Oral sodium phosphate and sodium picosulfates/magnesium citrate are commonly used to evacuate the colon and rectum before colonoscopy or colorectal surgery. These substances, however, are known to cause electrolyte abnormalities. Seizures caused by electrolyte abnormalities associated with bowel preparation have only rarely been reported. We report the cases of three patients with no prior history of seizures, who had their first seizure associated with hyponatremia following ingestion of sodium phosphate or sodium picosulfates/magnesium citrate combination. Care must be taken with patients with a low seizure threshold and those with possible chronic sodium depletion, such as patients on thiazide diuretics, who are undertaking bowel preparation with oral sodium phosphate or sodium picosulfates/magnesium citrate combination.

Objective To re-evaluate a national prospective study in for relapse in this series. Only one relapse occurred >28 months after orchidectomy. Despite poor comNew Zealand after 17 years to define whether orchidectomy alone and surveillance for nonseminoma pliance in some patients (12%) their survival was not prejudiced. Three patients died from disease despite germ cell testicular tumour (NSGCTT) is a sound policy and matches the results achieved by other chemotherapy at relapse. At 17 years and a median follow-up of 53 months, 242 of the 248 men are treatment protocols. Patients and methods Between 1980 and 1997, 248 disease-free and the disease-specific survival rate is 98%. men with stage I NSGCTT, from six New Zealand centres, were managed by orchidectomy alone and Conclusions This study shows that orchidectomy alone and treatment of relapses produces excellent longsurveillance, with treatment of relapses using combination chemotherapy.term results without the adverse eCects associated with retroperitoneal node dissection or elective chemoResults Seventy of the 248 patients (28%) relapsed; 42 of 92 (46%) with vascular and/or lymphatic invasion therapy for high-risk cases. Keywords Stage I nonseminoma, testicular cancer, (VLI) in the primary tumour relapsed, whereas only 26 of 151 (17%) without this feature relapsed surveillance, long-term follow-up, orchidectomy (P<0.001). VLI was the only identifiable risk factor ommend that high-risk patients should be oCered adju-

For assessment of the ototoxic potential of carboplatin [cis-diammine-1,1-cyclobutane dicarboxylate platinum(II); CBDCA], pure-tone audiograms were evaluated in 27 patients receiving a total of 119 doses of carboplatin in the range of 300-400 mg/m2. Pure-tone audiometry (PTA) was done immediately prior to and 4 weeks after the administration of 80 doses (67%). Defining carboplatin ototoxicity as an increase of greater than or equal to 30 dB in auditory thresholds that was unexplainable by other causes, we identified 5 examples (19%). Hearing loss tended to be cumulative with increasing dose and was always maximal at 8,000 Hz. Two patients had an increase in auditory thresholds at 1,000 Hz, but this only amounted to 10 dB in each case. Patients developing ototoxicity tended to be older. Sex, the pre-treatment creatinine clearance, the pretreatment audiogram, the number of doses, and the cumulative dose did not emerge as being reliable predictors of subsequent ototoxicity. We conclude that although carboplatin is ototoxic, clinically significant deafness does not occur with conventional dosing and routine audiometric monitoring is therefore unnecessary. However, we suggest that caution should be exercised when carboplatin is given either at higher doses or for longer periods when there is concomitant use of other potentially ototoxic agents or when there is significant pre-existing auditory impairment.

No abstract

Significant hyponatraemia occurred in 26 (31%) of the patients, but in 12 of these it appeared likely that pseudohyponatraemia caused by IVIG was a contributing factor. Hyponatraemic patients have a poorer prognosis than eunatraemic patients, however it is difficult to separate this factor from other poor prognostic factors (older age and PPV).

Forty-six patients with clinical stage I testicular non-seminomatous germ cell tumours were followed up according to a protocol of active surveillance between 1979 and 1987. The median follow-up time was 40+ months. Thirteen patients (28%) relapsed, predominantly in retroperitoneum and/or lung. Ten of these relapses (76%) occurred within 8 months of orchiectomy. Relapses occurred in 7/35 T1 tumours and 5/10 T2 to T4 tumours. No correlation was detected between the histological type and relapse rate. Three late relapses were diagnosed at 23, 29 and 36 months. Eleven of the relapsed patients remain in prolonged complete remission after PVB chemotherapy +/- surgery; one patient, who initially refused treatment at the time of relapse, has died. Another relapsed with predominant elements of rhabdomyosarcoma intermingled with malignant teratoma in a bone metastasis. He had a partial response to PVB chemotherapy but subsequently died. Thirty-four patients (74%) did not undergo lymphography (LG) and had a higher relapse rate (11/34) than those who had LG (2/12); this was not a statistically significant difference in this small series. The policy of active surveillance is not yet the "state of the art" and should be under constant scrutiny with respect to safety and practice.

Forty-six patients with clinical stage I testicular non-seminomatous germ cell tumours were followed up according to a protocol of active surveillance between 1979 and 1987. The median follow-up time was 40 + months. Thirteen patients (28%) relapsed, predominantly in retroperitoneum and/or lung. Ten of these relapses (76%) occurred within 8 months of orchiectomy. Relapses occurred in 7/35 T1 tumours and 5/10 T2 to T4 tumours. No correlation was detected between the histological type and relapse rate. Three late relapses were diagnosed at 23, 29 and 36 months. Eleven of the relapsed patients remain in prolonged complete remission after PVB chemotherapy +surgery; one patient, who initially refused treatment at the time of relapse, has died. Another relapsed with predominant elements of rhabdomyosarcoma intermingled with malignant teratoma in a bone metastasis. He had a partial response t o PVB chemotherapy but subsequently died. Thirty-four patients (74%) did not undergo lymphography (LG) and had a higher relapse rate (1 1/34) than those who had LG (2/12); this was not a statistically significant difference in this small series. The policy of active surveillance is not yet the "state of the art" and should be under constant scrutiny with respect to safety and practice.

A higher incidence of germ cell testicular cancer was found in Maoris (6.84/100,000) compared with non-Maoris (5.26/100,000) in New Zealand from 1975 to 1986, especially in the 15-49 year age group (Maoris 12.30/100,000, non-Maoris 9.47/100,000; P = 0.04). Previous studies have shown Whites to have the highest incidence of this malignancy. Possible reasons for this and some other epidemiological features are discussed.