Jump to comment:

• Re:How useful and how safe is a Datscan?
  Nin Bajaj
  Published on: 13 April 2016
• How useful and how safe is a Datscan?
  Paul Morrish
  Published on: 13 April 2016

• Published on: 13 April 2016

  Re:How useful and how safe is a Datscan?

  • Nin Bajaj, Neurologist
  • Other Contributors:
    • Robert A Hauser, Igor D Grachev

  We thank Dr Morrish for his letter and comments. We agree that in most cases of possible Parkinsonism, history and examination are sufficient to come to an accurate diagnosis, especially in expert hands. However, not all individuals have access to a Movement Disorders expert, and it is clear that there are more diagnostic errors early on or when signs and symptoms are unusual. The diagnosis does indeed typically become c...

  Show More

  We thank Dr Morrish for his letter and comments. We agree that in most cases of possible Parkinsonism, history and examination are sufficient to come to an accurate diagnosis, especially in expert hands. However, not all individuals have access to a Movement Disorders expert, and it is clear that there are more diagnostic errors early on or when signs and symptoms are unusual. The diagnosis does indeed typically become clearer over time, but in the interim, the patient may experience unnecessary disability through a delay in diagnosing Parkinson's disease (PD) or receive medications that cause adverse events with no real chance of affording benefit in a false positive diagnosis of PD. The difficulty with diagnosis of early PD has been highlighted in several recent clinical trials that included neuroimaging. In the CALM PD (pramipexole versus L dopa), REAL-PET (ropinirole versus L-dopa), and ELLDOPA (L-dopa vs placebo) studies [1-3] patients with a clinical diagnosis of early PD were enrolled by movement disorder physicians across centers in Europe and the US. All patients had functional imaging assessment at baseline (B-CIT SPECT in CALM-PD, REAL-PET, and ELLDOPA, 18F dopa in REAL-PET), after the clinical diagnosis was made and after subjects had been enrolled in the trials [1-3]. Across the 3 trials, between 4% and 15% of subjects were found to have normal imaging findings. In the ELLDOPA study of 142 subjects, 21 subjects (15%) examined by B-CIT- SPECT had normal imaging and were classified as subjects with scans without evidence of dopaminergic deficiency (SWEDDs)[3]. Marek et al followed these SWEDD patients and reported no deterioration in neurological features over time. These patients also had normal imaging results at 9 months (19/19 cases), 18 months (17/17), 36 months (12/12), and 48 months (10/10)[4]. Consensus discussion regarding these cases suggested that the original clinical diagnosis of PD was wrong[5]. A further study from Marshall et al showed an initial tendency to over- diagnose PD in uncertain parkinsonian cases at baseline: again, this was likely due to the considerable overlap in clinical features between some benign tremor cases and cases with tremulous PD[6]. These studies indicate that even the most rigorous application of the UK Brain Bank Criteria is subject to interpretation and can lead to diagnostic errors. Dr Morrish believes 6 years follow up in SWEDD cohorts is not long enough to rule out a diagnosis of PD. This might be true for the rare case of benign tremulous PD, but it is clear to those engaged in regular clinical practice, that the vast majority of PD cases will show clear progression over a few years of follow up. Furthermore, this apparent lack of clinical progression over many years would pose more of an issue for those relying on clinical diagnostic acumen alone as it would take many years of careful follow up to finally detect the clinical signs that fulfil the UK Brain Bank Criteria for PD. Whilst we agree with Dr Morrish that more work needs to be done in this area, possibly including assessment of olfaction, and response to levodopa, in a larger group of patients, current indications are that the vast majority of SWEDDs do not have PD. Although it would be ideal to know the exact false positive and false negative rate of DaTSCAN, the gold standard for diagnosis is pathology examination, and it is simply not feasible to delay evaluation of DaTSCAN in early, clinically uncertain Parkinsonism by waiting 10-30 years for pathological confirmation to emerge. Dr Morrish has further criticism over audit in DaTSCAN [7]. Although 50% participation of centres in a national audit of accuracy of read of DaTSCAN will not capture all data, it is a useful start on a rolling programme that will include more centres year on year. This programme will drive up standards of reporting and reduce error rates in the future. With regard to quantitation, beginning with its approval in the EU (2000), DaTSCAN image visual interpretation became the standard practice used by nuclear medicine physicians as an adjunct for clinical diagnosis. This is because DaTSCAN image visual interpretation results already provide a high level of sensitivity and specificity in conducted clinical trials[7]. We agree that there may be additional value for quantitative image interpretation. Soderlund et al. 2013[8] found numerical increases in inter-rater agreement (as measured by kappa scores) of three experienced readers when these readers combined visual analysis and quantitative parameters, compared to visual analysis alone. Quantification may assist in the assessment of visually borderline or difficult-to-read cases, particularly if measurements from an individual exam can be compared to measurement ranges from a normal database that allows adjustment for age- related loss of DaT receptors. Is a 1:5000 risk of cancer acceptable? Only the patient can appropriately answer this, with appropriate guidance from the clinician, and adequate pre-test counselling should always be undertaken. Alternatives to DaTSCAN include monitoring signs over time, obtaining a second opinion, and conducting medication trials. The potential pros and cons of various options should be evaluated in the context of the clinical situation. As was stressed in our article, DaTSCAN is not a replacement for clinical acumen or judgement, but provides extra data to inform that judgement. The clinician remains the best judge and arbiter of where to place the scan result in his/her final clinical view. In this regard, the use of DaTSCAN is no different from the use of surrogate imaging modalities in the work up of other complex neurodegenerative conditions, such as dementia.

  References: 1. Parkinson Study Group. Pramipexole with levodopa in early Parkinson's disease: design and methods of the CALM-PD Study. Clin Neuropharmacol 2000;23:34-44. 2. Whone AL, Watts RL, et al. Slower progression of Parkinson's disease with ropinirole versus levodopa: The REAL-PET Study. Ann Neurol 2003;54:93 -101. 3. Fahn S and Parkinson Study Group. Does levodopa slow or hasten the rate of progression of Parkinson's disease? J Neurol 2005; (Suppl 4):252:iv37- iv42. 4. Marek K, Jennings D, et al. Long-term follow-up of patients with scans without evidence of dopaminergic deficit (SWEDD) in the ELLDOPA study. Neurology 2005;64(Suppl 1): A274(Abstract). 5. Seibyl J, Jennings D, et al. The role of neuroimaging in the early diagnosis and evaluation of Parkinson's disease. Minerva Medica 2005;96:353-64. 6. Marshall VL, Reininger CB, et al. Parkinson's disease is overdiagnosed clincally at baseline in diagnostically uncertain cases: a 3-year European multicenter study with repeat [123I]-FP-CIT SPECT. Mov Disord 2009;24:500- 8. 7. Nin Bajaj, Robert A Hauser, Igor D Grachev. Clinical utility of dopamine transporter single photon emission CT (DaT-SPECT) with (123I) ioflupane in diagnosis of parkinsonian syndromes. J Neurol Neurosurg Psychiatry 2013; 84: 1288-1295. 8. Soderlund TA, Dickson JC, Prvulovich E, Ben-Haim S, Kemp P, Booij J, et al. Value of semiquantitative analysis for clinical reporting of 123I-2-B- carbomethoxy-3B-(4-iodophenyl)-N-(3-fluoropropyl)nortropane SPECT studies. J Nucl Med. Published online March 14, 2013. doi:10.2967/jnumed.112.110106.

  Conflict of Interest:

  NB has received honoraria and a previous grant from GE Healthcare

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top
• Published on: 13 April 2016

  How useful and how safe is a Datscan?

  • Paul Morrish, Physician

  In 2009 GE Healthcare revealed that 219,000 people had undergone the Datscan test [1]; by now it must be many more. The review article [2] by one employee of the company and two GE Healthcare sponsored clinicians considers the utility of the test. They begin by disparaging experienced clinical assessment. It is unarguable that expertise at a specialised centre doesn't extend to general practice (they cite a study from...

  Show More

  In 2009 GE Healthcare revealed that 219,000 people had undergone the Datscan test [1]; by now it must be many more. The review article [2] by one employee of the company and two GE Healthcare sponsored clinicians considers the utility of the test. They begin by disparaging experienced clinical assessment. It is unarguable that expertise at a specialised centre doesn't extend to general practice (they cite a study from North Wales in 1999 that found that 47% of patients diagnosed did not fulfil Parkinson's Disease Brain Bank Criteria (PDBBC)} but this is a comment on education and service delivery, not the accuracy of experienced clinical diagnosis. They cite a study in which a full clinical assessment, with history and examination, was replaced by a videotape of patients with tremor. They go on to damn clinical assessment with faint praise, and I quote, "in practice clinical diagnosis is sufficient for many patients with advanced and typical manifestations of PD", thus ignoring the many years of clinical experience that tell us that in practice clinical diagnosis at symptom onset is usually correct [3] and when it isn't, then time or a second opinion often help. Patients may be happy to wait, and to allow the clinical picture to become clearer, before they decide to take medication.

  The authors then consider why between 4 and 15% of PD patients fulfilling PDBBC have a normal Datscan. I repeatedly explained to the manufacturers (then Amersham International) and many of their sponsored clinicians many years ago that, based on experience with 18Fdopa PET scanning, I was concerned that the technique showed insufficient separation between the range of values in the healthy population and that in the de-novo PD population; one could however establish the rate of false negatives in a de-novo PD population by comparing mean and standard deviation in this group with mean and standard deviation of an age-matched healthy population, the larger the sample sizes the better. They didn't listen and that study has never been carried out. In fairness to them, regulations wisely prevent the exposure of too many healthy people to radio-pharmaceuticals. My estimate in 2005, based on the limited data then available, was that the majority of the so-called SWEDD's would be false negatives [4]. The quoted long-term follow-up of SWEDD's in the ELLDOPA study (that has appeared in abstract only) was for a maximum of six years. The heterogeneity of progression in Parkinsonism (in Hoehn and Yahr's study [5] the average duration of life after diagnosis varied between one and thirty three years) means that short term follow-up has little value. Until every SWEDD has been followed up for many years, ideally to post-mortem, then the authors shouldn't imply that SWEDD's are likely to represent misdiagnoses. But the de-novo population is one in which around 90% will, by clinical assessment, have PD. In a population in which only 50% have parkinsonism as, for example, when separating PD from ET then, using the same data sources as above, Bayes theorem tells that the chance of an erroneous diagnosis may be as high as 50%.

  There are other issues to consider. The quoted national Datscan audit showed that one in twenty five Datscans were misreported in the 50% of centres that took part in the audit; we don't know the rate of misreporting in the other centres. A further problem for what is essentially an in-vivo biochemical measurement is that few human studies have been carried out with commonly used medications (for example anti- psychotics, anti-hypertensives or anti-depressants) that could affect scan results. With regard to quantitation the authors state that "the routine clinical use of quantitation...... cannot be recommended for clinical practice without a visual read". The picture we see is a colour scale interpretation of the numbers generated by the scan's detectors and computer; if quantitation can't be recommended, how can the authors recommend assessing the resulting images by eye?

  Eventually, after undermining clinical assessment and hinting loudly that the visually normal scans in patients presenting as clinical PD are explained by clinicians' diagnostic errors, the authors acknowledge that the true accuracy of the test is unknown.

  For a test to have clinical utility in helping us confidently tell a patient whether they have PD or not, we need to know the false negative and false positive rate in de-novo patients, the rate of error in image reporting in all centres, and the effect of all medications that a patient might be taking. That it has been used so extensively without this information represents a triumph of persuasive marketing over clinicians' common sense. That so many people have been exposed to radiation (one quoted author [2] has put the risk of cancer at 1 in 5000) for a test with such limitations is a worry. Instead of the 100% reliable, very safe and inexpensive test that we and our patients might want, we have false negatives (number unknown), misreading of scan images (4% in 50% of centres), possible confounding effects of medication (unknown), expense (around 800 UK pounds or 1500 US dollars per scan) and a significant risk of cancer.

  When given these facts and figures, patients will surely prefer to trust a clinician who is aware of their own limitations than a scan that has more limitations than is publicly acknowledged.

  References: 1. http://www.fda.gov/downloads/AdvisoryCommittees/CommitteesMeetingMaterials/Drugs/PeripheralandCentralNervousSystemDrugsAdvisoryCommittee/UCM186404.pdf (accessed February 2014) 2. Bajaj N, Hauser RA and Grachev ID. Clinical utility of dopamine transporter single photon emission CT (DaT-SPECT) with (123I) ioflupane in diagnosis of parkinsonian syndromes. J Neurol Neurosurg Psychiatry 2013; 84:1288-1295 3. De la Fuente-Feernandez R. Role of DaTSCAN and clinical diagnosis in Parkinson's disease. Neurology 2012; 78:696-701 4. Morrish PK. Controversies in Neuroimaging. 2008 Chapter 28 in Parkinson's disease; Diagnosis and Clinical Management. Demos Publishing (New York). Eds Factor and Weiner 5. Hoehn MM, Yahr MD. Parkinsonism: onset, progression, and mortality. Neurology 1967;17:427-42.

  Conflict of Interest:

  In 2001 I was interviewed for a job with what was then Amersham international. I used the interview to tell the company of my concerns about the test (based on my experience with 18Fdopa PET scanning). Not surprisingly I decided that I didn't want the job and they decided that they didn't want to employ me.

  Show Less

  Conflict of Interest:
  None declared.

  • Back to top