Problems in defining cutoff points of continuous prognostic factors: example of tumor thickness in primary cutaneous melanoma

Opinion: on the importance of maintaining the functional form of explanatory variables

In medical research, continuous variables are often categorised into two or more groups before being included in the analysis; this practice often comes with a cost, such as loss of power in analysis, less reliable estimates, and can often leave residual confounding in the results. In this research report, we show this by way of estimates from a regression analysis looking at the association between acute kidney injury and post-operative mortality in a sample of 194 neonates who underwent the Norwood operation. Two models were developed, one using a continuous measure of renal function as the main explanatory variable and second using a categorised version of the same variable. A continuous measure of renal function is more likely to yield reliable estimates and also maintains more statistical power in the analysis to detect a relation between the exposure and outcome. It also reveals the true biological relationship between the exposure and outcome. Categorising a continuous variable may not only miss an important message, it can also get it wrong. Additionally, given a non-linear relationship is commonly encountered between the exposure and outcome variable, investigators are advised to retain a predictor with a linear term only when supported by data. All of this is particularly important in small data sets which account for the majority of clinical research studies.

Association between preoperative hematologic markers and aggressive behavior in meningiomas

INTRODUCTION

Meningiomas have varying degrees of aggressive behavior. Some systemic hematologic makers are associated with malignancy, but their value in predicting aggressive meningioma behavior is not fully understood.

OBJECTIVE

To evaluate the association between preoperative markers such as neutrophil-lymphocyte ratio (NLR), neutrophil-monocyte ratio (NMR), monocyte-lymphocyte ratio (MLR), platelet-lymphocyte ratio (PLR), and prognostic nutritional index (PNI), and diagnostic and prognostic factors including WHO grade, proliferation index, presence of edema on preoperative MRI, and tumor recurrence.

METHODS

A retrospective review of patients treated between 2000 and 2019 with a preoperative complete blood count (CBC) differential lab draw before intracranial meningioma resection was conducted. All preoperative steroid dosages were converted to dexamethasone equivalents. Primary outcomes included presence/absence of perilesional edema, WHO grade, Ki-67/MIB-index, and recurrence. Univariate and multivariable regression analyses were conducted.

RESULTS

A total of 209 meningioma patients were included. Of these, 143 (68 %) were WHO grade I, 61 (29 %) grade II and 5 (2 %) were grade III. Recurrence was reported in 19 (9.1 %) tumors. No hematologic markers were associated with recurrence. In separate multivariable logistic analyses, no biomarkers were associated with perilesional edema or WHO grade. MLR was associated with higher MIB-index (p = 0.018, OR 6.57, 95 % CI 1.37-30.91).

CONCLUSION

Most hematologic markers were not associated with meningioma invasiveness, grade, proliferative index, or aggressiveness. Preoperative MLR was associated with high proliferation index in patients undergoing surgery for intracranial meningioma. Higher MLR could be a surrogate for meningioma proliferation and has potential to be used as an adjunct for risk-stratifying meningiomas.

Subgroup analyses in randomized controlled trials frequently categorized continuous subgroup information

BACKGROUND AND OBJECTIVES

To investigate how subgroup analyses of published Randomized Controlled Trials (RCTs) are performed when subgroups are created from continuous variables.

METHODS

We carried out a review of RCTs published in 2016-2021 that included subgroup analyses. Information was extracted on whether any of the subgroups were based on continuous variables and, if so, how they were analyzed.

RESULTS

Out of 428 reviewed papers, 258 (60.4%) reported RCTs with a subgroup analysis. Of these, 178/258 (69%) had at least one subgroup formed from a continuous variable and 14/258 (5.4%) were unclear. The vast majority (169/178, 94.9%) dichotomized the continuous variable and treated the subgroup as categorical. The most common way of dichotomizing was using a pre-specified cutpoint (129/169, 76.3%), followed by a data-driven cutpoint (26/169, 15.4%), such as the median.

CONCLUSION

It is common for subgroup analyses to use continuous variables to define subgroups. The vast majority dichotomize the continuous variable and, consequently, may lose substantial amounts of statistical information (equivalent to reducing the sample size by at least a third). More advanced methods that can improve efficiency, through optimally choosing cutpoints or directly using the continuous information, are rarely used.

Serum levels of soluble programmed death-ligand 1 (sPD-L1): A possible biomarker in predicting post-treatment outcomes in patients with early hepatocellular carcinoma

BACKGROUND

There have been great advances in hepatocellular carcinoma management over the last years. However, there are still no prognostic biomarkers that can identify patients who will benefit the most from curative treatments. We aimed to investigate whether sPD-L1 levels measured before curative treatment is a prognostic biomarker of survival in patients with HCC.

METHODS

HCC patients from a prospectively collected database were selected and soluble programmed death-ligand1(sPD-L1) levels were determined. The association of sPD-L1 levels and overall survival (OS) and disease-free survival (DFS) was assessed.

RESULTS

One hundred twenty-one patients with HCC were included. The best cut-off value of sPD-L1 for both DFS and OS was 96 pg/mL. Patients with a high sPD-L1 value (>96 pg/mL) had a shorter disease free survival and OS (hazard ratio 5.42, 95% confidence interval 2.28-12.91, p < 0.001, and hazard ratio 9.67, 95% confidence interval 4.33-21.59, p < 0.001). High sPD-L1 levels were associated with mortality independently from other known survival predictors. We found a positive correlation between sPD-L1 and PD-L1 expression in cancer cells (p = 0.01). In 16 out of 38 patients, sPD-L1 levels decreased from baseline value on week 6 after treatment and in 22 out of 38 patients, sPD-L1 levels increased from the baseline value. However, fluctuations of sPD-L1 in time had no influence on survival (p = 0.148).

CONCLUSION

We conclude that a high sPD-L1 level is a biomarkerfor a poor outcome in HCC. The predictive value of sPD-L1 levels for a successful anti-PD1/PD-L1 therapy should be investigated in the future.

The Ki-67 Proliferation Index as a Marker of Time to Recurrence in Intracranial Meningioma

BACKGROUND

There are examples of incongruence between the WHO grade and clinical course in meningioma patients. This incongruence between WHO grade and recurrence has led to search for other prognostic histological markers.

OBJECTIVE

To study the correlation between the Ki-67 proliferative index (PI), risk of recurrence, and recurrence rates in meningioma patients.

METHODS

We prospectively collected pathological diagnosis of de novo consecutive meningiomas. In total, we followed 159 patients with clinical controls until recurrence, death, or emigration. We estimated the correlation between risk of recurrence and Ki-67 PI when adjusted for age at diagnosis, sex, WHO grade, extent of surgical resection, and tumor location. We estimated the cumulative incidence of recurrence when considering death without recurrence a competing risk. We report recurrence rates per 100 person-years.

RESULTS

A 1%-point increase of Ki-67 PI yielded a hazard ratio of 1.12 (95% CI: 1.01-1.24) in a multivariate analysis. The cumulative incidence of recurrence was 3% for Ki-67 0% to 4% vs 19% for Ki-67 > 4% meningiomas after 1 yr, but 24% vs 35%, respectively, after 10 yr. There was no significant difference in mean Ki-67 PI between nonrecurrent and recurrent meningioma in a 2-sample t-test (P = .08). The strongest relationship was detected between Ki-67 PI and time to recurrence: Ki-67 < 4% meningiomas recurred after median 4.8 yr, compared to 0.60 to 0.75 yr for patients with higher Ki-67 PI.

CONCLUSION

Ki-67 PI was a marker for time to recurrence rather than a predictor of recurrence. Ki-67 PI may be utilized for patient tailored follow-up.

Comparative Evaluation of Median Versus Youden Index Dichotomization Methods: Exposure-Response Analysis of Mycophenolic Acid and Acyl-Glucuronide Metabolite

BACKGROUND AND OBJECTIVES

Dichotomization of pharmacokinetic exposure measures in exposure-response relationship studies provides results that are interpretable in clinical care. Several methods exist in the literature on how to define the cut-off values needed for the dichotomization process. Commonly, the sample median is utilized to define the dichotomizing value; however, statistical methods based on the exposure metric and its association with the outcome are argued to result in a more proper definition of the optimal cut-point. The Youden index is a recommended statistical method to define the cut-off value. The current analysis objective is to compare the dichotomization results based on the Youden index versus median methods.

METHODS

Utilizing mycophenolic acid (MPA) exposure data and its related acute rejection and leukopenia outcome variables, the current study compared the MPA exposure-response relationship outcomes when MPA exposure is dichotomized via the Youden index versus median methods. Univariate logistic models were utilized to quantify the relationships between MPA exposure, including total MPA, unbound MPA, and the acyl-glucuronide metabolite of MPA, and the probabilities of acute rejection and leukopenia.

RESULTS

The overall trend of the results of the logistic models demonstrated a general similarity in the inferred exposure-response associations when considering either the Youden index-based or the median-based dichotomization methods.

CONCLUSION

The results demonstrated in this analysis suggest that both the Youden index and the median methods provide similar conclusions when dichotomization of a continuous variable is considered. However, confirmation of these conclusions comes from future powered studies that include a larger number of subjects.

Patient-related risk factors for unplanned 30-day readmission following total knee arthroplasty: a narrative literature review

Osteoarthritis is a debilitating condition as well as a growing global health problem, and total knee arthroplasty (TKA) is an effective treatment for advanced disease. Unplanned 30-day hospital readmission is an indicator of complications, which is a significant financial burden on healthcare systems. We reviewed the literature to better understand the patient-related factors associated with unplanned 30-day readmission following TKA. MEDLINE and EMBASE were searched for studies reporting on patient-related risk factors for unplanned 30-day readmission following primary or revision TKA for any indication. The impact of specific medical comorbidities on increasing the risk of 30-day readmission following TKA is quite well established. The following comorbidities are strongly associated with readmission: bleeding disorder, diabetes, chronic kidney disease and dialysis, chronic immunosuppressant use and history of cancer. Other significant comorbidities include: dementia; depression; haematological (coagulopathy and anaemia), cardiovascular (atrial fibrillation, cardiovascular disease, coronary artery disease and congestive heart failure), respiratory (chronic obstructive pulmonary disease) and liver diseases; and cerebrovascular accident/transient ischaemic attack (but only in revision TKA patients). The influence of variation in sex, age and body mass index each demonstrate a more complex pattern. A systematic review and meta-analysis is required to quantify the impact of the various patient-related factors on 30-day readmission following TKA. Clinicians can use this information in preoperative decision-making.

Validation, clinical utility and limitations of the Amsterdam-Oxford model for primary sclerosing cholangitis

BACKGROUND & AIMS

Recently the Amsterdam-Oxford model (AOM) was introduced as a prognostic model to assess the risk of death and/or liver transplantation (LT) in primary sclerosing cholangitis (PSC). We aimed to validate and assess the utility of the AOM.

METHODS

Clinical and laboratory data were collected from the time of PSC diagnosis until the last visit or time of LT or death. The AOM was calculated at yearly intervals following PSC diagnosis. Discriminatory performance was assessed by calculation of the C-statistic and prediction accuracy by comparing the predicted survival with the observed survival in Kaplan-Meier estimates. A grid search was performed to identify the most discriminatory AOM threshold.

RESULTS

A total of 534 patients with PSC and a mean (SD) age of 39.2 (13.1) years were included. The diagnosis was large duct PSC in 466 (87%), PSC with features of autoimmune hepatitis in 52 (10%) and small-duct PSC in 16 (3%). During the median (IQR) follow-up of 7.8 (4.0-12.6) years, 167 patients underwent LT and 65 died. The median LT-free survival was 13.2 (11.8-14.7) years. The C-statistic of the AOM ranged from 0.67 at baseline to 0.75 at 5 years of follow-up. The difference between the predicted and observed survival ranged from -1.6% at 1 year to + 3.9% at 5 years of follow-up. Patients that developed AOM scores >2.0 were at significant risk of LT or death (time-dependent hazard ratio 4.09; 95% CI 2.99-5.61).

CONCLUSIONS

In this large cohort of patients with PSC, the AOM showed an adequate discriminative performance and good prediction accuracy at PSC diagnosis and during follow-up. This study further validates the AOM as a valuable risk stratification tool in PSC and extends its utility.

LAY SUMMARY

In our study we assessed whether the Amsterdam-Oxford model (AOM) is able to correctly estimate the risk of liver transplantation or death in patients with primary sclerosing cholangitis (PSC). This model uses 7 objective and readily available variables to estimate prognosis for individual patients at the time of PSC diagnosis. The AOM may aid in patient counselling and timing of diagnostic procedures or therapeutic interventions for complications of liver disease. We confirm that the model works well at PSC diagnosis, but also when the AOM is recalculated at different timepoints during follow-up, greatly improving the applicability of the model in clinical practice and for individual patients.

Determination of Cut Point in the Age of Colorectal Cancer Diagnosis Using a Survival Cure Model

Background and Objectives:

Colorectal Cancer (CRC) is the fourth cancer-related cause of death worldwide. CRC is a multi-stage cancer, which is curable during the early stages of the disease. Therefore, determining the time of cut-point existence could improve treatment planning and help directly allocate resources. This study aims to determine the cut point in the age of CRC diagnosis.

Methods:

This study, covering the course 1985 to 2012, consisted of 345 colorectal cancer patients registered in Taleghani Hospital, Tehran, Iran and followed up to 2013. The cut-point in the age of CRC diagnosis was obtained using a mixture cure model. The data were analyzed using SPSS and R, V. 20 and 2.15.0, respectively.

Results:

The results showed that the cut point in the age of CRC diagnosis was 50 years. Based on our estimation, 65% of the patients diagnosed with CRC at or younger than 50 were cured, while 31% of them diagnosed older than 50 were cured, and the younger group had a better survival over the older group.

Conclusion:

Since access to a cut-point and analysis of created prognostic groups are important in screening and treatment planning, our results suggested that it is better to estimate the cut-point in the age of curable cancers in early stages via survival cure models, and the cure rate would increase by CRC timely screening.

Long-Term Survival of Patients with Thin (T1) Cutaneous Melanomas: A Breslow Thickness Cut Point of 0.8 mm Separates Higher-Risk and Lower-Risk Tumors

BACKGROUND

Counterintuitively, more deaths from melanoma occur among patients with thin (T1) primary melanomas (≤ 1 mm) than among those with thick primary melanoma because the great majority present with T1 tumors. Therefore, it is important to stratify their risk as accurately as possible to guide their management and follow-up. This study sought to explore the relationship between tumor thickness and prognosis for patients with thin primary melanomas.

METHODS

A retrospective, single-institution study investigated 6263 patients with cutaneous melanoma (including 2117 T1 cases) who had a minimum follow-up period of 10 years.

RESULTS

For the entire patient cohort, the 10-year melanoma-specific survival (MSS) rate ranged between 92% for the patients with primary melanomas up to 0.3 mm thick and 32% for those with melanomas thicker than 8 mm. When divided into 25-quantile-thickness groups there was a significant difference in 10-year MSS between the two consecutive groups 0.8 and 0.9 mm; the differences in survival were not significantly different for any other consecutive cut points within the less than or equal to 1 mm thickness range, indicating a biologically-relevant difference in outcome above and below 0.8 mm. For the patients treated initially at the authors' institution, the 10- and 20-year MSS rates for those with tumors up to 0.8 mm thick were respectively 93.4 and 85.7%, and for tumors 0.9 to 1.0 mm, the rates were respectively 81.1 and 71.4%. Only 29.3% of the T1 patients who died of melanoma were deceased within 5 years.

CONCLUSIONS

A naturally occurring thickness cut point of 0.8 mm predicts higher or lower risk for patients with thin primary cutaneous melanomas. Long-term follow-up assessment of patients with T1 melanoma is important because late mortality due to melanoma is more common than early mortality.

Alkaline phosphatase at diagnosis of primary sclerosing cholangitis and 1 year later: evaluation of prognostic value

BACKGROUND

Primary sclerosing cholangitis (PSC) is a slowly progressive liver disease. Reliable biomarkers to predict outcome are urgently needed to serve as surrogate endpoints and/or stratifiers in clinical trials. Reduction in serum alkaline phosphatase (ALP) has been proposed as prognostic surrogate marker in PSC. The aim of this study was to asses if ALP at diagnosis (T0), 1 year later (T1), and percentage change between both time points hold prognostic value, and to determine the optimal threshold.

METHODS

We retrospectively collected ALP levels at T0 and T1 for patients included in a large PSC cohort. The association of ALP at T0, T1, and percentage change with the combined endpoint (PSC-related death, liver transplantation) was analysed. Predictive value was determined using C-statistics.

RESULTS

A total of 366 patients were included, of whom 66 (18%) reached an endpoint: 26 (7%) PSC-related death, 40 (11%) liver transplantation. At T0 and T1, 84% used ursodeoxycholic acid. A positive association was observed between level of ALP at T0 and T1 and the hazard of reaching an endpoint, up to values around 2.5 times upper limit of normal (xULN). A larger decrease in ALP between T0 and T1 decreased the event rate. A range of thresholds (0.5-3×ULN) with about similar C-statistics was found. In this cohort, the optimal threshold was 1.3×ULN at T1.

CONCLUSION

ALP can be used to discriminate between PSC patients with a good and a poor prognosis. These findings indicate that ALP can serve as stratifier, and potentially as surrogate endpoint for clinical trials in PSC.

Predictive value of tumor thickness for cervical lymph-node involvement in squamous cell carcinoma of the oral cavity: a meta-analysis of reported studies

BACKGROUND

Tumor thickness (TT) appears to be a strong predictor for cervical lymph-node involvement in squamous cell carcinoma of the oral cavity (OSCC), but a precise clinically optimal TT cutoff point has not been established. To address this question, the authors conducted a meta-analysis.

METHODS

All relevant articles were identified from MEDLINE and EMBASE as well as from cross-referenced publications cited in relevant articles. Lymph-node involvement was confirmed and identified as positive lymph-node declaration (P(LN)D) by either pathologic positivity on immediate neck dissection or by neck recurrence identified after follow-up > or = 2 years. Odds ratios (OR) were calculated to quantify the predictive value of TT. Negative predictive values (and the percentage of patients falsely predicted to not have P(LN)D [FN-P(LN)D]) were compared to determine the optimal TT cutoff point.

RESULTS

Sixteen studies were selected from 72 potential studies, yielding a pooled total of 1136 patients. Data were examined for the following TT cutoff points: 3 mm (4 studies, 387 patients), 4 mm (9 studies, 778 patients), 5 mm (6 studies, 367 patients), and 6 mm (4 studies, 488 patients). The OR (95% CI) was 7.3 (5.3-10.1) for the overall group. The proportion of FN-P(LN)D was 5.3% (95% CI, 2.0-11.2), 4.5% (2.6-7.2), 16.6% (11.5-22.8), and 13.0% (9.7-16.9) for TT<3, <4, <5, and <6 mm, respectively. There was a statistically significant difference between the 4-mm and 5-mm TT cutoff points (P = .007).

CONCLUSIONS

TT was a strong predictor for cervical lymph-node involvement. The optimal TT cutoff point was 4 mm.

Melanoma epidemiology and trends

Rising incidence rates of cutaneous melanoma have been observed during the last four decades in white populations worldwide. The cancer statistics in the United States have revealed 6 cases per 100,000 and year at the beginning of the 1970s and 18 cases per 100,000 inhabitants and year at the beginning of 2000, demonstrating a threefold increase in incidence rates. Incidence rates in central Europe increased in the same time period from 3 to 4 cases to 10 to 15 cases per 100,000 inhabitants and year, which is very similar to the increase in the United States. Cohort studies from several countries indicate that the trend of increasing incidence rates will continue in the future for at least the next 2 decades; thus, an additional doubling of incidence rates is expected. The highest incidence rates have been reported from Australia and New Zealand, from 40 to 60 cases per 100,000 inhabitants and year. Mortality rates likewise slightly increased in the United States and in Europe during the 1970s and 1980s. In the 1990 s, however, a leveling off of mortality rates was observed in many countries. Simultaneously, a clear decrease of Breslow tumor thickness was reported in the United States and European countries. This development indicates improved early recognition of cutaneous melanoma, which is presently the main factor for a more favorable prognosis.

A systematic review of sociodemographic, physical, and psychological predictors of multidisciplinary rehabilitation-or, back school treatment outcome in patients with chronic low back pain

STUDY DESIGN

A systematic review.

OBJECTIVE

To determine predictors of outcome of multidisciplinary rehabilitation-or back school treatment for patients with chronic low back pain.

SUMMARY OF BACKGROUND DATA

Numerous reviews have been performed to gain insight into which patients benefit from which treatment. However, no review has systematically focused on predictors from multiple domains (i.e., sociodemographic, physical, and psychological), or on treatment outcome measured as activity limitation or participation restriction.

METHODS

Studies were found by searching medical and psychological databases, and screening references. Two reviewers independently assessed the methodological quality using standard criteria. Studies were only included if they met a predefined level of internal validity. A qualitative analysis was performed.

RESULTS

Heterogeneity among studies in patient characteristics, predictors, treatment, and outcomes limited evidence. All reviewed studies were descriptive or exploratory in nature. Consistent evidence was found for the predictive value of pain intensity (more pain--> worse outcome), several work-related parameters (e.g., high satisfaction--> better outcome), and coping style (less active coping--> better outcome). Other sociodemographic and physical variables consistently lacked predictive value. No consistent evidence was found for other psychological variables.

CONCLUSIONS

It is impossible to define a generic set of predictors of outcome of multidisciplinary rehabilitation and back schools for patients with chronic low back pain because the reviewed studies were descriptive or exploratory in nature, and most predictors were only studied once. Nevertheless, for several predictors, consistent evidence was found. Large confirmatory studies are needed to test the value of these predictors.

A Molecular Signature of the Nottingham Prognostic Index in Breast Cancer

The Nottingham prognostic index (NPI) is a widely used clinicopathological staging system for breast cancer prognostication. Using a step-wise classification approach where breast tumor expression profiles were first divided into general "molecular subtypes" [estrogen receptor (ER)+, ER-, ERBB2+], followed by an independent analysis of each subtype, we identified a 62-gene expression signature (NPI-ES) highly correlated to the NPI in ER+ tumors. The NPI-ES classified the majority of ER+ tumors into two distinct groups with high confidence and was significantly correlated to NPI status in two independent sets of ER+ tumors derived from different centers. The NPI-ES is comparable to the classical NPI in identifying patients likely to exhibit a poor clinical prognosis, as well as to a recently described "prognosis signature" for breast cancer. Our findings demonstrate how expression profiling can complement classical staging systems employing histopathological parameters scored over a continuous range of values.

Using a continuous transformation of the Breslow thickness for prognosis in cutaneous melanoma

Although the Breslow thickness provides the most important histologic information for prognosis in cutaneous melanoma, controversies and uncertainty remain about how best to use thickness. It is unclear whether cut points should be used, or, if they are used, which are optimal. We studied new data collected from more than 1,000 patients followed up for a relatively long period. From Cox proportional hazards models of survival we learned that more cut points provide more prognostic information than using, for example, just 1 cut point at 1.7 mm. Nevertheless, a continuous transformation provides an effective alternative that captures the information that thickness provides, and it avoids the pitfalls of using multiple cut points. In a multivariate model, this transformation provided strong prognostic information, and the result produced a prognostic score for cutaneous melanoma. This score provides a practical way that Cox model results can be used, and we believe it consolidates the prognostic information provided by traditional histologic and clinical variables. When newer prognostic variables are introduced, we suggest that they be used with this continuous transformation of thickness rather than with cut points in thickness.

Pathology characteristics that optimize outcome prediction of a breast screening trial

The ability of pathology characteristics to predict outcome was tested with the 1029 cancers accumulated in the Edinburgh Randomized Trial of breast screening after 14 years follow-up. The majority (55.7%) were in the screening arm, which also had more operable cases (81.3% vs 62.2%); the reduction in the proportion of inoperable breast cancers in a UK female population invited to mammographic screening is a notable effect of the trial. In the 691 operable invasive cases the size, histological type, grade, node status and node number group individually showed highly significant (P<0.001) association with survival. In multivariate analysis the Nottingham Prognostic Index (NPI) derived from these features showed highly significant association with survival (P<0.001). However, when first adjusted for NPI, combined addition of pathological size in 6 categories and histological type as special or not had an independent association with survival that was statistically firmly based (P<0.001). For operable breast cancer the gains are in smaller sizes, better histological features, and higher proportion node negative. The weighting factors applied to pathology indicators of survival in the NPI are not optimal for a population included in a trial of screening. In particular, a linear trend of the index with pathological size is not appropriate. Inclusion of histological type as special or not improves the index further.

The delimitation of tubular carcinoma of the breast

Five hundred consecutive breast carcinomas from the first screening round of the Breast Cancer Detection Demonstration Project were studied quantitatively and semiquantitatively for features relevant to the diagnosis of tubular carcinoma. Tubularity was defined as the proportion of tumor cells that were adjacent to an open lumen. Nuclear morphology and mitotic activity were graded 1 to 3, and the presence of apocrine snouts as absent, few, common, or prominent. In plots and statistical cluster analysis, tubular carcinoma appears as part of a continuous spectrum of morphologies and not as a distinct entity. In multivariate analysis, apocrine snouts had no significant association with either nodal status or deaths of breast cancer. Tumors with 70% or greater tubularity by our definition and mitosis and nuclear grades 1 were not associated with either nodal metastases or deaths of breast cancer. The question is raised whether tubular carcinoma at the benign end of a spectrum shades into benign glandular proliferations, with particular reference to microglandular adenosis. A uniform and precise definition of tubularity is needed for the attainment of sufficient collective experience to delimit tubular carcinoma both from more aggressive carcinomas and from benign proliferations.

Methodological challenges in the evaluation of prognostic factors in breast cancer

Many studies are carried out in an effort to find factors that help explain the large unexplained variation in prognosis of breast cancer patients. The principles of good study design and analysis are less well appreciated for prognostic factor studies than for therapeutic trials. The oncology literature is full of results from studies of varying quality, often with conflicting findings. As a consequence, despite the large number of studies, there is still uncertainty about the importance of most prognostic factors. Few recently proposed prognostic factors for breast cancer have become widely accepted. This paper reviews the important methodological issues underlying such research. These issues are illustrated with examples from published studies and recent reviews of papers published in cancer journals. Guidelines are proposed for conducting and evaluating prognostic factor studies which should improve the quality of research in this important area.