Kristian Aabo

Chemotherapy for adenocarcinoma of the lung (WHO III): A randomized study of vindesine versus lomustine, cyclophosphamide, and methotrexate versus all four drugs

Two hundred seventy-nine patients with previously untreated nonresectable adenocarcinoma of the lung (ACL) entered a prospective randomized trial, comparing vindesine (VDS) to a combination of lomustine (CCNU), cyclophosphamide (CTX), and methotrexate (MTX), and to a regimen including all four drugs. Response assessment was possible in 218 patients, while 259 were evaluable for survival. Response rates were similar (22%, 23%, and 27%, respectively) as were median durations of response (15 weeks overall) and survival (29 weeks overall). Patients with dose-limiting toxicity had significantly higher response rate and longer survival than patients without toxicity. The major toxicity was peripheral neuropathy with VDS treatment and myelosuppression with the other two regimens. The VDS single-agent activity in ACL was confirmed, but addition of VDS to the three-drug regimen did not increase activity. Future studies of VDS in combination with other active agents, and comparison to a matched control group on supportive care, are indicated.

In vitro and in vivo evaluation of the indoloquinone EO-9 (NSC 382 459) against human small cell carcinoma of the lung☆

As the indoloquinone EO-9 has previously shown activity in several tumor model systems it was evaluated against four human small cell lung cancer cell lines by the clonogenic assay. In two cell lines (Nyh and Tol), exponential dose-response curves were achieved with both 1 h and continuous exposure, whereas no cell kill was obtained in the other two cell lines (69 and 592) when tested with 1 h incubation up to 0.25 microgram/ml. When the cells were exposed to drug in vitro, flow cytometric DNA analysis showed perturbations in the cell cycle distribution of the most sensitive cell line (Tol) at a lower EO-9 concentration than in the less sensitive cel line (592). This in vitro predicted difference in EO-9 sensitivity between two of the cell lines (592 and Tol) was confirmed when the cell lines were heterotransplanted to nude mice.

A randomized study of single agent vs combination chemotherapy in FIGO stages IIB, III and IV ovarian adenocarcinoma

From 1977 until 1980, 179 patients with newly diagnosed FIGO stages IIB, III or IV ovarian adenocarcinoma were randomized in a two-armed clinical trial: dihydroxybusulfan (B) 600 mg/m2 p.o. for 4-6 weeks q 12 weeks or cyclophosphamide (C) 150 mg/m2 p.o. for 7 days q 4 weeks vs a combination of cyclophosphamide 400 mg/m2 i.v., doxorubicin 30 mg/m2 and 5-fluorouracil 400 mg/m2 i.v., days 1 and 8 q 4 weeks (CAF). In addition, stage IIB patients were randomised to +/- pelvic irradiation. The patients were stratified according to anatomic stage. The treatment groups were comparable with respect to performance status, age and histology. Twenty-three patients were excluded because of protocol violation, leaving 156 patients evaluable for survival with an observation period of 3-6 yr. Twenty patients were in stage IIB, while the remaining 136 patients were classified as stages III and IV. No statistically significant difference was found in survival or response between the two single agents. The overall median response rate (single drug: 27%; CAF: 47%) and the median response duration (single drug: 5 months; CAF: 10 months) were significantly superior for the CAF group compared to the single agent group. No statistical difference in median survival was observed between single-drug treatment (12 months) and CAF (14 months), despite the fact that responders lived significantly longer than nonresponders (17 vs 10 months). In stage IIB patients receiving chemotherapy no benefit of pelvic irradiation was found. Thirty patients (19%) underwent second-look laparotomy, with 15 (50%) being completely free of disease. So far only one patient (7%) has relapsed. Two additional patients, who had microscopic disease removed at second-look laparotomy, seem to have been rescued by this procedure.

Carcinoembryonic antigen (CEA) and alkaline phosphatase in progressive colorectal cancer with special reference to patient survival.

The prognostic value of serial CEA tests was evaluated in 175 consecutive patients with progressive colorectal cancer who subsequently died of their disease. The upper normal plasma CEA limit was determined to be 8 ng/ml from serial CEA determinations in 31 patients radically operated on for colorectal cancer and observed in median 40 months without evidence of recurrence. A CEA value of greater than 8 ng/ml was highly suggestive of residual disease or recurrence, even when no clinical evidence was present. Approximately 90% of the patients dying from colorectal cancer showed an increase in CEA to greater than 8 ng/ml during the course of the disease. In 63% of the patients CEA increase preceded clinical progression or relapse, with a median time period of 4 months. Sixty-eight per cent of the patients had rising CEA values over an extended time period of many months, 14% had a preterminal increase, 13% had constantly normal and 5% constantly elevated CEA. As 6/9 patients developed a drop in CEA in relation to initiation of chemotherapy without clinical response, it is concluded that CEA is not a reliable indicator of clinical response to chemotherapy. Patients with liver metastases had higher CEA and alkaline phosphatase levels than patients with only localized disease. However, no good statistical correlation between CEA and serum alkaline phosphatase was found in patients with liver metastases (coefficient of correlation r = 0.35). An increase in CEA from normal to above 8 ng/ml predicted a decrease in survival time of median 60% counted from the time of diagnosis. The numerical CEA value was predictive of shortening of survival only when greater than 3000 ng/ml. Such high values were observed only in a minority of the patients (12%). Greater than 1000 U/l (27% of the patients) alkaline phosphatase predicted an extremely poor prognosis, with a median survival of 1 month (range 0.5-4 months). It is concluded that a rise in CEA to greater than 8 ng/ml indicates with high degree of certainty relapse or disease progression in colorectal cancer patients. CEA is not a reliable indicator of clinical response to chemotherapy, and an increase in the CEA level is of little prognostic value concerning survival. Alkaline phosphatase seems to be a more valuable predictor of a worsening of prognosis.

Prostate cancer: Evaluation of response to treatment, response criteria, and the need for standardization of the reporting of results

Response criteria and the reporting of results in clinical trials on drug therapy of stage D prostate cancer were evaluated by examination of studies listed in the Index Medicus 1980-1984. During this 5-year period, 70 studies (51 phase II and 16 phase III) were listed, comprising 3184 evaluable patients. Among 346 patients reported as having evaluable disease according to the WHO criteria, 198 had well-defined evaluable disease. A variety of response criteria were used, the NPCP criteria being the most frequent. Only three studies included solely patients with evaluable disease according to the WHO criteria. Reporting of results was often inadequate. The value of the most frequently used response parameters such as acid phosphatase, bone scan, per-rectal ultrasound, CT scan, bone pain and performance status is discussed. A system to standardise the reporting of results is proposed.

High-dose ketoconazole to untreated stage D prostate cancer

Eleven previously untreated patients with stage D prostate cancer were treated with ketoconazole in a dosage of 400 mg p.o. every 8 h. s-Testosterone was used as a measure of antiandrogen effect. Nine patients had a reduction in s-testosterone to castrate levels (less than 2.9 nmol/l) within 3 days. In the remaining two patients, dose escalation of ketoconazole to 400 mg every 6 h did not lead to sufficient reduction in s-testosterone. Two patients had a complete response and four patients had a partial response of 6/11. Additionally, two patients had bone pain relief without normalization of acid phosphatase. Side-effects and adverse reactions were prominent, causing discontinuation of the treatment in nine patients. It is concluded that high-dose ketoconazole is effective in disseminated prostate cancer, but the high frequency of side-effects makes it less attractive than conventional hormone manipulations like castration or estrogens.

Clonal dominance between subpopulations of mixed small cell lung cancer xenografts implanted ectopically in nude mice

Clonal evolution of neoplastic cells during solid tumour growth leads to the emergence of new tumour cell subpopulations with diverging phenotypic characteristics which may alter the behaviour of a malignant disease. Cellular interaction was studied in mixed xenografts in nude mice and during in vitro growth of two sets of small cell lung cancer (SCLC) subpopulations (54A, 54B and NYH, NYH2). The tumour cell lines differed in cellular DNA content enabling flow cytometric DNA analysis (FCM) to be used to monitor changes in the fractional composition of the mixed cell populations. The progeny clone 54B was found to dominate the parent 54A clone when grown as mixed subcutaneous xenografts in nude mice, whereas no dominance was exerted during in vitro growth. The in vivo dominance could not be explained by differences in growth kinetics between the two tumour cell lines, and the interaction was not dependent on 54B being in excess in mixed tumours. The dominance was dependent on close in vivo contact as no remote effect on the growth of 54A was found when the dominating 54B cells were growing in the opposite flank of tumour-bearing mice. Irradiation inactivated 54B cells were unable to exert the dominating effect, indicating that the interaction required viable and proliferating cells. Clonal dominance was not found in mixed NYH-NYH2 tumours indicating that the dominance mechanism(s) may not always be operational between subpopulations in heterogeneous tumours. Recognition of interaction between tumour cell populations may result in a better understanding of the behaviour of heterogeneous human malignancies.

Interactions between three subpopulations of Ehrlich ascites tumor and a P388 murine leukemia in mixed solid tumors in immune competent mice

Cellular interactions between three subpopulations of Ehrlich ascites tumor and between these and the P388 murine leukemia were studied during growth of solid tumors obtained by mixtures of the cells in immune competent N/D mice. An immunogenic Ehrlich cell line (E1.15) induced an immunologically based growth inhibition of the two other Ehrlich cell lines (E1.80 and E1.95) which themselves were non‐immunogenic. E1.15 was, however, unable to induce an immunological response against the P388 cell line. It is therefore suggested that when in close contact, immunologically induced cellular responses imposed by an immunogenic cell line on other cell lines require genetic and thereby close immunogenic resemblance between the cell lines. Another type of interaction was found between the E1.95 cell line and the P388 line which showed nearly identical growth characteristics as determined by tumor weight day 14, tumor growth curves, cell cycle times (per cent labelled mitoses) and cell cycle distributions (flow cytometric DNA analysis). After 2 weeks of growth of mixed P388/E1.95 tumors, flow cytometric DNA analysis on fine‐needle tumor aspirates showed nearly total dominance of P388. This type of interaction required close cellular contact of viable cells, and no cellular immune response was elicited by the host animals. A third finding was that a faster growing Ehrlich cell line E1.95 dominated the tumors when inoculated in mixture with a slower growing subpopulation E1.80. This could be explained on the basis of the cell kinetic differences between these two cell lines.

Interaction between three subpopulations of Ehrlich carcinoma in mixed solid tumours in nude mice: evidence of contact domination.

Clonal interaction between three subpopulations of Ehrlich carcinoma were studied during growth as mixed solid tumours and as ascites tumours in immune-incompetent nude NMRI mice. The tumour cell lines differed in DNA content as determined by DNA flow cytometry (FCM). Tumour growth was evaluated by tumour growth curves including calculation of tumour volume doubling times, tumour weight on day 14, cell cycle times (per cent labelled mitoses) and cell cycle distributions (FCM). Two subpopulations (E1.15 and E1.95) showed nearly identical growth characteristics during both solid and ascites tumour growth. The third subpopulation (E1.80) grew more slowly. FCM on fine-needle tumour aspirates was used to determine the relative proportions of the cell populations in mixed solid tumours in which E1.95 showed a growth-dominating effect on E1.15. No such effect was demonstrated during single-cell tumour growth in ascitic fluid in which the cells had no intimate contact. Ascitic fluid from E1.95-bearing animals or radiation-killed E1.95 cells had no effect on the growth of E1.15, and no remote effect was seen when the two cell lines were growing in opposite flanks. This indicates that only viable E1.95 cells in close in vivo contact were able to induce growth inhibition of the E1.15 subpopulation. Both the E1.95 and the E1.15 cells dominated the E1.80 cells, but in these cases cell kinetic differences may have played a role as the E1.95 and the E1.15 lines grew faster than the E1.80. The E1.80 cell line had no dominating effect on the E1.15 or E1.95. It is concluded that non-immunologically mediated cellular dominance in heterogeneous tumours may contribute to the evolution of these tumours and may be involved in fundamental tumour biological phenomena.

cis-dichlorodiammineplatinum (II) and hexamethylmelamine in advanced ovarian carcinoma: A phase II study

Thirty-eight previously treated patients with ovarian carcinoma received a combination of cis-dichlorodiammineplatinum (II) (CDDP) and hexamethylmelamine (HMM). The schedule was CDDP 75 mg/m2 i.v. with forced diuresis on day 1, followed by HMM 200 mg/m2 p.o. on days 8-21, repeated every 4 weeks. In 29 evaluable patients an overall response rate (CR + PR) of 35% with a median response duration of 4.5 months was observed. The performance status seemed to be an important prognostic factor. The gastrointestinal- and neurotoxicities were severe and resulted in dose modification and/or drug discontinuation in half of the patients. In conclusion, CDDP and HMM is an active drug combination in advanced ovarian carcinoma resistant to conventional chemotherapy. A possible superiority of this combination compared with CDDP or HMM alone has to await randomized trials.