Contrastive diagnostic embedding (CDE) model for automated coding - A case study using emergency department encounters

BACKGROUND

Billing codes are utilized for medical reimbursement, clinical quality metric valuation and for epidemiologic purposes to report and follow disease trends and outcomes. The current paradigm of manual coding can be expensive, time-consuming, and subject to human error. Though automation of the billing codes has been widely reported in the literature via rule-based and supervised approaches, existing strategies lack generalizability and robustness towards large and constantly changing ICD hierarchical structure.

METHOD

We propose a weakly supervised training strategy by leveraging contrastive learning, contrastive diagnosis embedding (CDE) to capture the fine semantic variations between the diagnosis codes. The approach consists of a two-phase contrastive training for generating the semantic embedding space adapted to incorporate hierarchical information of ICD-10 vocabulary and a weakly supervised retrieval scheme. Core strength of the proposed method is that it puts no limit on the 70 K ICD-10 codes set and can handle all rare codes for coding the diagnosis.

RESULTS

Our CDE model outperformed string-based partial matching and ClinicalBERT embedding on three test cases (a retrospective testset, a prospective testset, and external testset) and produced an accurate prediction of rare and newly introduced diagnosis codes. A detailed ablation study showed the importance of each phase of the proposed multi-phase training. Each successive phase of training - ICD-10 group sensitive training (phase 1.1), ICD-10 subgroup sensitive training (phase 1.2), free-text diagnosis description-based training (phase 2) - improved performance beyond the previous phase of training. The model also outperformed existing supervised models like CAML and PLM-ICD and produced satisfactory performance on the rare codes.

CONCLUSION

Compared to the existing rule-based and supervised models, the proposed weakly supervised contrastive learning overcomes the limitations in terms of generalization capability and increases the robustness of the automated billing. Such a model will allow flexibility through accurate billing code automation for practice convergence and gains efficiencies in a value-based care payment environment.

Results of the c-TRAK TN trial: a clinical trial utilising ctDNA mutation tracking to detect molecular residual disease and trigger intervention in patients with moderate- and high-risk early-stage triple-negative breast cancer

BACKGROUND

Post-treatment detection of circulating tumour DNA (ctDNA) in early-stage triple-negative breast cancer (TNBC) patients predicts high risk of relapse. c-TRAK TN assessed the utility of prospective ctDNA surveillance in TNBC and the activity of pembrolizumab in patients with ctDNA detected [ctDNA positive (ctDNA+)].

PATIENTS AND METHODS

c-TRAK TN, a multicentre phase II trial, with integrated prospective ctDNA surveillance by digital PCR, enrolled patients with early-stage TNBC and residual disease following neoadjuvant chemotherapy, or stage II/III with adjuvant chemotherapy. ctDNA surveillance comprised three-monthly blood sampling to 12 months (18 months if samples were missed due to coronavirus disease), and ctDNA+ patients were randomised 2 : 1 to intervention : observation. ctDNA results were blinded unless patients were allocated to intervention, when staging scans were done and those free of recurrence were offered pembrolizumab. A protocol amendment (16 September 2020) closed the observation group; all subsequent ctDNA+ patients were allocated to intervention. Co-primary endpoints were (i) ctDNA detection rate and (ii) sustained ctDNA clearance rate on pembrolizumab (NCT03145961).

RESULTS

Two hundred and eight patients registered between 30 January 2018 and 06 December 2019, 185 had tumour sequenced, 171 (92.4%) had trackable mutations, and 161 entered ctDNA surveillance. Rate of ctDNA detection by 12 months was 27.3% (44/161, 95% confidence interval 20.6% to 34.9%). Seven patients relapsed without prior ctDNA detection. Forty-five patients entered the therapeutic component (intervention n = 31; observation n = 14; one observation patient was re-allocated to intervention following protocol amendment). Of patients allocated to intervention, 72% (23/32) had metastases on staging at the time of ctDNA+, and 4 patients declined pembrolizumab. Of the five patients who commenced pembrolizumab, none achieved sustained ctDNA clearance.

CONCLUSIONS

c-TRAK TN is the first prospective study to assess whether ctDNA assays have clinical utility in guiding therapy in TNBC. Patients had a high rate of metastatic disease on ctDNA detection. Findings have implications for future trial design, emphasising the importance of commencing ctDNA testing early, with more sensitive and/or frequent ctDNA testing regimes.

Adult Survivors of Childhood Cancer: Views on COVID-19 and Vaccination

Purpose/Objective(s)

Adult childhood cancer survivors (ACCS) are at increased risk of developing late effects because of their childhood cancer treatment, including cognitive delay, diabetes, metabolic syndrome, and organ damage. Consequently, many ACCS may be at increased risk for worse outcomes with COVID-19 infection. It is important to determine ACCS views on the COVID-19 pandemic and vaccination.

Materials/Methods

A non-validated survey was created using multi-disciplinary input. Prior to the widespread rollout of COVID-19 vaccinations in Canada, we emailed an online survey to 235 ACCS followed through the BC Cancer Late Effects and Follow-Up clinic who had provided informed consent to email contact, receiving 89 responses (37.9% RR) which were analyzed. Statistical analysis was calculated using Chi-Squared test of association.

Results

Survey respondents were majority female (61%). The most common age range was 30-39 (30%) followed by 20-29 (28%). Most were of European descent (47%) and lived in an urban center (75%). The vast majority completed high school (97%), as well as post-secondary education (PSE, 73%). Only 29% did not use Complimentary or Alternative medicines (CAM), with herbal products being most common (48%), as well as massage therapy or other bodywork (46%). The most reported sources of health information were primary care practitioners (PCP, 80%), traditional media (60%), and specialized hospital clinics (46%). Of all respondents, 67% believed that ACCS should be prioritized for vaccination, with 87% indicating they would receive a COVID-19 vaccination if available. 78% and 89% believed that COVID-19 was a serious health problem for themselves or others, respectively. Views were evaluated across multiple subgroups. Respondents who had completed PSE were more likely to see COVID-19 as a risk to themselves (80% vs 71%, p = 0.358), to others (94% vs 75%, p = 0.012), and more likely to receive a vaccine (89% vs 79%, p=0.217). Respondents who received health information from traditional media felt COVID-19 was more likely to harm themselves (79% vs 75%, p=0.638), to harm others (93% vs 83%, p=0.181) and more likely to receive a vaccine (93% vs 78%, p=0.047). A similar trend was seen in those who receive information from a PCP, but without statistical significance. The opposite is true if health information is received from family or friends, with these respondents being less likely to feel COVID-19 poses a risk of harm to themselves (76% vs 78%, p=0.79), less likely to harm others (86% vs 90%, p=0.595), and a lower likelihood to get a vaccine (76% vs 92%, p=0.041). There was no discernable trend when examining by further subgroups, including CAM usage, age group, location, or other sources of health information.

Conclusion

Many ACCS appear to underestimate their risk from COVID-19; whether patients had completed PSE and the location from which they receive health information appeared to correlate most strongly with these results.

Evolving Practice Patterns Over Two Decades (1993-2013) in the Management of Desmoid-type Fibromatosis in British Columbia

AIMS

Due to the rarity and varied natural history of desmoid-type fibromatosis, evidence-based treatment standards for this disease remain lacking. This study evaluated outcomes in patients with desmoid-type fibromatosis managed at a Canadian institution over two decades.

MATERIALS AND METHODS

Records of 227 patients with desmoid-type fibromatosis referred from 1990 to 2013 were retrospectively reviewed to investigate management strategies including active surveillance, surgery, radiation therapy, cryoablation, and systemic therapy, including tamoxifen and chemotherapy.

RESULTS

Thirty-two per cent of cases were men, median age 40 years, median tumour size 5.4 cm. Initial treatments were surgery (79%), tamoxifen (13%), radiation therapy (5.0%), chemotherapy (1.8%) and cryoablation (1.2%). Active surveillance was used upfront in 26% of cases, most after 2005. At a median follow-up of 77 months, one patient died of disease, 13 died of unrelated causes and the remainder were alive with no evidence of disease (56%), stable/responding disease (33%) or progressive disease (4%). The recurrence rate was 25% after upfront surgery. Response rates and disease control rates were 40% and 76% for active surveillance; 68% and 96% for radiation therapy; 31% and 67% for tamoxifen; and 53% and 80% for chemotherapy. On univariable analysis, factors associated with a higher recurrence after initial surgery were young age (P = 0.012), male gender (P = 0.012) and extremity location (P = 0.005). On multivariable analysis, only young age was significantly associated with recurrence risk (P = 0.010).

CONCLUSIONS

Active surveillance was associated with spontaneous regression and long-term disease control consistent with other studies. Primary radiation therapy appeared to provide a similar response and disease control compared with systemic treatments and may be a viable option for patients who are not candidates for surgery or active surveillance.

Abstract P4-01-02: Early detection of residual breast cancer through a robust, scalable and personalized analysis of circulating tumour DNA (ctDNA) antedates overt metastatic recurrence

Background:

Many breast cancer patients relapse after primary treatment but there are no reliable tests to detect distant metastases before they become overt. Here we show earlier identification of recurring patients through a scalable personalised ctDNA analysis. The method is applicable to all patients, and not limited to hot-spot mutations typically detected by gene panels.

Methods:

Forty-nine non-metastatic breast cancer patients were recruited following surgery and adjuvant therapy. Plasma samples (n=208) were serially collected semi-annually. Using the analytically validated SignateraTM workflow, we determined mutational signatures from primary tumour whole exome data and designed personalised assays targeting 16 variants with high sensitivity by ultra-deep sequencing (average >100,000X). The patient-specific assay was used to detect the presence of the mutational signature in the plasma.

Results:

In 16 of 18 (89%) clinically-relapsing patients, ctDNA was detected ahead of metastatic relapse being diagnosed by clinical examination, radiological and biochemical (CA15-3) measurements, and remained ctDNA-positive through follow-up. Of the 2 patients not detected by ctDNA, one had a small local recurrence only (now resected) and the other had three primary tumours. None of the 31 non-relapsing patients were ctDNA-positive at any time point (n=142). Metastatic relapse was predicted by Signatera with high accuracy and a lead time of up to 2 years (median=9.5 months).

Conclusions:

The use of a scalable patient-specific ctDNA-based validated workflow detects breast cancer recurrence ahead of clinical detection. Accurate and earlier prediction by ctDNA analysis could provide a means of monitoring breast cancer patients in need of second-line salvage adjuvant therapy in order to prevent overt life-threatening metastatic progression.

Citation Format: Coombes RC, Armstrong A, Ahmed S, Page K, Hastings RK, Salari R, Sethi H, Boydell A-R, Shchegrova SV, Fernandez-Garcia D, Gleason KL, Goddard K, Guttery DS, Assaf ZJ, Balcioglu M, Moore DA, Primrose L, Navarro SL, Aleshin A, Rehman F, Toghill BJ, Louie MC, Zimmermann BG, Lin C-HJ, Shaw JA. Early detection of residual breast cancer through a robust, scalable and personalized analysis of circulating tumour DNA (ctDNA) antedates overt metastatic recurrence [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-01-02.

Abstract P2-04-01: Withdrawn

This abstract was withdrawn by the authors.

Abstract P1-08-02: CYP2D6 gene variants and effectiveness of adjuvant tamoxifen in breast cancer: A population-based case-control study

Tamoxifen, a cornerstone of adjuvant therapy for hormone-receptor-positive breast cancer, is metabolized to the active metabolite endoxifen through enzymatic activity of cytochrome P450 2D6. CYP2D6 has numerous alleles that affect metabolizing phenotype. Among women who take tamoxifen, those homozygous for inactive alleles (poor metabolizers) have lower levels of serum endoxifen than those with two functional alleles (extensive metabolizers). Several studies have reported increased risk of breast cancer recurrence or death in women homozygous for CYP2D6 inactive alleles, but others have found no association between CYP2D6 function and outcome. We explored this question in the large member population of the Kaiser Permanente Northwest (KPNW) integrated health plan. We conducted a population-based case-control study to evaluate the hypothesis that, after adjuvant tamoxifen treatment for breast cancer, women with CYP2D6 genotypes associated with poor metabolism of tamoxifen have an elevated risk of breast cancer recurrence compared to women with CYP2D6 genotypes associated with extensive metabolism of tamoxifen. We further hypothesized that women with CYP2D6 genotypes associated with intermediate metabolism of tamoxifen are at intermediate risk of recurrence. Study subjects were women who were diagnosed from 1980 to 2011 with hormone-receptor positive breast cancer, who received at least 180 days of adjuvant tamoxifen treatment, and for whom stored formalin-fixed paraffin-embedded (FFPE) normal tissue was available for laboratory analysis. Cases (358) were women with breast cancer recurrence recorded in the KPNW Tumor Registry and validated by medical record review. Randomly selected controls (833), without recurrent breast cancer, were matched to cases on tumor stage, diagnosis year, diagnosis age, race/ethnicity, and patterns of health plan membership. We collected data from medical records and from pharmacy, laboratory, tumor registry, and membership health plan databases. The Oregon Health & Science University Molecular Genetics Laboratory extracted genomic DNA from stored FFPE tissue blocks and performed allelic discrimination assays and pyrosequencing to accurately determine CYP2D6 variant status for the alleles, *3, *4, *5, *10, *17, and *41. All assays have been completed and study subjects have been categorized according to CYP2D6 metabolizer phenotype (poor, intermediate, extensive) and activity score (0-2). Based on the ethnicities in our study population, the CYP2D6 allele frequencies are in Hardy-Weinberg equilibrium, and the frequencies of the predicted metabolizer phenotypes also fall within the expected range. Using multivariable logistic regression analysis, we will assess CYP2D6 functional status and activity score in relation to breast cancer recurrence, taking into account factors that may alter the association, including tamoxifen dose and duration of use, as well as concomitant medications that alter the activity of the CYP2D6 enzyme. Results will be available by 12/1/2015.

Citation Format: Weinmann S, Richert–Boe K, Goddard K, Chen C, Punj S, Schwarzkopf D, Kalter M, Richards CS. CYP2D6 gene variants and effectiveness of adjuvant tamoxifen in breast cancer: A population-based case-control study. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P1-08-02.

PEComa of the terminal ileum mesentery as a secondary tumour in an adult survivor of embryonal rhabdomyosarcoma

Perivascular epithelioid cell tumours (pecomas) are rare mesenchymal tumours that are characterized by perivascular epithelioid cell differentiation and immunoreactivity to myogenic and melanocytic markers. These tumours can be classified as benign, uncertain malignant potential, or malignant. Because of the rarity of pecomas, their cause and clinical prognosis remain unclear. To the best of our knowledge, no reports in the literature describe a pecoma of the terminal ileum mesentery as a secondary tumour in an adult survivor of childhood embryonal rhabdomyosarcoma, let alone any childhood cancer. Here, we present the case of a 27-year-old man with a pecoma involving the mesentery of the terminal ileum. At the age of 5, he had been treated with a combination of chemotherapy and high-dose pelvic radiation therapy for embryonal rhabdomyosarcoma, most likely arising from the posterior bladder wall. During routine follow-up 22 years after this patient's initial treatment, computed tomography imaging revealed a mass within the terminal ileum mesentery. The tumour was successfully treated with surgical resection, and pathology examination determined the mass to be a pecoma with uncertain malignant potential. This first case of a pecoma of the terminal ileum mesentery arising within a high-dose radiation therapy field as a secondary tumour in an adult survivor of childhood cancer highlights the importance of screening and surveillance in high-risk childhood cancer survivors treated with high-dose radiation therapy. Further research to build a better understanding of this remarkably rare tumour is warranted.

Reclassification of ICD-9 Codes into Meaningful Categories for Oncology Survivorship Research

Background. The International Classification of Disease, ninth revision (ICD-9) is designed to code disease into categories which are placed into administrative databases. These databases have been used for epidemiological studies. However, the categories used in the ICD9-codes are not always the most effective for evaluating specific diseases or their outcomes, such as the outcomes of cancer treatment. Therefore a re-classification of the ICD-9 codes into new categories specific to cancer outcomes is needed. Methods. An expert panel comprised of two physicians created broad categories that would be most useful to researchers investigating outcomes and morbidities associated with the treatment of cancer. A Senior Data Coordinator with expertise in ICD-9 coding, then joined this panel and each code was re-classified into the new categories. Results. Consensus was achieved for the categories to go from the 17 categories in ICD-9 to 39 categories. The ICD-9 Codes were placed into new categories, and subcategories were also created for more specific outcomes. The results of this re-classification is available in tabular form. Conclusions. ICD-9 codes were re-classified by group consensus into categories that are designed for oncology survivorship research. The novel re-classification system can be used by those involved in cancer survivorship research.

Childhood, adolescent, and young adult cancer survivors (CAYACS) research program of British Columbia: Data linkage: Results to date

9555

Background: Long-term survivors of childhood and adolescent cancers are at risk for late mortality and morbidity. Using database linkages we assessed the extent of these issues and health care utilization in a population based cohort in British Columbia. Methods: Retrospective cohorts of 3,483 survivors (>5 years from diagnosis), and representative comparison groups, have been identified from population-based registries. Linkages were made with administrative databases of risk factors and outcomes. Late mortality, second cancers, late morbidity, health services utilization, continuity of care, and educational outcomes, among those diagnosed before age 20 between 1970 to 1995, and followed to 2000, have been examined. Results: Survivors experienced a 9-fold increase in mortality (SMR 9.1, 95% CI 7.8–10.5). Risk of developing a second cancer was 5 times higher than in the general population (SIR 5.0, 95% CI 3.8–6.5). Survivors had three times the odds of being hospitalized (OR 2.97, 95% CI 2.56–3.45) in a three-year period (1998–2000). Survivors were significantly more likely than the population group to consult any physician (excluding oncologists) (adj. RR 1.61, 95% CI 1.51–1.70). Survivors were found to experience a drop in continuity of primary health care as they aged and transitioned into adult care. Survivors were significantly more likely than their peers to receive special education (32.5% vs. 14.1%), most significantly among CNS survivors who received cranial irradiation. Conclusions: Survivors of childhood and adolescent cancers have severe long term health issues and increased health care utilization. Survivors of CNS tumors were at highest risk of poor health and educational outcomes measured. Data linkage provides useful insights for survivorship research.

No significant financial relationships to disclose.

Research, policy and practice related to survivors of childhood, adolescent and young adult cancers in British Columbia (BC), Canada: A population-based approach

9539

Background and Purpose: Prevalence of long term survivors of pediatric, adolescent and young adult cancers is increasing. In BC we are using a multidimensional population approach to address research and care for these survivors. Information on trends in prevalence of survivors and use of physician services are presented. Methods: We have identified a cohort of patients under 25, diagnosed from population registries since 1970, and linked their records with person-based longitudinal records of health utilization. Results: Prevalence rate of cancer survivors aged 0–59 years, originally diagnosed under age 25, increased 5 fold between 1975 and 2005, from 26 per 100,000 to 147 per 100,000; 78% were aged 20 or older. A total subgroup of 3,787 individuals surviving five years or more, diagnosed between 1970 and 1995, were identified; 2,590 (68%) of these were linked to physician visit records (excluding visits within BC's Children's Hospital) from 1986 to 2000. In year 2000, 77% of survivors had at least one physician visit; 75% of survivors saw a family physician (FP), and 44% of survivors visited a specialist physician with a specialty that could be related to a late effect. Between 1986 and 2000 survivor prevalence increased 3.3% per year; total visits by survivors to FPs increased by 11.6% per year and by 10.6% per year to specialists. In 1986, 66% of physician visits were for patients older than 20 which increased to 79% by 2000. A provincial pediatric oncology network consisting of oncologists and community care representatives has been established to implement changes to long-term follow up policy. Conclusion: The prevalence rate and utilization of health services among survivors of childhood, adolescent and young adult cancers continues to escalate. Research results, along with other published literature, will inform the development of policy and practice within the BC Network.

No significant financial relationships to disclose.

Function and health status outcomes in a randomized trial comparing preoperative and postoperative radiotherapy in extremity soft tissue sarcoma

PURPOSE

Morbidity associated with wound complications may translate into disability and quality-of-life disadvantages for patients treated with radiotherapy (RT) for soft tissue sarcoma (STS) of the extremities. Functional outcome and health status of extremity STS patients randomized in a phase III trial comparing preoperative versus postoperative RT is described.

PATIENTS AND METHODS

One hundred ninety patients with extremity STS were randomized after stratification by tumor size dichotomized at 10 cm. Function and quality of life were measured by the Musculoskeletal Tumor Society Rating Scale (MSTS), the Toronto Extremity Salvage Score (TESS), and the Short Form-36 (SF-36) at randomization, 6 weeks, and 3, 6, 12, and 24 months after surgery.

RESULTS

One hundred eighty-five patients had function data. Patients treated with postoperative RT had better function with higher MSTS (25.8 v 21.3, P <.01), TESS (69.8 v 60.6, P =.01), and SF-36 bodily pain (67.7 v 58.5, P =.03) scores at 6 weeks after surgery. There were no differences at later time points. Scores on the physical function, role-physical, and general health subscales of the SF-36 were significantly lower than Canadian normative data at all time points. After treatment arm was controlled for, MSTS change scores were predicted by a lower-extremity tumor, a large resection specimen, and motor nerve sacrifice; TESS change scores were predicted by lower-extremity tumor and prior incomplete excision. When wound complication was included in the model, patients with complications had lower MSTS and TESS scores in the first 2 years after treatment.

CONCLUSION

The timing of RT has minimal impact on the function of STS patients in the first year after surgery. Tumor characteristics and wound complications have a detrimental effect on patient function.

New combination of the old drugs for elderly patients with small-cell lung cancer: a phase II study of the PAVE regimen

PURPOSE

A regimen of cisplatin, doxorubicin, vincristine, and etoposide (PAVE) was designed for patients with small-cell lung cancer (SCLC) who were older than 65 years, with the following objectives compared with standard chemotherapy regimens: maintain efficacy, diminish toxicity, enhance compliance, and improve chemotherapy administration convenience at an acceptable cost.

PATIENTS AND METHODS

The PAVE regimen consisted of cisplatin 30 mg/m2 intravenously (i.v.) day 1; doxorubicin 40 mg/m2 i.v. day 1; vincristine 1.0 mg/m2 i.v. day 1; and etoposide 100 mg/m2 i.v. day 1 and orally days 3 and 5. Cycles were repeated every 3 weeks for four cycles. Patients with limited-stage disease and selected patients with extensive-stage disease received thoracic irradiation delivered concurrently with etoposide-cisplatin (EP) at the time of the second chemotherapy cycle.

RESULTS

Sixty-six eligible patients were treated, which included 25 patients with limited-stage disease and 41 patients with extensive-stage disease. Median survival was 70 weeks and 5-year survival was 25% for limited-stage disease. Median survival was 46 weeks for extensive-stage disease. Only one treatment-related death occurred and severe toxicity was infrequent. The median delivered dose-intensity was according to protocol and the mean delivered total dose was 80% of intended.

CONCLUSION

The treatment outcome achieved with PAVE in a phase II study of elderly patients compared favorably with published results of standard regimens in patient populations with better prognostic factors. Because the PAVE regimen can be delivered with good compliance, has acceptable toxicity, and is associated with logistic advantages compared with standard regimens, this protocol is suitable for further investigative trials in elderly patients with SCLC.

Influence of apolipoprotein E genotype on the transmission of Alzheimer disease in a community-based sample

The epsilon 4 allele of the apolipoprotein E locus (APOE) has been found to be an important predictor of Alzheimer disease (AD). However, linkage analysis has not clarified the role of APOE in the transmission of AD. The results of the current study provide evidence that the pattern of transmission of memory disorders differs in nuclear families in which the AD-affected proband did carry an epsilon 4 allele versus those families in which the AD-affected proband did not carry an epsilon allele. Further, risk of AD due to APOE genotype in the probands is modified by family history of memory disorders, suggesting gene-by-gene interactions. Family history remained a significant predictor of AD for affected probands with some, but not all, APOE genotypes in a logistic regression analysis. Though nonadditive in the prediction of AD, APOE genotype and family history acted additively in the prediction of age at AD onset. The results of complex segregation analysis were inconsistent with Mendelian segregation of memory disorders both in families of affected probands who did or did not carry an epsilon 4 allele, yet these two groups had significantly different parameter estimates for their transmission models. These results are consistent with gene-by-gene interactions, but also could result from common elements in the familial environment.

Cytogenetic analysis of 109 pediatric central nervous system tumors

Reports of cytogenetic abnormalities in pediatric central nervous system (CNS) tumors are important for collection and comparison of large numbers of karyotypes of primary CNS neoplasms to produce statistically significant correlations. We report cytogenetic results of 119 samples of pediatric CNS tumors from 109 patients. Tumors included 33 low-grade astrocytomas, 18 high-grade astrocytomas, 14 gangliogliomas, 13 ependymomas, 17 primitive neuroectodermal tumors (PNET), three choroid plexus papillomas and carcinomas, and a miscellaneous group of 20 rare primary CNS tumors and metastases. In each group, cytogenetic results were correlated with histologic subtype and survival. The study indicated specific chromosome abnormalities in different groups of tumors. Low-grade astrocytomas showed mostly numeric abnormalities with gains of chromosome 7, high-grade astrocytomas showed differences from karyotypic changes observed in adults in lacking double minutes (dmin) and monosomy 10. The ependymoma group showed the largest proportion of abnormal karyotypes with frequent involvement of chromosome 6 and 16. Chromosome 6 was the single most common abnormal chromosome in this study, closely followed by chromosomes 1 and 11. Pediatric CNS neoplasms differ from adult tumors cytogenetically as well as histologically and biologically.

Intensive weekly chemotherapy for the treatment of extensive-stage small-cell lung cancer

The regimen of cisplatin, vincristine, doxorubicin, and etoposide (CODE) was designed to double the dose intensity of these drugs in comparison with a standard regimen (alternating cyclophosphamide, doxorubicin, and vincristine [CAV] and etoposide-cisplatin [EP]) for extensive-stage small-cell lung cancer (SCLC). The dose intensity was increased by more frequent treatments rather than by increasing the dose size. The structure of this outpatient protocol includes weekly administration of chemotherapy, alternation of myelosuppressive and nonmyelosuppressive treatments, supportive corticosteroids, gastroprotective agents, and prophylactic antibiotics. Although the duration of chemotherapy was brief (9 to 12 weeks), the total cumulative doses of drugs delivered were similar to the standard regimen. Patients with no residual disease outside the chest after chemotherapy received thoracic irradiation, and patients with complete responses (CRs) received prophylactic cranial irradiation. Eligible extensive-stage SCLC patients were ambulatory, younger than 66 years of age, and free of brain metastasis. Forty-eight extensive-stage SCLC patients were treated. Forty-five (94%) responded to chemotherapy, with 19 (40%) attaining CR. After consolidative thoracic irradiation, the CR rate was 56%. The median time to progression was 43 weeks, and the median survival was 61 weeks. The 2-year survival rate was 30%. The most common site of first relapse was brain (38%). Although two patients (4%) died of toxicity, overall toxicity was acceptable for an outpatient regimen. We conclude that the CODE regimen reliably produces palliative remissions for selected extensive-stage SCLC patients, and it may be associated with durable remissions for some patients. The results of this pilot study are sufficiently promising to justify a phase III trial of CODE versus standard (alternating CAV and EP) chemotherapy.

Posterior fossa medulloblastoma in childhood: treatment results and a proposal for a new staging system

Seventy-two children with posterior fossa medulloblastoma were diagnosed at the Hospital for Sick Children, Toronto, from 1977 to 1987 and treated by standard methods. The 5- and 10-year survival and disease-free survival rates were 71% and 63%, and 64% and 63%, respectively. Total tumor resection, as determined by the surgeon was the most significant favorable prognostic factor. Post-operative meningitis, a residual enhancing mass lesion on the post-operative, pre irradiation CT scan and dissemination to the brain or cord at diagnosis were unfavorable factors. These four easily definable factors were used to define a staging system with prognostic significance. Five-year disease-free survival rates were for Stage I (total resection, no adverse factor) 100%, Stage II (total resection with one or more adverse factor or less than total resection with no other adverse factor) 78%, and Stage III (less than total resection with one or more adverse factor) 18%. Evaluation of treatment results in medulloblastoma requires that these prognostic factors be known.