This brief study of ESRD patients following parathyroidectomy (PTX) investigates the relationships between changes in their bone mineral densities ( BMD ) and serum chemistries. Are serial BMD measurements useful for assessing the degree of succe.l's achieved by surgical intervention? Patients at the University of Missouri Medical Center, two males; thirteen females ages 20-71 were divided into four groups;
1 Dramatic and prolongedfall in Ca! P levelsfor longer than six months, requiring Rocaltrol, and without phosphate binders (N = 7).
2 Decrease in Ca! P levels for less than six months, receiving short course of Rocaltrol and with phosphate binders shortly after surgery (N = 6).
3 No change in Ca! P levels, with PTH levels remaining elevated (N = 1 ).
4 After each of several successive surgeries, slight decreases in Ca! P ratiofor less than 6 months (N = 1 ).
Phalangeal BMD (mass!cm2) by radiographic absorptiometry (RA) was determined from serial radiographs. Where possible, BMD rates of change before and after surgery were tallied and compared with concomitant changes in serum chemistries. 1n most instances of succesful surgeries as judged by serum chemistries, there was a profound improvement in BMD. Group 1 patients with low BMD suffer hypocalcemia and hypophosphatemia after successful PTX .Their. hungry' bones remineralize rapidly in contrast to Group 11 patients .
Key words
Bone mineral density, renal osteodystrophy, parathyroidectomy, serum calcium, alkaline phosphatase, parathyroid hormone
From the Clinical Radiology Testing Laboratory , Yellow Springs, Ohio* , and the University of Missouri Medical Center, Columbia, Missouri**.
Introduction
Phalangeal bone mineral density (BMD) in mass units/cm2 by radiographic absorptiometry (RA) developed by Colbert and Bachtell ( 1) is routinely used at the Univerisity of Missouri Medical Center (UMMC) at Columbia and neighboring dialysis units to assess the skeletal changes in ESRD patients on hemodialysis and CAPD (2, 3), and to compare progress of patients receiving these two treatment modalities. Other methods of assessing such rates of change in bone mineral include: single photon absorptiometry of the forearm (4) , dual photon absorptiometry of the spine (5), and computed tomography of the spine (6).
In this study of 15 parathyroidectomized patients, serial radiographs of the left hand which include a small aluminum alloy reference wedge in the image field were mailed from UMMC to Clinical Radiology Testing Laboratory (CRTL) in Ohio where they were scanned for phalangeal BMD. Findings were returned to UMMC by mail.
Casual examination of the longitudinal BMD records of many patients by one of us (ZJT) in light of laboratory findings seemed to show: that certain patients were experiencing dramatic increases in BMD that are unusual in ESRD patients who frequently suffer unremitting renal osteodystrophy; and that most of these same patients required careful monitoring to avoid life-threatening consequences of hypocalcemia and hypophosphatemia (7) .A review of patients' charts suggested that finger BMD might help the nephrologist decide whether to operate and might signal the degree to which surgical intervention had been successful. It was therefore decided to compare quantitative BMD measures pre and post-surgery with corresponding serum chemistries .
Method
Fifteen* ESRD patients at the University of Missouri, two males; thirteen females ages 20 71 were divided into four groups:
- Dramatic and prolonged fall in Ca/P levels for longer than six months, requiring Rocaltrol, and without phosphate binders (N = 7).
- Decrease in Ca/P levels for less than six months, receiving short course of Rocaltrol and with phosphate binders shortly after surgery (N = 6).
- No change in Ca/P levels, with PTH levels remaining elevated (N = I).
- After each of several successive surgeries, slight decreases in Ca/P levels for less than 6 months (N = 1).
Table I records the daily supplementation of elemental calcium, I ,25-dihydroxyvitamin D3 (Rocaltrol) and the administration of phosphate binders.
Phalangeal BMD (mass/cm2) by radiographic absorptiometry was determined from serial radiographies pre and post-surgery .The method has a precision error of :t 2% (c.v.). Observation times ranged from 2.2 to 7.4 years. In II of the 15 patients, it was possible to compute rates of BMD change before and after surgery and to compare these with concomitant changes in serum chemistries. In two other patients we obtained post surgery trends only.
To determine bone mineral density of the phalanges, a densitometer scans the x-ray images of middle phalanges of the index, middle and fourth fingers. As each phalangeal image is scanned in successive transverse slices, the densitometer , reads' the gray levels ( called picture elements or pixels) of the images as one would read a printed page, letter-by-letter and line-by-line. These pixel values are transmitted to a computer which calculates the weight (mass) of the mineral and the size of * Actually 14 individuals, one patient was counted twice, at first in Group II after an unsuccessful surgery , and later in Group I after a successful PTX. the bone ( cm2) from which the areal mineral density (mass/cm2) is determined (I). The average BMD of the patient's three middle phalanges is automatically compared with the age and sex-matched norms previously stored in the computer's memory. Correction terms obtained from a scan of the wedge image compensates for differences between films owing to any change in radiographic technique or processing procedure. To double check our findings, we use two films taken at different kilovoltage setting. The BMD findings are accepted only if they agree within :t2% of their mean.
Results
A typical bone mineral density report is shown in Figure I. From serial radiographs taken before surgery , we calculated the trend, i.e. , the rate of change of BMD in excess of the normal rate owing to aging. From serial radiographs taken after surgery, we calculated each patient's trend. Pre and post-surgery trends (% per year) are given in Table II along with the time spans over which the changes were observed. Less credence is given to those observations in which either the time span or number of data points is small. Table II also records the BMD Z-score for each patient just before surgery .The Z-score is the numbers of standard deviations or fractions thereof that the patient's BMD score is relative to normal.
Figure 2 shows three X-rays of the index finger of the patient whose BMD record is graphed in Figure I, taken when the patient was first seen, at the time of surgery and 1.7 years post surgery .Figure 3 summarizes the bone mineral changes of patients in Groups I and II before, at and after surgery .Figures 4-10 depict (by Group) the corresponding serum chemistries to the extent that data was available.
Figure 1 (patient 1, Group I) is an example of the dramatic effect that parathyroidectomy sometimes has on the restoration of phalangeal BMD. Table I shows that this patient's BMD rate of decline before PTX was 4.6%/yr. and his rate of improvement thereafter was 13.5%/yr. Four of the five patients in Group I (whose pre and post-surgery trends could be determined) had annual BMD loss rates (in excess of normal aging) ranging from 2.1 to 11.0%, and experienced annual gain rates ranging from 5.0% to 17.9% .Two of the four were observed over less than a one year span following surgery , making their post-surgery rates less reliable than we would have wished. Four of the five patients in Group II (whose pre and post-surgery trends could be determined) had annual BMD loss rates ranging from 0.9% to 9.2%. The other had a small presurgery gain rate which improved post-surgery trom 0.3% 3.4%. Group II patient No. 10 had no BMD record prior to surgery and did not seem to respond to the parathyroidectomy.
No. 14, the sole patient in Group 3 experienced an improvement in BMD (2.6%/yr). between ages 72.2 and 74.2 which is not accounted for by any effect of surgery since the PTX took place at age 76, about 5 months after the last x-ray was taken. She suffered severe BMD decline (10%) between ages 74.2 and 75.2, which probably persisted until her death at 77. She did not respond to surgery .At surgery four parathyroid glands were removed as confirmed by pathology. Yet her PTH levels continued to be elevated. Thus there had to be an ectopic gland which was isolated by PTH mapping as being in the mediastinum on the left side. This gland was never found, even on autopsy.
No.15, the sole patient in Group 4 had a series of seven partial parathyroidectomies between ages 48.0 and 53.9. This circumstance had made it difficult to assess the effect of any single surgery . However, the collective effect may explain why this patient held her own over the nearly six year span when normal female peers would be expected to lose about 1%/yr. in the early postmenopausal period (i.e. , after about age 50). In fact, this patient gained about 0.5%/yr. over and above the normal aging pattern .
Discussion
Although this small sample is not amenable to statistical analysis, these results suggest that the phalangeal BMD declines initially in many CAPD and HD patients and then reflects the effect of PTX either as a dramatic remineralization of the skeleton or at least an arrest of the downward trend. The concomitant serum chemistries are in consonance with the BMD findings. They are spot readings subject to temporal changes of a short term nature, whereas BMD changes are the cumulative effect of small day-to-day losses or gains and are therefore subject to slower variations.
TABLE II Bone mineral density findings
Annual BMD Prior to PTx
Patient Observation Intervals (yrs) Treatment No. of Data Points Rate of Change BMD Z-score
*Pt. #8 died 8/85
**No x-rays prior to PT x
***The arrest ofBMD loss does not occur until I yearpost-surgery. Patient's first PTx failed; second pTx. successful.
*Causes of BMD improvement between ages 72.2 and 72.4 and then decline until 75.2 are unexplained. No radiographs were taken post-surgery.
We call special attention to Figure 3 which summarizes the BMD response to PTX of patients in Groups I and II.
Group I patients on average have BMD values initially lower than their age and sex-matched peers and lose mineral at a rate much greater than normal aging, reaching deficits in the order of 20% below peer norms. These patients with severe renal osteodystrophy, are the ones who respond most readily to successful surgery. In these cases the bone matrix has not been lost, as is usually the case in osteoporosis of aging, but remains available even for complete remineralization under the right circumstances. These patients with low BMD and successful PIX are prone to develop life-threatening hypocalcemia (Figure 4) and hypophosphatemia (Figure 5) after surgery .They should be carefully monitored for levels of phosphorus (daily for one week and weekly thereafter), and should receive appropriate calcium, phosphorus and Rocaltrol supplementation.
Group II patients on average have BMD values initially at or near normal as compared to their peers. They lose mineral more slowly than Group I
Influence of Parathyroidectomy on Bone Mineral Density and Serum Chemistries in Dialysis Patients patients and do not suffer such severe deficits prior to PTX. Rather than responding dramatically, their loss rates tend to be arrested post-PTX. We note that these patients with high or normal BMD are not so likely to develop hypocalcemia or hypophosphatemia even after successful PTX as judged by PTH levels.
In short, Group I patients have' empty bones , which are. hungry , and remineralize rapidly, whereas Group II patients are modestly sub-mineralized and cannot take on much additional mineral.
Conclusion
We conclude that monitoring BMD is of help in deciding when to undertake surgery and of help in judging the degree of success following PTX .
Acknowledgments
We gratefully acknowledge support given to the Clinical Radiology Testing Laboratory by the following organizations: The American Kidney Fund, The National Kidney Foundation of the Miami Valley and The Fraternal Order of Eagles.
References
- Colbert C, Bachtell RS. Radiographic Absolptiometry (Photodensitometry). In: Non-Invasive Measurements of Bone Mass and Their Clinical Applications . Cohn SH, ed. Boca Raton: CRC Press 1981: 51-84.
- Colbert C, Jerison A, Bachtel1 RS. The Effect of CAPD on Phalangeal Bone Mineral Density. Peritoneal Dialysis Bulletin 1984; 4: 148-51.
- Colbert C, Jerison A, Bachtell RS. The Effect of CAPD on Phalangeal Bone Mineral Density, Part II. In: Khanna R, ed. Proceedings of the Fifth Annual CAPD Conference. Kansas City: Peritoneal Dialysis Bulletin, Inc (Publishers) Toronto 1985: 141-44.
- Cohn SH, ed. Non-Invasive Measurements of Bone Mass and Their Clinical Application. Boca Raton: CRC Press 1981: 85-88.
- Ibid., 85-88.
- Ibid., 121-49.
- Spitz, B, Korzets Z, Dinbar A. Immediate Postoperative Management of Parathyroidectomized HemodiaIyzed Patients. Dialysis and Transplantation 1986; 15: 507-13.