- Top of page
Summary: Purpose: Depression is common in temporal lobe epilepsy (TLE) and after temporal lobectomy, and its etiology is obscure. In nonepileptic depression (including depression associated with other neurologic disorders), a consistent PET imaging finding is frontal lobe hypometabolism. Many TLE patients have hypometabolism involving frontal regions. Thus in data available from routine clinical assessments in an epilepsy surgery unit, we tested the hypothesis that the pattern of hypometabolism, particularly in the frontal lobe, may be associated with the depression seen in patients with TLE and TLE surgery.
Methods: We studied 23 medically refractory TLE patients who underwent anterior temporal lobectomy and who had preoperative FDG-PET scanning. All patients had pre- and postoperative psychiatric assessment. By using statistical parametric mapping (SPM-99), patterns of hypometabolism were compared between patients who had a preoperative history of depression (n = 9) versus those who did not (n = 14) and between those in whom postoperative depression developed (n = 13) versus those in whom it did not (n = 10). A significant region of hypometabolism was set at p < 0.001 for a cluster of ≥20 contiguous voxels.
Results: Patients with a history of depression at any time preoperatively showed focal hypometabolism in ipsilateral orbitofrontal cortex compared with those who did not (t= 4.64; p < 0.001). Patients in whom depression developed postoperatively also showed hypometabolism in the ipsilateral orbitofrontal region (t= 5.10; p < 0.001).
Conclusions: Although this study is methodologically limited, and other explanations merit consideration, orbitofrontal cortex dysfunction, already implicated in the pathophysiology of nonepileptic depression, may also be relevant to the depression of TLE and temporal lobectomy.
Depression is a common comorbidity in temporal lobe epilepsy (TLE) (Lambert and Robertson, 1999). As well as the distress and impairment of the illness itself, depression contributes to the risk of suicide, to impaired quality of life, to subjective cognitive dysfunction, and possibly to the progression of TLE itself (Kanner and Balabanov, 2002). In general, depression rates are reduced after epilepsy surgery, particularly when seizure freedom is achieved; however, occasional cases of de novo depression do occur (Devinsky et al., 2005). The rate of depression in TLE is probably greater than that in other forms of epilepsy and, further, the elevated rate may be particularly for mesial TLE (MTLE) (Quiske et al., 2000). The clinical features of depression associated with MTLE are similar to those of depressive disorders occurring in nonepilepsy populations, although it has been proposed that an interictal dysphoric disorder also is specific to TLE (Blumer et al., 2004).
At present, only conjectures and fragmentary insights exist into the etiology and pathogenesis of the depression of MTLE. In studying mood disorders generally, one productive research approach has been neuroimaging, both structural and functional (Drevets, 2001; Mayberg, 2003). Studies of depression secondary to neurologic disorders such as Parkinson disease and stroke have particularly implicated paralimbic regions, especially orbital and inferior prefrontal cortex and temporal cortex (Mayberg, 2001; Drevets et al., 2004). An important finding is that, rather than the increased metabolism in orbitofrontal cortex observed in functional imaging studies of primary depression, in secondary depression, orbitofrontal metabolism (and blood flow) is reduced or normal; and dorsolateral prefrontal blood flow is apparently normal.
18F-fluorodeoxyglucose-positron emission tomography (FDG-PET) neuroimaging in TLE patients typically shows hypometabolism in the epileptogenic temporal region, which is usually more extensive than the underlying anatomic abnormality (Casse et al., 2002; Henry and Votaw, 2004; Mauguiere and Ryvlin, 2004). Moreover, even TLE patients with no evidence of a structural lesion may have marked hypometabolism covering large areas of temporal lobe, including the temporal pole, mesial structures, and parts of the lateral temporal cortex (Semah, 2002; Carne et al., 2004). Hypometabolism in TLE may also occur extratemporally, the commonest region involved being the frontal lobe (Semah, 2002).
Neuroimaging studies examining depression in TLE are logistically difficult to conduct and remain rare (Bromfield et al., 1992; Victoroff et al., 1994; Schmitz et al., 1997; Tebartz van Elst et al., 1999; Quiske et al., 2000; Giovacchini et al., 2005), and no firm conclusions can yet be drawn from the available evidence. The best-quality functional imaging study to date used FDG-PET imaging in partial epilepsy patients, defining a priori regions of interest to compare patients with a history of depression with those without (Bromfield et al., 1992). Both a structured psychiatric interview and a depression rating scale (the Beck Depression Inventory, BDI) were used. In patients with temporal lobe foci, the authors observed lower metabolism in inferior frontal cortex in patients with higher depression scores, compared with those with lower scores and with normal control subjects.
Here we report a study examining the patterns of FDG-PET metabolism, by using statistical parametric mapping (SPM), in patients undergoing surgery for medically refractory TLE to determine whether patients who had a preoperative history of clinical depression, or in whom postoperative depression developed, demonstrated differences from nondepressed patients. The analyses tested two primary hypotheses: (a) that TLE patients with a history of depression would show extratemporal regions of focal hypometabolism compared with TLE patients without a history of depression. Guided by the study of Bromfield et al. (Bromfield et al., 1992), we predicted abnormalities in the orbitofrontal cortex in particular; and (b) that the presence of extratemporal hypometabolism on the preoperative FDG-PET scan would predispose patients to the development of depression after a temporal lobectomy for medically refractory TLE.
- Top of page
In this study we found that medically refractory TLE patients with a history of depression had a region of focal hypometabolism on a presurgical FDG-PET involving the orbitofrontal cortex ipsilateral to the epileptogenic focus compared with patients who did not have a history of depression. A similar region of hypometabolism was found when the group of patients in whom depression subsequently developed after a temporal lobectomy were compared with the patients in whom it did not. The findings are consistent with a previous study that showed hypometabolism in the inferior frontal cortex in patients with complex partial seizures who had a history of depression (Bromfield et al., 1992). However, no previous studies demonstrated a relation between preoperative hypometabolism in orbitofrontal cortex and the development of depression after subsequent epilepsy surgery for temporal lobe epilepsy. Because the association was with previous and future, rather than currently active depressive illness, it is postulated that these metabolic changes may represent chronic changes in either neuronal structure or cellular metabolism rather than a reactive functional change to the depression.
In idiopathic (non–epilepsy-related) depression, orbitofrontal hypometabolism has been shown on FDG-PET (Mayberg, 2003), as well as hypometabolism in dorsolateral and ventrolateral prefrontal cortex. Structural MRI studies (Bremner et al., 2002; Lacerda et al., 2004) have shown evidence of reduction in gray matter volumes specifically in orbitofrontal cortex (i.e., with no reductions in other frontal areas). Hypometabolism in orbitofrontal cortex has also been shown in patients with other neurologic diseases [e.g., Parkinson's disease (Berding et al., 2001)]. The role of the orbitofrontal cortex in normal function and in depression has been extensively studied (Blair and Cipolotti, 2000; Cavada et al., 2000; Cavada and Schultz, 2000; Ongur and Price, 2000; Roberts and Wallis, 2000).
The results of the present study suggest the hypothesis that orbitofrontal hypometabolism may act as a predisposing risk factor for the development of depression in patients with TLE. If so, the relevant mechanisms are at present unknown, but several possibilities should be considered in future studies. First, the hypometabolism may reflect extension of the sclerosis and cell loss from the temporal lobe to extratemporal structures (Schwarcz et al., 2002; Semah, 2002). Alternatively, the hypometabolism may represent compensatory neuronal inhibition. Third, the hypometabolism may result from the effects of seizure spread to the orbitofrontal cortex, a common site of extratemporal propagation in TLE. Fourth, it may be a marker for general severity and cerebral dysfunction associated with TLE, rather than being indicative of a specific role of orbitofrontal cortex. Finally, the orbitofrontal hypometabolism may be a secondary functional consequence of the depression. This last explanation is unlikely, as most of the patients were not depressed at the time of the PET acquisition (although no depression measure was administered at the time of the PET scan, all assessments were conducted within 3.0 ± 0.6 months of the scan). Whatever the mechanism for its induction, the orbitofrontal hypometabolism seen in these TLE patients is likely to reflect underlying neuronal dysfunction in this brain region, and, if this is correct, such dysfunction may predispose to the development of depression in TLE patients.
Sixty-two percent of patients in whom depression developed after the surgery had a presurgical history of depression. This is consistent with previous evidence indicating that patients who have a history of depression before surgery are more likely to have depression develop after surgery (Kanner and Balabanov, 2002), although overall, the tendency is for depression rates to diminish. The patients had a variety of different pathologies, with one third having nonspecific pathology, whereas two thirds had MTS. Patients with a history of depression and those without were well balanced in the types of pathologies.
Given the great scarcity of functional imaging studies of TLE-associated depression, it would be valuable if this finding, which resembles that of Bromfield et al. (Bromfield et al., 1992), were to be followed up with a larger prospective study. Larger sample sizes would allow an examination of the relation to the nature of the underlying pathology, especially MTS versus other pathology (Quiske et al., 2000).
Future studies should also include structured psychiatric interview to obtain current and lifetime diagnoses of any type of depression, not just clinical major depression as was done in this study. To investigate the pathophysiologic underpinnings of the association observed between ipsilateral orbitofrontal hypometabolism, it would also be interesting to examine whether a relation exists between this and disturbances in the hypothalamic–pituitary–adrenal axis (Zobel et al., 2004) and 5HT1A receptor density in the medial temporal lobe (Giovacchini et al., 2005). Methods are available for volumetric study of orbitofrontal cortex (Lacerda et al., 2003), so correlation between orbitofrontal atrophy on MRI and hypometabolism on PET imaging could be examined. Where PET studies are initiated primarily for clinical rather than research purposes, it would be valuable to measure mood at the time of scanning. In this exploratory study, the early postoperative psychiatric assessments were driven and constrained by clinical need; longer-term psychiatric follow-up should be performed to characterize more carefully the course of postoperative depression and its imaging correlates.
Finally, it would be exceptionally useful to perform further functional imaging studies comparing samples of well-characterized depressed TLE patients, non-TLE epilepsy patients, and depressed nonepilepsy controls.