It is the little things that usually get you and they are the hardest to anticipate.
While I am not a neurosurgeon, it appears to me that Deep Brain Stimulation (DBS) lead implantation surgery has a lot of moving parts, any one of which can go wrong. One of the best descriptions of the surgery I heard was from Dr. Kathryn Holloway who described it as “fussy.” As a neurologist, both in and out of the operating room, I can see how the little things can go wrong and what that ultimately means postoperatively; which fortunately is relatively uncommon. But dealing with some of the postoperative problems, I have been able to compile a “wish list” of things I wish would have been done during surgery or soon thereafter. This list is described as hints and admittedly, the helpful part is for the DBS programmer afterward.
The need for certainty
There are hundreds of potential combinations of DBS electrode configurations, stimulation parameters, and pulse trains. The combination of electrode configurations will increase dramatically as segmented DBS leads become commercially available and the ability to apply different stimulation currents over multiple contacts simultaneously. Each combination takes time to implement and assess. Difficulties settling quickly on the optimal combination can be frustrating. Looming over the costs of time, effort and frustration, is the temptation to default back to medications. However, just defaulting back to the medications will surely fail, as the failure of medication is the reason for the patient having DBS.
It is unlikely that every conceivable combination will be tested and thus, eliminated as a possibility. Consequently, every DBS programmer is confronted with the decision of when to give up on the programming of a patient who fails to reach satisfactory results without compromising on what constitutes “satisfactory.” It is a slippery slope to re-define “satisfactory” in terms of what it means for the programmer, rather than for the patient. Certainly, as the number of combinations tried increases, the probability of there being a combination that will produce a satisfactory response becomes less.
With an increasing number of failed combinations, the programmer begins to construct a probability of there actually being a good combination. In terms of Bayes Theorem, this estimate would be a posterior probability. However, the posterior probability based on the programming experience with a particular patient must be weighted based on the prior probability. In this case, the prior probability is the likelihood that the DBS lead is in a location that is capable of providing a satisfactory response. Every programmer does a type of Bayesian analysis, even if only implicitly. If the programmer has confidence that the DBS lead is well placed, then the prior probabilities will be high and thus increase the possibility of finding a combination that will provide a satisfactory response. If however, the programmer has confidence that the DBS lead is not well placed, then the prior probability will be low. In either case, the programmer can arrive at a point of equanimity about how hard to pursue the DBS programming. The problem comes when the DBS programmer has no confidence one way or the other. The risk is giving up too soon with the DBS or continuing programming well past the point of diminishing returns.
The confidence of prior probabilities, as described above, depends greatly on the efforts in the operating room. Certainly, there are enormous pressures to complete the surgery as quickly as possible, for many reasons. However, the necessary confidence of the prior probability may require some additional effort. Perhaps, another microelectrode recording trajectory may help provide that confidence. Although one tries to assure that the optimal trajectory has been found prior to implanting the DBS lead, additional testing through the DBS lead and perhaps DBS testing in an additional tract may be necessary to achieve confidence. However, this first requires that the neurosurgeon appreciate the importance of confidence on the part of the post-operative DBS programmer.
The tyranny of partial benefit
What does the treating physician do when the patient achieves only 50% of the reasonably expected benefit, for example experiencing only a 30% improvement in the Unified Parkinson Disease Rating Scales when most clinical trials demonstrate an average on the order of 60%? Prior to the DBS lead implantation surgery, the patient balances the risks of the DBS lead implantation surgery against an improvement of 60%. Now the patient, family members, caregivers, physicians and healthcare professionals are faced with the issue of revising the DBS lead. The same risks associated with the first DBS surgery are confronted by the second surgery, but the benefit reduces from an improvement of 60% over the condition prior to the first surgery, to only an improvement of 30% compared to the state prior to the second surgery.
The risk-to-benefit ratio is significantly different for the second or revision surgery compared to the initial surgery. For some patients, the risk-to-benefit ratio for the revision surgery may not justify the second surgery. The consequence is that the patient is forced to accept only the partial benefit, hence the tyranny. Obviously, this situation is to be avoided and sometimes extra measures are needed to ensure optimal DBS lead placement.
Immediate post-operative imaging
Even after nearly 20 years of working in the operating room and the post-operative DBS clinics, I continue to learn. Fortunately, things generally go very well, but occasionally not. In those circumstances, one clearly needs to learn what did not go right. A significant factor in DBS leads not optimally placed is intra-operative brain shift. The brain has a specific gravity greater than cerebrospinal fluid, which means the brain will sink to the lowest point within the skull. That lowest point will differ between being supine, such as having a targeting MRI, and the semi-recumbent position in the operating room. Brain shift due to the brain sinking will be exacerbated if there is more room within the skull for the brain to sink, such as in elderly patients.
In my experience, one of the most frequent causes of misplaced DBS leads results from marked brain shift associated with intracranial air, particularly when a tension pneumocephalus is produced. The tension pneumocephalus often results in DBS lead placement anteriorly due to posterior shift of the brain. However, the cerebrum pivots on the brainstem so that posterior displacement also can produce a torsion that will result on the DBS lead being placed too medially or laterally dependent on the trajectory, relative to the axis of rotation. The risk for tension pneumocephalus can be reduced by a scrupulous surgical technique, such as copious irrigation when the intra-dural cavity is exposed and the use of agents, such as fibrin glue, to seal the burr hole.
As intra-cranial air is reabsorbed, the only way to know if there was significant brain shift due to intracranial air is if a MRI or CT scan is obtained very soon after surgery. If three months later the DBS programmer is confronted with a difficult patient and imaging performed then shows poorly placed leads, it will not be possible to know whether a tension pneumocephalus or some other factor was involved unless scans are done very soon after surgery. One will not learn from the failure and therefore will not be in a position to reduce the probability of a future failure.
Low tech to the rescue
An acute loss of DBS efficacy can have serious consequences. This is particularly true for disorders, such a Parkinson’s disease, where other treatments, medications for example, have been reduced. In patients with Parkinson’s disease, DBS in the vicinity of the subthalamic nucleus results in an approximately 50% reduction in anti-Parkinson medications. Thus, if the DBS fails, the patient only has a fraction of the medications previously need for some level of control. It is not infrequent that such a patient comes into the emergency room often in the evenings, nights and weekends.
One potential cause of acute loss of DBS efficacy is DBS lead migration. If the patient had a routine AP and lateral skull x-ray post-operatively at a time when the patient was benefiting from DBS, an easy test would be to simply obtain another set of skull x-rays. Comparison of the skull x-rays may readily demonstrate a lead migration without having to call in the MRI tech. Also, the skull x-ray, performed with a chest x-ray may also demonstrate hardware failures such as a fracture of the DBS lead or extension of an uncoupling of the connectors.
One could obtain a skull x-ray very soon after DBS lead implantation. However, if there is substantial intra-cranial air, the skull x-ray may be unhelpful as the lead may migrate over the next one to two weeks that it takes for the intra-cranial air to be absorbed. Further, obtaining the skull x-rays as soon as the patient demonstrates a satisfactory response allows the inference that the DBS lead is in good position. Thus, if the skull x-ray at the time of DBS failure shows that the DBS lead has not migrated, then one can turn attention to other potential causes of acute DBS failure.