In their 1977 paper Branston et al.  referred to spontaneous, transient increases in extracellular potassium ion concentration (Ke) which occurred in the ischaemic penumbra following experimental MCAO. Similar, spontaneous events were later reported in another MCAO preparation, also in a gyrencephalic species . It was suggested then that such events, PIDs  or HSDs , were ''not necessarily benign'' , and specific studies have confirmed this page 24: Evolution of PID Patterns with Time Pathogenic Potential and Recruitment of Penumbra into Core Territory). The critical points of difference between CSD and PIDs are that CSD in completely healthy cortex requires an initiating stimulus and does not damage normally perfused and metabolising grey matter, whereas PIDs are spontaneous and do cause damage, and in the case of the ischaemic penumbra, appear to play a large part in recruiting this zone of tissue into the expanding core infarct until this reaches what appears to be a ''predestined'' size (assuming no treatment).
Detection with Electrodes, and Characteristics of PIDs in Experimental in Vivo Models
PIDs have usually been documented from recordings of the cortical DC potential, and traditionally this has been regarded as a reference detection method. Such electrodes need to be non-polarisable, and usually consist of a glass micropipette filled with physiologically neutral electrolyte and inserted into the cortex, or a chlorided silver ball placed on the cortical surface. Twin-barrelled surface contact or glass microelectrodes allow the signal from an ion-selective barrel (most often to K+) to be compared with that from an adjacent electrode, both of them referenced to a remote ground electrode. The time course of Ke as recorded from such an electrode during a PID resembles that of CSD in respect of onset and peak amplitude, but may differ in that the recovery phase may be more prolonged. In baboons, a linear, direct relationship of Ke clearance half time with degree of ischaemia was shown, and interpreted as indicating that clearance was no longer by Na-K ATPase (energy-dependent), but relied instead on passive elution by residual perfusion . Studying MCAO in rats, Gill and colleagues  distinguished ''small'' (duration @1 minute) and ''big'' PIDs, both recorded with DC electrodes, the latter having much longer time courses. In the same study, this group showed that the time course of depletion of extracellular calcium mirrored that of the DC potential, indicating that ''big'' PIDs were associated with protracted increases in intracellular calcium, likely to be cytotoxic.
The Response of CBF to a Peri-Infarct Depolarisation
The hyperaemic response to a CSD wave is well recognised from observation of cortical vessels , serial section autoradiography in the rat brain (generating a time series as the event propagates along the hemisphere) , laser Doppler flowmetry , and, by inference, from monitoring of transient increases in tissue pO2 . Following MCAO, the CBF response is greatly attenuated, or even reversed; thus laser Doppler flowmetry in a deteriorating patient with an intracerebral haematoma at first revealed transient increases in perfusion coupled to probable CSD episodes, but the perfusion responses reversed to transient hypoperfusion as brain swelling progressed  (Fig. 8). Back et al. showed that the positive tissue hyperoxia of CSD becomes a transient decrease in tissue pO2 in focal ischaemia .
In open-skull animal models of stroke, it is usually necessary to leave electrodes at a fixed location rather than probing different cortical areas sequentially, and it is also not possible to determine the extent of propagation of a presumed PID wave with one or more electrodes in the cortex. The use of a method that acquires sequential images of the exposed core and penumbral areas offers a solution if the variable being imaged is affected by the pathophysiology. When illuminated with fluorescent light at 370 nm, the cortex will fluoresce blue, emitting light in the range 445470 nm; the fluorochrome responsible is the reduced species of the nic-otinamide adenine dinucleotide redox couple (NAD/NADH), the coen-zyme for succinic dehydrogenase in the mitochondrial respiratory chain. Only NADH - the reduced species - fluoresces, so that oxidation of the couple leads to a fall in fluorescence, whereas reduction causes an increase. Interpretation of such images needs to take account of the capacity of haemoglobin, particularly when oxidised, to absorb or quench blue light (hence its colour!). This method was applied in non recovery MCAO studies in cats , and revealed spontaneous increases in 450 nm fluorescence that appeared almost always to originate near the core territory and propagate outwards into the penumbra at rates in the range 1-4 mm cortex per minute and hence very characteristic of CSD (Figs. 4-6). Propagation is invariably around the walls of a sulcus, with no evidence that the event can spread directly between gyri lying in contact at the surface. Time courses of the events could be classified into (1) fluorescence increases that did not reverse, (thus closely resembling the time course of terminal depolarisation as recorded with a Ke or with a DC-potential electrode), and which occurred on penumbral cortex close to the core, (2) more peripheral
transient increases in fluorescence that had propagated centrifugally from cortex affected by PIDs with the first pattern, and (3) transient decreases in fluorescence, occurring in cortex close to the anterior cerebral artery input, and probably lying just outside penumbra (Figs. 4-6). In some cases, a single PID was seen to propagate from penumbra into anterior cerebral territory, changing its polarity from increase to decrease as an unseen interface was crossed (Fig. 5).
Increases in fluorescence may represent either reduction of the redox couple or a decrease in haemoglobin at the same locus, or a combination of the two, although also not excluding a small increase in haemoglobin outweighed by a larger NADH increase. Whichever the explanation, the observed increase in raw fluorescence indicates either vascular or metabolic compromise, and the method has been used largely to confirm propagation of the events, and to detect them. The depression in crude fluorescence grey level in normally perfused cortex outside the penumbra accords well with the depression of compensated fluorescence during CSD as shown by Rosenthal and Somjen .
Fig. 5. Sequence of digital images illustrating initiation and propagation of a peri-infarct depolarisation in the penumbra following experimental middle cerebral artery occlusion. (For orientation of the image field in relation to the whole hemisphere please see Fig. 4) After exposure of the brain a sequence of grey scale fluorescence images was acquired 53 minutes after occlusion of the middle cerebral artery. The baseline image acquired at time zero was subtracted from each subsequent image and the difference image calculated and displayed in pseudocolour. Green background indicates no change in fluorescence while colours up through the rainbow spectrum to red, pink, white represent increases in fluorescence, and changes into blue, purple or black, decreases respectively. Panel 1: EG ectosylvian gyrus (ischaemic core). SG Suprasylvian gyrus (inner penumbra). MG Marginal gyrus (outer penumbra). Principal middle cerebral input is from lower right of the field, and anterior cerebral from upper right (panel 6) (see also Fig. 4). White line in panel 1 represents the anterior margin of the craniectomy exposing the cortex. Red lines represent sulci, and blue line, the line of the sagittal sinus medial to MG. Shortly before the image shown in panel 1, an area of increased fluorescence emerges from the lower sulcus and propagates outwards (from MCA input) throughout the SG (panels 2-3). After an interval between panels 3 and 4, the depolarisation (verified by potassium-selective electrode on posterior SG) emerges onto the MG and propagates forwards and medially (panel 5) but on reaching cortex perfused by anterior cerebral artery (ant. cer.), the event dissipates, represented only by a decrease in fluorescence in panel 6 (upper right of panel). Thus, the white line drawn on MG in panels 5 and 6 represents an apparent interface between middle and anterior cerebral territory. In this example, fluorescence has returned to baseline in the suprasylvian gyrus, but after one or more subsequent similar events, fluorescence increases on this gyrus often culminate in a permanent increase, probably indicating terminal depolarisation (Fig. 6). (Reproduced with permission from Strong et al. 1996 )
Initiation of PIDs
Experience with in vivo imaging suggests that the great majority of PIDs originate at the edge of core territory , and the high levels of Ke present in core areas are a probable cause [11, 27], but the same considerations apply as in CSD, and glutamate or other factors liberated from ischaemic tissue might contribute.
In the core infarct territory established soon after experimental MCAO, the DC potential rapidly becomes negative, but, unlike a PID, does not then resolve, instead becoming increasingly negative and reaching a plateau that is interpreted as indicating complete depolarisation of all cellular elements. Terminal depolarisation - effectively a failure to repolarise spontaneously (as does a PID) - is commonly taken to imply complete depletion of the ATP pool required for repolarisation, and will lead inevitably to infarction unless the ATP pool can be restored promptly by reperfusion.
Evolution of PID Patterns with Time, Pathogenic Potential, and Recruitment of Penumbra into Core Territory
PIDs - however detected - recur at irregular intervals during ischaemia, and observation over 10-12 hours reveals that, at least under chloralose anaesthesia, the pattern of recurrence eventually culminates in terminal depolarisation in outer areas of penumbra, similar to the sequence that occurs earlier in more central penumbra . A feature of the progression, when Ke is monitored, is that resolution of each Ke PID transient towards the pre-transient baseline becomes steadily less complete with time, leading to a gradually increasing Ke baseline. Harris et al. showed that in the case of Ke, there is a striking acceleration of Ke increase when it reaches 13 mmol, suggesting a specific change in a membrane conductance ; terminal depolarisation follows, and the area of penumbra affected is thus recruited into the core infarct. This sequence of events suggests that number or frequency of PID events in the penumbra is a principal determinant of infarct size, and three pieces of evidence support this. First, Gill et al. showed that when the number of PIDs was restricted with the non-competitive NMDA antagonist dizocilpine in rats subjected to MCAO, infarct size was reduced . Secondly, Mies and colleagues reported findings closely similar to those of Gill's group . The association of a larger infarct with increasing PID number may simply reflect the operation of a different, underlying mechanism determining both infarct size and PID frequency. However, thirdly and conclusively, Busch et al. were able to increase infarct size in rats by inducing CSD events outside the penumbra which propagated into it and caused enlargement of the definitive core infarct .
Arising from the original demonstration that loss of evoked potential amplitude could be reversed upon reperfusion, the initial concept of the ischaemic penumbra was of a ''sleeping beauty'' - a zone of cortex whose function was reversibly suppressed in a stable fashion, so that function could be restored at a much later time point by the magical touch of a vascular neurosurgeon carrying out an extra-intracranial vascular bypass procedure . The study of PIDs and manipulations of their frequency has demonstrated instead that - without early reperfusion - the ischaemic penumbra is a maturation phenomenon in which the core infarct gradually expands into penumbra, thus ''recruiting'' it. The time course of this progression is probably shortest in rats - perhaps 3 hours, extending to 12 to 24 hours in cats, and is believed in humans to extend to perhaps 48 hours. The factors which might influence PID frequency and hence the rate of progression need to be considered.
Tower and Young's observation of a relationship of cerebral cortical glia : neuron ratio with brain mass is relevant to brain injury since, as mentioned earlier, glial buffering of potassium ion concentration and uptake of neurotransmitters, especially glutamate, are important mechanisms for homeostasis of the extracellular space. It is therefore not a matter of surprise that the frequency of PIDs following MCAO in rats should be high , but much less so in cats . CSD is also difficult to induce in monkeys . Efforts to make a direct comparison of PID frequency between cats and primates were frustrated by considerable inter-experiment variability in frequency within a species, but variations in plasma glucose emerged from these experiments as a cause of this variability; this is discussed below (some page: Relationship of Cortical Glucose Availability with PID Frequency). The inference from such comparisons is that PIDs in humans might be rarer still - perhaps vanishingly so - and the relevant, new evidence is described later.
Effects of Drugs and Anaesthetic Agents on PID Frequency
The beneficial effects of NMDA-type glutamate receptor blockade on PID frequency and on infarct size have been reviewed above (page 18: Drugs and Anaesthetic Agents). The AMPA/kainate-type glutamate receptor antagonist NBQX has been shown to reduce PID frequency and volume of ATP depletion in rats subjected to MCAO , and this agent has also been shown to reduce ischaemic lesion volume . It is of interest that, unlike MK-801, NBQX does not prevent induction of CSD in the normal brain . The volatile anaesthetic agent halothane may achieve its experimental neuroprotective effect by reducing PID numbers , and can, like propofol, block CSD . The fact that halothane also blocks gap junctions in cultures of astrocytes  supports the argument for a role of glial gap junctions in the propagation of CSD .
Relationship of Cortical Glucose Availability with PID Frequency
As CBF progressively falls in focal ischaemia, a shift to anaerobic gly-colysis is inevitable once oxygen extraction is maximal. At that point, a dramatic loss in efficiency of glucose utilisation is equally inevitable, with a fall in net ATP yield per mole glucose utilised from 38 to 2 moles. Glucose utilisation increases to compensate ; this is possible despite presence of ischaemia, due to the remarkable effectiveness of the capillary glucose uptake/transport mechanism. This concept is based on several lines of evidence. Hansen showed that following cardiac arrest in rats, delay before terminal ischaemic depolarisation was proportional to plasma glucose, indicating an inverse relationship between depolarisation rate (the dependent variable) and glucose availability in the brain . In 1986, Nedergaard and Astrup showed in rats (MCAO) that hyperglycaemia reduced the frequency of PIDs (although a plasma level in excess of 30 mmol/L was needed to achieve this) . They also showed an increase in phosphorylation of [14C]2-deoxyglucose (an index of metabolic rate) that was related to frequency of PIDs, and predicted that with ischaemia accompanied by PIDs the brain free glucose pool would tend towards zero as delivery and extraction from plasma would quickly become inadequate, given the high, anaerobic utilisation rate. In cats (MCAO), dependence of homeostasis on plasma glucose is demonstrable at glucose levels that are frequently encountered in clinical practice: thus Strong et al. showed a striking increase in PID frequency in this situation when mean postocclusion plasma glucose fell below 4.5 mmol/L (the lower limit of normal quoted for clinical plasma glucose assays in our institution is 3.3 mmol/L) . Our subsequent, unpublished work suggests that the threshold may be nearer 6.5 to 7 mmol/L. This is of potential importance for clinical management since insulin is used to control hyperglycaemia in many intensive care units, with the target range varying in different units. At least one trial of glucose and insulin (to restrict ischaemic acidosis) in acute stroke is under way . There is also striking (and influential) evidence favouring the use of insulin in the intensive care of systemic critical illness .
In summary, the initiation of a PID appears to be a random event in which an elevated Ke level at the edge of core infarct territory causes depolarisation of neighbouring tissue because membrane homeostasis there is partially impaired. The impairment is due to a combination of factors in which reduction of glucose availability (the multiple of perfusion (absolute, ml/100 g/min) and plasma glucose levels) as ischaemia deepens becomes particularly important. It seems that reduction of glucose availability increases the probability of initiation of a PID.
The Metabolic ''Signature'' of PIDs
The transient hypoperfusion or reduction in tissue pO2 that occurs in association with a PID has been described above. Given the likelihood of transient tissue glycopenia during recovery from a PID, and the critical dependence of the ATP pool on the balance between on the one hand,
ATP utilisation for restitution of cation gradients during PID recovery, and glucose availability on the other, it becomes valuable to measure the available tissue glucose pool with sufficient time resolution to detect the effects on it of a PID. This has recently been achieved with the use of cerebral microdialysis coupled with rapid sampling of dialysate by means of an online, automated flow-injection assay [105, 106]. The technology allows enzymatic assay of microlitre dialysate samples for glucose and lactate at intervals of 30 seconds each. When dialysate was sampled from penumbral tissue closely adjacent to the core area after MCAO in cats, a PID arriving at the microdialysis probe was associated with complete disappearance of glucose from the dialysate within approximately 3 minutes. In more peripheral penumbra, PIDs were accompanied by transient, stereotyped increases in lactate and decreases in glucose, superimposed in the case of recurrent PIDs on decreasing glucose and increasing lactate baselines (Fig. 7) . This reproducible combination of transient metabolite changes may be taken as a typical metabolic "signature" for a PID, of potential value for the monitoring of patients with severe TBI or acute cerebral ischaemia.
Was this article helpful?
All you need is a proper diet of fresh fruits and vegetables and get plenty of exercise and you'll be fine. Ever heard those words from your doctor? If that's all heshe recommends then you're missing out an important ingredient for health that he's not telling you. Fact is that you can adhere to the strictest diet, watch everything you eat and get the exercise of amarathon runner and still come down with diabetic complications. Diet, exercise and standard drug treatments simply aren't enough to help keep your diabetes under control.