to prevent a second "peak" of action, at the time of maximum effect of the sustained effect opioidt that y to ined is higher than the peak caused by the rapid onset opioid, which allows the patient to feel side effects while he or she is administering the drug. When a rapid onset opioid is combined with an opioid with slow-onset and sustained effect, the concentration of both opioids is adjusted so that the self-limiting effects of the rapid-onset opioid serves to limit exposure of the patient to the slow-onfiet opioid. The rapid onset opioid acts as an early warning system of sorts, triggering side effects in ari adequate timeframe. We have found mat side effects are experienced before toxicity is reached. More specifically, subjects that experienced side effects at the end of dosing or shortly af :er completion of dosing did not progress to toxic side effects whereas subjects that experienced side effects during dosing and continued to inhale drug progressed to toxicity, specifically, hypoxia. t As can be appreciated by the above description, creation of the invention requires (1) thorough understanding of the pharmacokinetics and pharmacodynamics of one or more opioids, and (2) thorough understanding of the relationship between opioids, cafbbn dioxide production and elimination, and ventilation, (3) careful selection of one or mqre opioids, and (4) precise determination of the optimal concentration of each opioid in the final formulation in order to achieve the desired clinical profile of the drug. The fir.al formulation is determined by pharmacokinetic and pharmacodynamic modeling of tie system parameters, with dose optimization performed to find the dose that exhibits tie best patient safety profile while still providing an adequate analgesic response. In the Drawings Figure 1 is a flow diagram which represents a computer simulation model ibr sedation. In all flow diagrams, squares represent amounts, arrows represent rate (amounts per unit time), and circles represent either a calculation, rate, or constant. Figure 2 is a flow diagram which represents a computer simulation model fjar ventilatory depression. Figure 3 is a flow diagram which represents a computer simulation model for inhalation device. Figure 4 is a flow diagram which represents a computer simulation model fcr the pharmacokinetic profiling of opioid as administered to a patient through a pulmonary route. Figure 5 is a flow diagram which represents the Stella™ computer simulation of the pharmacokinetics of the administration of a single opioid. Figure 6 is a graph showing output of the Stella™ computer simulation of F: gure 5 (ventilatory depression and sedation models disabled) expressed as a time course of quantity of opioid in the inhalation device, and quantity of opioid in the lung o patient. The X axis shows time in minutes. The Y axis shows dose units of formulation, in mg. The amount of drug in the inhaler dropped steadily over the first 10 minut;s of stimulation. The amount of drug in the lungs reflects the net processes of inhalation of drug into the lungs and absorbtion of drug from the lungs into the systemic circulation. Figure 7 is a graph showing the time course of ventilatory depression in the Stella™ computer simulation of Figure 5 (ventilatory depression and sedation models disabled). Ventilatory depression (expressed as a fraction of baseline ventilation) was expressed over time of simulation (in minutes). Figure 8 is a graph showing the time course of quantity of opioid in the inhalation TR/f device and in the lung of the patient, in the Stella computer simulation of Figure 5 (ventilatory depression model enabled, sedation model disabled). The X axis shows time hi minutes. The Y axis shows dose units of formulation, in mg. Patient ventil the dropped to approximately 25% of baseline ventilation, such depression persisting approximately 5-10 minutes. Figure 9 is a graph showing the time course of. ventilatory depression ir TM tion for the Stella computer simulation of Figure 5 (ventilatory depression model enabled, sedation model disabled). Ventilatory depression (expressed as a fraction of baseline ventilation) was expressed over time of simulation (in minutes). Change in ventilation caused by the self-limitation of opioid uptake offers considerable safety to the patient (compared to figure 7). Figure 10 is a graph showing the time course of quantity of opioid in the inhalation device and in the lung of the patient, in the Stella™ computer simulaticin of Figure 5 (ventilatory depression and sedation models enabled). The X axis shows time in minutes. The Y axis shows dose units of formulation, hi mg. Drug inhalation stopped completely at approximately 8 minutes, due to a sedation state being reached and selflimitation of drug intake. Figure 11 is a graph showing the time course of ventilatory depression in the TTkif Stella computer simulation of Figure 5 (ventilatory depression and sedation models enabled). Ventilatory depression (expressed as a fraction of baseline ventilation was expressed over time of simulation (in minutes). Change in ventilation caused by thelselflimitation of opioid uptake from sedation offers considerable safety to the patient (compared to figure 7 or 9) Figure 12 is a flow diagram which represents a computer simulation model for the administration of two opioids. Figure 13 is a flow diagram which represents the Stella™ computer simulation of the pharmacokhietics of the administration of two opioids. Figure 14 is a graph showing the output of Stella™ computer simulation of Figure 13 expressed as a time course of total quantity of opioid in the inhalation device ana in the lung of the patient (ventilatory depression and sedation models enabled). Y axis shows fentanyl equivalents of formulation in the inhaler (I), of the rapid-onset opioid in the lung (2), and the sustained-effect opioid hi the lung (3), expressed in ng/ml (fentanyl equivalents) of drug over time (hi ininutes). After approximately 12 minutes, the patil stopped inhaling more opioid, reflecting opioid-induced sedation. Figure 15 is a graph showingjhe time course of concentration of each opioid akid of total opioid at the effect site in the Stella™ computer simulation of Figure 13 (ventilatory depression and sedation models enabled). Amount of rapid-onset opioid ( ), sustained-effect opioid (2) arid the combination effect of both the rapid-onset opioid and the sustained-effect opioid (3) at the site of effect were shown, in ng/ml of fentanyl equivalents, over time (hi minutes). Figure 16 is a graph showing the time course of ventilatory depression during and "TTUf • ' ' after delivery of opioids in the Stella computer simulation of Figure 13 (ventilatory depression'and sedation models enabled). Ventilatory depression (expressed as a fractioii of baseline ventilation) was expressed over time of simulation (in minutes). The combination g£ the two opioids reaches a peak during the administration of the opioid. first mof eing anil, Figure 17 is a flow diagram which represents the Stella computer simiilati the pharmacokinetics of the administration of two opioids, where the two opioids administered are alfentanil and morphine. Figure 18 is a graph showing tile time course of conbentration of alfen morphine^ arid combined opioid at the effect site in the Stella"*1 computer simulation of Figure 11 (ventiiatory depression and sedation models enabled). Line 1 shows concentration of alfentanil; line 2 shows concentration of morphine, and line 3 shows combined concentration. All drug levels are shown at the site of effect, and expressed in ng/ml of fentanyl equivalents over time (in minutes). Drug administration was terminated after delivery of 90% of the drug because of patient sedation. As seen in line 3 the highest opioid exposure occurs during inhalation. s Figure 19 is a graph showing the time course of ventiiatory depression was Stella computer simulation of Figure 17 (ventiiatory depression and sedation m enabled). Ventiiatory depression (expressed as a fraction of baseline ventilation expressed over time of simulation (in minutes). Ventilation decreases to about 65% of baseline during drug administration. the dels sma oid. Figure 20 is a graph showing maximum concentration of opioid in the pi; against end of dose concentration of opioid in the plasma of patients administered op Figure 20A shows patients administered with a combination of fentanyl and liposori rally encapsulated fentanyl through a pulmonary route. Figure 20B shows patients administered with fentanyl intravenously. Maximum concentration of opioid was not significantly higher than the concentration at end of dose, indicating that if the "er d of dose" amount is non-toxic, the maximum concentration of opioid taken by the subj ct is likely also non-toxic. Figure 21 is a graph showing time to side/toxic effect versus time to end of dose for the side effects and toxic effects of patients administered a combination of fentanyl and liposomally encapsulated fentanyl through a pulmonary route. In all cases, tinje to toxicity was equal to or longer than time to side effect. Figure 22 is a table showing the statistical correlation of side effect to toxic effect Side effect is correlated to toxic effect at ap, 04. Examples . The examples below are designed to demonstrate but not limit the embodiments of the present invention. Example 1: Theoretical Model for Opioid Delivery Examples 2-4 are based on a theoretical model for opioid delivery; this theoretical model is described for greater certainty here in Example 1. The theoretical model for opioid delivery was programmed into the computer simulation package "Stella" (High Performance Systems, Lebanon, NH). The elements shown in this example, both in figures and in text, are adapted from the Stella model representation, and explain both the programming of the simulation, sand how the simulation works. In the figures, rectangles represent variables that indicate accumulation of a substance (with exceptions noted below). Open arrows represent flow into or out of me accumulators, and closed arrows represent the elements that control the flow. Some closed arrows are omitted for simplicity of representation. Ovals represent model parameters (inputs) and time-independent calculations. Many model parameters and constants were obtained from the prior art (see Scott JC, Stanski DR Decreased fentanyl and alfentanil dose requirements with age. A simultaneous phaimacokinetic and pharmacodynamic evaluation. J Pharmacol Exp Ther. 1987 Jan;240(l): 159-66). (a) Sedation Model A model for opioid induced sedation was designed (Fig. 1 — Sedation Mode Opioid in Effect Site 1010 was used as a variable denoting the concentration of opioid kt the site of drug effect. If more than 1 opioid was present at the site of drug effect, Opioid in Effect Site 1010 was built to represent the sum of the opioids present, each normalizejd to their relative potency (for example, in Examples 3 and 4, below). Sedation Threshold 1020 was defined as the Opioid Concentration 1010 thdt would render the patient unable to use the inhaler. Sedation Threshold 1020 wafi determined either through experimentation or through the known pharmacokinetics if the opioid. Sedation Eyaluator 1030 was a test of whether Opioid Concentration 1010 exceeded Sedation Threshold 1020. If Opioid Concentration 1010 exceeded Sedation Threshold 1020, Sedation Evaluator converted the value of Sedation State 1040 from 0 to 1. Sedation State 1040 was an exception to the rule that rectangles represent accumulation of a substance: Instead, the role of Sedation State 1040 within the rJiodel was that of a memory component, which would remember that the opioid had exceeded the sedation threshold. In subsequent models, data from Sedation State 1040 functioned to turn off further administration of opioids, simulating patient sedation and the resulting removal of the inhaler from the mouth. (b) Ventilatory Depression Model A Ventilatory Depression simulation was programmed (Fig. 2). In this model, C02 was produced by the metabolic activity of the body at a rate CO2 Production ,010, flowed into the plasma, (Plasma CO2 2020). CO2 Production 2010 was either determined experimentally, or known from prior art (see, for example, Bouillon T, Schmidt Garstka G, Heimbach D, Staffbrst D, Schwilden H, Hoeft A. Pharmacokir eticpharmacodynamic modeling of the respiratory depressant effect of alfenranil. Anesthesiology. 1999 Jul;91(l):144-55 and Bouillon T, Bruhn J, Radu-RadulesJu L, Andresen C, Cohane C, Shafer SL. A model of the ventilatory depressant potency of remifentanil in the non-steady state. Anesthesiology. 2003 Oct;99(4):779-87.). PlLma CO2 2020 equilibrated with the CO2 hi the brain (Brain CO2 2040) at a rate (Bkin- Plasma CO2 Equilibrium 2030). C02 was eliminated from the plasma in a manner simulating the exhalation of air from the lungs, at a rate CO2 Elimination 2050 thatj was mediated by the parameter Ventilatory Depression 2060. Ventilatory Depression 2060 increased as the opioid concentration at the si:e of drug effect (Opioid in Effect Site 1010) increased. Ventilatory Depression decreased the elimination of C02 from the lungs (C02 Elimination 2050), causing CO2 to rise im the brain, (Brain CO2 2040). As Brain C02 2040 increased, it stimulated ventilation through a negative effect on Ventilatory Depression 2060, offsetting in part the depressant effects of Opioid in Effect Site 1010, which has a positive effect on Ventilatory Depression 2J060. Other parameters were designed to effect Ventilatory Depression 2060; the suna of these parameters were illustrated in this model as Model Parameters 2070; parameters comprising Model Parameters 2070 were described in greater detail in Figure 5. These Model Parameters 2070 effect Ventilatory Depression 2060, which in turn effects GO2 Elimination 2050 and Brain C02 2040. Although the programming of this simulation into Stella is novel, the Ventilatory Depression Model is known in the art, and is referred to as an "Indirect Response Modejl." (c) Device Model A model for the inhalation device is shown in Figure 3. Dose 3050 represents ithe total amount of opioid added to the Inhaler. Opioid Dose 3050 is added to the inhaler at a rate Fill Inhaler 3010. This rate is required for the working of the simulation, but is calculated at an instantaneous rate. Formulation In Inhaler 3020 represents the opioid contained within the inhaler. The patient inhales the formulation at a rate of inhalation (Inhalation 3030) into the lungs, (Formulation in Lungs 3040). Inhalation 3030 is effected by Ventilatory Depression 2060 and Sedation State 1040. Specifically, Inhalation 3030 is slowed by the increase of Ventilatory Depression 2060. For example, if Ventilatory Depression 2060 was 50% of baseline, then drug was inhaled at half the baseline rate (Inhalation 3030 was half baseline). However, if Sedation State 1040 - 1, then inhalatiqn of drug into the lungs ends, and no further drug is inhaled. (d) PharmacoJdnetic Model A Pharmacokinetic Model for systemic opioid was programmed. Formulation fli Lungs 3040 was absorbed systemically at a rate Systemic Absorption 4010 into the blood plasma (Opioid in Plasma 4020). Opioid in Plasma 4020 equilibrated at a rate Plasma- Effect Site Drug Equilibrium 4030 with opioid at the site of drug effect (Opioid in Effecjt Site 1010). Opioid also redistributed into tissue Opioid in Tissue 4060 at a rate Opioi Redistribution 4050 or was eliminated from the plasma at a rate Opioid Eliminatio: 4070. Opioid in Tissue 4060 and Opioid Redistribution 4050 were programmed optional parameters that could be used or not used depending on the pharmacokinetii model of the particular opioid utilized. The rates Systemic Absorption 4010, Plasma-) Effect Site Drug Equilibrium 4030, Opioid Elimination 4070, and Opioid Redistribution^ 4050 were all determined by a vector of pharmacoldnetic parameters of the particulaii opioid being administered, represented in the model as Opioid Pharmacokijietic Parameters 4080, and calculated by pharmacokinetic modeling. Although ttie programming of this simulation into Stella was novel, the Pharmacokinetic Model is known in the aft, and is referred to as a "Mammillary Pharmacokinetic Model With An Effect Site." Mammillary models as represented a'sove typically have 0, 1 or 2 tissue compartments, yielding models referred to as 1, 2, or 3 Compartment Models with an effect site, respectively. Example 2: Administration of a Single Opioid This example is an application of Example 1: Theoretical Model for Opioid Delivery. This example is meant to illustrate the Theoretical Model for Opioid Delivery in use; the model parameters do not reflect any specific opioid. Instead, the model parameters in this example have been designed to clearly demonstrate the self-limiting aspect of the proposed system of opioid delivery. This Example shows the integration of the four simulations as described in Example 1, and output from the model when the simulation is run. (a) Integration of the Model Figure 5 shows the elements of the model as described in Example 1, whersin a single opioid is administered through inhalation. Figure 5 encompasses: a Device Model 5010 that is equivalent to the Device Model shown and explained in Example 1 as the whole of Figure 3; a Pharmacokinetic Model 5020 that is equivalent to the Pharniacokinetic Model shown and explained in Example 1 as the whole of Figure 4 (with the exception of the exclusion of optional parameters Opioid in Tissue 4060 and Opioid Redistribution 4050, and with the further exception that Opioid Pharmacokinetic Parameters 4080 were built into Systemic Absorbtion 4010, Opioid Elimination 4070, and Plasma; Effect Site Equilibration 4030, and not shown as a separate parameter - see source code for more information); a Ventilatory Depression Model 5030, which was equivalent to the Ventilatory Depression Model-shown and explained in Example 1 is the whole of Figure 2 (with the exception that Model Parameters 2070 are shovra hi 'expanded' form, with various elements comprising Model Parameters 2070, namely PAC02@0 2071, KelCO2 2072, keOC02 2073, C50 2074, Gamma 2075, and F £076, shown; and a Sedation Model 5040, that is equivalent to the Sedation Model shoW in Figure 1. The mechanics of these four models were described in depth in Example 1, with the exception of the expansion of Model Parameters 2070, the mechanics of which are explained as follows: Baseline CO2 2071 is the CO2 at baseline, prior to administration of opioid. kel CO2 2072 is the elimination rate relating Plasma C02 2020 to CO2 Elimination 2050, so that at baseline (i.e., in the absence of ventilatory depression): CO2 Elimination 2050 - kel C02 2072 x Plasma CO2 2020 It follows that at baseline, carbon dioxide in the body is at steady state, and hence the CO2 Elimination 2050 = CO2 Production 2010. This permits calculation of the rate of C02 production (which is constant) in terms of Baseline C02 2071 and kel CO2 2072 as: CO2 Production 2010 = kel CO2 2072 x Baseline Plasma CO2 2071. § The rate of Brain Plasma Equilibration 2020 is determined,by the parameter keO CO2 2073, so that: Brain Plasma Equilibration 2020 = keO CO2 2073 x (Plasma CO2 2020 - Brain CO2 2040) Opioids depress ventilation as a sigmoidal function of the Opioid in the Effect Site, 1030, and the parameters C50 2074, the opioid concentration associated with 50% of maximum effect, and gamma 2075, the steepness of the concentration vs. response relationship, with the contribution of the opioid to ventilatory depression expressed as: Opioid in the effect site loso63™*2075 C50 20748amtra2075 + Opioid in the effect site 1030samma 2075 Conversely, carbon dioxide stimulations ventilation. The increase in ventilation can be modeled as a function of Baseline C02 2071, Brain CO2 2040, and F 2076, a parameter describing the steepness of the relationship: ( Brain C02 2040 Y VBaselineCO2207lJ Putting these together, Ventilatory Depression 2060 can be described as: Ventilatory depression 2060;= Opioid in the effect site lOSO* "*20-, "| / Brain CQ2 20 C50 2074BOTma2075 + Opioid in the effect site loSO*"™1*2075 J ^Baseline C02 With ventilatory depression 2060 now defined, we can fully define CO2 Elimination 2050 in the presence of opioid induced ventilatory depression as: CO2 Elimination 2050 = kel CO2 2072xPlasma C02 2020x Ventilatory Depression 2060 completing the description of tne model. In this manner, the models from Example 1 were combined into one model of opioid effect. This model, shown in Figure 5, can also be described by the following mathematical model, as represented in the Stella programming language (source code): Brain_CO2_2040(t) = Brain_CO2_2040(t- dt) + (Btain_Plasma_CO2_EquilibrationJ2020) N dt rNITBrain_CO2_2040 = Baseline_CO2_2071 « INFLOWS: Brain_Plasma_CO2_Equilibration_2020 = keO_CO2_2073*(Plasma_CO2_2020- Brain_CO2_2040) Formulation_in_Inhaler_3020(t) = Fommlation_in_lnlialer_3020(t - dt) 4- (Fill_lnlialer_3010 Inhalation_3030) * dt INTT Formulatioa_m_Inhaler_3020 = 0 INFLOWS Fill_Inhaler_3010 = if time = 0 then Dose_3050/DT else 0 OUTFLOWS: Inhalation_3030 = If Sedation_State_1040 = 0 then .5*(Ventilatory_Depression_2060) else 0 Fonnulatiori_in_Lung_3040(t) = Formulation_in_Lung_3040(t - dt) + (Inhalation_3030 - Systemic_Absorption_4010) * dt INTT Fonrnilation_in_Lung_3040 = 0 INFLOWS: Inhalation_3030 = If Sedation_State_1040 = 0 then .5*(Ventilatory_Depression_2060) else 0 OUTFLOWS: Systemic_Absorption_4010 = FormuIation_in_Lung_3040*.693/l Opioid_in_Effect_Site_1010(t) = Opioid_in_Effect_Site_1010(t - dt) -f (Plasma_Effect_Sfte_Equilibration_4030) * dt INIT Opioid_in_Effect_Site_1010 = 0 INFLOWS: Plasma_Effect_Site_EquiUbration_4030 = (Opioid_in_Plasma_4020- Opioid_in_Effect_Site_1010)*.693/l Opioid_in_Plasma_4020(t) = Opioid_inJPlasma_4020(t - dt) + (Systemic_Absorption_4010 - Opioid_Elimination_4070 - Plasma_Effect__Site_Equilibration_4030) * dt INIT Opioid_in_PIasma_4020 = 0 INFLOWS: Systemic_Absoiption_4010 = Fonnulation_in_Lung_3040*.693/l OUTFLOWS: Opioid_Elimination_4070 = Opioid_in_Plasma_4020*.693/10 Plasma_Eflfect_Site_EquiIibration_4030 = (Opioid_ia_Plasma_4020- Opioid_in_E£fect_Site_1010)*.693/l Plasma_CO2_2020(t) =Plasma_CO2_2020(t - dt) + (CO2_Productioii_2010 - Brain_Plasma_CO2_Eqailibration_2020 - CO2_ElimiDation._2050) * dt INTT Plasma_CO2_2020 = Baseline_CO2_2071 INFLOWS: CO2_Production._2010= Baseline_CO2_2071*kelCO2_2072 OUTFLOWS: Brain_Plasma_C02_Equilibration_2020 = keO_CO2_2073 *(Plasma_CO2_2020- Brain_CO2_2040) GO2_Elimmation_2050=Plasim_CO2_2020*kelC02_2072*VentOatoryJDepression_20^^ Sedation_State_1040(t) = Sedati6n_State_1040(t - dt) + (Sedation_Evaluator_1030) * dt INIT Sedation_State_l 040 = 0 INFLOWS: Sedation_Evaluator_1030 = if(Opioid_in_Effect_Site_1010>Sedation_Threshhold_1020) then 1 elseO . Baseline_CO2_2071 =40 C50 2074 = 3 Dose_3050 = 5 F_2076 = 4 Gamna_2075 = 1.2 keO_CO2_2073 = 0.92 kelC02_2072 = 0.082 Sedation_Threshhold_1020 = 1.5 Ventilatory_Depression_2060 =.(!• Opioid_m_Effect_Site_1010AGani^ 10AGaimna_2075))*(Brain_CO2_2040/Baseline_CO2_2071)AF_2076 (b) Output of the model when run with Ventilatory Depression Model and Section Model Disabled The model designed and described in (a) was run as a simulation of opioid effect, using the following initial parameters: Formulation In Inhaler 3020 = 5 milliliters ai: time = 0. The model was allowed to run over a time course of two hours. For this simulation, the feedback loop on drug uptake aspects of the Ventilatory Depression Model (i. s. the feedback of the effect of Ventilatory Depression 2060 on Device Model 5010), aid the Sedation Model were disabled. Output of the model, when run, was plotted for various parameters in Figures 6 and 7. Figure 6 shows the output of the model as run in the absence of patient selflimiting inhalation of opioid (i.e. with the Ventilatory Depression Model and the Sedation Model disabled). Figure 6 shows the time course of drug in the inhaler (Formulation In Inhaler 3020 - line 1), and in the lungs (Formulation in Lungs 3040 - line 2) i absence of the self-limiting aspects of the invention. The amount of drug in the inhaler dropped steadily over the first 10 minutes of simulation, at a rate Inhalation 3030. n the The amount of drug in the lungs reflected the net processes of inhalation of drug into the lungs, and absorption of drug from the lungs into the systemic circulation. Figure 7 shows Ventilatory Depression 2060 over time, for the same simulation* (Ventilatory Depression Model and Sedation Model disabled). The graph output indicated that patient's ventilation dropped to approximately 25% of baseline ventilation in this simulation. The Ventilatory depression persisted for approximately 5-10 mijiutes. The drop in ventilation was reversed as carbon dioxide built up in the patient's pl&sma, and, at the same rate, the patient's lungs (not simulated), counteracting the depressant effect of the opioid on ventilation. This drop in ventilation exposed the patient to risk from injury from hypoxia. . • • • (c) Output of the model when run with Ventilatoiy Depression Model enabled The simulation used in (b) was modified by enabling the Ventilatoiy Depression Model, and run again with the same initial parameters of Formulation In Inhaler 3020 F= 5 milliliters at time 0. Output of various parameters were plotted over time. Figure 8 shows Formulation hi Inhaler 3020 (line 1), depicting the amount of drug that is left in the inhaler, and Formulation la Lungs 3040 (line 2), depicting the amount of drug in the lungs, in the presence of ventilatory depression, one of the two self-limiting aspects of I invention (the other being sedation). As compared to Example 2(b), as expected, it took longer to inhale the drug when the simulation was run with the Ventilatory Depression Model enabled - inhalation of drug in Figure 8 took place over approximately 17 minutes as opposed to the 10 minutes in Figure 6. This was due to a reduction in ventilation caused by ventilatory depression, which limited the patient's exposure to the opioid. This reduction in ventilation was best illustrated in Figure 9, which plotted Ventilatqry Depression 2060 over time for the same simulation. Ventilatory Depression 2060 was depressed by 50% in Figure 9. When compared with the simulation shown in Figure 7, the patient was breathing half as much (in Figure 9) as when simulation was run with the Ventilatory Depression Model deactivated (in Figure 7). This simulation shows that me change in ventilation caused by the self-limitation of opioid uptake offers considerable safety to the patient. (d) Output of the model when run with Ventilatory Depression Model and Sedation Model enabled The same simulation (Formulation hi Inhaler 3020 = 5 milliliters at time = 0) was run, this time with both the Ventilatory Depression Model 5030 and the Sedation Model 5040 enabled. Output of various parameters were plotted, over time. Figure 10 shows the time course of Formulation hi Inhaler 3020 (Line 1) and Formulation In Lungs 3040 (Line 2) in the presence of ventilatory depression and sedation. As seen in the figure^ after 8 minutes drag inhalation stopped completely. The reason was that the patient has become sedated, and could no longer hold the inhaler to the mouth (simulated here a£ Sedation State 1040 turning from 0 to 1). At this time, approximately 2 milliliter£ remained in Formulation In Inhaler 3020, and therefore, approximately 40% of the opioid dose remained in the inhaler and was not inhaled. Figure 11 plots Ventilatory Depression 2060 during the time course of this simulation. The maximum depression of ventilation in Figure 11 was approximately 60%. When compared with Figure 9, the improved; safety from the opioid-induced sedation is evident. Thus, Example 2, as illustrated in Figures 5 through 11, demonstrate* through simulation the effects and advantages of the self-limiting system of opioid delivery, as described herein. Example 3: Administration of two opioids In this simulation, the model parameters do not reflect any specific opioids, but have been adjusted to demonstrate clearly the self-limiting aspect of the proposed system of opioid delivery. The simulation models and measures the same variables, this time for an opioid composition comprising of two different opioids with different pharmacoldnetics. (a) Building a two opioid model. Figure 12 addresses how two opioids are combined into a single opioid concentration for the model. In the two opioid simulation, Rapid Opioid In Effect Site 12010 represents the concentration of rapid onset opioid; Slow Opioid In Effect Site 12020 represents the slow onset opioid. Each of these is determined in parallel and in the same manner as in the one opioid model (Example 2). However, each is determined separately, then combined to determine Combined Opioid Effect Site Concentration 12030. Combined Opioid Effect Site Concentration 12030 is calculated using the known relative potency of each opioid, Relative Potency 12040. Combined Opioid Effec Site Concentration 12030 is equal to, and depicted as, Opioid in effect Site 1010 in th two opioid models illustrated in Figures 13 and 17. Figure 13 illustrates the algorithm for the two opioid model simulation. It encompasses: a Device Model 13010, equivalent to and illustrated as Device Model 5010 and as described in Examples 1 and 2; a Pharmacokinetic Model 13020 comprising a combination of two instances of the Pharmacokinetic Model 5020 (one for the irapid opioid, and one for the slow opioid), each as illustrated in Figure 4 and Figure 5, aijid as described in Examples 1 and 2, and each running in parallel, then combined using the Two Drug Model 13050, as described in Figure 12; a Ventilatory Depression Model 5030, as illustrated in Figure 2 and Figure 5, and as described in Examples 1 and 2; and a Sedation Model, 5040, as illustrated in Figure 2, Figure 5, and as described in Examples 1 and 2. The model shown in figure 13 can also be ^described by the following mathematical model, as represented in the Stella programming language (source code)| Brain_CO2_2040(t) = Brain_C02_2040(t - dt) + (Brain_Plasma_CO2_Equilibration_2020) * - INIT Brain_C02_2040 = BaseIine_C02_2071 INFLOWS: Brain_Plasma_CO2_Equih'bration_2020 = keO_C02_2073*(Plasma_CO2_2020- Brain_CO2_2040) Forraulation_in_Inhaler_3020(t) = Formdation_in_Inhaler_3020(t- dt) + (Fill_In|aler_3010 - Iahalation_I_3031 - Inhalation_2_3032) * dt INIT FonnuMon_inJnhaler_3020 = 0 INFLOWS: Fill_Inhaler_3010 = if time = 0 then Dose_3050/DT else 0 OUTFLOWS: inhalation_l_3031 = if Sedation_State_1040 = 0 then 0.25*VentilatoryvDepression else 0 Inhalation_2_3032 = if Sedation_State_1040 = 0 then 0.25*Ventilatory_Depression_2060 else Opioid_inJ3ffect_Site_1010(t) = Opioid_in_Effect_Site_1010(t - dt) INIT Opioid_in_Effect_Site_1010 = 0 Plasma_CO2_2020(t) = Plasma_CO2_2020(t - dt) + (CO2_Production_2010 - Brain_Plasma_CO2_EquM>ration_2020 - CO2_Elimination_2050) * dt INIT Plasma_CO2_2020 = Baseline_CO2_2071 INFLOWS: C02_Production_2010 = { Place right hand side of equation here... } OUTFLOWS: Brain_Plasma_CO2_Equilibration_2020 = keO_CO2_2073*(Plasma_CO2_2020- Braui_CO2_2040) C02_Eh'mination_2050 = Plasma_CO2_2020*kelC02_2072*Ventilatory_Depression_2060 RapidJDrag_Effect_Site(t) == Rapid_Drag_Effect_Site(t - dt)+ (Rapid_Drug_Plasma_Effect_Site_Equilibration) * dt JNTT RapidJDmg_Effect__Site = 0 INFLOWS: Rapid_Drug_Plasma_Effect_Site_Equilibration = (Rapid_Drug_In_Plasma- RapidJDrugr_Effect_Site)*.693/l Rapid_Drug_In_Plasma(t) = Rapid_Drug_In_Plasma(t - dt) 4- (Rapid_Drug_Absorption - Rapid_Drug_Clearance - Rapid_Drug_Piasma_Effect_Site_EquiKbration) * dt INIT Rapid_DragJn_Plasma = 0 INFLOWS: Rapid_Drug_Absorption = Rapid_Fonnulation_in_Lung.693/l Rapid_Drag_Concentratio OUTFLOWS: Rapid_Drug_Clearance = Rapid_Drug_In_Plasma.693/10 ® Rapid_Drug_Plasma_Effect_Site_Equilibration=(Rapid_Drag_In_Plasma- Rapid_Drug_Effect_Site)*.693/l Rapid_Fornmlation_i[i_Limg(t) = Rapid_Formulation_in_Lang(t - dt) + (Inhalation_l_3031 Rapid_Drug_Absorption) * dt INTT Rapid_Formulation_in_Lung = 0 INFLOWS: inhalation_l_3031 = if Sedation_State_1040 = 0 then 0.25*Ventilatory_Depression else 0 OUTFLOWS: Rapid_Drug_Absorption = Rapid_Fonriulation_iaJLung*.693/l *Rapid_Drug_Concentrati Sedation_State_1040(t) = Sedation_State_1040(t- dt) + (Sedation_Evaluator_1030) * dt INIT Sedation_State_1040 = 0 INFLOWS: Sedation_Evaluator_1030 = i^Opioid_in_Effect_Site_1010>Sedation_Threshhold_1020) elseO Slow_Drug_Effect_Site(t) = Slow_Drug_Effect_Site(t - dt) + (Slow_Drug_Plasma_EfFect_Site_Equilibration) dt INIT Slow_Drag_Effect_Site = 0 INFLOWS: hen 1 Slow_Drug_Plasma_Effect_Site_EquiIibration = (Slbw_Drug_In_Plasma- SlowJDrug_Effect_Site)*.693/10 Slow_Drug_Ih_Plasma(t) = Slow_Drag_In_Plasma(t - dt) + (Slow_Drag_Absorption - Slow_Drug_Clearance - SIow_Drug_Plasma_Effect_Site_Equilibration) * dt INIT Slow_Drug_In_Plasma = 0 INFLOWS: Slow_Dnig_Absorption = Slow_Fonnulation_In_Lung*.693/12*Slow_Drag_Concentration OUTFLOWS: Slow_Drug_aearance = Slow_Drug_In_Plasma*.693/300 SIow_Drag_Plasma_Effect_Site_Equilibration = (Slow_Drug_In_Plasma- Slow_Drug_Efifect_Site).693/10 Slow_Formuktion_Ih_Lung(t) = Siow_Formulation_Li_Lung(t - dt) + (Jnhalation_2_3032 - Slow_Drag_Absorption) dt s INTT Slow_Fonnulation_In_Lung = 0 INFLOWS: Inhalation_2_3032 = if Sedation_State_1040 = 0 then 0.25*Ventilatory_Depression_2060 else (|) OUTFLOWS: Slow_Drag_Absorption = Slow_Formulation_In_Luiig*.693/12*Slow_Drug_Coiiceiitration Baseline_CO2_2071 =40 C50_2074 = .3 Dose_3050 = 5 F_2076 = 4 Gamma_2075 = 1.2 keO_CO2_2073 = 0.92 kelCO2_2072 = 0.082 Opioid_in_EfFect_Site_1010 = Rapid_Drug_Effect_Site+Slow_Drug_Effect_Site Rapid_Drug_Concentration = 1 Sedation_Threshhold_1020=1.5 Slow_Drug_Concentration =1 Ventilatory_Depression__206Q = (1- OpioidJn_Effect_Site_1010^aimna_2075/(C50^ ma_2075))*(Brain_C02_2040/BaseIine_CO2_2071)AFJ076 . (b) Output of mo"del when run with. Ventilatory Depression Model and Sedation model enabled The same simulation (Formulation In Inhaler 3020 = 5 milliliters at time = 0| was run in the two opioid model as illustrated in Example 3(a) and Figure 13. Figure 14 shows the time course of Formulation In Inhaler 3020 (Line 1), Formulation In ping (Rapid Opioid) 3040 (Line 2), and Formulation In Lung (Slow Opioid) 3040 (Line 3), in the presence of ventilatory depression and sedation. The simulation showed that, over 12 minutes of run, the drug was inhaled by the patient. The rate of fall in the amount of] drug in the inhaler was not perfectly linear, reflecting the slowed breathing with opioidinduced ventilatory depression. After approximately 12 minutes, the patient stopped inhaling more opioid, reflecting opioid-induced sedation. The rapidly acting opioia was quickly taken up into the systemic circulation, which limited how much accumulated in the lung, and produced a quick drop in concentration in the lung when the patient sto sped inhaling more opioid. The slowly acting opioid was taken up slowly by the lung, uhich permitted more drag to accumulate in the lung during inhalation^ and the administration of opioid into the systemic circulation for over two hours following the end of opioid delivery to the patient. Figure 15 shows different variables for the same simulation. In Figure 15, line 1 indicates the rapidly acting opioid concentration in the effect site (Rapid Drug Effect Site), over time, and demonstrates the rapid rise owing to quick absorption and rapid plasma-effect site equilibration, and a rapid drop owing to rapid metabolism. Line 2 is the slowly acting opioid concentration in the effect site (Slow Drag Effect Site), over time, and demonstrated a slow rise in concentration owing to slow absorption and slow plasmaeffect site equilibration, and a slow decrease over time owing to slow metabolism. Lrie 3 shows the combined concentration of rapid and slow onset drug (Combined Opioid EJfect Site Concentration) As can be seen, the combination reaches a peak during the administration of the first opioid. Figure 15 and Figure 14 show different variables for the same simulation run[ on the same X axis (time). One can therefore refer back to Figure 14 to see that the patient stopped self-administering drug at approximately 12 minutes. When Figure 14 is interposed with Figure 15, we can see that this reflected the patients' response to the rapidly acting opioid, as the concentration of the slowly acting opioid was negligible at 12 minutes. However•, the overall opioid concentration remained fairly steady over time. This reflected the slowly acting opioid gradually replacing the rapid acting opioip in Opioid In The Effect Site as the rapidly acting opioid was eliminated from the system through Rapid Drug Clearance. Figure 16 shows the time course of Ventilatory Depression 2060 during and after delivery of opioids with the two-opioid delivery system, in the same simulation Figure 16 illustrates an initial decrease in ventilation to approximately 60% of baseline. As mentioned previously (in the description for Figure 11), this is well tolerated] by patients. As the CO2 builds up, ventilation was stimulated. Note that there was very little decrease in ventilation after this initial drop. The reason is that there is now adequate CO2 accumulation in the patient to continue driving ventilation. As demonstrated by figures 13, 14, and 15, in the two drug embodiment of me device, the first drug acts as a 'probe' of the patient's sensitivity to opioids, and limits the dose of both the first and the second opioid. In this manner, the patient can receive an slowly acting opioid without receiving an excessive dose. A combination of two opioids, one of them fast acting, can therefore be used to increase the safety profile of either opioid alone, or, more particularly, of the slow acting opioid. Example 4: Alfentanil and Morphine as examples of opioids in the two drug model This example shows an application of Example 3 to two specific drugs, name alfentanil and morphine, wherein alfentanil is the rapidly acting opioid and morphine is the slowly acting opioid. Figure 17 encompasses: a Device Model 17010, comprising 2 Device Modbl 5010's, as described in Figure 5 and explained in Example 1, and each running in parallel, but each modified and re-labeled for the specific known parameters of the opioids alfentanil and morphine; a Ventilatory Depression Model 5030, as described in Figure 2, a Sedation Model 5040, as described in Figure 1, and a Two Drug Model, 17050, ajs described in figure 12 but re-labeled to reflect the specific drags alfentanil and morphine^ Figure 17 exposes all of the parameters 2070 of the ventilatory depression model, 1703d39 The; parameters 4080 of the pharmacokinetic models for morphine and alfentanil, 17020, are now fully exposed: Alfentanil and morphine are each represented by a 3 compartment mammillary model with an effect site. . The model shown in figure 17 can also be described by the following mathematical model, as represented in the Stella programming language. The constants for alfentanil arid morphine are based on existing literature for these drugs. Alfentanttjnjnhaler(t) = Alfentana_in_Inhaler(t - dt) + (- InhaIe_Alfentanil) * dt INIT Alfentanil_in_Inhaler = Alferitanil_Dose _ ug OUTFLOWS: Innale_Alfentanil = If Sedation_State = 0 then Alfentanfl_Dose__ug/Dose_Duratioii*Ventilatory_Depressioii else 0 Alfentanfljn_Lung(t) = Alfentanil_in_Lung(t - dt) + (Inhale_Alfentanil - AlfentanilJJptake INTT Alfentanil_in_Lung = 0 INFLOWS: Inhale_Alfentanil = If Sedation_State = 0 then Alfentanil Dose ug/Dose Duration Ventilatory Depression else 0 OUTFLOWS: Alfentanil_Uptake = Alfentanil_in_Limg.693/Alfentaiul_Absorption_Half_Life Alfentanil_Xl(t) = AlfentaniI_Xl(t - dt) + (Alfentanil_CI2 + Alfentanil_CB Alfentanil_Uptake - AlfentaniI_Cll) dt INIT Alfentanil_Xl = 0 INFLOWS: Alfentanil_C12 = Alfentanil_X2 Alfentanil_K21-Alfentanil_Xl*AIfentanil_K12 Alfentanil_CB = Alfentanil_X3*Alfentanil_K3 l-Alfentanil_Xl*Alfentanil_K13 AlfentaniI_CLe = Alfentanil_Xeffect*Alfentanil_KeOAlfentanil_ Xl *Alfentanil_KeO*.001/Alfentanil_Vl Alfentanil_Uptake = Alfentanu_inJLung*.693/Alfentanil_Absorption_Half_Life OUTFLOWS: Alfentanil_Cll = Alfentanil_Xl*Alfentanil_K10 Alfentanil_X2(t) = Alfentanil__X2(t - dt) + (- Alfentanil_C12) * dt INIT Alfentanil X2=0 dt OUTFLOWS: AlfentaniI_C12 = AIfentanil_X2*Alfentanil_K21-Alfentanil_Xl*Alfentanil_K12 Alfentanil_X3(t) =. Alfentanil_X3(t - dt) + (- Alfentanil_C13) * dt INIT Alfentanil_X3 = 0 OUTFLOWS: Alfentarul_C13 = AJfentanil_X3*AlfentanU_K31-Alfentanil_Xl*Alfentanil_K13 Alfentanil_Xeffect(t) = Alfentanil_Xeffect(t - dt) + (- Alfentanil_CLe) * dt INIT Alfentanil_Xeflfect = 0 OUTFLOWS: Alfentanil_CLe = Alfentanil_Xeffect*Alfentaml_KeOAIfentanil_ Xl*Alfentaml_KeO*.001/Alfentanil_Vl Morphine_in_Inhaler(t) = Moiprtine_in_Inhaler(t - dt) + (- Inhale_Morphine) * dt ® INTT Morphine_in_lnhaler = Moiphine_Dose__mg* 1000 OUTFLOWS: Inhale_Morphine = If sedation_state = 0 then Morphine_Dose__rng*1000/Dose_Duration*Ventilatory_Depressionelse 0 Morphine_in_Lung(t) = Morphine_in_Lung(t - dt) + (Innale_Morphine - Morphine_Uptake) * cit INTT Morphine_in_Lung = 0 INFLOWS: Inhale_Morphiae = If sedation_state = 0 then Morphine_Dose__mg*1000/Dose_Duration*VentiIatory_Depression else 0 OUTFLOWS: Morphine_Uptake = Morphine_in_Liing*.693/Morphine_Absorption_Half_Life Morphine_Xl(t) = Morphine_Xl(t - dt) + (Morphine_C12 + Morphine_C13 + Morphine_CLe + Morphine_Uptake - Morphine_Cll) * dt INIT Morphine_Xl = 0 INFLOWS: Morphrne_C12 = Morphine_X2*Moiphine_K21-Morphine_Xl*Morphine_K12 Moiphine_Q3=Morphine_X3*Moiphine_Ol-Morphine_Xl*Morphine_K13 Morphine_CLe = Morphine_Xeffect*Morphine_KeOMorphineJX: Morphine_KeO*.001/Morphine_Vl Morphinejjptake = Mo^hine_in_Limg*.693/MorphinejVbsorption_HaIf_Life OUTFLOWS: Morphine_Cll = MorphineJQ*Morphine_K10 * Morphine_X2(t) = Morphine_X2(t - dt) + (- Morphine_C12) * dt INIT Morphine_X2 = 0 OUTFLOWS: Morphine_C12 = MorpIrme_X2*Morphiae_K21-Morphir^_Xl*]V[orphineJK:i2 Morphine_X3(t) = Morphine_X3(t - dt) + (- Morphine_CI3) * dt INrr Morphine_X3 = 0 OUTFLOWS: Morphine_C13 = Morphine_X3*Morphine_K31-Morphirie_Xl*Morpliine_K13 Morphine_XetTect(t)=Morphine_Xet&ct(t - dt) 4- (- Morphine_CLe) * dt INTT Morpbrne_XerTect = 0 OUTFLOWS: Morphine_CLe = Morphine_Xeffect*Morplrine_KeOMorphine_ Xl*Morpliine_KeO*.OOl/Morp1iine_Vl PaCO2(t) =PaCO2(t - dt) + (CO2_Accuraulation - CO2Equilb) * dt INTT PaCO2 = PaCO2@0 INFLOWS: CO2_Accumulation = KelCO2*PaC02@0-KfilC02*VentilatotyJDepression*PaCO2 OUTFLOWS: CO2Equilb = keOCO2*(PaCO2-PeCX)2) PeCO2(t) =PeC02(t - dt) + (CO2Equilb) * dt INIT PeCO2 = PaOO2@0 INFLOWS: GO2Equilb = teOCO2*(PaC02-PeCO2) Sedation_State(t) = Sedation_State(t - dt) + (Sedation_Evaluator) * dt INIT Sedation_State = 0 INFLOWS: SedationJBvaluator = if Combined_Opioid_Effeci_Site_Concentration 0 else 1 Alfentanil_Absorption_Half_Life = 1 Alfentanil_Ce = Alfentanil_Xeffect/.001 Alfentanil_Cp = Alfentanil_X l/Alfentanil_Vl AlfentanilJDose ug = 1500 AlfentanOjaO = 0.090957 Alfentanil_K12 = 0.655935 Alfentanil_K13 = 0.112828 Alfentanil_K21= 0.214 AlfentanilJOl = 0.017 Alfentauil_KeO = 0.77 Alfentanil_Vl = 2.18 C50 = l.l Combined_Opioid_EfFeci_Site_Concentratioa = Alfailaiul_Ce/60+Morphine_Ce/70 Dose_Duration= 12 F = 4 Gamma =1.2 keOCO2 = 0.92 KelCO2 = 0.082 Moiphine_Absorption_Half_Life = 2 Moq3hine_Ce=Moiphine_Xefiect/.001 Morphine_Qp=Morphine_X 1 /Moiphine_Vl Morphine_D6se__mg = 20 Morphine_K10 = 0.070505618 Moiphine_K12 = 0.127340824 Moiphme_K13 = 0.018258427 Moiphine_K21 = 0.025964108 Moiphine_K31 =0.001633166 MorphineJKeO == 0.005 MoiphineJVl = 17.8 PaCO2@0 = 40 Sedation_Threshold= 1.5 VentilatoryJDepression^l- CombinedJ pioidJEffeciJSit Concentration Site_ConcentrationAGanima))*(PeCO2/PaC02@0)AF The simulation was run with a starting parameter of 700 meg of bioavailable alfentanil and 67 meg of bioavailable morphine in the inhaler at time 0 (Alfentanil hi Inhaler = 700 meg at time = 0; Morphine In Inhaler = 67 meg at time 0). Figure 18 and 19 shows the concentrations of various parameters when the simulation was run. Figure 18 showed concentration of alfentanil (in ng/ml, line 1), morphine (in ng/ml, line 2) and combined opioid (in ng/ml of fentanyl equivalents, line 3) over time (in minutes) at the effect site following inhalation of the combined product. In this ^example, drug administration has terminated after 90% of the inhaled drug was delivered because of patient sedation. As can be seen, the alfentanil concentration rises quickly in the effect site (line 1) producing a rapid drug effect. The morphine drug effect rises quite slowly in the effect site (line 2), producing a slowly rising drug effect. Line 3 is shows the combined opioid effect site concentrations, where each drug has been adjusted for its potency relative to fentanyl. All three lines have different Y scales, as can be seen on the Y axis, to normalize the effect site concentrations for relative potency. As can be seen in line 3, the highest opioid exposure occurs at the time of inhalation, and is almost entirely due to the alfentanil. However, as the alfentanil washes out of the effect site, it is almost exactly replaced by the influx of morphine into the effect site. A concentration at the site of effect of less than 25 ng/ml on the alfentanil scale (equivalent to 37.5 ng/ml on the morphine scale and 0.5 ng/ml on the fentanyl scale due to their relative potency) is considered sub-therapeutic; a patient will typically feel analgesic effects between 50 ng/ml and 100 ng/ml (on the alfentanil scale), side effects between 75 and 125 ng/ml (on the alfentanil scale) and toxic effects above 125 ng/ml (on the alfentanil scale). Figure 19 showed the ventilatory depression from the inhalation of an alfentanil morphine combination opioid delivery system. As shown in figure 19, the ventilation decreases to about 65% of baseline during drug administration, and then recovers to approximately 80% of baseline as CO2 accumulates. Ventilation is maintained at 80% of baseline throughout the next 4 hours, as the morphine effect is sustained. As demonstrated in figures 17,18, and 19, in the alfentanil morphine combination opioid delivery system, based on simulations using parameter values taken from the literature, the patient self-limiting opioid delivery system prevents administration of a toxic dose of opioid, and provides for the safe delivery of a slowly acting opioid by combining the slowly acting opioid with a rapidly acting opioid, and using the effects of the rapidly acting opioid to limit the total opioid exposure. ; Example 5: Clinical testing of Fentanvl preparations in human subjects ; fa) Method of Preparation of Free and Liposome Encapsulated Fentanyl Preparation^. Preparations containing a mixture of free fentanyl and liposome encapsulated fentanyl were prepared by mixing an ethanolic phase with an aqueous phase. The ethanolic phase comprised ethanol, fentanyl citrate, phosphatidylcholine and cholesterol. The aqueous phase comprised water for injection. Before mixing, both phases were heated to a temperature of about 56 to 60 degrees centigrade. The two phases were mixed and the mixture was stirred for a further 10 minutes at 56-60 degrees centigrade. The mixture was then allowed to cool to room temperature over approximately two hours. Typically, each ml of the final aqueous formulation contained 500 meg fentanyl, 40 rag phosphatidylcholine, 4 mg cholesterol, and 100 mg ethanol. After filling, preparations were autoclaved for final sterilization. Final preparations contained between 35 to 45 of the fentanyl as free drug with the remainder in the encapsulated fraction. (b) Treatment protocol The procedure of the following example shows how the administration of] a mixture of free and liposome encapsulated fentanyl through the lungs of a patient delivers therapeutically effective concentrations to the bloodstream and that side effects of hypoxia are generally (but not always) preceded by somnolence, dizziness or sedation during the administration period. Healthy volunteer subjects were treated with single or multiple doses of a mixture of free and liposome encapsulated fentanyl using the AeroEclipse™ Nebulizer breatiiactuated unit with compressed air set at 8 litres/minute. During each dosing period the nebulizer was charged with a 3 ml of the mixture of free (40%) and liposome encapsulated (60%) fentanyl and the subjects were instructed to inhale nebulized drug until the device no longer generated aerosol for inhalation. Subjects that become drowsy, sleepy or dizzy during the inhalation period were encouraged to continue to selfadminister the drug until the nebulizer was no longer generated aerosol. Plasma samples were collected through the administration period and for tile 12 hours following initiation of administration to monitor plasma fentanyl concentrations. Patients were monitored for any adverse events, including changes in respiratory rate and hypoxia. Control subjects were given intravenous fentanyl. (c) Measurement of maximum plasma concentration and end of dose plasma concentration In order to determine whether patients could prevent toxic levels of drug by selflimiting the drug before a toxic effect was exhibited, maximum plasma concent -ation (Cmax) was plotted against plasma concentration at end of dosing (Ceod) (Figure 20A). Ceod was found in most cases to be within 80% of Cmax, indicating that the max mum concentration of opioid was not significantly higher than the concentration at the time the subject stopped taking the opioid. This is in stark contrast to the control subjects (figure 20B) where patients given intravenous fentanyl exhibited Cmax concentrations significantly higher than Ceod. This indicates an increased safety in titration of drug by the subjects, since the concentration (and resulting toxic effects) of the opioid will not increase significantly after the subject stops taking the opioid. This indicates that, in an inhaled opioid formulation of the disclosed concentration over a relatively long period of time (2-20 minutes), if the "end of dose" amount is non-toxic, the maximum concentration of opioid taken by the subject is likely also non-toxic. (d) Determination of time points for side effects and toxic effects In order for subjects to effectively self-titrate, side effects of the drug such as drowsiness, dizziness or ventilatory depression should occur before the onset of toxic effects. Toxic effects were denned in this experiment as blood hypoxia resulting in a blood oxygen saturation lower than 90% of normal for the subject. In order to deteimine whether side effects occur before toxic effects, time to a side effect, and time to a toxic effect, were plotted against time to end of dose (Figure 21). For any end of dose time, time to side effect was equal or shorter than time to toxic effect, indicating that drowsiness, dizziness or ventilatory depression always took place before or at the time of toxic effect. • (e) Determination of.correlation between toxic effect and side effect. In order for subjects to effectively self-titrate, a toxic effect should be almost always preceded by a side effect causing (or signaling) the cessation of administration of drug. Figure 22 shows, for the total population of the study, that side effect is closely correlated to toxic effect, indicating that it is extremely likely that a subject exhibiting a toxic effect will have also exhibited a side effect. This example shows, in a controlled trial of human subjects, that (1) a toxic ef is almost always preceded by a side effect, and that (2) Cmax of inhaled opioid, in the dose profile given in Ms example, is approximately Geod. Therefore, a subject who stops administration of opioid when a side effect is felt will likely not reach opioid concentration levels required for toxic effect. WE CLAIM: 1. An opioid formulation for use in a method of providing analgesia to a patient; wherein the formulation comprises of at least one rapid-onset opioid; of atleast one sustained relief opioid and a pharmaceuticaHy acceptable carrier as herein described, wherein the rapid-onset opioid is chosen from fentanyl, alfentanil, sufentanil and remifentanil, preferably from fentanyl and alfentanil, wherein the concentration and type of each opioid in the formulation is selected so that, during inhalation, analgesia is achieved before the onset of said side effect, and the onset of said side effect occurs before the onset of toxicity, and so that the maximum total opioid plasma concentration does not reach toxic levels, whereby the onset of said side effect can be used by the patient to terminate inhalation to avoid toxicity, wherein the said formulation is dispensed by the pulmonary drug delivery device at a mass medium aerodynamic diameter of from 1 to 5 microns; wherein the ratio of concentration of total sustained-effect opioid to concentration of total rapid-onset opioid is from 1:2 to 6:1 and the said one sustained-effect opioid is liposomally encapsulated fentanyl and the rapid-onset opioid is fentanyl. 2. The formulation as claimed in claim 1 wherein the formulation is dispensed by the pulmonary drug delivery device at a mass medium aerodynamic diameter of from 1 to 3 microns. 3. The formulation as claimed in claim 2 wherein the formulation is dispensed by the pulmonary drug delivery device at a mass medium aerodynamic diameter of from 1.5 to 2 microns. 4. The formulation as claimed in any of claims 1 to 3 wherein the maximum total opioid plasma concentration at the onset of side effect is no less than 66% of the maximum total opioid plasma concentration. 5. The formulation as claimed in claim 4 wherein the maximum total opioid plasma concentration at the onset of side effect is no less than 80% of the maximum total opioid plasma concentration 6. The formulation as claimed in claim 1 wherein the ratio of concentration of liposomally encapsulated fentanyl to fentanyl is from 1:1 to 5:1 7. The formulation as claimed in claim 1 wherein the ratio of concentration of liposomally encapsulated fentanyl to fentanyl is from 2:1 to 4:1. 8. The formulation as claimed in claim 1 wherein the ratio of concentration of liposomally encapsulated fentanyl to fentanyl is from 3:1. 9. The opioid formulation as claimed in claim 1 wherein the total opioid concentration is from 250 to 1500 mcg/ml. 10. The opioid formulation as claimed in claim 9 containing liposomally encapsulated fentanyl in a concentration of from 100 to 750 mcg/ml. 11. The opioid formulation as claimed in claim 9 containing fentanyl in a concentration of from 100 to 750 mcg/ml. 12. The opioid formulation as claimed in claim 1 wherein the total opioid concentration is 500 mcg/ml, the fentanyl concentration is 200 mcg/ml and the liposomally encapsulated fentanyl concentration is 300 mcg/ml. 13. An opioid formulation as claimed in any one of claims 1 to 12 as and when used in the manufacture of a medicament for providing analgesia to a patient.

Documents:

4177-delnp-2005-abstract.pdf

4177-DELNP-2005-Assignment-(11-06-2009).pdf

4177-DELNP-2005-Assignment-(29-06-2009).pdf

4177-DELNP-2005-Assignment-11-04-2008.pdf

4177-DELNP-2005-Claims-(26-06-2009).pdf

4177-delnp-2005-claims.pdf

4177-DELNP-2005-Correspondence-Others-(11-06-2009).pdf

4177-DELNP-2005-Correspondence-Others-(26-06-2009).pdf

4177-DELNP-2005-Correspondence-Others-(29-06-2009).pdf

4177-DELNP-2005-Correspondence-Others-11-04-2008.pdf

4177-delnp-2005-correspondence-others.pdf

4177-DELNP-2005-Correspondence-PO-(11-06-2009).pdf

4177-delnp-2005-description (complete).pdf

4177-delnp-2005-drawings.pdf

4177-DELNP-2005-Form-1-(11-06-2009).pdf

4177-delnp-2005-form-1.pdf

4177-delnp-2005-form-18.pdf

4177-delnp-2005-form-2.pdf

4177-DELNP-2005-Form-26-(11-06-2009).pdf

4177-DELNP-2005-Form-26-11-04-2008.pdf

4177-delnp-2005-form-3.pdf

4177-delnp-2005-form-5.pdf

4177-DELNP-2005-Others-Documents-(11-06-2009).pdf

4177-delnp-2005-pct-210.pdf

4177-delnp-2005-pct-237.pdf

4177-delnp-2005-pct-308.pdf

4177-delnp-2005-pct-311.pdf

4177-DELNP-2005-Petition-137-(29-06-2009).pdf