Characterizing Relative and Disease-Specific Survival in Early-Stage Cancers | Oncology | JAMA Internal Medicine | JAMA Network
[Skip to Navigation]
Figure.  Three Relationships Between Relative and Disease-Specific Survival in Patients With Early-Stage Cancer
Three Relationships Between Relative and Disease-Specific Survival in Patients With Early-Stage Cancer

The solid lines indicate relative survival, and the dashed lines indicate disease-specific survival. A, In some cancer types, such as stage 1 cancers of the lung/bronchus, bladder, and oral cavity/pharynx, relative survival is worse than disease-specific survival owing to upstream factors such as smoking. B, In cancers where upstream factors are less relevant, such as stage 1 pancreas and testicular cancer, relative and disease-specific survival are similar and reflect the low (pancreas) or high (testicular) curability of the cancer. C, In yet other cancer types, such as early-stage cancers of the thyroid, skin (melanoma), prostate, and breast, relative survival is higher than disease-specific survival and exceeds 100%, owing to the healthy-user effect. Numbers represent 10-year relative survival (RS) and disease-specific survival (DSS) for American Joint Committee on Cancer Staging Manual, 6th edition, stage 1 cancers (Gleason 3 + 3 or lower for prostate) and ductal carcinoma in situ (DCIS). Data are from the National Cancer Institute’s Surveillance, Epidemiology and End Results 9 Research Data set, April 2019 release, based on data from the November 2018 submission (version 8.3.6), using years 2004 to 2015.

1.
Henson  DE, Ries  LA.  The relative survival rate.  Cancer. 1995;76(10):1687-1688. doi:10.1002/1097-0142(19951115)76:10<1687::AID-CNCR2820761002>3.0.CO;2-IPubMedGoogle ScholarCrossref
2.
Shrank  WH, Patrick  AR, Brookhart  MA.  Healthy user and related biases in observational studies of preventive interventions: a primer for physicians.  J Gen Intern Med. 2011;26(5):546-550. doi:10.1007/s11606-010-1609-1PubMedGoogle ScholarCrossref
3.
Davies  L, Petitti  DB, Martin  L, Woo  M, Lin  JS.  Defining, estimating, and communicating overdiagnosis in cancer screening.  Ann Intern Med. 2018;169(1):36-43. doi:10.7326/M18-0694PubMedGoogle ScholarCrossref
4.
Welch  HG, Kramer  BS, Black  WC.  Epidemiologic signatures in cancer.  N Engl J Med. 2019;381(14):1378-1386. doi:10.1056/NEJMsr1905447PubMedGoogle ScholarCrossref
5.
Lu  D, Andersson  TML, Fall  K,  et al.  Clinical diagnosis of mental disorders immediately before and after cancer diagnosis: a nationwide matched cohort study in Sweden.  JAMA Oncol. 2016;2(9):1188-1196. doi:10.1001/jamaoncol.2016.0483PubMedGoogle ScholarCrossref
6.
Withrow  DR, Pole  JD, Nishri  ED, Tjepkema  M, Marrett  LD.  Choice of relative or cause-specific approach to cancer survival analysis impacts estimates differentially by cancer type, population, and application: evidence from a Canadian population-based cohort study.  Popul Health Metr. 2017;15(1):24. doi:10.1186/s12963-017-0142-4PubMedGoogle ScholarCrossref
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words
    Research Letter
    Less Is More
    December 9, 2019

    Characterizing Relative and Disease-Specific Survival in Early-Stage Cancers

    Author Affiliations
    • 1Department of Surgery, Head and Neck Service, Memorial Sloan Kettering Cancer Center, New York, New York
    • 2Weill Cornell Breast Center, Divisions of Breast and Endocrine Surgery, Department of Surgery, New York Presbyterian/Weill Cornell Medical Center, New York, New York
    • 3Urology Service, Department of Surgery, Memorial Sloan Kettering Cancer Center, New York, New York
    • 4The Dartmouth Institute for Health Policy and Clinical Practice, Lebanon, New Hampshire
    JAMA Intern Med. 2020;180(3):461-463. doi:10.1001/jamainternmed.2019.6120

    When explaining cancer survival statistics to patients, physicians may use net measures such as disease-specific survival—the proportion of patients not dying of their cancer in a given time frame. Another complementary statistic is relative survival, which compares survival of a cancer population to a similar population without cancer, matched for age, race, and sex.1 These measures are often numerically similar, although relative survival also incorporates characteristics—beyond cancer—that distinguish these patients. If patients with particular cancers often also tend to engage in lifestyle or health behaviors that impair survival, such as smoking, their relative survival will be lower. The converse is also true: patients with certain cancers might have higher relative survival, if the cancer is primarily detected incidentally or through screening, because patients receiving preventive care also engage in other healthy behaviors (the healthy-user effect).2,3 This phenomenon has been recognized in early-stage breast, prostate, thyroid, and skin (melanoma) cancers.4 Our objective was to characterize disease-specific and relative survival rates for patients with early-stage cancers, to better understand which cancers may be of lower risk and more likely to be diagnosed in otherwise healthy, health-conscious individuals.

    Methods

    Using 2004 to 2015 data from the Surveillance Epidemiology End Results (SEER) registry, we calculated relative and disease-specific survival rates for patients with 10 early-stage cancers: lung/bronchus, oral/pharyngeal, bladder, pancreatic, testicular, prostate, breast (invasive and ductal carcinoma in situ [DCIS]), thyroid, and melanoma. We determined stage with American Joint Committee on Cancer 6th edition or Gleason score (prostate). Relative survival was calculated as the ratio of observed to expected survival, based on national annual life tables. Disease-specific survival was calculated as the probability of survival, censoring noncancer causes of death. This study was determined to be exempt from review by the Memorial Sloan Kettering Cancer Center institutional review board because all data used were deidentified and publicly available. Analyses were performed using SeerStat statistical software (version 8.3.6, National Cancer Institute) in October, 2019. Additional methods are in eMethods in the Supplement.

    Results

    From 2004 to 2015, 281 970 patients diagnosed with 10 early-stage cancers were identified in the SEER registry. Relative survival was lower than disease-specific survival among patients with early-stage lung, oral/pharyngeal, and bladder cancers (Figure, A), but no different from disease-specific survival among patients with early-stage pancreatic and testicular cancers (Figure, B). In contrast, 10-year relative survival rates were higher than disease-specific survival, and more than 100% among patients with Gleason grade group 1 (≤3+3) prostate cancer (109.6%; 95% CI, 109.0%-110.1%), DCIS (104.4%; 95% CI, 103.6%-105.1%), stage I melanoma (102.8%; 95% CI, 102.1%-103.4%) and stage 1 cancers of the thyroid (101.7%; 95% CI, 101.3%-102.1%), and breast (101.1%; 95% CI, 100.7%-101.6%) (Figure, C).

    Discussion

    We identified 5 early-stage cancers for which patients had relative survival rates higher than disease-specific survival rates, and more than 100%, suggesting that these patients were likely living longer than age-, sex-, and race-matched counterparts without cancer. These are cancers in which screening or incidental detection often identifies very low-risk disease among otherwise healthy, health-conscious individuals.

    In the appropriate context, these statistics can help physicians explain to certain patients that their actions prior to diagnosis may mean more for their survival than the diagnosis itself. This may be reassuring and mitigate the risk of depression and anxiety associated with cancer diagnosis.5 Where appropriate, these statistics might open doors to considering active surveillance—a treatment option accepted in prostate cancer, beginning to be offered in thyroid cancer, and under investigation for DCIS. Understandably, when evaluating survival rates, many patients anchor their expectations at 100%, and strongly desire intervention when they believe their health is at risk. Relative survival may help patients reorient their expectations by providing a comparison to a similar person without cancer, revealing that not every cancer diagnosis necessarily portends an earlier death.

    In some other early-stage cancers, relative survival rates were lower than disease-specific survival rates, likely attributable to an associated risk factor, such as smoking, that increases mortality risk from cardiovascular disease or other causes.

    Although these statistics may provide physicians and patients a better understanding of early-stage cancer prognosis, they have important limitations. Disease-specific survival rates may be miscalculated if cause of death is misclassified, and relative survival requires accurate life tables for the background population.6

    Back to top
    Article Information

    Corresponding Author: Luc G. T. Morris, MD, MSc, Department of Surgery, Memorial Sloan Kettering Cancer Center, 1275 York Ave, C-1077, New York, NY 10065 (morrisl@mskcc.org).

    Accepted for Publication: October 20, 2019.

    Published Online: December 9, 2019. doi:10.1001/jamainternmed.2019.6120

    Author Contributions: Dr Morris had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Davies and Morris were co–senior authors.

    Study concept and design: Marcadis, Hakimi, Morris.

    Acquisition, analysis, or interpretation of data: Marcadis, Marti, Ehdaie, Davies, Morris.

    Drafting of the manuscript: Marcadis, Hakimi, Morris.

    Critical revision of the manuscript for important intellectual content: All authors.

    Statistical analysis: Marcadis, Morris.

    Obtained funding: Morris.

    Administrative, technical, or material support: Marcadis, Morris.

    Study supervision: Ehdaie, Hakimi, Morris.

    Conflict of Interest Disclosures: Dr Morris reported research funding (no personal payments) from Bristol-Myers Squibb, AstraZeneca, and Illumina, unrelated to this work. No other disclosures were reported.

    Funding/Support: We acknowledge funding from the Jayme Flowers Fund, NIH K08 DE024774 and R01 DE027738 (Dr Morris), and NIH 5T32CA009685 (Dr Marcadis).

    Role of the Funder/Sponsor: The Jayme Flowers Fund and the National Institutes of Health had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

    References
    1.
    Henson  DE, Ries  LA.  The relative survival rate.  Cancer. 1995;76(10):1687-1688. doi:10.1002/1097-0142(19951115)76:10<1687::AID-CNCR2820761002>3.0.CO;2-IPubMedGoogle ScholarCrossref
    2.
    Shrank  WH, Patrick  AR, Brookhart  MA.  Healthy user and related biases in observational studies of preventive interventions: a primer for physicians.  J Gen Intern Med. 2011;26(5):546-550. doi:10.1007/s11606-010-1609-1PubMedGoogle ScholarCrossref
    3.
    Davies  L, Petitti  DB, Martin  L, Woo  M, Lin  JS.  Defining, estimating, and communicating overdiagnosis in cancer screening.  Ann Intern Med. 2018;169(1):36-43. doi:10.7326/M18-0694PubMedGoogle ScholarCrossref
    4.
    Welch  HG, Kramer  BS, Black  WC.  Epidemiologic signatures in cancer.  N Engl J Med. 2019;381(14):1378-1386. doi:10.1056/NEJMsr1905447PubMedGoogle ScholarCrossref
    5.
    Lu  D, Andersson  TML, Fall  K,  et al.  Clinical diagnosis of mental disorders immediately before and after cancer diagnosis: a nationwide matched cohort study in Sweden.  JAMA Oncol. 2016;2(9):1188-1196. doi:10.1001/jamaoncol.2016.0483PubMedGoogle ScholarCrossref
    6.
    Withrow  DR, Pole  JD, Nishri  ED, Tjepkema  M, Marrett  LD.  Choice of relative or cause-specific approach to cancer survival analysis impacts estimates differentially by cancer type, population, and application: evidence from a Canadian population-based cohort study.  Popul Health Metr. 2017;15(1):24. doi:10.1186/s12963-017-0142-4PubMedGoogle ScholarCrossref
    ×