- Email: [email protected]
Hemolysis falsely decreases intraoperative parathyroid hormone levels
The American Journal of Surgery (2009) 197, 222–226
Clinical Surgery-American
Hemolysis falsely decreases intraoperative parathyroid hormone levels Jacob Moalem, M.D.*, Daniel T Ruan, M.D., Rachel L. Farkas, M.D., Wen T. Shen, M.D., Jessica E. Gosnell, M.D., Steve Miller, M.D., Quan-Yang Duh, M.D., Orlo H. Clark, M.D., Electron Kebebew, M.D. University of California, San Francisco, CA, USA KEYWORDS: False-negative results; False-positive results; Failed parathyroidectomy; Intraoperative parathyroid hormone; Parathyroidectomy; Parathyroid hormone
Abstract BACKGROUND: Intraoperative parathyroid hormone (IOPTH) measurement is used to confirm biochemical cure during parathyroidectomy. Falsely decreased IOPTH measurements could result in false-negative or false-positive results and lead to failed parathyroidectomy or unnecessary additional exploration. STUDY DESIGN: The records of all patients who underwent parathyroidectomy with IOPTH between May and August 2007 were retrospectively reviewed, and the frequency of hemolysis of IOPTH samples was determined. Separately, 10 split-samples were hemolyzed using the freeze-thaw technique. RESULTS: Forty-seven patients underwent parathyroidectomy, and 226 IOPTH samples were sent. Seventeen (7.5%) specimens from 9 (18.8%) patients were hemolyzed. In 8 split-samples, the range of decrease caused by hemolysis was 24.5% to 53.8% compared with nonhemolyzed controls. CONCLUSIONS: Hemolysis of IOPTH samples occurs commonly and falsely decreases IOPTH levels. Unrecognized hemolysis in pre-excision specimens could result in false-negative IOPTH results and lead to unnecessary continued exploration. Unrecognized hemolysis in postexcision specimens could lead to false-positive IOPTH results and lead to failed parathyroidectomy and the need for reoperation. Thus, hemolysis may be an easily preventable cause of erroneous IOPTH measurements. © 2009 Elsevier Inc. All rights reserved.
Primary hyperparathyroidism, the most common indication for parathyroidectomy, has an estimated annual incidence of 1/100,000 in the United States.1 It results from inappropriate and autonomous secretion of parathyroid hormone (PTH). In 80% of cases, primary hyperparathyroidism may be caused by a single parathyroid adenoma. The remaining 20% cases are caused by multigland disease, approximately 75% of which involve parathyroid hyperplasia and 25% of which involve 2 hyperfunctioning glands (ie, * Corresponding author. Tel.: ⫹1-585-276-4633; fax: ⫹1-585-2731251. E-mail address: [email protected] Manuscript received May 21, 2008; revised manuscript July 7, 2008
0002-9610/$ - see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2008.07.020
double adenoma).2 Parathyroid carcinoma is rare and accounts for ⬍1% of the reported cases of primary hyperparathyroidism.2 Effective surgical treatment of hyperparathyroidism requires resection of hyperfunctioning parathyroid tissue(s). Traditionally, this is accomplished by bilateral exploration and inspection of all 4 glands, with resection of all abnormal-appearing parathyroid tissue. When diffuse hyperplasia is identified, the equivalent of 1 normal parathyroid gland is left after subtotal resection. This approach is associated with a ⬎95% success rate in many centers,3–7 although considerably lower success rates have been reported.8 The use of the intraoperative “quick PTH” (intraoperative PTH [IOPTH]), first described in 1988 by Nussbaum et
J. Moalem et al.
Hemolysis falsely decreases parathormone levels
al9 and later popularized by Irvin et al in 1993,10 contributed to the widespread use of focused (or “limited”) parathyroidectomy. Although some surgeons have reported excellent results without using IOPTH,6,11–14 many, but certainly not all, surgeons consider it to be either helpful or critical for a successful focused parathyroidectomy.15–17 Several criteria exist for IOPTH to confirm that all hyperfunctioning parathyroid tissue has been excised. These criteria were recently compared in a retrospective study of 352 patients who underwent parathyroidectomy for primary hyperparathyroidism. In that study,18 a ⬎50% decrease from preincision IOPTH at 10 minutes after excision19,20 was the most reliable criterion for predicting long-term biochemical cure. All published criteria, however, rely on the percentage change or on the absolute values of several IOPTH measurements to confirm successful parathyroidectomy, and assume that all measurements are accurate and reproducible. In April 2007, Roche Diagnostics released a memorandum about falsely decreased IOPTH results in hemolyzed specimens when analyzed by the Elecsys 1010 (Roche Diagnostics, Indianapolis, Indiana) PTH assay (unpublished data). Although according to this warning the false decrease was only apparent for specimens with a true PTH value of ⱕ100 ng/L, the majority of postexcision specimens fall within this range in the investigators’ practice. Because they use the Elecsys 1010 PTH assay, for quality assurance they conducted a study to validate the claim that hemolysis falsely decreases IOPTH levels and to verify that this effect is only seen with PTH values ⬍100 pg/mL.
223
not normocalcemic, 1 had no glands resected at the time of her surgery, and subsequently she underwent sternotomy for resection of a deep mediastinal adenoma. One patient with familial hyperparathyroidism remains hypercalcemic after bilateral exploration. Neither of his lower glands could be resected at the time of initial surgery and subsequently have been localized to the mediastinum. He awaits reoperation at this time. One patient (with MEN 1; 3.5 glands removed) had temporary hypocalcemia and is recovering.
IOPTH measurement
Patients and Methods
Two pre-excision IOPTH blood samples were collected, the first at time zero (before dissection of the parathyroid gland) and a second immediately before clamping the parathyroid adenoma’s vascular pedicle. Post-excision IOPTH samples were collected at 5 minutes and 10 minutes after the parathyroid adenoma was removed. Additional samples were collected if multiple glands were to be resected or if they were deemed helpful by the operating surgeon. According to attending surgeon preference, samples were drawn either by direct venipuncture from the ipsilateral internal jugular vein or from an indwelling large-bore intravenous catheter in the patient’s foot (the first 10 mL of blood were discarded). The difference in frequency of hemolysis between the 2 blood collection methods was tested using chi-square test. Specimens were immediately transferred to test tubes containing 7.2 mg ethylenediaminetetraacetic acid [EDTA], stored on ice, and sent to the laboratory en masse after the final IOPTH specimen was collected. IOPTH measurement results were generally not relied on for intraoperative decision making if the patient had 2 concordant localizing studies (neck ultrasound and sestamibi scan).21
Patient selection
Specimen processing
All IOPTH results between March 1, 2007, and August 8, 2007, were downloaded directly from the Elecsys1010 PTH assay machine. During the study period, 47 parathyroidectomies were performed at the University of California San Francisco/Mount Zion Medical Center, of which 40 (85%) were for primary sporadic hyperparathyroidism and 7 (15%) were for familial disease (2 with multiple endocrine neoplasia [MEN]: 1 with MEN 1 and 1 with MEN 2). Eight patients (16%) had recurrent or persistent disease at presentation. Thirty-two patients (68%) had a single gland resected; 11 (25%) had 2 glands removed; and 3 (6%) had a subtotal (3.5) gland resection for parathyroid hyperplasia or tertiary hyperparathyroidism. One patient had no glands found in the neck. The patients had a mean age of 61 years (SD 13.4) and 38 (81%) were women. Two patients who underwent surgery by a surgeon not in this group were excluded from the analysis. All but 3 patients were normocalcemic at their most recent follow-up appointment, and no injuries to the recurrent laryngeal nerve occurred. Of the 3 patients who were
EDTA whole-blood samples were centrifuged, and the plasma was visually inspected for evidence of hemolysis (pink or red discoloration, which begins to be apparent at .1 to .2 g/dL).22 No quantitative measure of hemolysis was performed. Specimens were subsequently assayed using the Elecsys 1010, a 1-step electrochemiluminescence assay that uses 2 monoclonal antibodies to detect PTH. The first is directed against the N terminus and detects amino acids 26 to 32, but it also requires codetection of amino acids 1 to 8. The second antibody is directed against the C terminus and detects amino acids 55 to 64.
Hemolysis of IOPTH samples To assess the quantitative impact of hemolysis on PTH measurements, 10 blood samples from patients who were not part of this study (7 patients with primary hyperparathyroidism and 3 patients without hyperparathyroidism) were split. The first portion was assayed for PTH as described, and the second half (whole blood) was frozen at
224 Table 1 samples
The American Journal of Surgery, Vol 197, No 2, February 2009 PTH change and extent of hemolysis of 10 split-
Specimen no.
Baseline PTH (pg/mL)
Heme PTH (pg/mL)
Decrease (%)
Free Hg (g/dL)
1 2 3a 4a 5 6 7 8 9 10 Mean SD
114.1 35.0 181.0 240.0 22.0 17.0 182.0 152.0 343.0 115.0 122.5 108.3
86.1 24.8 50.2 97.2 14.6 7.8 87.3 98.7 193.5 55.5 71.0 60.7
24.53 29.03 72.29 59.51 33.82 53.88 52.02 35.07 43.59 51.78 40.5 .1
.1 .1 17.9 14.2 .2 .1 .7 .2 .2 .2 .2 .2
a Sample nos. 3 and 4 were excluded because of nonphysiologic degrees of hemolysis.
⫺20°C overnight to cause lysis of red blood cells. This sample was then thawed and centrifuged, and the plasma fraction was assayed as previously described. Paired Student t tests were used to determine whether hemolysis had a significant impact. Free hemoglobin was measured in all specimens (hemolyzed and controls) as an index of hemolysis. To assess whether PTH degrades when plasma is frozen overnight, plasma from 10 additional patients who were not in this study was split. One portion was immediately assayed for PTH, and the other was assayed for PTH after being frozen at ⫺20°C for 24 to 48 hours. Paired Student t tests were used to compare these 2 groups. The precision of Elecsys 1010 was determined by serial (⫻10) assays of 3 blood samples with different PTH values (52.9, 172.7, and 705.0 pg/mL). Means, SDs, and coefficients of variation were then determined. All patients agreed to have their medical records reviewed as part of this study, and the study was approved by the Committee on Human Research at the University of California San Francisco. None of the investigators have any financial interest in the product investigated.
(43%) of these had values ⬍100. Free hemoglobin measurements were not obtained from these intraoperatively collected specimens. Of the 73 samples obtained by direct aspiration from ipsilateral internal jugular veins, 9 (12.3%) were hemolyzed; of the 153 samples drawn from patients’ feet, 8 (5.2%) were hemolyzed (P ⫽ .10).
Effect of hemolysis Ten split-samples had baseline PTH levels between 17 and 343 pg/mL (Table 1). After these 10 specimens were split and half of each were frozen to artificially induce hemolysis, it was determined that 8 of the hemolyzed samples had a free hemoglobin level within the range that is ordinarily found from blood draws (minimal or mild hemolysis, ie, ⬍1,000 mg/dL). These samples were included in the analysis (mean .23, range .14 to .66, SD .18 g/dL); 2 specimens that were hemolyzed to a great extent (free hemoglobin 17.2 and 14.9 g/dL) were excluded. PTH values in the hemolyzed samples were decreased by 24.5% to 53.8% (mean 40%, median 39%, SD 11%) compared with their nonhemolyzed controls (P ⫽ .022, Fig. 1). The observed percent decrease in IOPTH was not significantly associated with the extent of hemolysis (R2 ⫽ .03) or the baseline PTH level (R2 ⫽ .13). Baseline PTH values (⬎ or ⬍100 pg/dL) did not affect the percentage decrease caused by hemolysis (P ⫽ .79). When pure plasma (no cells) was frozen, PTH decreased by 12.3% at 24 hours and by 16.8% at 48 hours (P ⫽ .014 and .002, respectively). The magnitude of this decrease was significantly less than the observed decrease caused by hemolysis (P ⱕ.0001). Coefficient of variation varied directly with baseline PTH: it was 1.5% at PTH of 52.9, increased to 1.6% at a PTH of 172, and increased again to 2.5% at a PTH of 705 pg/mL.
Comments While some investigators have suggested that focused parathyroidectomy with IOPTH should be adopted as the new
Results IOPTH data Two hundred twenty-six IOPTH measurements were obtained for 47 patients (mean 4.8 specimens/patient). Slightly more than half (58%) of the 130 specimens had IOPTH levels ⬍100 pg/mL and were therefore susceptible to the effect of hemolysis as reported by the manufacturer. IOPTH results ranged from 10 to 740 pg/mL (median 78.5, mean 120, and SD 121). Overall, 17 of the 226 (7.5%) specimens that were collected from 9 different patients were hemolyzed. Nine
Figure 1
The impact of hemolysis on 8 IOPTH samples.
J. Moalem et al.
Hemolysis falsely decreases parathormone levels
standard of care for the treatment of hyperparathyroidism, others disagree.23 There have been a large number of investigations on the most accurate use of this adjunct6,18 –20,24 –27 and a few reports on possible causes of false-positive and falsenegative results. To the investigators’ knowledge, however, this is the first study to suggest that there might be a cause for erroneous IOPTH results that is intrinsic to the assay itself and not to the interpretation of the results. The use of IOPTH to confirm successful parathyroidectomy is predicated on the short (⬍4 minutes26) half-life of PTH. The successful removal of a hyperfunctioning, enlarged parathyroid gland or glands is confirmed by an appropriate decrease in PTH level. At least 6 different criteria for IOPTH assays have been reported to refine the use of IOPTH6,19,20,24 –27 and thereby minimize operative failures. Nevertheless, all of these criteria rely on an adequate decrease in IOPTH level. The differences among them are primarily those in (1) which measurement is used as the baseline (preincision,19,20 pre-excision of the gland,6 or the highest preexcision24,26,27) value from which the decrease in IOPTH is measured or (2) whether or not the postexcision IOPTH level is required to be normal. Because these published criteria all rely on absolute changes in IOPTH values, reproducibility and accuracy of the individual measurements is of paramount importance. If taken as a baseline, a falsely decreased pre-excision measurement could mask a sufficient decrease in IOPTH after the abnormal gland is resected, thus yielding a false-negative result. This could cause the surgeon to unnecessarily continue exploring the neck for abnormal glands that are unlikely to be found, adding unnecessary morbidity, time, and frustration to the operation. Conversely, a falsely decreased postexcision IOPTH measurement (ie, false-positive result) could erroneously satisfy criteria for successful parathyroidectomy. In patients with unsuspected multigland disease, this error could contribute to failed parathyroidectomy and necessitate re-exploration, also subjecting the patient to unnecessary morbidity. The present findings suggest that hemolysis is relatively common: During the 4-month study period, 7.5% of IOPTH specimens were hemolyzed, and 19% of patients who underwent parathyroidectomy had at least 1 IOPTH specimen that was hemolyzed. One possible criticism of this article might be that no free hemoglobin measurements were made on all of the specimens to confirm that these were truly hemolyzed and that no additional specimens were hemolyzed. Although this cannot be definitively proven given the retrospective design of the study, hemolysis with a free hemoglobin level ⬎.1 mg/dL causes plasma to appear orange and is visible.22 Another criticism is that the effect of hemolysis induced by the freeze-thaw mechanism might not be identical to the purely mechanical effect seen during handling of the IOPTH specimen. To evaluate this, the investigators are conducting a prospective study wherein split samples are being hemolyzed mechanically. The mechanism by which hemolysis causes a decrease in PTH levels is unknown.
225
The present findings also indicate that hemolysis of a specimen could cause both false-negative (if a pre-excision specimen is hemolyzed) and false-positive (if a postexcision specimen is hemolyzed) IOPTH results. With a median decrease of 39% in IOPTH measurement, the impact of hemolysis alone may be sufficient to fulfill criteria for successful parathyroidectomy if a postexcision specimen is hemolyzed. Conversely, if the hemolyzed specimen is a pre-excision specimen, the false result will always lessen the subsequent relative decrease in PTH and make it less likely that a sufficient decrease in PTH will be seen, even after successful parathyroidectomy. Therefore, hemolysis may be an unappreciated cause of failed parathyroidectomy or of unnecessary conversion of a focused parathyroidectomy to a traditional bilateral exploration. Although the frequency of hemolysis was higher when specimens were drawn directly from the internal jugular vein, this trend was not statistically significant. These findings also contradict previously published results that demonstrate a higher hemolysis rate when blood draws are obtained from indwelling catheters as opposed to direct venipuncture.28,29 Hemolysis could occur at any time during the processing of the sample: at the time of collection, at the time of transfer to a test tube, during storage, or when the sample is introduced to the assay machine.22,30 The present study was not designed to discern where the most common source of hemolysis might occur. In summary, although parathyroidectomy, with or without the use of IOPTH, is a highly successful operation, the use of IOPTH is increasingly common and relied on as one of the factors used to guide the extent of cervical explorations. In this study, hemolysis was found in 7.5% of 17 IOPTH specimens from 9 patients and erroneously decreased the IOPTH levels by a mean of 40%. Furthermore, the extent of decrease in PTH level was independent of the initial PTH value or the extent of hemolysis. The investigators recommend that IOPTH specimens be handled with care to avoid excessive pressurization during aspiration and transfer between containers and tubes. Moreover, the investigators advise that specimens be screened visually, at the very least, for the presence of hemolysis, and if present, that repeat draws be performed. This simple added step adds no time to the processing of the specimen and bypasses this avoidable source of error. If objective measures are desired, free hemoglobin measurements could be made at the same time as IOPTH measurements. Most importantly, the investigators recommend that IOPTH results of hemolyzed specimens should be disregarded and not relied on for intraoperative decision making.
Acknowledgments This study was supported in part by the Mount Zion Health Fund, the Hellen and Sanford Diller Family Foundation, the Gerold Diller Family Foundation, and the
226 Friends of Endocrine Surgery. We would also like to thank Pamela Derish for editorial review of the manuscript.
References 1. Kebebew E, Clark OH. Parathyroid adenoma, hyperplasia, and carcinoma: Localization, technical details of primary neck exploration, and treatment of hypercalcemic crisis. Surg Oncol Clin North Am 1998; 7:721– 48. 2. Clark OH. Current management of patients with hyperparathyroidism. Adv Surg 1996;30:179 – 87. 3. Clark OH, Duh QY. Primary hyperparathyroidism. A surgical perspective. Endocrinol Metab Clin North Am 1989;18:701–14. 4. Grant CS, Thompson G, Farley D, et al. Primary hyperparathyroidism surgical management since the introduction of minimally invasive parathyroidectomy: Mayo Clinic experience. Arch Surg 2005;140: 472–9. 5. Clark OH, Duh Q-Y, Kebebew E. Textbook of Endocrine Surgery, 2nd ed. Philadelphia, PA: W. B. Saunders; 2005. 6. Weber CJ, Ritchie JC. Retrospective analysis of sequential changes in serum intact parathyroid hormone levels during conventional parathyroid exploration. Surgery 1999;126:1139 – 434. 7. Irvin GL, III, Sfakianakis G, Yeung L, et al. Ambulatory parathyroidectomy for primary hyperparathyroidism. Arch Surg 1996;131:1074 – 8. 8. Malmaeus J, Granberg PO, Halvorsen J, et al. Parathyroid surgery in Scandinavia. Acta Chir Scand 1988;154:409 –13. 9. Nussbaum SR, Thompson AR, Hutcheson KA, et al. Intraoperative measurement of parathyroid hormone in the surgical management of hyperparathyroidism. Surgery 1988;104:1121–7. 10. Irvin GL, III, Dembrow VD, Prudhomme DL. Clinical usefulness of an intraoperative “quick parathyroid hormone” assay. Surgery 1993;114: 1019 –23. 11. Jacobson SR, van Heerden JA, Farley DR, et al. Focused cervical exploration for primary hyperparathyroidism without intraoperative parathyroid hormone monitoring or use of the gamma probe. World J Surg 2004;28:1127–31. 12. Sprouse LR, 2nd, Roe SM, Kaufman HJ, et al. Minimally invasive parathyroidectomy without intraoperative localization. Am Surg 2001; 67:1022–9. 13. Gil-Cardenas A, Gamino R, Reza A, et al. Is intraoperative parathyroid hormone assay mandatory for the success of targeted parathyroidectomy?. J Am Coll Surg 2007;204:286 –90. 14. Gawande AA, Monchik JM, Abbruzzese TA, et al. Reassessment of parathyroid hormone monitoring during parathyroidectomy for primary hyperparathyroidism after 2 preoperative localization studies. Arch Surg 2006;141:381– 484.
The American Journal of Surgery, Vol 197, No 2, February 2009 15. Hanif F, Coffey JC, Romics L, Jr, et al. Rapid intraoperative parathyroid hormone assay—More than just a comfort measure. World J Surg 2006;30:156 – 61. 16. Barczynski M, Konturek A, Cichon S, et al. Intraoperative parathyroid hormone assay improves outcomes of minimally invasive parathyroidectomy mainly in patients with a presumed solitary parathyroid adenoma and missing concordance of preoperative imaging. Clin Endocrinol 2007;66:878 – 85. 17. Irvin GL, III, Solorzano CC, Carneiro DM. Quick intraoperative parathyroid hormone assay: Surgical adjunct to allow limited parathyroidectomy, improve success rate, and predict outcome. World J Surg 2004;28:1287–92. 18. Chiu B, Sturgeon C, Angelos P. Which intraoperative parathyroid hormone assay criterion best predicts operative success? A study of 352 consecutive patients. Arch Surg 2006;141:483– 8. 19. Gordon LL, Snyder WH, III, Wians F, Jr, et al. The validity of quick intraoperative parathyroid hormone assay: An evaluation in 72 patients based on gross morphologic criteria. Surgery 1999;126:1030 –5. 20. Libutti SK, Alexander HR, Bartlett DL, et al. Kinetic analysis of the rapid intraoperative parathyroid hormone assay in patients during operation for hyperparathyroidism. Surgery 1999;126:1145–51. 21. Kebebew E, Hwang J, Reiff E, et al. Predictors of single-gland vs multigland parathyroid disease in primary hyperparathyroidism: A simple and accurate scoring model. Arch Surg 2006;141:777– 82. 22. Dugan L, Leech L, Speroni KG, et al. Factors affecting hemolysis rates in blood samples drawn from newly placed IV sites in the emergency department. J Emerg Nurs 2005;31:338 – 45. 23. Meurisse M, Hamoir E, Defechereux T, et al. Bilateral neck exploration under hypno sedation: A new standard of care in primary hyperparathyroidism?. Ann Surg 1999;229:401– 8. 24. Carneiro DM, Solorzano CC, Nader MC, et al. Comparison of intraoperative iPTH assay (QPTH) criteria in guiding parathyroidectomy: Which criterion is the most accurate?. Surgery 2003;134:973– 81. 25. Irvin GL, III, Molinari AS, Figueroa C, et al. Improved success rate in reoperative parathyroidectomy with intraoperative PTH assay. Ann Surg 1999;229:874 –9. 26. Yang GP, Levine S, Weigel RJ. A spike in parathyroid hormone during neck exploration may cause a false-negative intraoperative assay result. Arch Surg 2001;136:945–9. 27. Irvin GL, III, Deriso GT, III. A new, practical intraoperative parathyroid hormone assay. Am J Surg 1994;168:466 – 8. 28. Grant MS. The effect of blood drawing techniques and equipment on the hemolysis of ED laboratory blood samples. J Emerg Nurs 2003; 29:116 –21. 29. Kennedy C, Angermuller S, King R, et al. A comparison of hemolysis rates using intravenous catheters versus venipuncture tubes for obtaining blood samples. J Emerg Nurs 1996;22:566 –9. 30. Bush V. The hemolyzed specimen: Causes, effects, and reduction. BD Lab Notes 2003;13:2–5.
Clinical Surgery-American
Hemolysis falsely decreases intraoperative parathyroid hormone levels Jacob Moalem, M.D.*, Daniel T Ruan, M.D., Rachel L. Farkas, M.D., Wen T. Shen, M.D., Jessica E. Gosnell, M.D., Steve Miller, M.D., Quan-Yang Duh, M.D., Orlo H. Clark, M.D., Electron Kebebew, M.D. University of California, San Francisco, CA, USA KEYWORDS: False-negative results; False-positive results; Failed parathyroidectomy; Intraoperative parathyroid hormone; Parathyroidectomy; Parathyroid hormone
Abstract BACKGROUND: Intraoperative parathyroid hormone (IOPTH) measurement is used to confirm biochemical cure during parathyroidectomy. Falsely decreased IOPTH measurements could result in false-negative or false-positive results and lead to failed parathyroidectomy or unnecessary additional exploration. STUDY DESIGN: The records of all patients who underwent parathyroidectomy with IOPTH between May and August 2007 were retrospectively reviewed, and the frequency of hemolysis of IOPTH samples was determined. Separately, 10 split-samples were hemolyzed using the freeze-thaw technique. RESULTS: Forty-seven patients underwent parathyroidectomy, and 226 IOPTH samples were sent. Seventeen (7.5%) specimens from 9 (18.8%) patients were hemolyzed. In 8 split-samples, the range of decrease caused by hemolysis was 24.5% to 53.8% compared with nonhemolyzed controls. CONCLUSIONS: Hemolysis of IOPTH samples occurs commonly and falsely decreases IOPTH levels. Unrecognized hemolysis in pre-excision specimens could result in false-negative IOPTH results and lead to unnecessary continued exploration. Unrecognized hemolysis in postexcision specimens could lead to false-positive IOPTH results and lead to failed parathyroidectomy and the need for reoperation. Thus, hemolysis may be an easily preventable cause of erroneous IOPTH measurements. © 2009 Elsevier Inc. All rights reserved.
Primary hyperparathyroidism, the most common indication for parathyroidectomy, has an estimated annual incidence of 1/100,000 in the United States.1 It results from inappropriate and autonomous secretion of parathyroid hormone (PTH). In 80% of cases, primary hyperparathyroidism may be caused by a single parathyroid adenoma. The remaining 20% cases are caused by multigland disease, approximately 75% of which involve parathyroid hyperplasia and 25% of which involve 2 hyperfunctioning glands (ie, * Corresponding author. Tel.: ⫹1-585-276-4633; fax: ⫹1-585-2731251. E-mail address: [email protected] Manuscript received May 21, 2008; revised manuscript July 7, 2008
0002-9610/$ - see front matter © 2009 Elsevier Inc. All rights reserved. doi:10.1016/j.amjsurg.2008.07.020
double adenoma).2 Parathyroid carcinoma is rare and accounts for ⬍1% of the reported cases of primary hyperparathyroidism.2 Effective surgical treatment of hyperparathyroidism requires resection of hyperfunctioning parathyroid tissue(s). Traditionally, this is accomplished by bilateral exploration and inspection of all 4 glands, with resection of all abnormal-appearing parathyroid tissue. When diffuse hyperplasia is identified, the equivalent of 1 normal parathyroid gland is left after subtotal resection. This approach is associated with a ⬎95% success rate in many centers,3–7 although considerably lower success rates have been reported.8 The use of the intraoperative “quick PTH” (intraoperative PTH [IOPTH]), first described in 1988 by Nussbaum et
J. Moalem et al.
Hemolysis falsely decreases parathormone levels
al9 and later popularized by Irvin et al in 1993,10 contributed to the widespread use of focused (or “limited”) parathyroidectomy. Although some surgeons have reported excellent results without using IOPTH,6,11–14 many, but certainly not all, surgeons consider it to be either helpful or critical for a successful focused parathyroidectomy.15–17 Several criteria exist for IOPTH to confirm that all hyperfunctioning parathyroid tissue has been excised. These criteria were recently compared in a retrospective study of 352 patients who underwent parathyroidectomy for primary hyperparathyroidism. In that study,18 a ⬎50% decrease from preincision IOPTH at 10 minutes after excision19,20 was the most reliable criterion for predicting long-term biochemical cure. All published criteria, however, rely on the percentage change or on the absolute values of several IOPTH measurements to confirm successful parathyroidectomy, and assume that all measurements are accurate and reproducible. In April 2007, Roche Diagnostics released a memorandum about falsely decreased IOPTH results in hemolyzed specimens when analyzed by the Elecsys 1010 (Roche Diagnostics, Indianapolis, Indiana) PTH assay (unpublished data). Although according to this warning the false decrease was only apparent for specimens with a true PTH value of ⱕ100 ng/L, the majority of postexcision specimens fall within this range in the investigators’ practice. Because they use the Elecsys 1010 PTH assay, for quality assurance they conducted a study to validate the claim that hemolysis falsely decreases IOPTH levels and to verify that this effect is only seen with PTH values ⬍100 pg/mL.
223
not normocalcemic, 1 had no glands resected at the time of her surgery, and subsequently she underwent sternotomy for resection of a deep mediastinal adenoma. One patient with familial hyperparathyroidism remains hypercalcemic after bilateral exploration. Neither of his lower glands could be resected at the time of initial surgery and subsequently have been localized to the mediastinum. He awaits reoperation at this time. One patient (with MEN 1; 3.5 glands removed) had temporary hypocalcemia and is recovering.
IOPTH measurement
Patients and Methods
Two pre-excision IOPTH blood samples were collected, the first at time zero (before dissection of the parathyroid gland) and a second immediately before clamping the parathyroid adenoma’s vascular pedicle. Post-excision IOPTH samples were collected at 5 minutes and 10 minutes after the parathyroid adenoma was removed. Additional samples were collected if multiple glands were to be resected or if they were deemed helpful by the operating surgeon. According to attending surgeon preference, samples were drawn either by direct venipuncture from the ipsilateral internal jugular vein or from an indwelling large-bore intravenous catheter in the patient’s foot (the first 10 mL of blood were discarded). The difference in frequency of hemolysis between the 2 blood collection methods was tested using chi-square test. Specimens were immediately transferred to test tubes containing 7.2 mg ethylenediaminetetraacetic acid [EDTA], stored on ice, and sent to the laboratory en masse after the final IOPTH specimen was collected. IOPTH measurement results were generally not relied on for intraoperative decision making if the patient had 2 concordant localizing studies (neck ultrasound and sestamibi scan).21
Patient selection
Specimen processing
All IOPTH results between March 1, 2007, and August 8, 2007, were downloaded directly from the Elecsys1010 PTH assay machine. During the study period, 47 parathyroidectomies were performed at the University of California San Francisco/Mount Zion Medical Center, of which 40 (85%) were for primary sporadic hyperparathyroidism and 7 (15%) were for familial disease (2 with multiple endocrine neoplasia [MEN]: 1 with MEN 1 and 1 with MEN 2). Eight patients (16%) had recurrent or persistent disease at presentation. Thirty-two patients (68%) had a single gland resected; 11 (25%) had 2 glands removed; and 3 (6%) had a subtotal (3.5) gland resection for parathyroid hyperplasia or tertiary hyperparathyroidism. One patient had no glands found in the neck. The patients had a mean age of 61 years (SD 13.4) and 38 (81%) were women. Two patients who underwent surgery by a surgeon not in this group were excluded from the analysis. All but 3 patients were normocalcemic at their most recent follow-up appointment, and no injuries to the recurrent laryngeal nerve occurred. Of the 3 patients who were
EDTA whole-blood samples were centrifuged, and the plasma was visually inspected for evidence of hemolysis (pink or red discoloration, which begins to be apparent at .1 to .2 g/dL).22 No quantitative measure of hemolysis was performed. Specimens were subsequently assayed using the Elecsys 1010, a 1-step electrochemiluminescence assay that uses 2 monoclonal antibodies to detect PTH. The first is directed against the N terminus and detects amino acids 26 to 32, but it also requires codetection of amino acids 1 to 8. The second antibody is directed against the C terminus and detects amino acids 55 to 64.
Hemolysis of IOPTH samples To assess the quantitative impact of hemolysis on PTH measurements, 10 blood samples from patients who were not part of this study (7 patients with primary hyperparathyroidism and 3 patients without hyperparathyroidism) were split. The first portion was assayed for PTH as described, and the second half (whole blood) was frozen at
224 Table 1 samples
The American Journal of Surgery, Vol 197, No 2, February 2009 PTH change and extent of hemolysis of 10 split-
Specimen no.
Baseline PTH (pg/mL)
Heme PTH (pg/mL)
Decrease (%)
Free Hg (g/dL)
1 2 3a 4a 5 6 7 8 9 10 Mean SD
114.1 35.0 181.0 240.0 22.0 17.0 182.0 152.0 343.0 115.0 122.5 108.3
86.1 24.8 50.2 97.2 14.6 7.8 87.3 98.7 193.5 55.5 71.0 60.7
24.53 29.03 72.29 59.51 33.82 53.88 52.02 35.07 43.59 51.78 40.5 .1
.1 .1 17.9 14.2 .2 .1 .7 .2 .2 .2 .2 .2
a Sample nos. 3 and 4 were excluded because of nonphysiologic degrees of hemolysis.
⫺20°C overnight to cause lysis of red blood cells. This sample was then thawed and centrifuged, and the plasma fraction was assayed as previously described. Paired Student t tests were used to determine whether hemolysis had a significant impact. Free hemoglobin was measured in all specimens (hemolyzed and controls) as an index of hemolysis. To assess whether PTH degrades when plasma is frozen overnight, plasma from 10 additional patients who were not in this study was split. One portion was immediately assayed for PTH, and the other was assayed for PTH after being frozen at ⫺20°C for 24 to 48 hours. Paired Student t tests were used to compare these 2 groups. The precision of Elecsys 1010 was determined by serial (⫻10) assays of 3 blood samples with different PTH values (52.9, 172.7, and 705.0 pg/mL). Means, SDs, and coefficients of variation were then determined. All patients agreed to have their medical records reviewed as part of this study, and the study was approved by the Committee on Human Research at the University of California San Francisco. None of the investigators have any financial interest in the product investigated.
(43%) of these had values ⬍100. Free hemoglobin measurements were not obtained from these intraoperatively collected specimens. Of the 73 samples obtained by direct aspiration from ipsilateral internal jugular veins, 9 (12.3%) were hemolyzed; of the 153 samples drawn from patients’ feet, 8 (5.2%) were hemolyzed (P ⫽ .10).
Effect of hemolysis Ten split-samples had baseline PTH levels between 17 and 343 pg/mL (Table 1). After these 10 specimens were split and half of each were frozen to artificially induce hemolysis, it was determined that 8 of the hemolyzed samples had a free hemoglobin level within the range that is ordinarily found from blood draws (minimal or mild hemolysis, ie, ⬍1,000 mg/dL). These samples were included in the analysis (mean .23, range .14 to .66, SD .18 g/dL); 2 specimens that were hemolyzed to a great extent (free hemoglobin 17.2 and 14.9 g/dL) were excluded. PTH values in the hemolyzed samples were decreased by 24.5% to 53.8% (mean 40%, median 39%, SD 11%) compared with their nonhemolyzed controls (P ⫽ .022, Fig. 1). The observed percent decrease in IOPTH was not significantly associated with the extent of hemolysis (R2 ⫽ .03) or the baseline PTH level (R2 ⫽ .13). Baseline PTH values (⬎ or ⬍100 pg/dL) did not affect the percentage decrease caused by hemolysis (P ⫽ .79). When pure plasma (no cells) was frozen, PTH decreased by 12.3% at 24 hours and by 16.8% at 48 hours (P ⫽ .014 and .002, respectively). The magnitude of this decrease was significantly less than the observed decrease caused by hemolysis (P ⱕ.0001). Coefficient of variation varied directly with baseline PTH: it was 1.5% at PTH of 52.9, increased to 1.6% at a PTH of 172, and increased again to 2.5% at a PTH of 705 pg/mL.
Comments While some investigators have suggested that focused parathyroidectomy with IOPTH should be adopted as the new
Results IOPTH data Two hundred twenty-six IOPTH measurements were obtained for 47 patients (mean 4.8 specimens/patient). Slightly more than half (58%) of the 130 specimens had IOPTH levels ⬍100 pg/mL and were therefore susceptible to the effect of hemolysis as reported by the manufacturer. IOPTH results ranged from 10 to 740 pg/mL (median 78.5, mean 120, and SD 121). Overall, 17 of the 226 (7.5%) specimens that were collected from 9 different patients were hemolyzed. Nine
Figure 1
The impact of hemolysis on 8 IOPTH samples.
J. Moalem et al.
Hemolysis falsely decreases parathormone levels
standard of care for the treatment of hyperparathyroidism, others disagree.23 There have been a large number of investigations on the most accurate use of this adjunct6,18 –20,24 –27 and a few reports on possible causes of false-positive and falsenegative results. To the investigators’ knowledge, however, this is the first study to suggest that there might be a cause for erroneous IOPTH results that is intrinsic to the assay itself and not to the interpretation of the results. The use of IOPTH to confirm successful parathyroidectomy is predicated on the short (⬍4 minutes26) half-life of PTH. The successful removal of a hyperfunctioning, enlarged parathyroid gland or glands is confirmed by an appropriate decrease in PTH level. At least 6 different criteria for IOPTH assays have been reported to refine the use of IOPTH6,19,20,24 –27 and thereby minimize operative failures. Nevertheless, all of these criteria rely on an adequate decrease in IOPTH level. The differences among them are primarily those in (1) which measurement is used as the baseline (preincision,19,20 pre-excision of the gland,6 or the highest preexcision24,26,27) value from which the decrease in IOPTH is measured or (2) whether or not the postexcision IOPTH level is required to be normal. Because these published criteria all rely on absolute changes in IOPTH values, reproducibility and accuracy of the individual measurements is of paramount importance. If taken as a baseline, a falsely decreased pre-excision measurement could mask a sufficient decrease in IOPTH after the abnormal gland is resected, thus yielding a false-negative result. This could cause the surgeon to unnecessarily continue exploring the neck for abnormal glands that are unlikely to be found, adding unnecessary morbidity, time, and frustration to the operation. Conversely, a falsely decreased postexcision IOPTH measurement (ie, false-positive result) could erroneously satisfy criteria for successful parathyroidectomy. In patients with unsuspected multigland disease, this error could contribute to failed parathyroidectomy and necessitate re-exploration, also subjecting the patient to unnecessary morbidity. The present findings suggest that hemolysis is relatively common: During the 4-month study period, 7.5% of IOPTH specimens were hemolyzed, and 19% of patients who underwent parathyroidectomy had at least 1 IOPTH specimen that was hemolyzed. One possible criticism of this article might be that no free hemoglobin measurements were made on all of the specimens to confirm that these were truly hemolyzed and that no additional specimens were hemolyzed. Although this cannot be definitively proven given the retrospective design of the study, hemolysis with a free hemoglobin level ⬎.1 mg/dL causes plasma to appear orange and is visible.22 Another criticism is that the effect of hemolysis induced by the freeze-thaw mechanism might not be identical to the purely mechanical effect seen during handling of the IOPTH specimen. To evaluate this, the investigators are conducting a prospective study wherein split samples are being hemolyzed mechanically. The mechanism by which hemolysis causes a decrease in PTH levels is unknown.
225
The present findings also indicate that hemolysis of a specimen could cause both false-negative (if a pre-excision specimen is hemolyzed) and false-positive (if a postexcision specimen is hemolyzed) IOPTH results. With a median decrease of 39% in IOPTH measurement, the impact of hemolysis alone may be sufficient to fulfill criteria for successful parathyroidectomy if a postexcision specimen is hemolyzed. Conversely, if the hemolyzed specimen is a pre-excision specimen, the false result will always lessen the subsequent relative decrease in PTH and make it less likely that a sufficient decrease in PTH will be seen, even after successful parathyroidectomy. Therefore, hemolysis may be an unappreciated cause of failed parathyroidectomy or of unnecessary conversion of a focused parathyroidectomy to a traditional bilateral exploration. Although the frequency of hemolysis was higher when specimens were drawn directly from the internal jugular vein, this trend was not statistically significant. These findings also contradict previously published results that demonstrate a higher hemolysis rate when blood draws are obtained from indwelling catheters as opposed to direct venipuncture.28,29 Hemolysis could occur at any time during the processing of the sample: at the time of collection, at the time of transfer to a test tube, during storage, or when the sample is introduced to the assay machine.22,30 The present study was not designed to discern where the most common source of hemolysis might occur. In summary, although parathyroidectomy, with or without the use of IOPTH, is a highly successful operation, the use of IOPTH is increasingly common and relied on as one of the factors used to guide the extent of cervical explorations. In this study, hemolysis was found in 7.5% of 17 IOPTH specimens from 9 patients and erroneously decreased the IOPTH levels by a mean of 40%. Furthermore, the extent of decrease in PTH level was independent of the initial PTH value or the extent of hemolysis. The investigators recommend that IOPTH specimens be handled with care to avoid excessive pressurization during aspiration and transfer between containers and tubes. Moreover, the investigators advise that specimens be screened visually, at the very least, for the presence of hemolysis, and if present, that repeat draws be performed. This simple added step adds no time to the processing of the specimen and bypasses this avoidable source of error. If objective measures are desired, free hemoglobin measurements could be made at the same time as IOPTH measurements. Most importantly, the investigators recommend that IOPTH results of hemolyzed specimens should be disregarded and not relied on for intraoperative decision making.
Acknowledgments This study was supported in part by the Mount Zion Health Fund, the Hellen and Sanford Diller Family Foundation, the Gerold Diller Family Foundation, and the
226 Friends of Endocrine Surgery. We would also like to thank Pamela Derish for editorial review of the manuscript.
References 1. Kebebew E, Clark OH. Parathyroid adenoma, hyperplasia, and carcinoma: Localization, technical details of primary neck exploration, and treatment of hypercalcemic crisis. Surg Oncol Clin North Am 1998; 7:721– 48. 2. Clark OH. Current management of patients with hyperparathyroidism. Adv Surg 1996;30:179 – 87. 3. Clark OH, Duh QY. Primary hyperparathyroidism. A surgical perspective. Endocrinol Metab Clin North Am 1989;18:701–14. 4. Grant CS, Thompson G, Farley D, et al. Primary hyperparathyroidism surgical management since the introduction of minimally invasive parathyroidectomy: Mayo Clinic experience. Arch Surg 2005;140: 472–9. 5. Clark OH, Duh Q-Y, Kebebew E. Textbook of Endocrine Surgery, 2nd ed. Philadelphia, PA: W. B. Saunders; 2005. 6. Weber CJ, Ritchie JC. Retrospective analysis of sequential changes in serum intact parathyroid hormone levels during conventional parathyroid exploration. Surgery 1999;126:1139 – 434. 7. Irvin GL, III, Sfakianakis G, Yeung L, et al. Ambulatory parathyroidectomy for primary hyperparathyroidism. Arch Surg 1996;131:1074 – 8. 8. Malmaeus J, Granberg PO, Halvorsen J, et al. Parathyroid surgery in Scandinavia. Acta Chir Scand 1988;154:409 –13. 9. Nussbaum SR, Thompson AR, Hutcheson KA, et al. Intraoperative measurement of parathyroid hormone in the surgical management of hyperparathyroidism. Surgery 1988;104:1121–7. 10. Irvin GL, III, Dembrow VD, Prudhomme DL. Clinical usefulness of an intraoperative “quick parathyroid hormone” assay. Surgery 1993;114: 1019 –23. 11. Jacobson SR, van Heerden JA, Farley DR, et al. Focused cervical exploration for primary hyperparathyroidism without intraoperative parathyroid hormone monitoring or use of the gamma probe. World J Surg 2004;28:1127–31. 12. Sprouse LR, 2nd, Roe SM, Kaufman HJ, et al. Minimally invasive parathyroidectomy without intraoperative localization. Am Surg 2001; 67:1022–9. 13. Gil-Cardenas A, Gamino R, Reza A, et al. Is intraoperative parathyroid hormone assay mandatory for the success of targeted parathyroidectomy?. J Am Coll Surg 2007;204:286 –90. 14. Gawande AA, Monchik JM, Abbruzzese TA, et al. Reassessment of parathyroid hormone monitoring during parathyroidectomy for primary hyperparathyroidism after 2 preoperative localization studies. Arch Surg 2006;141:381– 484.
The American Journal of Surgery, Vol 197, No 2, February 2009 15. Hanif F, Coffey JC, Romics L, Jr, et al. Rapid intraoperative parathyroid hormone assay—More than just a comfort measure. World J Surg 2006;30:156 – 61. 16. Barczynski M, Konturek A, Cichon S, et al. Intraoperative parathyroid hormone assay improves outcomes of minimally invasive parathyroidectomy mainly in patients with a presumed solitary parathyroid adenoma and missing concordance of preoperative imaging. Clin Endocrinol 2007;66:878 – 85. 17. Irvin GL, III, Solorzano CC, Carneiro DM. Quick intraoperative parathyroid hormone assay: Surgical adjunct to allow limited parathyroidectomy, improve success rate, and predict outcome. World J Surg 2004;28:1287–92. 18. Chiu B, Sturgeon C, Angelos P. Which intraoperative parathyroid hormone assay criterion best predicts operative success? A study of 352 consecutive patients. Arch Surg 2006;141:483– 8. 19. Gordon LL, Snyder WH, III, Wians F, Jr, et al. The validity of quick intraoperative parathyroid hormone assay: An evaluation in 72 patients based on gross morphologic criteria. Surgery 1999;126:1030 –5. 20. Libutti SK, Alexander HR, Bartlett DL, et al. Kinetic analysis of the rapid intraoperative parathyroid hormone assay in patients during operation for hyperparathyroidism. Surgery 1999;126:1145–51. 21. Kebebew E, Hwang J, Reiff E, et al. Predictors of single-gland vs multigland parathyroid disease in primary hyperparathyroidism: A simple and accurate scoring model. Arch Surg 2006;141:777– 82. 22. Dugan L, Leech L, Speroni KG, et al. Factors affecting hemolysis rates in blood samples drawn from newly placed IV sites in the emergency department. J Emerg Nurs 2005;31:338 – 45. 23. Meurisse M, Hamoir E, Defechereux T, et al. Bilateral neck exploration under hypno sedation: A new standard of care in primary hyperparathyroidism?. Ann Surg 1999;229:401– 8. 24. Carneiro DM, Solorzano CC, Nader MC, et al. Comparison of intraoperative iPTH assay (QPTH) criteria in guiding parathyroidectomy: Which criterion is the most accurate?. Surgery 2003;134:973– 81. 25. Irvin GL, III, Molinari AS, Figueroa C, et al. Improved success rate in reoperative parathyroidectomy with intraoperative PTH assay. Ann Surg 1999;229:874 –9. 26. Yang GP, Levine S, Weigel RJ. A spike in parathyroid hormone during neck exploration may cause a false-negative intraoperative assay result. Arch Surg 2001;136:945–9. 27. Irvin GL, III, Deriso GT, III. A new, practical intraoperative parathyroid hormone assay. Am J Surg 1994;168:466 – 8. 28. Grant MS. The effect of blood drawing techniques and equipment on the hemolysis of ED laboratory blood samples. J Emerg Nurs 2003; 29:116 –21. 29. Kennedy C, Angermuller S, King R, et al. A comparison of hemolysis rates using intravenous catheters versus venipuncture tubes for obtaining blood samples. J Emerg Nurs 1996;22:566 –9. 30. Bush V. The hemolyzed specimen: Causes, effects, and reduction. BD Lab Notes 2003;13:2–5.