Total urinary and free serum hydroxyproline in metastatic bone disease.

The present study examines the possibility of correlation between free serum and urinary total hydroxyproline and whether this correlation can be applied to clinical conditions. The correlation between the two indices was 0.80 (P less than 0.001) in 18 patients, mostly suffering from malignant disease. On comparing the same measurements in 37 patients, all with known metastatic bone disease, we found 29/37 normal results for free serum hydroxyproline, whereas only 2/37 values of urinary total hydroxyproline were normal. The authors therefore conclude that urinary total hydroxyproline, measured as the ratio hydroxyproline/creatinine in a fresh specimen of early-morning urine, is the best index of collagen breakdown in metastatic bone disease and preferable to measurement of free serum hydroxyproline.

TOTAL URINARY and free serum hydroxyproline have been claimed to be useful for the early diagnosis of metastatic bone disease (Kontturi et al., 1974;Powles et al., 1976). Total urinary hydroxyproline excretion (urinary Hyp) has been used for the estimation of the efficiency of a therapeutic regimen on female breast cancer in the presence of bone metastasis (Powles et al., 1975). However, it is often more convenient to collect blood samples than urine. If free serum hydroxyproline (serum Hyp) was as efficient as urinary total hydroxyproline in the diagnosis of skeletal metastases, it might be therefore preferred. If however blood samples are not as good, the ratio of total hydroxyproline to creatinine in a fresh morning urine (Hyp/Cr) is more convenient than making a 24 h urine collection. This latter measurement should be made on a constant diet for at least 48 h, while the Hyp/Cr may be collected after an overnight fast (Nordin et al., 1976). For these reasons it is important to know the relative diagnostic accuracies.
The present studies have been designed to answer the following questions: (a) Are the results of urinary Hyp and free serum Hyp correlated in a given situation?   The patients received a collagen-free diet for 24 h. Thereafter a fasting blood sample and a specimen of fresh morning urine were collected.
Chemical methods (Studies A and B) Blood was allowed to clot for 1 h at room temperature. Serum was collected and frozen at -20°C until the analyses were made. The urine was collected on thymol crystals at +4°C. It was then acidified with HCI and stored.
Hydroxyproline was measured by an automated method ( r-0 79 in a spot urine). All correlation coefficients are significant at the 0-001 level.

Study B
The results are given in Fig. 3. There are only 2/37 normal results for Hyp/Cr whereas 29/37 serum Hyp values are normal. The two indices are correlated (r-0 49, P<0-001). Laitinen (1966)  ,.X undertook the two studies in order to examine the possibility of replacing the measurement of urinary Hyp by serum Hyp. From our results it is clear that, although there is a good and highly significant correlation between the two measures, urinary Hyp is a better disre x 100 criminator between normal and metastatic disease. For urinary Hyp the percentage of raised levels is the same as that published by Powles et al. (1975) for the same condition. We cannot confirm the results * of Kontturi et al. (1974), who found an earlier response of serum Hyp than of urinary Hyp in patients with prostatic cancer with scintigraphic evidence of skeletal metastases, neither in this study with various cancer types, or in a study reported elsewhere, comprising only proy static carcinoma (Jeannet, 1978).

DISCUSSION
A good correlation between urinary Hyp, expressed either as its 24 h excretion + or as Hyp/Cr in 24 h urine or in a spot urine has been shown (Powles et al., 1976 between serum Hyp and urinary Hyp, the measurement of urinary Hyp in either a 24 h or in a spot urine is of more clinical relevance. In view of the results reported elsewhere (Gasser et al., in preparation) the little (though significant) influence of collagen intake on Hyp/Cr in a spot urine makes this parameter the measurement of choice in screening patients for skeletal involvement of malignant disease. The same parameter may be useful for monitoring the effect of treatment in this group of patients.