Purpose
To evaluate the quality of evidence about effectiveness of regular periodic surveillance testing for metastasis in patients with primary uveal melanoma (PUM) following treatment of the primary tumor in prolonging survival.
Design
Literature review and personal perspective of the authors.
Methods
Identification and analysis of peer-reviewed articles on human PUM published between 1980 and 2009 that reported on “screening,” “surveillance,” or “systemic follow-up evaluation” for metastasis in patients with PUM following treatment of primary tumor.
Results
Of 4222 identified articles, only 31 were considered satisfactory for inclusion in this study. Satisfactory articles reported levels of specific biomarkers when metastasis was first confirmed (14), percentage of patients with abnormal results on surveillance testing (13), values of diagnostic markers (eg, sensitivity, specificity) associated with evaluated components of a surveillance regimen (7), survival time after first detection of metastasis from primary uveal melanoma (7), total survival time after initial diagnosis or initial treatment of primary uveal melanoma (3), percentage of patients whose metastatic tumors were detected by presymptomatic testing (5), surveillance regimens employed by different groups (1), and relationship with generally accepted clinical and histopathologic prognostic factors for primary uveal melanoma metastasis (1). However, none of these articles reported survival times of comparable subgroups of patients in which regular periodic surveillance for metastasis vs no surveillance was performed.
Conclusion
Available evidence from the peer-reviewed literature does not provide any compelling evidence of survival benefit for any regimen or frequency of surveillance for metastasis relative to no such testing. In view of this, advisability of periodic surveillance for metastasis in routine clinical practice must be questioned.
When an ophthalmologist encounters a patient with a primary uveal melanoma, he or she typically first arranges for the patient to undergo a limited baseline systemic staging evaluation to rule out overt metastatic disease that might alter treatment plans. Assuming that this systemic testing proves negative for metastasis, the ophthalmologist then provides or arranges for some local therapy (eg, enucleation, plaque radiotherapy, proton beam irradiation) for the intraocular tumor that he or she hopes will be curative. Subsequent to this treatment, the ophthalmologist (or medical oncologist, or internist, or family practitioner) frequently arranges for the patient to undergo periodic systemic follow-up testing designed to detect emerging metastatic disease. Unfortunately, published peer-reviewed articles on treatment of metastatic uveal melanoma to date do not provide compelling scientific evidence of a survival benefit of treatment for any subgroup of patients with metastasis. In addition, no currently available method of adjuvant therapy following initial treatment of the intraocular tumor in patients without clinically apparent metastasis has been shown to prevent or delay the emergence of metastasis in patients destined to develop this problem.
In view of these observations, one may legitimately question why surveillance testing for metastasis is recommended almost routinely by many physicians for patients who have undergone definitive local treatment of their primary uveal melanoma. The conventional responses to this question include 1) reassurance of patients who do not have detectable metastasis that they are not in any imminent danger of dying from metastasis, and 2) detection of metastasis, if present, at an asymptomatic, limited substage (ie, when fewer and smaller metastatic lesions are present and only 1 organ is overtly involved) amenable to certain treatments that might not be options for a more advanced substage. Inherent in this latter response is a belief that some form of aggressive therapy for limited substage metastasis will improve the patient’s survival prognosis substantially. As mentioned above, however, such benefit has yet to be confirmed by appropriately designed prospective comparative clinical trials.
The purpose of this study was to evaluate the quality of evidence contained in the published peer-reviewed literature during the past 30 years regarding the impact of regular periodic surveillance testing for metastasis on overall survival time of patients who have undergone some definitive treatment for their primary uveal melanoma and to determine the strength of that evidence in support of routine surveillance testing.
Methods
Using PubMed, the authors performed a literature search for all peer-reviewed articles on primary uveal melanoma concerning human subjects that were published during the interval January 1, 1980 through December 31, 2009. The terms employed for this search were “uveal melanoma,” “choroidal melanoma,” “ciliary body melanoma,” and “iris melanoma.” The authors then searched this group of articles for the additional term “metastasis.” They reviewed the titles and abstracts of all articles identified by these first 2 search steps and identified those that mentioned “screening” or “surveillance testing” for metastasis from the primary uveal melanoma in the title, abstract, or both.
Using the “related articles” function of PubMed, the authors generated lists of the first 20 related articles for each of the articles identified by the preceding steps. They reviewed the titles and abstracts of all of these related articles and identified those that mentioned some form of systemic follow-up testing to detect metastasis from primary uveal melanoma but did not contain the terms “screening” or “surveillance testing.”
The authors obtained a hard copy of each of the articles identified by the preceding steps and reviewed them in detail. They excluded all of the articles that did not contain any original patient data and those that reported only baseline systemic testing at the time of diagnosis or initial treatment of the primary intraocular tumor. The authors classified each of the remaining articles after these exclusions in terms of: 1) its type (review article, letter to the editor, case report, descriptive case series report, or comparative clinical trial report); 2) the size of the reported case group; 3) the category of the reported study (retrospective or prospective, descriptive or comparative, and cross-sectional or longitudinal); and 4) the outcome or outcomes evaluated in the study.
Finally, to specifically evaluate the level of evidence regarding the impact of surveillance testing on survival provided by the published peer-reviewed articles, the authors classified each article that reported total survival following treatment of the primary intraocular tumor into ones that did vs did not contain comparative survival data for subgroups that did and did not undergo surveillance testing after initial treatment of the primary uveal melanoma.
Results
Our literature search on the terms “uveal melanoma,” “choroidal melanoma,” “ciliary body melanoma,” and “iris melanoma” identified 4222 published articles on this topic during the 30 years encompassed by this study. Search of this set of articles on the supplemental term “metastasis” reduced the number of articles to 682, and search of this reduced set of articles on the terms “screening” and “surveillance” further reduced the number of articles to 16. Review of the titles and abstracts of the first 20 articles linked to each of these 16 satisfactory articles by the “related articles” function of PubMed identified 28 additional articles that mentioned some form of follow-up testing or systemic evaluation to detect metastasis from primary uveal melanoma but did not specifically mention the terms “screening” or “surveillance testing”.
Four of the 44 articles identified by the preceding steps were eliminated because they were literature reviews that did not contain any original or relevant patient data, and 9 others were eliminated because they reported only baseline data on the evaluated patients. This left us with a final set of 31 satisfactory articles. Detailed review of these 31 published articles showed 29 to be case series reports and 1 each to be a letter to the editor containing original patient surveillance data and a survey of practice patterns. None of the articles was a report of a randomized or nonrandomized comparative clinical trial of total post-treatment survival in subgroups assigned to regular periodic surveillance for metastasis vs no surveillance testing. The size of the reported patient group was ≥100 patients in 12 articles, ≥30 but <100 patients in 12 articles, and ≥4 but < 30 patients in 6 articles. The 1 article that reported a survey of practice patterns did not contain any original patient-specific data. Eighteen of the 31 reported studies were retrospective while 13 were prospective (at least in part). Twenty-seven were descriptive, and 22 were comparative for at least some aspect of the evaluated outcomes (see next paragraph). Twenty-two of the reported studies were longitudinal (at least in part) and 14 were cross-sectional in nature.
The outcomes evaluated in the reported studies included levels of specific biomarkers (14 articles), percentage of patients with abnormal results on surveillance testing (13 articles), values of diagnostic markers (eg, sensitivity, specificity, positive predictive value) associated with the evaluated components of a surveillance regimen (7 articles), survival time after first detection of metastasis from primary uveal melanoma (7 articles), total survival time after initial diagnosis or initial treatment of the primary intraocular tumor (3 articles), percentage of patients who developed metastasis whose metastatic tumors were detected by presymptomatic testing (5 articles), surveillance regimens employed by different groups (1 article), and relationship with generally accepted clinical and histopathologic prognostic factors for uveal melanoma metastasis (1 article). None of the identified articles reported survival times of comparable subgroups of patients in which regular periodic surveillance for metastasis vs no surveillance had been performed.
Of the 3 articles that reported total survival time between initial treatment of the primary uveal melanoma and death from any cause, only 1 reported subgroup survival data for asymptomatic vs symptomatic patients. Unfortunately, the authors of this article did not report comparative baseline clinical prognostic factors for metastasis in the symptomatic and asymptomatic subgroups, so one cannot comment on the similarity vs dissimilarity of these subgroups in terms of these factors. Although the authors reported that detection of metastasis by surveillance testing occurred in 93 of 138 cases in which this information was available, they did not report comparative overall subgroup survival for the patients whose metastasis was detected by surveillance vs post-symptomatic testing.
Results
Our literature search on the terms “uveal melanoma,” “choroidal melanoma,” “ciliary body melanoma,” and “iris melanoma” identified 4222 published articles on this topic during the 30 years encompassed by this study. Search of this set of articles on the supplemental term “metastasis” reduced the number of articles to 682, and search of this reduced set of articles on the terms “screening” and “surveillance” further reduced the number of articles to 16. Review of the titles and abstracts of the first 20 articles linked to each of these 16 satisfactory articles by the “related articles” function of PubMed identified 28 additional articles that mentioned some form of follow-up testing or systemic evaluation to detect metastasis from primary uveal melanoma but did not specifically mention the terms “screening” or “surveillance testing”.
Four of the 44 articles identified by the preceding steps were eliminated because they were literature reviews that did not contain any original or relevant patient data, and 9 others were eliminated because they reported only baseline data on the evaluated patients. This left us with a final set of 31 satisfactory articles. Detailed review of these 31 published articles showed 29 to be case series reports and 1 each to be a letter to the editor containing original patient surveillance data and a survey of practice patterns. None of the articles was a report of a randomized or nonrandomized comparative clinical trial of total post-treatment survival in subgroups assigned to regular periodic surveillance for metastasis vs no surveillance testing. The size of the reported patient group was ≥100 patients in 12 articles, ≥30 but <100 patients in 12 articles, and ≥4 but < 30 patients in 6 articles. The 1 article that reported a survey of practice patterns did not contain any original patient-specific data. Eighteen of the 31 reported studies were retrospective while 13 were prospective (at least in part). Twenty-seven were descriptive, and 22 were comparative for at least some aspect of the evaluated outcomes (see next paragraph). Twenty-two of the reported studies were longitudinal (at least in part) and 14 were cross-sectional in nature.
The outcomes evaluated in the reported studies included levels of specific biomarkers (14 articles), percentage of patients with abnormal results on surveillance testing (13 articles), values of diagnostic markers (eg, sensitivity, specificity, positive predictive value) associated with the evaluated components of a surveillance regimen (7 articles), survival time after first detection of metastasis from primary uveal melanoma (7 articles), total survival time after initial diagnosis or initial treatment of the primary intraocular tumor (3 articles), percentage of patients who developed metastasis whose metastatic tumors were detected by presymptomatic testing (5 articles), surveillance regimens employed by different groups (1 article), and relationship with generally accepted clinical and histopathologic prognostic factors for uveal melanoma metastasis (1 article). None of the identified articles reported survival times of comparable subgroups of patients in which regular periodic surveillance for metastasis vs no surveillance had been performed.
Of the 3 articles that reported total survival time between initial treatment of the primary uveal melanoma and death from any cause, only 1 reported subgroup survival data for asymptomatic vs symptomatic patients. Unfortunately, the authors of this article did not report comparative baseline clinical prognostic factors for metastasis in the symptomatic and asymptomatic subgroups, so one cannot comment on the similarity vs dissimilarity of these subgroups in terms of these factors. Although the authors reported that detection of metastasis by surveillance testing occurred in 93 of 138 cases in which this information was available, they did not report comparative overall subgroup survival for the patients whose metastasis was detected by surveillance vs post-symptomatic testing.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
