Institutional Review Boards (IRB) approval was obtained for this study. The Patient Treatment File (PTF) database and the national archiving system for inpatient data in the DVA medical system were accessed to retrieve data of interest. Security clearance was obtained to use the DVA Data Processing Center in Austin, TX, USA, which has data from all the administrative databases of DVA facilities. Its computer components include the IBM Z900 (2064-2CS) and IBM Z890 (2086-320)\Sun E10000 HD-UX mainframes coupled with EMC storage computers (EMC 8730 and 8830; IBM, Armonk, NY, USA). Information on the PTF database may be retrieved from the Veterans Information Resource Center (VIREC) website at http://www.virec.research.va.gov.

National DVA datasets using computer codes were queried for a pre-existing diagnosis of schizophrenia and a later diagnosis of breast carcinoma. The PTF database was searched to identify patients with the International Classification of Diseases, Ninth Revision, Clinical Modifications (ICD-9-CM) (www.cdc.gov/nchs/icd/icd9.htm) diagnostic codes for schizophrenia or schizoaffective disorder (codes 295.00-295.85) and subsequently diagnosed with breast cancer (codes 174-174.9, 198.2, 198.81, 217, 233.0, 611.72) during the fiscal years 1999–2005. The PTF database identified appropriate patients based on these computer codes and yielded demographic information as well as data on comorbidities.

For additional data, chart-based information was requested from the DVA Medical Centers, where patients had received their breast carcinoma care. These data included discharge summaries, progress notes, operative reports, medical oncology consultation summaries, psychiatric consultation notes and pathology reports. To obtain further information regarding the psychiatric history of each patient, medical records were requested predating the admission for breast cancer. From this pool of patients, we selected patients diagnosed with MBC were selected. The patients were then grouped into those initially presenting with MBC (n=10) and those with MBC after prior curative-intent treatment (n=7).

Descriptive statistics were utilized to analyze the demographic information, clinical course and outcomes of the patients. Patients were excluded from further review according to pre-specified criteria: medical chart data not available to provide adequate information; apparent coding errors that, upon chart review, demonstrated no evidence of schizophrenia or schizoaffective disorder; coding errors that demonstrated no evidence of a breast carcinoma; diagnosis of breast cancer prior to diagnosis of schizophrenia and patients assessed by their physicians as inappropriate candidates for adjuvant endocrine and/or cytotoxic chemotherapy, for any reason. We assumed a variable of interest was present when mentioned as present in the medical record. We assumed it was absent when not mentioned in the record. We recognized that this would lead to a conservative estimate of the presence of most variables.

Approximately 25 million veterans are eligible for VA healthcare, while ∼5 million veterans receive care in DVA medical centers each year (Congressional Budget Office) (17). During the time period of our study, 10 patients were identified with an initial diagnosis of MBC as well as seven patients with MBC after initial curative-intent treatment.

There were 10 patients with no prior diagnosis of (or treatment for) breast carcinoma, who presented initially with MBC. Of these 10 patients, two lived alone, five with a family member, while one was homeless. Data on the remaining two were not available. Five patients were documented not to have a family history of breast cancer. Additional data is summarized in Table I.

Four of the 10 patients were diagnosed with paranoid, one with catatonic and two with undifferentiated schizophrenia, while three had schizoaffective disorder. Based on the diagnostic criteria for schizophrenia (Table IV), seven patients had a concurrent history of other Axis I psychiatric disorders, such as depression, bipolar disorder and post-traumatic stress disorder (PTSD) (18). One of the seven patients had major depression, one had PTSD, two had bipolar disorder, two had a psychotic disorder-not otherwise specified (NOS) and one had a psychotic disorder-NOS with PTSD.

Affect is often abnormal in schizophrenic patients. The affect of two patients was described as full-range and appropriate, of one patient as constricted, one blunted/appropriate, while two patients affect was described as blunted/guarded. One patient had affect described as dull and one patient as labile. Delusions are very common in schizophrenic patients. Of the 10 patients, one had delusions of grandiosity, two had persecutory delusions and one had thought-insertion delusions. Two patients had a history of disruptive behavior, one of whom set fire to her own clothes after being denied her holistic medicines. Additional evidence of abnormal thoughts and behaviors are shown in Tables II and III.

Three of the 10 patients were taking phenothiazines (haloperidol, perphenazine and trifluoperazine). Seven patients were taking atypical antipsychotics (most commonly loxapine, olanzapine and risperidone), with six of the seven experiencing adverse reactions to their antipsychotic medication.

Seven of the 10 patients inappropriately delayed seeking medical attention for the MBC. For four of the seven, the mean time from onset of MBC symptoms to initial treatment was 259 days and two of these seven waited >1 year. The mean waiting time from symptom onset to palliative surgery was three years, with two of the seven waiting >4 years. Three of the 10 patients could not give a coherent history. One patient believed that someone had put an egg into her right breast and a radio transmitter into her left one. Three patients were non-compliant with their antipsychotic medication regimen when offered medications for MBC. Of these three patients, none were compliant with the chemotherapy regimen.

Of the 10 patients in this group, seven underwent a computed tomography (CT), six a bone scan, three a chest ultrasound examination, six a chest radiograph and seven a mammogram during the extent-of-disease studies.

Seven of the 10 patients were offered cytotoxic chemotherapy, six received the initial doses of chemotherapy, although only two of the six completed ≥3 cycles. One patient refused treatment saying ‘God will help me through this’. Five of the 10 patients (50%) were offered endocrine therapy, two refused therapy and three were compliant. Five of the 10 patients (50%) were offered radiation therapy, two refused it and three complied with treatment. One patient’s guardian refused to permit radiation treatment because she believed that the patient would not comply. The same patient drank sassafras tea as ‘alternative treatment’ and covered her fungating breast mass with a bandage. Seven of the 10 patients were offered palliative breast surgery, two of whom refused and five underwent the surgery. One patient who refused to take any psychiatric medication because hospital staff denied her shark cartilage for the breast cancer. Another patient refused the recommended treatment and took only herbal teas and vitamins as treatment.

There were seven patients with prior curative-intent treatment, who presented with MBC. Of the seven patients, four lived alone and two were in a nursing home. One of the seven was institutionalized. Three of the seven patients had no documented family history of breast cancer. Additional information is summarized in Table I.

Two of these seven were diagnosed with paranoid schizophrenia, one had residual schizophrenia, three had schizoaffective disorder and one was unspecified. Four of the seven had Axis I personality disorders (three had major depression and one had bipolar disorder with psychosis).

Four of the seven had auditory, while two had visual hallucinations. One patient noted spiders crawling up his arm and said that people around him are in ‘league with the devil’. One of the seven (14%) had somatic hallucinations. Only two of the seven had full-range, appropriate affect. One had affect described as constricted, one described as blunted/inappropriate and one described as labile. Two had affect described as unspecified. One had somatic and two had grandiose delusions. Two had a history of disruptive behavior and one of removing intravenous lines. One had a history of self-injury. Additional evidence of abnormal thoughts and behaviors can be found in Tables II and III.

Two of the seven were taking phenothiazines (most often chlorpromazine and fluphenazine). Five of the seven (71%) were taking atypical agents (most often olanzapine, quetiapine and risperidone). Four of the seven (57%) had experienced adverse reactions to their antipsychotic medication.

Four of the seven inappropriately delayed seeking medical attention. For these four, the mean time from symptom onset to initial treatment was 102 days. None of the four waited >1 year. However, the mean wait time from symptom onset to palliative surgery was 130 days. Three of the seven (43%) could not give a coherent history. All seven were compliant with their antipsychotic regimen during the course of treatment for MBC, while five were compliant with their MBC therapy.

All seven had a CT, six had a bone scan and three had a chest ultrasound. Six had a chest radiograph, and six had a mammogram during their extent of disease studies.

Five of the seven were offered cytotoxic chemotherapy. Four received initial chemotherapy, while only two completed ≤4 cycles. Five of the seven were offered endocrine therapy, three refused this therapy and only two were compliant. One patient refused tamoxifen, noting that ‘I want to see if I am mortal or immortal’. Four of the seven were offered radiation therapy, one refused and three complied with treatment. All seven patients were offered palliative breast surgery, two of whom refused, while five underwent the surgery.