What happens when doctors rely on invalid data?

On April 15, 2009, the government announced that a 302 million dollars “qui tam” whistleblower settlement against Quest Diagnostics, Inc. The relator (whistleblower), Tom Cantor, will receive approximately $45 million dollars. Cantor is the owner of Scantibodies. This lawsuit was based upon parathyroid hormone assay kits sold to third party laboratories from 2000 through 2006 by now defunct Nichols Institute Diagnostics, Inc. Nichols was purchased and then shut down by Quest. The False Claims Act dates back to the Civil War, and permits private citizens to sue on behalf of the government and receive a reward if it can be demonstrated that the government has been defrauded.

The relator was able to successfully allege that Nichols continued to distribute these kits despite problems leading to inaccurate test results. In September 2005, Cantor published that as a consequence of a drift in parathyroid hormone (PTH) values, the dose of vitamin D analog required to suppress parathyroid hormone in ESRD patients and the number of subtotal parathyroidectomies increased.

The observations that Dr. Cantor demonstrated in Seminars in Dialysis, September, 2005 (1) are a sad example of just how frail our science can be when one relies just on a laboratory value and does not correlate it with the clinical picture. We are always taught, and continue to teach that one should look at the patient first and if the lab tests do not fit the clinical picture, repeat them. We should also rely on ranges that are assay specific and laboratory specific. What is normal in one lab may not be normal in another. This is not only true for PTH, but for many other tests – even the serum creatinine level!!

Since his article, the PTH assay has continued to be challenging. Serum biomarkers have always been unreliable as markers of bone pathology in renal disease (2, 3). Most doctors realize that PTH levels are neither accurate nor precise. PTH is cleaved to an 84 amino acid peptide within the parathyroid gland and stored in secretory granules until released. Non intact fragments are also secreted from the gland. Once released, PTH has a half life is only 2 to 4 minutes, and it is quickly fragmented further then metabolized. Bioassays cross react with these fragments, which may or may not have bioactivity. The NKF KDOQI Guidelines relied on the Nichol’s Allegro intact PTH assay and the range of 150 to 300 pg/m was best correlated with normal turnover. The allegro assay used an immunoreactive “sandwich” with two antibodies each directed against a different portion of the peptide. This assay did not specifically target the active N terminal. A third generation bio intact assay used an antibody which did target amino acids in the first 6 to 8 N terminal residues. Other assays have used a ratio between the 1-84 and 7-84 residues. The widely used 3^rd generation bio-intact PTH Advantage by Nichols Institute was ultimately withdrawn because of assay reliability. (4)The large dialysis organizations now use a variety of tests – the Bayer AdviaCentaur, DPC Immulite, Roche Elecys PTH and Diasorin LIAISON are all second generations. Ca-PTH IRMA by Scantibodies is a third generation assay. When compared with the original Allegro-intact PTH level, median bias or in plain English, assay variability, can be very large.

These assays were compared with one another in a very well written article from France, published in July, 2006 in Kidney International(5), and confirm what Dr. Cantor said several years ago – that there continues to be variation.

With the development of newer medications like cinacalcet to suppress PTH, the need for surgery has been drastically diminished, but the risk of oversuppression became a great problem.

It was shown in a correlation between bioassay and bone biopsy that at levels deemed “normal” by the original Allegro intact PTH assay 150-300 pg/ml, African Americans had over suppression of their bones, and that the cut off for this population should be higher. Bone turnover assessment may be more valid when using the ratio of PTH-(1-84)/N-terminally truncated fragments in African Americans, where “normal” PTH levels may lead to oversuppression of bone (6). Recent observational studies indicate that 300 pg/ml may be too low of a target. DOPPS data suggests that mortality risks do not increase until the PTH values are over 600 pg/ml (7).

There is a valuable lesson here: Physicians must depend upon their clinical judgment when assessing patients, and not just the laboratory. When we do so, we are in a sense validating laboratory values. As we expand beyond the patient, laboratory accuracy and precision becomes even more critical. Clinical judgment uses skills that cannot be quantitated, leaving only laboratory data for systems, analytics and trends, as the safeguards of clinical assessment are stripped away. Laboratory values are essential in analyzing trends, judging performance, assessing reimbursement, public reporting and sanctioning providers who fall outside accepted norms. If we are expected to rely so heavily upon this data, should we not assure it will be accurate?

The Nichols debacle demonstrates what happens when data validation fails. Laboratory and data validation are our only safeguards against a repeat of this disaster.

References

1. Cantor, T: Parathyroid hormone assay drift: an unappreciated problem in dialysis patient management. Semin Dial, 18:359-64, 2005.

2. Gal-Moscovici, A & Popovtzer, MM: New worldwide trends in presentation of renal osteodystrophy and its relationship to parathyroid hormone levels. Clin Nephrol, 63:284-9, 2005.

3. Malluche, HH, Langub, MC & Monier-Faugere, MC: The role of bone biopsy in clinical practice and research. Kidney Int Suppl, 73:S20-5, 1999.

4. Fadem, SZ & Moe, SM: Management of chronic kidney disease mineral-bone disorder. Adv Chronic Kidney Dis, 14:44-53, 2007.

5. Souberbielle, JC, Boutten, A, Carlier, MC, Chevenne, D, Coumaros, G, Lawson-Body, E, Massart, C, Monge, M, Myara, J, Parent, X, Plouvier, E & Houillier, P: Inter-method variability in PTH measurement: implication for the care of CKD patients. Kidney Int, 70:345-50, 2006.

6. Fehmi, H, Osman, Y, Bhat, S, Nguyen, K, Daramola, O, Cantor, T, Monier-Faugere, MC, Yee, J & Malluche, HH: Absence of adynamic bone disease in African-Americans with CKD stage 5 after 3 years of vitamin D therapy guided by iPTH and the PTH-(1-84)/N-terminally truncated PTH fragments ratio. Clin Nephrol, 71:267-75, 2009.

7. Tentori, F, Blayney, MJ, Albert, JM, Gillespie, BW, Kerr, PG, Bommer, J, Young, EW, Akizawa, T, Akiba, T, Pisoni, RL, Robinson, BM & Port, FK: Mortality risk for dialysis patients with different levels of serum calcium, phosphorus, and PTH: the Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Dis, 52:519-30, 2008.