Recovery pattern of patients with osteitis fibrosa cystica in primary hyperparathyroidism after successful parathyroidectomy

Recovery pattern of patients with osteitis fibrosa cystica in primary hyperparathyroidism after successful parathyroidectomy Gaurav Agarwal, MS, DNB, PDC (Endocr Surg), Saroj K. Mishra, MS, FACS, Dilip K. Kar, MS, PDC (Endocr Surg), Anil K. Singh, MD, DM (Endocrinol), Vivek Arya, MD, DM (Endocrinol), Sushil K. Gupta, MD, DM (Endocrinol), and Ambrish Mithal, MD, DM (Endocrinol), Lucknow, India

Background. After parathyroidectomy, recovery of osteitis fibrosa cystica, which continues to dominate presentation of primary hyperparathyroidism in India has not been documented objectively. Methods. We followed up clinical recovery, biochemic markers of bone turnover, bone mineral density, and skeletal radiology in 51 patients with primary hyperparathyroidism and osteitis fibrosa cystica for 9 to 124 months (median, 32 months). Results. After parathyroidectomy, 46 patients had hypocalcemia. During postoperative week 1, bone pain improved in 71%. During 3 months, appendicular fractures healed in all 33 such patients, and 6 of 7 patients who were bedridden could walk. Mean bone mineral density increments (percent change/y) seen at various sites at 1 week, 3, 6, and 12 months were distal forearm — 37, 28, 23, 21; lumbar spine — 165, 104, 101, 106; and total hip — 168, 157, 166, 133. Follow-up radiographs demonstrated prompt recovery though disorderly remineralization. Brown tumors and fractures showed hyperdensities within 3 months. Brown tumors regressed partially in 6 of 27 patients after 6 months. Conclusions. After parathyroidectomy, patients with primary hyperparathyroidism have early, marked, and sustained recovery of osteitis fibrosa cystica. Early (1 week) bone mineral density increments of > 100%/y hint at the skeleton’s ability to promptly restore itself. Densitometric recovery is prompt at cancellous (lumbar spine), but not at cortical (forearm) bone sites. Contour defects and bony tumors persist, and may need corrective osteotomies. (Surgery 2002;132:·1075-85.) From the Departments of Endocrine Surgery and Endocrinology, Endocrine Sciences Centre, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India

PRIMARY HYPERPARATHYROIDISM in India is typically seen as a severe, symptomatic disorder with skeletal, muscular, and renal manifestations at a relatively young age.1 Skeletal manifestations of primary hyperparathyroidism (PHPT) account for most of the morbidity associated with this disease. Calcium and vitamin D nutrition determine the severity of skeletal disease in PHPT, but additional pathogenetic factors also contribute.1 In nutritionally sufficient populations where biochemic screening for hypercalcemia is used routinely, Supported in part by Sanjay Gandhi Postgraduate Institute of Medical Sciences intramural grant No. 545/98 (Dr Agarwal). Presented at the 23rd Annual Meeting of the American Association of Endocrine Surgeons, Banff, Alberta, Canada, April 7-9, 2002. Reprint requests: Gaurav Agarwal, MS, DNB, PDC, Department of Endocrine Surgery, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Raebareli Road, Lucknow 226014, India © 2002, Mosby, Inc. All rights reserved. 0039-6060/2002/$35.00 + 0 11/6/128484 doi:10.1067/msy.2002.128484

overtly symptomatic PHPT and osteitis fibrosa cystica (OFC), the classic form of skeletal disease in PHPT, have all but disappeared. In such societies, subtle skeletal manifestations can be demonstrated using highly sensitive techniques such as bone densitometry and histomorphometric analysis.2 However, overt manifestations and OFC still dominate the presentation of PHPT in India.1 After successful parathyroidectomy (PTx), most manifestations of PHPT, including the skeletal disease, are readily reversed.3 Using precise methods of evaluation of bone mineral density (BMD) and reliable biochemic markers of bone turnover, objective assessment of loss of bone mass and its recovery after treatment can be made. Such methods have been applied to document recovery of skeleton in patients with asymptomatic PHPT.2,3 The recovery of OFC after PTx has not been documented comprehensively, though there are some anecdotal reports of improvement in BMD in patients with OFC after PTx.4,5 We prospectively followed up a cohort of patients with PHPT preSURGERY 1075

1076 Agarwal et al

Fig 1. Plain-film radiographs of pelvis. A, Preoperative film showing marked generalized osteopenia, multiple bone cysts, and brown tumors. B, Three months postoperatively, there is significant remineralization, but bone cysts and tumors persist.

senting with OFC to study the extent and pattern of recovery of OFC using objective biochemic, radiologic, and densitometric methods. METHODS Fifty-one patients with PHPT treated at the department of endocrine surgery at the Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India, during 1990 to 2001 were included in this study. All patients presented with overt skeletal disease and OFC, and underwent successful PTx. Patients treated during the study period with asymptomatic (n = 3) or recurrent/persistent PHPT (n = 2), and those with less than 6 months of follow-up information (n = 3, including 2 who died during perioperative period) were excluded. In addition to detailed clinical evaluation, preopera-

Surgery December 2002

tive assessment included estimation of serum total and ionized calcium, inorganic phosphorus, intact parathyroid hormone by immunoradiometric assay (DSL, Webster, Tex); calcidiol and caclitriol by radioimmuno assay (Incstar, Still Water, Minn); and urinary calcium and creatinine excretions. Biochemic bone turnover studies included serum total alkaline phosphatase (ALP) activity in all, and serum osteocalcin level by coated-tube immunoradiometric assay (DSL) and urinary excretions of type-I collagen specific peptides (Crosslaps) by enzyme-linked immunosorbent assay (DSL) in a subset of 22 consecutive patients. Skeletal radiographs including sites of suspected fractures or brown tumors were taken. BMD by dual-energy xray absorptiometry (Hologic 4500QDR; Hologic Inc, Bedford, Mass) was studied in the 32 patients treated since 1997, the time when this investigative modality became available at our hospital. All patients underwent successful PTx, after which they received calcium and calcitriol supplements. Subjective improvement of symptoms, such as pain and muscle weakness, and physical findings were recorded during postoperative week 1 and at each follow-up visit. Early biochemic changes (serum-calcium and phosphorus concentrations, and ALP daily for 7 days; serum-intact parathyroid hormone on day 1 and 7; serum-osteocalcin and urinary crosslaps on day 7) and BMD (on day 7) were studied. Patients were discharged 7 to 10 days after PTx with oral calcium and vitamin D supplements. Clinical, biochemic, radiologic, and densitometric evaluations were carried out on follow-up visits at 3 weeks, 3 monthly during the first year, and annually thereafter whenever possible. All BMD evaluations were carried out on the same densitometer and coefficient of variation was < 1%. Values are displayed as arithmetic mean and range. RESULTS Patient characteristics. Mean age of the patients was 36 years (range, 13–62). There were 11 men, 29 premenopausal women, and 11 postmenopausal women. Average duration of symptoms was 29 months (7 months to 26 years). All patients had history of bone pain. Thirty-two patients had proximal muscular weakness, and 7 were bedridden as a result of multiple fractures and muscular weakness. Thirty-three patients had 1 or more appendicular fractures. Twenty-seven patients had clinically evident bony deformities or brown tumors, and 8 had palpable parathyroid tumors. The preoperative biochemic indices are shown in Tables I and II. One patient had normocalcemic PHPT. Mean serum-calcidiol and calcitriol concentrations were

Surgery Volume 132, Number 6

Agarwal et al 1077

Fig 2. Plain-film radiographs of left femur in a 17-year-old girl with parathyroid carcinoma. A, Preoperative film with almost invisible femur. B, Three months postoperatively, marked remineralization, though with persistent brown tumors that have become radiodense. C, Twelve months postoperatively, the brown tumors and bone cysts remain unchanged.

11.3 ng/mL (range, 00-57; reference range, 9-45) and 47.7 pg/mL (range, 22-89; reference range, 10.5-55), respectively. Urinary calcium excretion was normal in the majority of patients. Radiologic examination of the skeleton demonstrated generalized osteopenia in all patients, with characteristic cortical bone loss and subperiosteal resorption seen in hands and skull (Fig 1 to 5). In 4 patients, some of the long bones were almost invisible on radiography (Fig 4, A). Brown tumors of long and flat bones were seen in 32 patients. The upper end of the tibia was the most common site of brown tumors (Fig 5), and other sites included the femur, pelvic bone, mandible, ribs, and vertebral body. The upper end of the tibia, neck or shaft of femur, and humerus were the more common sites of fracture. Two patients had fractures of the vertebral bodies. Significant demineralization of cancellous bone sites (lumbar spine and hip) was also appreciated. These findings were confirmed in the 32 recent patients in whom BMD at cortical and cancellous bone sites were found to be grossly subnormal, with an average z score of -4.8 at the distal third of radius and -3.9 at lumbar spine (Table III). All but 12 recent patients underwent PTx using a standard bilateral neck exploration. In these 12 patients with localization of pathology to a single parathyroid gland on preoperative 99mTc-sestamibi scanning and neck ultrasonography, a targeted, minimally invasive PTx was carried out. A single parathyroid adenoma was excised in 30 patients, and 2 underwent excision of 2 adenomas. Six patients underwent subtotal PTx for primary

parathyroid hyperplasia. In 3 patients, en bloc excision of parathyroid tumor along with the ipsilateral thyroid lobe was carried out for parathyroid cancer. Post-PTx course Clinical recovery. Improvement in bone pain and muscular weakness were evident in postoperative week 1 in 36 patients. Appendicular fractures healed in all 33 patients with fractures over median time of 3 months. Six of 7 patients who were bedridden could be ambulated more than 3 months after PTx. Bony deformities and swellings as a result of brown tumors regressed only partially in 6 of 27. Biochemic changes. Table I provides a comparative account of biochemic indices before and after PTx. Postoperatively, all patients had biochemic hypocalcemia. Symptomatic hypocalcemia occurred in 46 patients within a median period of 22 hours, which was managed with intravenous calcium gluconate infusion at 1 to 2 mg/kg/h for 36 to 168 hours (mean 93). Those patients who were hypocalcemic also had low postoperative serum phosphorus and magnesium concentrations (assessed in 31 patients), indicative of hungry bone syndrome. In the follow-up, ranging 9 to 124 months (median 32 months), all patients remained normocalcemic. Calcium and vitamin D supplements were continued until normalization of serum-ALP levels, and were required for a mean duration of 6.5 months (range, 4-13). Two patients had increased serum intact parathyroid hormone levels up to 3 months in follow-up, though they were both normocalcemic.

1078 Agarwal et al

Surgery December 2002

Fig 3. Plain-film radiographs of both bones of leg in same patient as in Fig 2. A, Preoperative films show extreme osteopenia, with barely visible cortical margins. B, Twelve months postoperatively, all bones have near normal radiologic appearance.

Fig 4. Plain-film radiographs showing brown tumors at upper shaft of right femur in a 35-year-old woman. A, Preoperative film showing 2 expansile lesions with grossly destructed cortex. B, Three months postoperatively, both the brown tumors appear hyperdense and corresponded to clinically palpable swellings. C, Twelve months postoperatively the tumors persist, though have become much more rare.

Changes in biochemic markers bone turnover. Table II shows serum ALP and osteocalcin, markers of bone formation, increased in most patients soon after PTx and normalized after 9 and 6 months, respectively. PTx resulted in decrease in urinary crosslaps, a biochemic marker of bone resorption, as early as 1 week later. Urinary crosslaps levels normalized in the majority of patients by the end of 3 months. In 3 young patients who had extreme osteopenia, the biochemic markers of bone turnover have remained high 1 year after a curative PTx. Densitometric recovery. Post-PTx BMD was studied at 1 week in 12 patients, at 3 and 6 months in 26, and at 12 months in 21; results are summarized in Table III. BMD showed remarkable, early, and sus-

tained improvement after PTx. Although the increments in BMD at hip and lumbar spine (mean +166 and +101%/y, respectively, at 6 months) were remarkable, those at the distal third of radius were only moderate (mean +23%/y at 6 months). Also, the rate of increase in BMD varied widely from patient to patient: from +28%/y to +203%/y at hip; and from +22%/y to +289%/y at lumbar spine at 6 months. In the 12 patients in whom BMD was studied at 1 week, an increment in hip bone density of up to 0.039 gm/cm2 in absolute terms, and up to 332.9%/y was seen so early after the PTx. This improvement in BMD was sustained, and by end of 1 year after PTx, the BMD at hip and lumbar spine was observed to have more than doubled in most

Agarwal et al 1079

Surgery Volume 132, Number 6 Table I. Biochemic indices in patients with PHPT with OFC before and after PTx After operation

Serum Calcium (mg/dL) Inorganic phosphorus (mg/dL) iPTH (pg/mL)

Normal

Preoperative

8.2 - 10.5

12.6 (10.8-17.2) 2.1 (1.6-3.1)

3.5 - 5.5

9 - 55

824 (117-1890)

1 day

7 days

6.9 8.2 (5.8-7.9) (6.7-9.6) 2.6 3.2 (1.9-3.3) (2.7-3.7) 11 (00-76)

29 (3-68)

3 weeks

3 months

6 months

12 months

9.6 (8.3- 10.2) 4.4 (3.6- 5.7)

9.2 (8.8-10.1) 4.6 (3.9-5.8)

9.2 (8.7-10.5) 4.4 (3.85.7)

9.1 (8.8-10.3) 4.0 (3.8-5.6)

28 (12-55)

37 (18-72)

40 (19-53)

33 (23-48)

Values are mean (range). iPTH, intact parathyroid hormone.

patients. In absolute terms, however, the BMD of 13 patients remained lower than that of the healthy population at the end of 1 year, even though they all have had BMD increments of around 100%. The incremental trend in BMD has continued beyond 1 year of follow-up in the 17 patients for up to a maximum period of 4 years. Radiologic recovery. Follow-up radiographs showed recovery of subperiosteal resorption and osteoporosis as can be appreciated in Fig 1 to 5. Though remineralization was early and vigorous, its pattern was disorderly, conforming to the distorted bony contours and brown tumors, and to fracture sites (Fig 1 to 5). The extent of remineralization varied from patient to patient and according to site in the same patient. In patients who had almost invisible long bones on radiographs, the deformed bony outlines became visible at 3 months, and the bones progressively became denser with time (Figs 1, 2, and 3). Cortical and appendicular fractures healed in all patients within a median period of 3 months (range, 2-7) after PTx. The fracture sites had abnormal hyperdensity in all, which persisted until the last follow-up as much as 6 years after PTx. Brown tumors that could not be appreciated on preoperative radiographs as a result of extreme osteopenia became evident on follow-up radiographs, and persisted even after 6 years of follow-up (Figs 2 and 4). The sites of bone cysts, brown tumors, and fractures appeared abnormally radiodense on radiographs within 3 months of PTx (Figs 4, B, and 5, B). The hyperdensity of these sites regressed partially at between 6 and 12 months (Figs 4, C, and 5, C). Bony deformities and brown tumors regressed only partially in 6 of the 27 patients who had such deformities. DISCUSSION The fact that bones are subject to abnormalities in PHPT has been well-appreciated since the disease was first recognized. Though OFC was a dominant

manifestation of PHPT in the past, by the early 1980s, overt skeletal involvement had become extremely rare in most developed countries.6 Currently, the only evidence of skeletal effects of this disease is subtle cortical osteopenia seen on highly sensitive BMD studies.3 Trabecular bones such as vertebrae are spared because parathyroid hormone is known to have a protective effect on cancellous bone.2 A large number of asymptomatic patients with PHPT are treated nonsurgically.7 In patients with specific indications for PTx, after a successful operation, early and sustained recovery of BMD occurs, more at the cancellous compared with cortical bone sites.2,3 There is a large geographic variation in the profile of PHPT, and severe skeletal disease continues to be the predominant mode of presentation in some parts of the world.8 In India, PHPT continues to ravage the skeleton, leading to tremendous loss of bone mineral, resulting in multiple fractures, deformity, and often a disabled state. Inadequacies of calcium and vitamin D nutrition, coupled with a delay in seeking medical attention, can lead to such severe osteopenia that some of the bones are almost invisible on routine radiographs.1 In such severe disease, cancellous bone is not spared, as was evident in our patients who had a mean lumbar spine density of 0.45 gm/cm2 (average z score of -3.9); some also had vertebral fractures or brown tumors. There have been reports on the positive impact of successful PTx on the bone health of patients with PHPT from India and other developing nations.1,4,9 However, there have been no comprehensive, prospective reports objectively documenting the post-PTx course of OFC. Sensitive, reliable, and noninvasive quantification of bone mass by dual-energy x-ray absorptiometry, and reliable biochemic markers of bone turnover have facilitated objective evaluation of extent of bone loss in PHPT.2,3,10,11 Moreover, these techniques have enabled clinicians to objectively follow-up and monitor the recovery of skele-

1080 Agarwal et al

Surgery December 2002

Table II. Sequential changes in biochemic markers of bone turnover in patients with PHPT undergoing PTx for OFC After operation 1 day

Serum ALP (U/L) Osteocalcin (ng/mL) (n = 22) Urine Crosslaps (µg/mmol creatinine) (n = 22)

7 days

Normal

Preop

Value

% of preop

Value

% of preop

< 125

1786 (340-8767)

202 (137-238)

2-12

895 (177-7240) 47.3 (17-66)

—

—

1889 (246-2387) 49.4 (22-68)

236 (122-289) 103 (102-132)

< 1035

4495 (540-8900)

—

—

1893 (650-4500)

56 (18-153)

Values are mean (range). Preop, preoperative. Crosslaps, type-I collagen specific peptides.

Table III. Sequential changes in BMDs in patients with PHPT undergoing PTx for OFC After operation Before operation (n = 32)

7 days (n = 12)

Site

BMD

(gm/cm2)

z score (SD)

BMD

Distal third radius

0.346 (0.303-0.623) 0.434 (0.413-0.529) 0.454 (0.341-0.549)

-4.3 (-2.8--6.7) -2.6 (-2.2- -3.5) 3.9 - (-2.6- -4.2)

0.350 (0.318-0.626) 0.449 (0.423-0.536) 0.470 (0.352-0.591)

Total hip Lumbar spine (lateral)

(g/cm2)

% change/y* 36.5 (-3.2-39.2) 168.2 (93-332.9) 165.2 (80.4-198.4)

Values are mean (range). * Compared with preoperative BMD.

tal effects of PHPT after successful treatment. The pattern and extent of recovery of skeletal disease in patients with asymptomatic PHPT and in secondary hyperparathyroidism are well-studied and documented. In patients with asymptomatic PHPT, various studies have reported an increase of 5% to 10% in the bone mass after PTx. The greatest gains tend to occur in the first postoperative year; however, in some patients the reversal of bone loss may not be complete. The benefits of PTx extend even to the cancellous bones, as has been shown by Silverberg et al.2 However, recovery of overtly symptomatic skeletal disease in advanced PHPT, especially in vitamin D-deficient populations, is ill-studied. There have been anecdotal reports of improved BMD after PTx in patients with OFC;4,12-14 however, little is known about how much bone mass can be restored after successful PTx in such patients. In 51 patients with symptomatic PHPT and bone disease, using objective densitometric and biochemic markers of bone turnover, we could demonstrate remarkable, early, and sustained recovery of OFC after a successful PTx. The biochemic markers of bone formation, serum ALP

and osteocalcin, were found to increase soon after PTx and normalize during the subsequent 6 to 9 months. Urinary type-I collagen specific peptides, markers of bone destruction, started decreasing 1 week after PTx, however, they normalized only at about 3 months. Sequential BMD studies in early post-PTx period and follow-up could demonstrate spectacular remineralization of the skeleton as early as 1 week after PTx. The increments in BMD were 300%/y in the early post-PTx period, and this trend was found to be sustained beyond 1 year. Postoperative plain-film radiographs showed interesting, vigorous, but uniquely disorderly remineralization of the skeleton. Fractures healed promptly, but the fracture and bony tumor sites appeared abnormally hyperdense on early postoperative radiographs. An interesting observation in our patients was a near disappearance of bone pain and regaining of muscular strength during the first post-PTx week. Almost 90% of the patients had symptomatic hypocalcemia that could be attributed to hungry bone syndrome as they had low serum calcium, phosphorus, and magnesium levels. These patients

Agarwal et al 1081

Surgery Volume 132, Number 6

After operation 3 weeks Value

3 months % of preop

768 (234-4534)

98 (45-136)

—

—

Value

6 months

% of preop

12 months

Value

% of preop

Value

% of preop

—

569 (209-2289) 29.2 (6-72)

73 (56-102) 68 (47-89)

345 (178-2230) 13.3 (5-33)

54 (32-103) 33 (20-56)

117 (78-1287) 12.4 (5-23)

33 (12-53) 27 (22-48)

—

998 (350-2850)

32.7 (6-111)

668 (200-2200)

23 (3-93)

710 (150-2900)

21.3 (4-83)

After operation 3 months (n = 26) BMD

(g/cm2)

0.378 (0.312-0.632) 0.561 (0.422-0.610) 0.576 (0.368-0.649)

% change/y* 28.3 (1.2-36.7) 156.8 (32.3-298.1) 103.9 (89.4-199.2)

6 months (n = 26) BMD

(g/cm2)

12 months (n = 21)

% change/y*

0.392 (0.341-0.704) 0.785 (0.624-1.117) 0.696 (0.387-0.937)

required up to 3 mg/kg/h of calcium gluconate infusion during the early postoperative period, in addition to oral calcium carbonate (2-4 gm/day) and calcitriol (0.5 to 2 µg/ day). The manifestations of this post-PTx sequel were much more severe in our patients than previously reported.15 In 4 such patients with hungry bone syndrome, BMD studies on postoperative day 2 showed increments of up to 300%/y at lumbar spine whereas the increments were only up to 12%/y at distal radius (data not shown). The marked increase in BMD of up to 332%/y as early as 1 week after PTx in these patients with hungry bone syndrome demonstrates the anabolic effects of PTx on the skeleton and the remarkable capacity of the skeleton to restore itself after PTx.4 To the best of our knowledge, there are no reports suggesting such early BMD changes after PTx, and this needs further evaluation in a larger cohort of patients. Biochemic markers of bone turnover are known to change soon after PTx in patients with asymptomatic PHPT.10,11,16 However, there are no studies documenting such changes in patients with OFC. Our findings of early decrease in urinary

23 (8.2-28.9) 166 (28.3-202.8) 101 (22.4-289.1)

BMD

(g/cm2)

0.421 (0.387-0.768) 0.976 (0.694-1.212) 0.925 (0.738-1.135)

% change/y* 21 (12-38.5) 133 (67.8-178) 106 (54.2-176.3)

z score (SD) -3.4 (-1.9-4.6) -1.3 (-0.6-2.1) 1.8 -(-0.8-2.6)

crosslaps levels, signifying a decrease in the osteoclastic activity in the first week after PTx, are in agreement with similar observations in asymptomatic PHPT.10,11,16 The markers of bone formation, serum ALP and osteocalcin, were observed to increase during the early post-PTx phase and decrease 6 to 9 months after PTx, similar to reports on patients with asymptomatic PHPT.10,11,16 The persistently high levels of markers of bone formation in the face of decreasing markers of bone resorption indicated that a true increase in bone mass was taking place.4 This fact was further substantiated by occurrence of hungry bone syndrome, and early and sustained increments in BMD. Our results showed that biochemic markers of skeletal metabolism provide an objective evaluation of state of bone turnover, and can be used to monitor the same in patients with OFC after PTx. Whether or not such studies can be used to predict fracture risk or provide prognostic information remains to be seen. Successful PTx results in recovery of BMD in patients with asymptomatic PHPT.2,3,16 Gains in BMD are quicker and more substantial at predomi-

1082 Agarwal et al

Surgery December 2002

Fig 5. Plain-film radiographs showing large brown tumor arising from upper end of tibia. A, Preoperative film. B, Three months postoperatively, the tumor seems radiodense, as if “filled in” with minerals. C, Twelve months postoperative film showing persistent tumor, though with slightly lesser osteodensity

nantly cancellous bone sites (lumbar spine and ultra distal radius) compared with the cortical bone (distal third of radius).2,16,17 Our results and observations by Kulak et al4 and Minisola et al5 in small numbers of patients with OFC have shown similar preferential restoration of cancellous bone as against cortical bone, similar to patients with asymptomatic PHPT. Besides, we have been able to demonstrate that, depending on the baseline BMD at cortical bone sites, even cortical bones show remarkably early and rapid remineralization, with the mean distal third radius BMD increment of 36%/y at 7 days and 21% at 1 year. The magnitude of increments in BMD are exponentially higher in presence of severe osteopenia as seen in our patients compared with the asymptomatic PHPT.2 In contrast to this, Suzuki et al18 reported that longitudinal increase in BMD after PTx was more in patients with asymptomatic PHPT than in those with renal calculosis or OFC. Possible explanations of discrepancy in the BMD gains at lumbar spine as compared to the radius include differences in the rate of bone turnover as a result of difference in marrow composition. Thus there is a greater rate of bone remodeling at cancellous versus cortical bone sites, and consequently, there is faster restoration of bone mass at sites of greater turnover dynamics.2,4,19 In young patients, remineralization and remodeling may be aided by concomitant acquisition of peak bone mass with skeletal maturity. Further, in PHPT, there is a decrease in osteoid-covered surfaces soon after PTx, resulting in increased mineralization.20

Greater remineralization at cancellous bone sites can thus be explained on the basis that in PHPT, the remodeling space in cancellous bone is increased to a far greater extent (up to 100% above normal) than at cortical sites.2 In the presence of OFC, given the extensive subperiosteal and cortical resorption, differences in expansion of remodeling spaces between cancellous and cortical bones would accentuate the different rates at which bone would recover. However, another influence could be the different mechanical forces, such as weight bearing, to which these bones might be subjected.4,19 Our results, and those of Kulak et al,4 have demonstrated that the rapidity and extent of remineralization after PTx are directly proportional to the extent of bone loss at baseline. Similar trends have been reported by others in presence of OFC5 and in asymptomatic PHPT.2,16 Follow-up plain-film radiographs brought out some interesting observations. Bones that were almost invisible owing to severe osteopenia became readily visible. Appreciable radiologic differences could be seen as early as 1 week after PTx. Most noticeable was a dramatic “lighting up” of the brown tumors, as if they were filled up with bone minerals (Figs 4 and 5). The pelvis and spine were the earliest to show generalized recovery of the osteopenia at 3 weeks to 3 months. Recovery of subperiosteal and cortical resorption at skull and hands occurred later, 6 to 9 months after PTx. Sequential radiologic examination during 4 years in 17 patients showed continued remodeling of the bones, evident by ini-

Agarwal et al 1083

Surgery Volume 132, Number 6

tial abnormally radiodense bone cysts, tumors, and fracture sites in the background of generalized osteopenia, followed by gradual decrease in the density of these tumors or cysts, and improved density of the rest of the skeleton. However, even after 4 years, the abnormal densities at these sites, and the defects of bone contour and brown tumors, have persisted. CONCLUSIONS A spectacular early and sustained recovery of OFC followed a successful PTx in patients with PHPT in India, though the skeletal remineralization is disorderly. Prompt recovery of biochemic markers of bone resorption and increments of BMD > 100%/y as early as 1 week after PTx hint at the skeleton’s remarkable ability to restore itself. The densitometric recovery was prompt and brisk at cancellous (spine), but not so at cortical (forearm) bone sites. Many patients with OFC did not regain normal BMD even after a few years. Increased levels of biochemic markers of bone formation for 6 to 9 months and sustained gains in BMD for up to 4 years indicated that the process of bone remodeling and repair after PTx is lengthy. Contour defects and bony tumors persisted, as did other radiologic deformities, and may need corrective osteotomies. Using objective and reliable densitometric and biochemic tools, the state of skeletal metabolism can be closely monitored after a curative PTx.

6. 7. 8.

9.

10.

11.

12.

13.

14.

15.

The authors wish to acknowledge the help of the following individuals: Dr Amit Agarwal, Dr Perumalla Rajasekhar, and Dr Preeti Dabadghao for patient treatment, data collection, and critical input relating to methodology; Dr Jonas Rastad, Uppsala, Sweden, for critical input relating to the concept and methods of this study; Mr Manoj Shukla and Mrs Meera Srivastava for help with crosslaps and osteocalcin assays; and Mr Anil Kumar for photography support. REFERENCES 1. Mishra SK, Agarwal G, Kar DK, Gupta SK, Mithal A, Rastad J. Unique clinical characteristics of primary hyperparathyroidism in India. Br J Surg 2001,88:708-14. 2. Silverberg SJ, Gartenberg F, Jacobs TP, Shane E, Siris E, Staron RB, et al. Increased bone mineral density after parathyroidectomy in primary hyperparathyroidism. J Clin Endocrinol Metab 1995;80:729-34. 3. Syed Z, Khan A. Skeletal effects of primary hyperparathyroidism. Endocr Pract 2000;6:385-8. 4. Kulak CAM, Bandeira C, Voss D, Sobieszczyk SM, Silverberg SJ, Bandeira F, et al. Marked improvement in bone mass after parathyroidectomy in osteitis fibrosa cystica. J Clin Endocrinol Metab 1998;83:732-5. 5. Minisola S, Rosso R, Romagnoli E, Pacitti MT, Scarnecchia L, Carnevale V, et al. Trabecular bone mineral density in primary hyperparathyroidism: relationship to clinical pre-

16.

17.

18.

19.

20.

sentation and biomarkers of bone turnover. Bone Miner 1993;20:113-23. Silverberg SJ, Bilezikian JP. Primary hyperparathyroidism: still evolving? J Bone Miner Res 1997;12:856-62. Consensus development conference statement. J Bone Miner Res 1991;6:s9-13. Mithal A, Bandeira C, Meng XW, Silverberg SJ, Shi SJ, Mishra SK, et al. Clinical presentation of primary hyperparathyroidism: India, Brazil, and China. In: Bilezikian JP, Marcus R, Levine M, editors. The parathyroids: basic and clinical concepts. 2nd ed. San Diego: Academic Press; 2001. p. 375-86. Cheung PS, Boey YH, Wang CC, Ma JT, Lam KS, Yeung RT. Primary hyperparathyroidism: its clinical pattern and results of surgical treatment in Hong Kong Chinese. Surgery 1988;103:558-62. Guo CY, Holland PA, Jackson BF, Hannon RA, Rogers A, Harrison BJ, et al. Immediate changes in biochemical markers of bone turnover and circulating interleukin-6 after parathyroidectomy for primary hyperparathyroidism. Eur J Endocrinol 2000;142:451-9. Steiniche T, Christensen P, Vesterby A, Ullerup R, Hessov I, Mosekil LE, et al. Primary hyperparathyroidism: bone structure, balance, and remodeling before and 3 years after surgical treatment. Bone 2000;26:535-43. Agarwal G, Mishra SK, Mithal A, Arya, V. Remarkable but disorderly recovery of osteitits fibrosa cystica following parathyroidectomy. J Jpn Surg Soc 1999;100:164. Brossard JH, Garon J, Lepage R, Gascon-Barre M, D’Amour P. Inhibition of 1,25(OH)2D production by hypercalcemia in osteitis fibrosa cystica: influence on parathyroid hormone secretion and hungry bone disease. Bone Miner 1993;23:15-26. Knee TS, Drake AJ, Turton D, Shakir KM. Effect of parathyroid adenoma resection on bone density in primary hyperparathyroidism and osteitis fibrosa cystica. Orthopedics 2001;24:1000-2. Brasier AR, Nussbaum SR. Hungry bone syndrome: clinical and biochemical predictors of occurrence after parathyroid surgery. Am J Med 1988;84:654-60. Abe Y, Ejima E, Fujiyama K, Kiriyama T, Ide A, Sera N, et al. Parathyroidectomy for primary hyperparathyroidism induces positive uncoupling and increases bone mineral density in cancellous bones. Clin Endocrinol (Oxf) 2000; 52:203-9. Christensen P, Steiniche T, Brixen K, Hessov I, Melsen F, Heickendorf L, et al. Primary hyperparathyroidism: effect of parathyroidectomy on regional bone mineral density in Danish patients: a three year follow-up study. Bone 1999; 25:589-95. Suzuki S, Fukushima T, Ami H, Asahi S, Onogi H, Nakamura I, et al. Pre- and postoperative bone metabolism of primary hyperparathyroidism. Biomed Pharmacother 2000;54:s90-6. Minisola S, Rosso R, Romagnoli E. Comment on marked improvement in bone mass after parathyroidectomy in osteitis fibrosa cystica. J Clin Endocrinol Metab 1998;83:3760. Christiansen P, Steiniche T, Mosekilde L, Hessov I, Melsen F. Primary hyperparathyroidism: changes in trabecular bone remodeling following surgical management. Bone 1990;11:75-9.

DISCUSSION Dr Richard E. Goldstein (Nashville, Tennessee). I wondered if you might comment on your experience with the onset and the management of bone hunger. I

1084 Agarwal et al

have had a number of patients in whom the onset of severe bone hunger has really not been apparent for 48 or 72 hours. And in this country, we are pushed to get patients out of the hospital quickly. I wondered if you might comment on what you saw in these patients, the way you manage bone hunger, and at what point in time you saw it. Dr Agarwal. As I said, 46 of these 51 patients had manifestations of hungry bone syndrome. In our patients we often see manifestations of hungry bone syndrome as early as 6 to 7 hours after PTx. During the last 3 years, it has been our policy to put these patients on prophylactic calcium infusion to prevent any symptomatic hypocalcemia from occurring. We have done some studies to look at predictors of hungry bone syndrome, and have found that the extent of skeletal involvement, the tumor weight, and the vitamin D levels are 3 important predictors of the extent of hungry bone syndrome. The way we manage the hungry bone syndrome is, these patients are put on prophylactic calcium infusion, starting with a dose of 1 mg/kg body weight/h. In patients who have signs or symptoms of hypocalcemia or have extremely low serum calcium levels that we estimate at 6 hours and 12 hours, the doses are increased further. At the same time, the patients are given 2 g of calcium and 0.5 µg of calcitriol orally on the same evening as the operation. The doses are built up as needed. The maximum dose we have needed to use was 3 µg or oral calcitriol and 6 g/day of oral calcium. The way we get these patients off the calcium is, most of our patients who have severe hungry bone syndrome require calcium infusion for a period of up to 3 days, and the calcium infusion is gradually tapered off by halving the dose over a period of 2 to 3 days. All the oral supplements are continued at the same dose they were started in the patient until 3 weeks. And during these 3 weeks, we monitor the serum ALP level, which is a very easily available biochemic marker of bone turnover, and patients are put off calcium and calcitriol supplements on a tapering basis during a period of 3 to 6 months if the markrs of bone turnover show expected decline. Some patients have required supplements, up to 1 year, depending on persistently raised biochemical markers of biochemical bone turnover. Dr Ian Gough (Brisbane, Australia). This is an interesting and remarkable series. I am interested in the role of vitamin D, which you just touched on to some extent. What do you think is the role of vitamin D in the pathogenesis of the OFC and what role does vitamin D have in its healing after a successful parathyroid operation? Dr Agarwal. Thanks for bringing up that point. That is a very intriguing thing that we have been studying for almost a decade now. In these patients, we have been able to demonstrate an inverse correlation of vitamin D levels with parathyroid tumor weight and a direct correlation with the BMD. We are in the process of initiating some molecular studies to see if the vitamin D receptors and other molecular markers can correlate vitamin D nutritional and metabolic status of these patients with

Surgery December 2002 the pathogenesis of OFC. We are quite convinced, on the basis of our clinical studies, that vitamin D nutrition does play an important role in the pathogenesis of OFC. Regarding the management of OFC, vitamin D plays an important part in the treatment of these patients after successful PTx. There have been reports that these patients, if treated before PTx with calcitriol, might experience lesser severity of hungry bone syndrome. We have not used this management policy so far because we are afraid of causing hypercalcemic crisis in the presence of such severe bone disease. Dr Paolo Miccoli (Pisa, Italy). Apparently, the carcinoma rate is quite high in your series. Do you have any information about the outcome? In other words, were they really parathyroid carcinomas according to the standard definition, meaning was there an invasion or distant metastases in spite of the short follow up interval? What can you say? Dr Agarwal. Thanks for your comments and question. Three of the 51 patients, who constituted little less than 6% of all patients, had parathyroid cancers. And the diagnosis of parathyroid cancer in these patients was on the basis of gross operative findings. There was evidence of infiltration of the thyroid lobe in 2 patients; of strap muscles in 1 patient; and 1 patient each had invasion of the esophagus and trachea. So the diagnosis of parathyroid cancer was quite evident even before histology. Histologic features were quite consistent with the diagnosis of parathyroid cancer, and included gross capsular invasion, with or without vascular invasion and presence of wide fibrovascular septae. However, ultimate proof of the diagnosis of malignancy in these patients is occurrence of a recurrent or metastatic disease. Unfortunately, all these 3 patients have had metastatic or recurrent disease in the follow-up at 26, 39, and 64 months after PTx. The most interesting patient of these 3 was 1 who, after 3 years of PTx, presented in hypercalcemic crisis in the emergency department. She subsequently had rapid deterioration in her higher functions and consciousness, went into respiratory failure, was intubated, and ventilated. We found that she was evidently hyperparathyroid, and had multiple cerebral metastases. And this was only, to the best of my knowledge, the second known case of intracranial metastases of parathyroid cancer. Dr Collin J. Weber (Atlanta, Georgia). I, too, am absolutely astonished by the youth of some of your patients who presented with parathyroid cancer. My question is, why do you think this is happening in your patient population? Is it a result of diet? Is it a result of the environment? Do these 51 patients, if I understood the number correctly, represent a small cohort of all the patients with hyperparathyroidism you have seen during a decade? Or are these the only patients with hyperparathyroidism you have operated on in that span? Dr Agarwal. To answer the last part first, these 51 patients constitute a majority of the total of 63 patients who we have treated during the last decade. So, what we are

Surgery Volume 132, Number 6

treating is just the tip of the iceberg. We are not diagnosing the majority of the patients. The reasons why 12 of our patients were not included in this study are that we excluded any patients who had asymptomatic diseases; any patients who had recurrent or persistent hyperparathyroidism, which were 2; and any patients for whom we did not have at least 6 months of follow up data, because they were either being followed-up in different institutions, or 2 patients who died during perioperative period.

Agarwal et al 1085

Regarding the reasons why this kind of a disease is being seen in our patients at such a young age, this seems to be a result of a combination of many factors. Firstly, we are only picking up the extreme cases. We are not diagnosing the asymptomatic cases because we do not have a policy of routine screening for calcium. At the same time, our clinical studies have shown that vitamin D and calcium nutritional levels do play an important role in parathyroid oncogenesis and progression. And that probably also reflects in the higher percentage of parathyroid cancer in these young patients.