Pulmonary function in patients with trophoblastic disease treated with low-dose methotrexate.

The Sheffield Trophoblastic Disease Centre treats about 25 patients with persistent trophoblastic disease each year. A total of 75% of patients are classified as low risk according to the Charing Cross Hospital prognostic scoring system and receive methotrexate (MTX) 50 mg, i.m., on days 1, 3, 5, 7 with folinic acid 7.5 mg orally 24 h after each methotrexate injection. There is a 7-day rest between treatment cycles. Remission is achieved in 85% of cases. Approximately 20% of patients experienced pleuritic chest pain and dyspnoea. We have evaluated prospectively lung function in 16 low-risk patients receiving methotrexate. All patients had pulmonary function tests [spirometry-forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC), peak expiratory flow rate (PEFR), and transfer factor - TLCO, kCO] performed before and after completed treatment. A mean of 7.5 cycles of MTX were administered (range 4-11). There was a significant reduction in the mean TLCO (mean pre/post 8.15/7.38 mmol min-1 kPa-1, P = 0.01), but there were no other statistically significant changes. Three patients experienced respiratory symptoms and were found to have a 39%, 28%, and 11% reduction in TLCO from baseline, improving on follow up to pretreatment levels. Low-dose MTX is an effective therapy but may cause troublesome pulmonary toxicity.

The low-dose methotrexate regimen is usually well tolerated. The most common side-effects of treatment are conjunctivitis and oral mucositis, which are usually ameliorated with appropriate therapy. However, approximately 20% of patients (Sheridan et al, 1993) experience the potentially disabling side-effects of pleuritic chest pain and dyspnoea. If severe, this can lead to the discontinuation of methotrexate therapy and alteration of treatment to other chemotherapeutic agents.
In a number of conditions, pulmonary toxicity with the clinical and pathological features of pneumonitis has been described after methotrexate administration (Sostman et al, 1976). However, the association between pleurisy and pulmonary function has not been previously investigated. The objective of this study was to prospectively assess pulmonary function in patients with persistent trophoblastic disease treated with single-agent methotrexate.

MATERIALS AND METHODS Subjects
This report describes our findings in 16 consecutive patients with low-risk persistent trophoblastic disease who required treatment with methotrexate between February 1994 and August 1995. The median age of the group was 25.5 years (range 15-35), median risk score 4 (range 2-6) and median number of treatment cycles 8 (range 4-11).
Only 2 out of the 16 patients had pulmonary metastases detectable by chest computerized tomography (CT) scan.

Treatment
The low-dose methotrexate regimen consists of methotrexate 50 mg, i.m., administered on days 1, 3, 5 and 7 with folinic acid 7.5 mg orally 24 h after each methotrexate injection. There is a Pulmonary function analysis was performed in each subject before commencement of the first treatment cycle and at the conclusion of the last cycle. The technicians performed the tests in a singleblind fashion, being unaware of the patient's condition or treatment. Those patients who experienced respiratory symptoms during treatment were recalled for follow-up analysis. The following pulmonary function tests were used in the assessment: (a) Forced expiratory volume in 1 (FEVy), forced vital capacity (FVC) and peak expiratory flow rate (PEFR) were measured by spirometer (Vitalograph Compact).
(b) Single breath transfer factor (TLCO) was measured on the Gould Diffumatic.
The transfer factor is a measure of gas transfer from the alveoli to the blood in the pulmonary capillaries. Transfer is measured for the whole lungs -7nCO, and the coefficient per unit of alveolar volume (kCO) can be calculated. The transfer factor has become established as a sensitive and reproducible routine pulmonary function test. In our laboratory, standard methods are adopted for measuring transfer factor (Guidelines for the Measurement of Respiratory Function, 1994). The coefficient of variation for the test is approximately 5% -although this is reduced in our laboratory by repeating the test and reporting the mean value.

Statistical analysis
All pulmonary function tests yield biological data that are normally distributed. As the number of observations in our study is relatively small, paired t-tests for means with the appropriate degrees of freedom have been used for statistical analysis of the data. Table 1 shows the results for the 16 subjects in the group as a whole. It can be seen that there is no significant change in PEFR, FEVy, FVC, kCO, alveolar volume or haemoglobin. There is, however, a significant reduction in transfer factor (TLCO) from a mean of 8.15 mmol min-' kPa-' before treatment to 7.38 mmol min-' kPa-1 after treatment (P = 0.01). experienced severe symptoms of dyspnoea and pleuritic chest pain during the second, third and fourth treatment cycles. Her chemotherapy was therefore altered to an oral etoposide regimen. The patient experienced no further respiratory problems and completed treatment 2 months after biochemical remission. (A treatment summary is shown in Figure   By studying the group as a whole, it is impossible to analyse the effects of treatment on individual patientsand it is these data that may prove to be far more relevant. Figure 1 shows pre-and posttreatment kCO values for all subjects studied. The effect of treatment is varied, with some patients having an increase and others a decrease in kCO. It is, however, of note that the three patients experiencing the largest falls in kCO were those who developed respiratory symptoms during treatment, suggesting that such patients may have measurable lung function impairment. These cases are described below.

RESULTS
Case 1 A 23-year-old woman with persistent trophoblastic disease was referred for assessment in February 1994. Despite two attempts to evacuate her uterus, she had persistent bleeding per vaginum and an elevated 3hCG. Staging investigations were performed; the patient had a risk score of 4 and was started on the lowdose methotrexate regimen. The patient achieved biochemical remission by the commencement of the third cyclehowever, she Case 2 After the diagnosis of a molar pregnancy, this 28-year-old patient had three attempted uterine evacuations. However, her f-hCG remained persistently elevated and she was assessed here in November 1994. The patient had a Charing Cross risk score of 4 and was therefore treated with methotrexate. After two courses the patient was in biochemical remission and went on to complete five cycles of therapy. However, during the fourth and final cycles she experienced severe dyspnoea and pleuritic chest pain. The patient experienced no further respiratory symptoms after the conclusion of treatment.

Case 3
An 18-year-old woman presented to her local gynaecologist with bleeding per vaginum in early pregnancy. After a pelvic ultrasound scan, a diagnosis of suspected molar pregnancy was made and the patient underwent a suction curettage. Histological analysis of the products of conception confirmed the diagnosis. Despite a further uterine evacuation, the patient's bleeding persisted and ,BhCG remained elevated. On review at the treatment centre, in December 1994, the patient had a risk score of 4 and was commenced on methotrexate. The patient had a haemoglobin count of 8 g dl-I and required a blood transfusion during her first cycle of treatment. After three cycles the patient was in biochemical remission and treatment was concluded after seven cycles. During the fourth and fifth cycles she experienced pleuritic chest pain and dyspnoea. Simple analgesia was used prophylactically in subsequent cycles to good effect.
None of the three patients had chest metastasis detectable on CT scans before commencing chemotherapy. All had numerous investigations performed when they experienced their respiratory symptoms. Of note, no infectious agents were identified on sputum culture and there were no abnormal findings on electrocardiogram or chest radiograph. A ventilation-perfusion scan was performed on patient number 2 and no abnormality was detected. Simple analgesia was used to provide pain reliefopiates being used in case 1 only when this provided inadequate symptom relief. Table 2 shows the percentage changes in TLCO and kCO experienced by these three symptomatic patients. No significant changes were seen in any other parameter measured.
At the conclusion of the study we recalled these patients and again performed pulmonary function analysis. As shown in Table  2, in all patients the TLCO and kCO changes have resolved or appear to be returning to baseline.

DISCUSSION
Methotrexate is a commonly used chemotherapeutic agent. In addition to its use in a wide variety of neoplastic conditions, it is also used to treat severe unresponsive psoriasis and other severe immunologically mediated disease states. In a number of these British Journal of Cancer (1997) 76(10) I   x x v x lic conditions, pulmonary complications have been described following methotrexate administration (Sostman et al, 1976).
The dosage of methotrexate received by a patient is dependent on the condition being treated. Pneumonitis is experienced by those receiving both high and low dosesindeed some earlier investigators found the complication more common in those receiving small oral doses (Clarysse et al, 1969). A conclusion was drawn that the route of administration was the dependent factor (Zurek et al, 1968). This, however, was not supported by subsequent studies (Nesbit et al, 1976), and pharmacological data suggest that this particular complication is a schedule-related, not doseor route-related, phenomenon (Louis et al, 1970;Huffman et al, 1973). This may explain the high incidence of methotrexateinduced pneumonitis in our patients, in whom low doses of the drug were administered relatively frequently.
The pathogenesis of methotrexate-induced pneumonitis is not fully understood. Previous investigators have performed lung biopsies in symptomatic patients, histological analysis being typified by alveolar damage, interstitial inflammation, alveolar space infiltrates and hyaline membrane formation (Sostman et al, 1976). Others suggest that this condition is the result of an immunological cell-mediated mechanism (Akoun et al, 1987;White et al, 1989), although clinical data suggest that repeat exposure to methotrexate seldom leads to a worsening of the pneumonitis (Sostman et al, 1976;Willson, 1978).
It has previously been shown that the transfer factor may alter with methotrexate-induced pneumonitis (Arnett et al, 1973). Our results show that transfer factor is significantly reduced in a cohort of patients receiving methotrexate for persistent trophoblastic disease. It should be noted, however, that this alteration of lung function in our study group is heavily influenced by the results in our symptomatic patients, and if all observations recorded by these patients are removed from the statistical analysis there is no significant change in any parameter.
The transfer factor (7LCO) is a measure of gas transfer from the alveoli to the blood in the pulmonary capillaries. In human lungs the diffusion pathway consists of the alveolar space, alveolar epithelium, alveolar basement membrane, tissue space, capillary basement membrane, capillary endothelium, capillary lumen and the erythrocyte. The transfer of oxygen is by means of simple diffusion. It is influenced by the surface area over which transfer takes place, the length and permeability of the diffusion pathway, the partial pressure gradient of oxygen and the rate of oxygen uptake by haemoglobin in the erythrocyte. Predicted values are standardized for age, height and sex.
Dividing TLCO by alveolar volume gives the coefficient of transfer (kCO). This assesses if a reduced transfer factor is caused by a diminished alveolar volume with a normal diffusion pathway or a normal alveolar volume with an impaired diffusion pathway. Figure 1 shows the alteration of kCO with methotrexate treatment in our study group. The effect of treatment on the kCO is not consistent; however, large falls were demonstrated in the three symptomatic patients. These patients (and some others to a lesser degree) have apparently undergone a methotrexate-induced deterioration in diffusion pathway function. One could postulate that the critical level of loss of function is required for the patients to experience symptoms. In our study, we did not perform bronchoalveolar lavage or lung biopsy on any patients. It is not, therefore, possible to state definitely the pathological process responsible for the falls in kCO. However, either an immune alveolitis or an inflammatory infiltrate into the tissue space would result in an impairment of diffusion pathway functionand so could explain our findings. The effect on function would appear to be temporary, as all symptomatic patients have kCO values that revert towards baseline at long-term follow-up.
The treatment of methotrexate-induced pneumonitis can be difficultin some cases it leads to the alteration of the chemotherapeutic regimen despiie adequate tumour response. There is anecdotal evidence that adequate hydration and increasing the dosage of folinic acid rescue may reduce the incidence of this complication. For those who develop pleurisy, simple paracetamol-containing analgesia may provide adequate symptomatic relief; if not, the use of tramodol hydrochloride may reduce the need to resort to conventional opiate analgesia. Non-steroidal antiinflammatory drugs are not recommended as they reduce renal methotrexate excretion. The role of corticosteroids in the treatment of this complication is unproven.
In conclusion, methotrexate is a highly effective first-line chemotherapeutic agent for treatment of low-risk persistent trophoblastic disease. It is the drug of choice because of its efficacy and relative lack of short-and long-term toxicity (Bagshawe et al, 1989). Unlike other agents active in the treatment of this disease, methotrexate has not, so far, been associated with subfertility or second tumour induction (Rustin et al, 1987(Rustin et al, , 1996. Treatment with low-dose methotrexate is usually well tolerated; however, up to 20% of patients will experience methotrexateinduced pulmonary toxicity. This study shows that such toxicity can be associated with significant, though temporary, changes in pulmonary function. Consideration should be given to altering therapy in patients who develop respiratory problems and/or show a deterioration in, pulmonary function. Low-dose methotrexate is not an innocuous treatment, however, it should still be regarded as standard treatment for patients with persistent low-risk gestational trophoblastic disease.