Thursday, December 22, 2011

In Support of Sham Surgery

A control group seems to be an undisputed cornerstone to a strong study. So what happens when people say the control group is muddling the conclusions? This is the opinion dividing researchers who develop surgical treatments for Parkinson’s disease. In the last decade, several interventions that appeared promising in Phase I trials failed to have a significant impact in Phase II (sham-controlled) trials and subsequently were abandoned [1,2,3,4]. Many patients who benefited from the early trials say that the sham controls are obscuring the efficacy of the much-needed treatments. Meanwhile, proponents of the sham controls claim that the controls are necessary to demonstrate the efficacy of a treatment, especially in light of the robust placebo effect in the Parkinson’s population. Katsnelson’s article “Why Fake It?: How ‘Sham’ Brain Surgery Could Be Killing Off Valuable Therapies for Parkinson’s Disease,” presents an interesting discussion of this issue [5]. The ethical concerns of the two groups are as follows:

Fans of Shams:
Investigators in favor of sham controls are driven by an ethical imperative to produce the highest quality research possible. A Phase I open-label trial is a safety test that may suggest whether an intervention works; the Phase II trial truly establishes efficacy by controlling for confounds such as investigator bias and placebo effect. This control is particularly important in the Parkinson’s population, where the placebo effect is robust: believing in the efficacy of a treatment triggers dopamine release [6], which directly mitigates Parkinson’s symptoms. Researchers are adamant that these confounds should be eliminated with a control group, and sham controls are the most thorough type of controls. These arguments have the support of 94% of polled Parkinson’s researchers, who “believe sham-surgery controls are better than unblinded controls for testing the efficacy of neurosurgical interventions” [7]. Supporters also note that adverse events in sham control groups are very rare and that participants receiving the sham treatments can receive the experimental treatment if it is approved (although the studies cited in Katsnelson’s article suggest that approval is rare) [5].

Contesting the Controls:
An opposing group of scientists (many from Europe, where sham surgery is more controversial than in the US) believe that experiments featuring sham controls are not necessarily the best evaluation of new neurosurgical interventions for Parkinson’s disease. Dr. Robert Barker explains that inconsistencies in treatment administration account for part of the lackluster Phase II results and indicate a need for protocol refinement during Phase I trials. In this situation, the use of sham controls is premature. Barker also notes that the high cost of this design (an estimated $10 million for a 50 patient study [5]) limits the number and size of studies that can be performed. Furthermore, the pool of eligible patients is already small – assigning patients to be sham controls limits the size of the experimental group and “provides results of limited statistical utility” [5].

Critics also point to the risks of sham surgery. There are physical risks, including those associated with anesthesia, installation of a halo, drilling of holes into the skull and in some cases, even penetration beyond the dura mater. And there are unknown risks stemming from “psychological effects of unblinding” (Perry Cohen, quoted in [5]). Cohen mentions a patient who experienced dramatic improvements in the control group, but rapidly deteriorated after the unblinding.

Finally, investigators opposing the inclusion of a sham control group believe it actually compromises the ecological validity of the experiment. They say that a placebo effect is an inherent part of the treatment – if the therapy is approved, future patients will know they received it, and may gain some benefit from the ensuing dopamine release (Perry Cohen, quoted in [5]). That physiological change is still a part of the treatment.

Clinical investigators are not alone in their concerns about the use of sham controls. Many Phase I participants are frustrated that seemingly effective treatments (even as far out as 10 years) failed in Phase II and were subsequently abandoned (Peggy Willocks, quoted in [5]). Investigators such as Cohen agree and some of those treatments (neurturin and GDNF) are now being reconsidered [5].

To Sham or Not to Sham?
I love a good, tidy control group as much as the next scientist, but after reading Katsnelson’s article, I think investigators should reconsider the use of sham controls, at least for the time being. Sham controls are included to produce the highest quality science possible, but their inclusion seems to be hindering the development of some truly promising treatments.

It seems that further refinement of the therapies and protocols should be done during Phase I trials to reduce variation in the Phase II studies. And once Phase II begins, a sham control may not be a prerequisite for a rigorous clinical study. Investigators in Europe and researchers in other areas of medical research (rehabilitation, for example) publish effective studies without the use of sham controls. They control variables in other ways: having blinded raters evaluate patient progress and using control groups of patients receiving ‘usual and customary care’ (established treatments such as deep brain stimulation that are already in use). Finally, a control group may not even be the best way to establish efficacy – long-term (3-5 years) studies may be better. Most studies end after a year, quite possibly before the therapy has had maximal impact.

It is important to consider not only the ethical imperative to conduct quality research, but also researchers’ ethical responsibilities to patients. Clinical investigators are not just scientists – they are also clinicians and should respect patients’ wishes (which are often wishes for careful consideration of potential therapies). Investigators must recognize the fact that using sham controls to eliminate false positives may have inadvertently generated false negatives and the discarding of effective therapies (to the disappointment of many patients).

I recognize that my opinion on the sham controls is certainly not the only one. Other people can read the same facts and draw a very different conclusion. But as long as researchers consider their ethical responsibilities to both quality science and their patient population, I feel they can conduct an ethical investigation.

--Kim Lang
 Neuroscience Graduate Program

Want to cite this post?
Lang, K. (2011). In Support of Sham Surgery. The Neuroethics Blog. Retrieved on , from

[2] Gross RE, Watts RL, Hauser RA, Bakay RA, Reichmann H, von Kummer R, Ondo WG, Reissig E, Eisner W, Steiner-Schulze H, Siedentop H, Fichte K, Hong W, Cornfeldt M, Beebe K, Sandbrink R; Spheramine Investigational Group. (2011) Intrastriatal transplantation of microcarrier-bound human retinal pigment epithelial cells versus sham surgery in patients with advanced Parkinson's disease: a double-blind, randomised, controlled trial. Lancet Neurol. 10(6):509-19.
[2] Marks WJ Jr, Bartus RT, Siffert J, Davis CS, Lozano A, Boulis N, Vitek J, Stacy M, Turner D, Verhagen L, Bakay R, Watts R, Guthrie B, Jankovic J, Simpson R, Tagliati M, Alterman R, Stern M, Baltuch G, Starr PA, Larson PS, Ostrem JL, Nutt J, Kieburtz K, Kordower JH, Olanow CW (2010) Gene delivery of AAV2-neurturin for Parkinson's disease: a double-blind, randomised, controlled trial. Lancet Neurol. 9(12):1164-72.
[3] Lang AE, Gill S, Patel NK, Lozano A, Nutt JG, Penn R, Brooks DJ, Hotton G, Moro E, Heywood P, Brodsky MA, Burchiel K, Kelly P, Dalvi A, Scott B, Stacy M, Turner D, Wooten VG, Elias WJ, Laws ER, Dhawan V, Stoessl AJ, Matcham J, Coffey RJ, Traub M (2006) Randomized controlled trial of intraputamenal glial cell line-derived neurotrophic factor infusion in Parkinson disease. Ann Neurol. 59(3):459-66.
[4] LeWitt PA, Rezai AR, Leehey MA, Ojemann SG, Flaherty AW, Eskandar EN, Kostyk SK, Thomas K, Sarkar A, Siddiqui MS, Tatter SB, Schwalb JM, Poston KL, Henderson JM, Kurlan RM, Richard IH, Van Meter L, Sapan CV, During MJ, Kaplitt MG, Feigin A (2011). AAV2-GAD gene therapy for advanced Parkinson's disease: a double-blind, sham-surgery controlled, randomised trial. Lancet Neurol. 10(4):309-19.
[5] Katnelson A (2011) Why Fake It?: How ‘Sham’ Brain Surgery Could Be Killing Off Valuable Therapies for Parkinson’s Disease. Nature 476:142-144.
[6] de la Fuente-Fernández R, Ruth TJ, Sossi V, Schulzer M, Calne DB, Stoessl AJ (2001) Expectation and dopamine release: mechanism of the placebo effect in Parkinson's disease. Science 293(5532):1164-6
[7] Kim SY, Frank S, Holloway R, Zimmerman C, Wilson R, Kieburtz K (2005) Science and ethics of sham surgery: a survey of Parkinson disease clinical researchers. Arch Neurol 62(9):1357-60.

No comments: