People who receive a cancer diagnosis are at markedly increased risk for suicide and for fatal cardiovascular events in the following weeks, according to a large Swedish cohort study reported in the April 5 issue of the New England Journal of Medicine.
"This spike in risk was particularly prominent among patients in whom cancers with a poor prognosis were diagnosed, and was not explained by preexisting psychiatric or cardiovascular conditions," said Dr. Fang Fang of the department of medical epidemiology and biostatistics, Karolinska Institutet, Stockholm, and her associates.
The dramatic increase in mortality risk from suicide or cardiovascular causes also cannot be attributed to cancer treatment or cancer progression, since it peaks during the first week after diagnosis and declined over time rather than increasing as treatment intensified or disease worsened, they noted.
Previous studies have demonstrated that cancer patients are at increased risk for suicide and CV events, but most have attributed it to "the burden of living with a progressing cancer." Very few have explored the period immediately following a cancer diagnosis, Dr. Fang and her colleagues said.
They examined this time period using data from a Swedish national registry of cancers and causes of death, both of which must be reported by law in Sweden. The study period, 1991 through 2006, covered more than 6 million Swedes aged 30 and older at baseline.
A total of 534,154 people received a first cancer diagnosis during this interval, including 95,786 with prostate cancer, 74,977 with breast cancer (among women), 62,719 with colorectal cancer, 47,169 with skin cancer, 36,648 with lymphatic or hematopoietic cancer, 34,743 with lung cancer, and 13,447 with CNS cancer. Another 26,335 patients were pooled in a single category of "highly fatal cancers of the esophagus, liver, and pancreas," and the remaining 142,330 patients had other types of cancer.
During follow-up, there were 786 completed suicides among patients with any type of cancer, for a rate of 0.36 per 1,000 person-years. This was double the rate among Swedish adults who did not have cancer: 13,284 completed suicides for a rate of 0.18 per 1,000 person-years (N. Engl. J. Med. 2012;366:1310-8).
The rate of completed suicides peaked during the first week after cancer diagnosis, at 2.50 per 1,000 person-years.
"We found a relative risk of 4.8 during the first 12 weeks after diagnosis (110 patients; incidence rate, 0.95 per 1,000 person-years), with the highest relative risk observed for highly fatal cancers of the esophagus, liver, or pancreas, followed by lung cancer," they added.
The magnitude of elevated risk dropped rapidly after that, but remained higher than average beyond the first year for all cancers.
"Focusing on the first 52 weeks after a diagnosis of any cancer, we found a relative risk of 3.1 for suicide (260 patients; incidence rate, 0.60 per 1,000 person-years). The expected number of suicides, adjusted for all demographic factors, during these 52 weeks was 87, leaving 173 cases associated with cancer diagnoses," the researchers said.
The relative risks of suicide were stronger among patients who had no concomitant psychiatric disorders than among those with psychiatric disorders.
Turning to fatal cardiovascular events, there were 48,991 CV deaths among people who received a diagnosis of any cancer, for an incidence of 23.10 per 1,000 person-years. This rate is approximately three times higher than that for people who did not have cancer (543,144 deaths; incidence rate, 7.53 per 1,000 person-years.
As with suicide, the highest relative risk of CV death (5.6) peaked during the first week after diagnosis (1,318 patients; incidence rate, 116.8 per 1,000 person-years). The risk elevation was highest for central nervous system cancers, followed by highly fatal cancers of the esophagus, liver, or pancreas, and then by lung cancer.
Also as with suicide, the magnitude of the increased risk of CV death dropped rapidly after the first several weeks. It did not persist beyond the first year after diagnosis.
"Focusing on the first 4 weeks after a diagnosis of any cancer but CNS tumors, we found a relative risk of 3.3 for cardiovascular death (2,641 patients; incidence rate, 65.81 per 1,000 person-years). The adjusted expected number of CV deaths was 766 during these 4 weeks, leaving 1,875 deaths associated with a cancer diagnosis," the investigators said.
Of note was the finding that the increased hazards were seen in both men and women.
To adjust for unmeasured confounders, the researchers performed an additional case-crossover analysis among all cancer patients in the cohort who died from suicide or CV causes. The results "further allayed the concern" that there may be some alternative explanation for the observed associations, such as an unknown factor that causes cancer, suicide, and CV death.